Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Lateral sound flanking transmission at curtain wall mullions: an empirical investigation to identify controlling mechanisms
(USC Thesis Other)
Lateral sound flanking transmission at curtain wall mullions: an empirical investigation to identify controlling mechanisms
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
LATERAL SOUND FLANKING TRANSMISSION AT CURTAIN WALL MULLIONS:
AN EMPIRICAL INVESTIGATION TO IDENTIFY CONTROLLING MECHANISMS
By
Elizabeth Valmont
Bachelor of Architecture
Master of Building Science
University of Southern California
______________________________________________________________________________________
A Dissertation Proposal Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
In Partial Fulfillment of the Requirements for the Degree
DOCTOR OF PHILOSOPHY IN ARCHITECTURE
August 2015
Copyright 2015 Elizabeth Valmont
2
ACKNOWLEDGMENTS
The dissertation has brought about great learning and life challenges. Those who have been part of this
journey have contributed to my life in ways that are beyond the expression of these acknowledgements.
I would like to firstly thank my dissertation committee who have steered my work to completion. Most
especially Professor Douglas Noble, the committee chair, for his mentoring, guidance and persistent
encouragement to achieve my best. Professor Marc Schiler, one my first mentors in the building sciences,
academic teaching and profound thinking, always supported the direction of my work. Jerry Christoff, my
first acoustic mentor, always advocated for me to pursue the PhD and this topic, and made time to see
me succeed to the end. Professor James Moore provided mentoring to help me delineate the big picture
of the research study and clarify.
Additionally I especially thank those at Enclos and WEAL who contributed to the research project. Mic
Patterson kick started the process with his motivation and belief in me, and presented TJ Deganyar at
Enclos. TJ Deganyar was responsible for realizing the laboratory test circuits and was also very passionate
about this research, providing both supportive time and funding. Gary Mange, the former lab director at
WEAL, provided technical counsel and research support during all the test phases. Raul Martinez, the
laboratory technician at WEAL, labored with me and provided innovative ways to rig the full scale building
test specimens.
Hooshang Khosrovani, from Veneklasen, provided considerable time and great mentorship and advice to
help me think critically about the research structure and questions. My great appreciation goes to Karen
Kensek who thoughtfully reviewed and critically edited the dissertation. Adam Foxwell from Arup took
the time to review my work and gave me great inspiration. Rachid Abu Hassan, also from Arup, helpfully
reviewed the initial idea with me.
The many laboratory measurements, documentation, photography, equipment, and moral support would
not have been possible without the assistance of my fellow colleagues and friends. I owe an enormous
thank you to the MBS and PhD colleagues (Eve Shih-Hsin Lin, Jae Yong Suk, Yara Masri, Andrea Martinez-
Arias, and Ed Losch), Enclos colleagues (James Casper and Daniel Bettenhausen), and Veneklasen
colleagues. I particularly appreciate Eve, generous with her knowledge and time. Rick Lasser from Arup
helped with photography and Patrick Masson was the first to help me on the first day of testing.
I appreciate all financial grants, funding and awards from Enclos Corp, WEAL, Thornton Thomasetti, Arup,
Newman Medal, and the Architectural Research Centers Consortium.
I would also like to acknowledge the Ph.D. program and the Chase L. Leavitt Graduate Building Science
program at the University of Southern California. They provided a great scholarly community of professors
and students conducting inspiring and outstanding research. I am honored to contribute to the seminal
research of the Ph.D. program in Architecture at the University. Professor Douglas Noble is responsible
for pioneering this unique program that targets the advancement of multidisciplinary research in façade
tectonics and without whom I would have never entered this program.
Most importantly, I thank my parents and family. Their love and support helped me to persevere and
strengthen my awareness.
3
TABLE OF CONTENTS
ACKNOWLEDGMENTS 2
LIST OF TABLES 8
LIST OF FIGURES 10
ABSTRACT 16
TERMS & ABBREVIATIONS 18
ACOUSTIC TERMS/TEST .................................................................................................................. 18
ACOUSTIC ABBREVIATIONS ............................................................................................................. 18
CHAPTER 1 INTRODUCTION TO CURTAIN WALL SOUND FLANKING
TRANSMISSION 20
1.1 ACOUSTICRELEVANCE TO GLASS CURTAIN WALL DESIGN ............................................................. 20
1.2 PROBLEMSTATEMENT AND HYPOTHESIS .................................................................................... 23
1.2.1 Building Sound Flanking Paths .............................................................................. 25
1.2.2 Composite Transmission Loss ................................................................................ 26
1.3OVERVIEW OF RESEARCH .......................................................................................................... 27
1.3.1 Research Objectives ............................................................................................... 28
1.3.2 Research Method ................................................................................................... 28
1.3.3 Approach to Measuring the Test Specimens ........................................................ 29
1.4 DISSERTATIONOUTLINE ........................................................................................................... 30
CHAPTER 2 FAÇADE BACKGROUND REVIEW AND PROFESSIONAL APPLICATIONS
IN ACOUSTICS 32
2.1 ACOUSTICDETAILING AT CURTAIN WALL MULLIONS .................................................................... 32
2.1.1 The Anatomy of the Glass Curtain Wall ................................................................ 32
2.1.2 Multidisciplinary Design Provisions....................................................................... 39
2.1.3 Acoustic Detail Considerations .............................................................................. 41
2.2 ACOUSTICMITIGATIONPRACTICES AND PRODUCTS FOR MULLIONS ............................................... 42
2.2.1 Practice-Based Mullion Modifications .................................................................. 42
2.2.2 Product-Based Mullion Modification .................................................................... 45
2.2.3 Summary of Practice and Product Solutions......................................................... 48
2.3 SOUNDISOLATIONMETRICS AND MEASUREMENT METHODS ........................................................ 49
2.3.1 ASTM Standard Test Methods and Rating Procedures ........................................ 49
2.3.2 ISO Standard Test Methods and Rating Procedures ............................................ 52
4
2.3.3 Sound Isolation Criteria ......................................................................................... 55
2.4 ACOUSTICPRECEDENTRESEARCHSTUDIES .................................................................................. 58
2.4.1 Laboratory Entities and Investigations of Façade Sound Isolation ...................... 58
2.4.2 Sound Flanking Prediction Methods ..................................................................... 60
2.4.3 Precedent Acoustic Test Measurements of Curtain Wall Elements ..................... 61
2.5 SUMMARY OF PROFESSIONAL BACKGROUND AND PRECEDENT RESEARCH ....................................... 66
CHAPTER 3 UNITIZED VERTICAL MULLION RESEARCH METHODOLOGY 67
3.1 INTRODUCTION ....................................................................................................................... 67
3.2 REVIEW OF ACOUSTIC PRACTICES AND PROCEDURES AT CURTAIN WALL FACADES ............................ 68
3.3 LABORATORYTESTPROCEDURE ................................................................................................. 68
3.3.1 Test Constants and Variables ................................................................................ 70
3.3.2 Curtain Wall System Specimen.............................................................................. 70
3.3.3 Test Specimen Components .................................................................................. 73
3.3.4 Laboratory Test Chambers .................................................................................... 75
3.3.5 Test Experiment Factors ........................................................................................ 82
3.4 LABORATORYRESULTANALYSIS ................................................................................................. 82
3.4.1 Sound Transmission Correlations and Comparisons ............................................ 82
3.4.2 Composite Transmission Loss Predictions............................................................. 82
CHAPTER 4 TEST RESULTS AND ANALYSIS OF THE UNITIZED VERTICAL MULLION
MEASUREMENT PHASES 84
4.1 INTRODUCTION ....................................................................................................................... 84
4.2 PHASE1–MULLIONELEMENT(ISOLATED) ................................................................................. 85
4.2.1 Phase 1 Specimen and Test Chamber Description ................................................ 85
4.2.2 Testing Classifications ........................................................................................... 88
4.2.3 Material descriptions ............................................................................................. 89
4.2.4 Phase 1 Class A Test Sequence .............................................................................. 91
4.2.5 Phase 1 Class B Test Sequence .............................................................................. 93
4.2.7 Phase 1 Class C1 Test Sequence ............................................................................ 98
4.2.8 Phase 1 Class C2 Test Sequence .......................................................................... 101
4.2.9 Phase 1 Class C3 Test Sequence .......................................................................... 104
4.2.10 Phase 1 Class C4 Test Sequence ........................................................................ 106
4.2.11 Phase 1 Summary .............................................................................................. 110
4.3 PHASE2A–COMPOSITESEAL AND CONNECTION ELEMENTS (WITH MULLION) ............................. 113
4.3.1Phase 2A Test Specimen Description ................................................................... 114
4.3.2 Phase 2A Testing Classifications ......................................................................... 116
4.3.3 Phase 2A Class A Test Sequence ......................................................................... 119
5
4.3.4 Phase 2A Class B Test Sequence .......................................................................... 121
4.3.5 Phase 2A Class C Test Sequence .......................................................................... 124
4.3.6 Phase 2A Class D Test Sequence ......................................................................... 128
4.3.7 Phase 2A Class E Test Sequence .......................................................................... 131
4.3.8 Phase 2A Conclusion ............................................................................................ 132
4.4 PHASE2B–CONNECTIONELEMENT(WITHOUTMULLION) .......................................................... 136
4.4.1 Phase 2B Test Specimen Description................................................................... 138
4.4.3 Phase 2B Testing Classifications ......................................................................... 141
4.4.4 Phase 2B Class A Test Sequence .......................................................................... 143
4.4.5 Phase 2B Class B Test Sequence .......................................................................... 151
4.4.6 Phase 2B Class C Test Sequence .......................................................................... 156
4.4.7 Phase 2B Observations and Conclusion .............................................................. 160
4.5 PHASE3–GLAZINGASSEMBLYELEMENT ................................................................................. 161
4.5.1 Phase 3 Test Specimen Description ..................................................................... 162
4.5.2 Phase 3 Test Chamber Construction ................................................................... 163
4.5.3 Phase 3 Test Sequence ........................................................................................ 167
4.5.4 Phase 3 Summary ................................................................................................ 168
4.6 UVMMEASUREMENTSUMMARY ........................................................................................... 169
CHAPTER 5 ANALYSIS OF CONTROLLING MECHANISMS AND COMPOSITE
TRANSMISSION LOSS 171
5.1 INTRODUCTION ..................................................................................................................... 171
5.2 CONNECTED VERSUS UNCONNECTED MULLION CONDITIONS ....................................................... 171
5.2.1 Comparison: Single Figure Rating ....................................................................... 172
5.2.2 Noise Reduction Comparison: Phase 1 and Phase 3........................................... 174
5.2.3 Mechanisms Limiting Sound Isolation Performance .......................................... 181
5.3 COMPOSITETRANSMISSIONLOSSPREDICTIONS ......................................................................... 186
5.3.1 Calculation Variables and Descriptions .............................................................. 186
5.3.2 Composite TL with Low Performing UVM Elements (without glass) ................. 189
5.3.3 Composite TL with Low Performing UVM Elements (with glass) ....................... 191
5.3.4 Composite TL with High Performing UVM Elements (without glass) ................ 192
5.3.5 Composite TL with High Performing UVM Elements (with glass) ...................... 193
5.3.6 Summary of Composite TL Analysis .................................................................... 194
5.4 RANKINGRELATIVEPERFORMANCE .......................................................................................... 196
5.5 ANALYSISSUMMARY .............................................................................................................. 199
CHAPTER 6 CONCLUSION 200
6.1 INTRODUCTION ..................................................................................................................... 200
6
6.2 PRACTICES AND PROCEDURES ................................................................................................. 201
6.3 TESTMETHODOLOGY FOR UNITIZED MULLIONS ........................................................................ 202
6.4 CONTROLLINGELEMENTS AT THE UNITIZED VERTICAL MULLION .................................................. 204
6.4.1Improvements at the Unconnected Unitized Mullion ......................................... 206
6.4.2 Improvements at the Connected Unitized Mullion ............................................. 207
6.5 COMPOSITETRANSMISSIONLOSSPERFORMANCE ...................................................................... 207
6.6 LIMITATIONS ......................................................................................................................... 208
CHAPTER 7 FUTURE WORK 209
7.1 FUTURETESTING AND DESIGN INVESTIGATIONS ......................................................................... 209
7.1.1 Future Laboratory Tests Measurements ............................................................. 209
7.1.2Future Curtain Wall Design Concept Studies ....................................................... 210
7.1.3 Future Research Analytical Studies...................................................................... 210
7.2 CONCLUSION ........................................................................................................................ 211
BIBLIOGRAPHY 212
APPENDIX A TERMINOLOGY 217
A.1 LABORATORYTESTSTANDARDS ............................................................................................... 217
A.2 ACOUSTIC AND ARCHITECTURAL TERMS ................................................................................... 218
APPENDIX B UVM LABORATORY TEST RESULTS 229
B.1 INTRODUCTION ..................................................................................................................... 229
B.1.1Test Measurement Standards .............................................................................. 229
B.2 PHASE1WEALTESTRESULTS ................................................................................................ 230
B.3 PHASE2AWEALTESTRESULTS ............................................................................................. 252
B.4 PHASE2BWEALTESTRESULTS.............................................................................................. 276
B.5 PHASE3WEALTESTRESULTS ................................................................................................ 307
APPENDIX C ANCILLARY SOUND ANALYSIS 313
C.1 INTRODUCTION ..................................................................................................................... 313
C.2 PHASE1MULLIONCONTROLS ................................................................................................ 314
C.3 UVMTESTELEMENTCOMPARISON WITH A 4” AIRSPACE ........................................................... 315
C.4 UVMTESTELEMENTCOMPARISON WITH A 3” AIRSPACE ........................................................... 316
C.5 UVMTESTELEMENTCOMPARISON WITH A 3” AIRSPACE, BATT INFILL ........................................ 317
C.6 UVMTESTELEMENTCOMPARISON WITH A 3” AIRSPACE, GYPSUM OVERCLAD ............................. 318
C.7 PHASE2BCLASSASUMMARYGRAPHS .................................................................................... 319
7
C.8 PHASE2BPARALLELPLATE WITH VARIED AIR SPACE .................................................................. 320
C.9 COMPARISONS BETWEEN PHASE 1 AND PHASE 3 ....................................................................... 321
C.10 UNCONNECTED AND CONNECTED (HOLLOW AND FILLED) ......................................................... 324
APPENDIX D ANCILLARY VIBRATION ANALYSIS 326
D.1 INTRODUCTION ..................................................................................................................... 326
D.2 VIBRATIONMEASUREMENTSETUP ......................................................................................... 327
D.2.1 Measurement and Chamber Set Up ................................................................... 327
D.2.2 Equipment and Procedure .................................................................................. 329
D.3 VIBRATIONMEASUREMENTRESULTS ....................................................................................... 330
D.3.1 [A
v
] Vibration Measurement Results .................................................................. 331
D.3.2 [B
v
] Vibration Measurement Results .................................................................. 332
D.3.3 [C
v
] Vibration Measurement Results .................................................................. 333
D.3.4 [D
v
] Ph3 - Vibration Measurement Results ......................................................... 334
D.4 VIBRATIONANALYSIS ............................................................................................................. 335
D.4.1 Measurement EV1 [A
v
] and EV2 [B
v
]. ................................................................. 335
D.4.2 Calculation Procedure ......................................................................................... 336
D.4.2.1 Acceleration Level (dB) to Acceleration (m/s
2
) ................................................ 338
D.4.3 Graphed Results of Conversions ......................................................................... 339
D.4.4 Analysis Summary ............................................................................................... 342
D.5 VIBRATIONMEASUREMENTSUMMARY .................................................................................... 343
8
LIST OF TABLES
TABLE1-1: DISSERTATION OUTLINE ..................................................................................................... 31
TABLE2-1: MULLION PRODUCTS TO REDUCE SOUND FLANKING TRANSMISSION AT THE CURTAIN WALL ......... 47
TABLE2-2: CORRELATION BETWEEN ASTM AND ISO SOUND ISOLATION INDICES ..................................... 52
TABLE2-3: INTERNATIONALSTANDARDS FOR SOUND ISOLATION AND FLANKING TRANSMISSION ................ 54
TABLE2-4: AMERICANSTANDARDS FOR SOUND ISOLATION AND FLANKING TRANSMISSION ....................... 54
TABLE2-5: APPROXIMATE COMPARISON OF SOUND ISOLATION CRITERIA FOR INTERNATIONAL RESIDENTIAL
BUILDING CODE, STC ........................................................................................................ 55
TABLE3-1: OVERVIEW OF APPROACH TO RESEARCH STUDY ................................................................... 67
TABLE3-2: UVMMETHODDESCRIPTION ............................................................................................ 69
TABLE3-3: FILLER WALL TESTS, PERFORMANCES AND DESCRIPTION S ....................................................... 77
TABLE 4-1: TEST PHASE OUTLINE OF DATE, UVM, AND RESULTS ........................................................... 84
TABLE4-2: PHASE1CLASSA,STCRESULTS, AND SPECIMEN DESCRIPTION .............................................. 91
TABLE4-3: PHASE1CLASSB,STCRESULTS, AND SPECIMEN DESCRIPTION .............................................. 93
TABLE4-4: CLASSB STANDARD DEVIATION .......................................................................................... 96
TABLE4-5: PHASE1CLASSC1,STCTESTRESULTS AND SPECIMEN DESCRIPTION...................................... 98
TABLE4-6: CLASSC1 STANDARD DEVIATION ...................................................................................... 100
TABLE4-7: PHASE1CLASSC2,STCTESTRESULTS AND SPECIMEN DESCRIPTION.................................... 101
TABLE4-8: CLASSC2 STANDARD DEVIATION ...................................................................................... 102
TABLE4-9: PHASE1CLASSC3,STCTESTRESULTS, AND SPECIMEN DESCRIPTION .................................. 104
TABLE4-10: CLASSC3 STANDARD DEVIATION ...................................................................................... 105
TABLE4-11: PHASE1CLASSC4,STCTESTRESULTS AND SPECIMEN DESCRIPTION.................................... 106
TABLE4-12: CLASSC4 STANDARD DEVIATION ...................................................................................... 108
TABLE4-13: STANDARD DEVIATION BETWEEN ALL PHASE 1 TEST SPECIMENS ............................................ 111
TABLE4-14: PHASE1COMPARISONS WITH MC-1 (TL13-311) ............................................................. 112
TABLE4-15: PHASE2ACLASSATESTCONFIGURATIONDESCRIPTION,PLANDRAWING ............................. 116
TABLE4-16: PHASE2ACLASSBTESTCONFIGURATIONDESCRIPTION,PLANDRAWING ............................. 116
TABLE4-17: PHASE2ACLASSCTESTCONFIGURATIONDESCRIPTION,PLANDRAWING ............................. 117
TABLE4-18: PHASE2ACLASSDTESTCONFIGURATIONDESCRIPTION,PLANDRAWING ............................. 117
TABLE4-19: PHASE2ACLASSETESTCONFIGURATIONDESCRIPTION,PLANDRAWING ............................. 118
TABLE4-20: PHASE2ACLASSA,STCRESULTS AND SPECIMEN DESCRIPTION........................................... 119
TABLE4-21: PHASE2ACLASSB,STCRESULTS AND SPECIMEN DESCRIPTION ........................................... 122
TABLE4-22: PHASE2ACLASSC,STCRESULTS AND SPECIMEN DESCRIPTION ........................................... 125
TABLE4-23: PHASE2ACLASSD,STCRESULTS AND SPECIMEN DESCRIPTION .......................................... 129
TABLE4-24: PHASE2ACLASSE,STCRESULTS, AND SPECIMEN DESCRIPTION .......................................... 131
TABLE4-25: RESULT SUMMARY OF STC RANGE FOR SMALL AND LARGE VEC ........................................... 133
TABLE4-26: PHASE2BCLASSASOLIDPLATEDESCRIPTION AND PLAN DRAWING ..................................... 141
9
TABLE4-27: PHASE2BCLASSASTAGGEREDPLATEDESCRIPTION AND PLAN DRAWING ............................ 141
TABLE4-28: PHASE2BCLASSCPRODUCT TEST DESCRIPTION AND PLAN DRAWING .................................. 142
TABLE4-29: PHASE2BCLASSA,STCTESTRESULTS, AND SPECIMEN DESCRIPTION .................................. 145
TABLE4-30: PHASE2BCLASSB,STCTESTRESULTS AND SPECIMEN DESCRIPTION ................................... 153
TABLE4-31: PHASE2BCLASSC,STCTESTRESULTS AND SPECIMEN DESCRIPTION ................................... 156
TABLE4-32: PHASE3STCTESTRESULTS ............................................................................................ 167
TABLE5-1: STC AND NR RATING COMPARISON OF SELECT MULLION TESTS FROM PHASE 1 AND PHASE 3 ... 172
TABLE5-2: OBSERVATIONS OF THREE COMPARABLE TEST SPECIMENS FROM PHASE 1 AND PHASE 3 ........... 174
TABLE5-3: TL13-316 AND TL14-168 ............................................................................................. 183
TABLE5-4: SUMMARY OF STC EXTRAPOLATION BETWEEN UNCONNECTED AND CONNECTED MULLIONS .... 184
TABLE5-5: SUMMARY OF PREDICTED COMPOSITE TRANSMISSION LOSS ................................................ 194
TABLE5-6: SUMMARY OF PREDICTED COMPOSITE TRANSMISSION LOSS ................................................ 194
TABLE6-1: RELATIVESTC IMPROVEMENTS FOR VARIOUS MULLION RETROFIT OPTIONS ............................ 206
TABLE6-2: RELATIVESTC IMPROVEMENTS FOR VARIOUS MULLION RETROFIT OPTIONS ............................ 207
TABLE6-3: SUMMARY OF PREDICTED COMPOSITE TL PERFORMANCES ................................................... 207
TABLE7-1: FUTUREWORK .............................................................................................................. 209
TABLE A- 1: DECIBELS ...................................................................................................................... 221
TABLE A-2: PHYSICAL DESCRIPTIONS AND AUDITIVE PERCEPTION ........................................................... 224
TABLE A-3: SUBJECTIVE EFFECT OF CHANGES IN SOUND PRESSURE LEVEL ............................................... 224
TABLE D- 1: PHASE 3 CONFIGURATION OF VIBRATION MEASUREMENT ................................................... 327
TABLE D- 2: PHASE 3 CONFIGURATION OF VIBRATION MEASUREMENT ................................................... 329
TABLE D- 3: VIBRATION MEASUREMENT EQUIPMENT ........................................................................... 329
TABLE D- 4: [A
V
] SOURCE AND RECEIVER RESULTS, 5 SEC DB L,LEQ ....................................................... 331
TABLE D- 5: [B
V
] SOURCE AND RECEIVER RESULTS, 5 SEC DB L,LEQ ....................................................... 332
TABLE D- 6: [C
V
] SOURCE AND RECEIVER RESULTS, 5 SEC DB L,LEQ ....................................................... 333
TABLE D- 7: [D
V
] SOURCE AND RECEIVER RESULTS, 5 SEC DB L,LEQ ....................................................... 334
TABLE D- 8: SELECTED VIBRATION MEASUREMENTS FROM TEST EV1 AND EV2 ........................................ 335
TABLE D- 9: SELECTED VIBRATION MEASUREMENTS FROM TEST EV1 AND EV2 ....................................... 342
10
LIST OF FIGURES
FIGURE 1-1: DIAGRAM OF SOUND TRANSMISSION PATHS AT THE CURTAIN WALL FACADE .......................... 20
FIGURE 1-2: DESIGN CONSIDERATIONS REQUIRED AT THE CURTAIN WALL SYSTEM INCLUDING ACOUSTICS .... 21
FIGURE 1-3: L.A. LIVE TOWER AND RESIDENCES, LOS ANGELES, CA ......................................................... 22
FIGURE 1-4: PLAN DIAGRAM, LATERAL SOUND FLANKING PATH AT THE INTERFACE BETWEEN A CURTAIN WALL
SYSTEM AND INTERIOR DEMISING PARTITION ........................................................................ 23
FIGURE 1-5: PLAN DIAGRAM OF A UNITIZED MULLION SYSTEM CONNECTED TO AN INTERIOR WALL PARTITION;
THREE SOUND PATHS IDENTIFIED. ...................................................................................... 24
FIGURE 1-6: DIAGRAM SHOWING FOUR JUNCTIONS AT A DEMISING WALL ................................................ 25
FIGURE 1-7: DIAGRAM OF RESEARCH DESIGN ....................................................................................... 27
FIGURE 1-8: PLAN DIAGRAM OF THE EXPERIMENT TEST PHASES DEFINED BY THREE CURTAIN WALL
SPECIMENS ...................................................................................................................... 29
FIGURE 2-1: SOUND TRANSMISSION PATHS ALONG A CURTAIN WALL FACADE........................................... 33
FIGURE 2-2: UNITIZED VERTICAL MULLION EXTRUSION DISASSEMBLED FROM THE GLASS INFILL:
THE CONNECTED MULLION (A) AND UNCONNECTED MULLION (D).......................................... 34
FIGURE 2-3: SECTION OF PLASTIC MULLION MOCKUP WITH ANTI-BUCKLING CLIPS- ...................................... 34
FIGURE2-4: PLAN DIAGRAM OF FIVE INDICATIVE SOUND PATHS AT THE GLASS CURTAIN WALL ....................... 36
FIGURE2-5: INDICATIVE DIAGRAMS SHOWING THE COMPOSITE PERFORMANCE BETWEEN A HIGH STC WALL
RATING AND LOW STC RATING OF A MULLION. ..................................................................... 37
FIGURE2-6: SECTION THROUGH A CURTAIN WALL TRANSOM WHERE IT IS CONNECTED AT THE STRUCTURAL
SLAB. .............................................................................................................................. 39
FIGURE2-7: CATEGORIES OF MULLION DETAIL MODIFICATIONS CONSIDERED TO IMPROVE SOUND ISOLATION . 42
FIGURE2-8: STC REFERENCE CONTOUR AGAINST THE TRANSMISSION LOSS OF A TESTED SPECIMEN ............... 50
FIGURE2-9: DEFINITION OF SOUND TRANSMISSION PATHS BETWEEN ROOMS PER ISO 12354 .................... 54
FIGURE2-10: PLAN DRAWING OF “SPLIT” MULLION (LEFT) AND DNF PERFORMANCE (RIGHT) ......................... 61
FIGURE2-11: SIMULATED FLOOR SLAB CONDITION AND TEST CHAMBER CONFIGURATION (LEFT) AND
PHYSICAL TEST SETUP (RIGHT) ............................................................................................ 63
FIGURE2-12: TEST SPECIMEN OF AN OVERCLAD MULLION ELEMENT AND LOCATION IN THE FILLER WALL
APERTURE ...................................................................................................................... 64
FIGURE2-13: TL OF THE MODIFIED MULLIONS ......................................................................................... 64
FIGURE3-1: PLAN DIAGRAM OF BUILDING ELEMENTS AND TEST PHASES ASSOCIATED WITH THE CURTAIN
WALL ASSEMBLY ............................................................................................................... 69
FIGURE3-2: UNITIZEDCURTAINWALLSYSTEM SHOP DRAWINGS, PLAN AND ELEVATION ............................. 71
FIGURE3-3: UNITIZEDCURTAINWALLSYSTEM SHOP DRAWINGS, DETAIL AT THE JOINT .............................. 72
FIGURE 3-4: PLAN DIAGRAM OF THE MULLION USED IN PHASE 1. ................................................................. 73
FIGURE 3-5: PLAN DIAGRAM OF A CONCEPT CONNECTION USED PHASE 2. ..................................................... 73
FIGURE 3-6: PLAN DIAGRAM OF GLASS CURTAIN WALL USED IN PHASE 3. ...................................................... 73
FIGURE3-7: BASIC SETUP OF THE TEST SPECIMEN LOCATED IN THE TEST CHAMBER FILLER WALL. ................... 73
11
FIGURE3-8: DIAGRAM PLAN DRAWINGS OF THE TRANSMISSION LOSS CHAMBER AT WEAL AND THE TEST
ELEMENTS FROM THE CURTAIN WALL SYSTEM. ...................................................................... 75
FIGURE3-9: CHAMBER FILLER WALL CONSTRUCTION (DETAIL COURTESY OF WEAL) .................................... 76
FIGURE3-10: APERTURE IN CHAMBER FILLER WALL IS OPEN WITHOUT A TEST SPECIMEN ................................ 77
FIGURE3-11: APERTURE IN CHAMBER FILLER WALL IS FILLED WITH A TEST SPECIMEN
(SPECIFICALLY FILLED WITH A MULLION AND SILICONE CONNECTION USED IN PHASE 2A) ............ 77
FIGURE3-12: SOUND TRANSMISSION PLOTS OF FILLER WALL ...................................................................... 78
FIGURE3-13: PLAN DRAWING OF AUXILIARY SEMI-ANECHOIC CHAMBERS 3S AND 3R CUSTOM BUILT FOR
THE PHASE 3 TEST MEASUREMENTS .................................................................................... 79
FIGURE3-14: INTERSECTIONDETAIL AT THE WEAL FILLER WALL AND THE CURTAIN WALL BAY ...................... 80
FIGURE 3-15: SEMI-ANECHOIC CHAMBER 3S AT THE SOURCE SIDE FILLED WITH BATT INSULATION.................... 80
FIGURE 3-16: SEMI-ANECHOIC CHAMBER 3S AT THE SOURCE SIDE ............................................................... 81
FIGURE 3-17: SEMI-ANECHOIC CHAMBER 3R AT THE SOURCE SIDE .............................................................. 81
FIGURE4-1: PHASE1:PLAN DIAGRAM OF UVM ELEMENTS AND ASSOCIATED TEST PHASES .......................... 85
FIGURE4-2: PLAN DRAWING OF UNITIZED VERTICAL MULLION PROFILE ...................................................... 86
FIGURE4-3: IMAGE OF HOLLOW EXPOSED MULLION PROFILE WITH ANTI-BUCKLING CLIPS, WOOD SPACES
AND SILICONE GASKET ....................................................................................................... 86
FIGURE4-4: IMAGE OF HOLLOW EXPOSED MULLION PROFILE WITH ANTI-BUCKLING CLIPS, WOOD SPACES
AND SILICONE GASKET ....................................................................................................... 87
FIGURE 4-5: A ¼” GAP BETWEEN MULLION PERIMETER AND FILLER WALL .................................................. 87
FIGURE 4-6: PUTTY APPLIED TO SPECIMEN PERIMETER EDGE TO SEAL ACOUSTIC LEAKS ................................ 87
FIGURE4-7: ELEVATION OF THE MULLION IN THE FILLER WALL WITH SEALED PUTTY PERIMETER ..................... 88
FIGURE4-8: CLASSA–MULLION(HOLLOW AND EXPOSED) ..................................................................... 90
FIGURE4-9: CLASSB–FILLEDMULLIONS ............................................................................................. 90
FIGURE4-10: CLASSC–OVERCLADMULLIONS ....................................................................................... 90
FIGURE4-11: PHASE1-ATRANSMISSIONLOSSCURVES ............................................................................ 92
FIGURE4-12: CLASSBFILLEDMULLION(A) PEA GRAVEL (B) SAND (C) MINERAL WOOL (D) MLV PILLOWS ....... 94
FIGURE4-13: PHASE1-BTRANSMISSIONLOSSCURVES ............................................................................ 95
FIGURE4-14: CLASSCOVERCLADMULLIONIMAGES FROM TEST SERIES C1, C2, C3, AND C4 ....................... 97
FIGURE4-15: PHASE1-BTRANSMISSIONLOSSCURVES ............................................................................ 99
FIGURE4-16: PHASE1-C2TRANSMISSIONLOSSCURVES ........................................................................ 102
FIGURE4-17: PHASE1-C3TRANSMISSIONLOSSCURVES ........................................................................ 105
FIGURE4-18: PHASE1-C4TRANSMISSIONLOSS .................................................................................... 107
FIGURE4-19: TRANSMISSIONLOSSOVERLAY OF PHASE 1-C4 AND PAC INTERNATIONAL® RSIC SPECIMEN ... 108
FIGURE4-20: WEALTL13-329STC48,LABFILLER WALL STC 74 FILLER WALL ....................................... 108
FIGURE4-21: RALTL05-167,STC58, COMPOSITE WALL PARTITION STC 64 © PAC INTERNATIONAL® ..... 108
FIGURE4-22: TRANSMISSIONLOSS SPECTRA OF ALL PHASE 1 E90 LABORATORY TESTS ................................ 110
FIGURE4-23: MULLIONCONSTANT1(MC1) SHOWN IN PLAN (LEFT) AND SPECIMEN PHOTO (RIGHT) ........... 111
FIGURE4-24: MULLIONCONSTANT2(MC2) SHOWN IN PLAN (LEFT) AND SPECIMEN PHOTO (RIGHT) ........... 111
FIGURE4-25: PHASE2A:PLAN DIAGRAM OF UVM ELEMENTS AND ASSOCIATED TEST PHASES...................... 113
12
FIGURE4-26: HEAD OF NARROW APERTURE WIDTH OF 7-3/4” ............................................................... 114
FIGURE4-27: HEAD OF WIDE APERTURE WIDTH OF 8-1/4” ..................................................................... 114
FIGURE4-28: IMAGE OF THE PCS-1 SILICONE COMPRESSION SEAL® PROFILE; UNCOMPRESSED PROFILE
DIMENSION 2-1/2” X 7/8” ............................................................................................. 115
FIGURE4-29: ISOMETRIC DRAWING OF THE PCS-1 SILICONE COMPRESSION SEAL® INSTALLATION. ............... 115
FIGURE4-30: PHASE2A-A,TL OF MULLIONS WITH ½” VERTICAL SEAL ..................................................... 120
FIGURE4-31: PHASE2A-B,TL OF MULLIONS WITH 1/2”AND 3/4” SEALS ................................................ 123
FIGURE4-32: PHASE2A-C,TL OF HIGHER PERFORMING MULLION WITH SILICONE PARTITION CLOSURE® .... 126
FIGURE4-33: PLANDRAWING OF TL13-413 (LEFT) AND TL 416 (RIGHT) ................................................. 126
FIGURE4-34: SOUNDTRANSMISSIONLOSS CURVES OF TL13-413 AND TL13-416 .................................... 127
FIGURE4-35: TL OF HOLLOW/EXPOSED MULLION CONNECTED TO RIZZA SILICONE PARTITION CLOSURE® .... 129
FIGURE4-36: PLANDRAWING OF TL13-418 (LEFT) AND TL 420 (RIGHT) ................................................. 130
FIGURE4-37: SOUNDTRANSMISSIONLOSS CURVES OF TL13-418 AND TL13-420 .................................... 130
FIGURE4-38: TLSPECTRA OF ALL PHASE 2A TESTS ................................................................................ 132
FIGURE4-39: PLANDRAWING OF TL13-413 (LEFT), TL13-417 (MIDDLE) AND TL 422 (RIGHT) .................. 133
FIGURE4-40: PHASE2AMC-1 COMPARED TO A COMPOSITE OF MC-1 + SMALL DEFLECTION CONNECTION... 134
FIGURE4-41: PHASE2AMC-2 COMPARED TO A COMPOSITE OF MC-2 + SMALL DEFLECTION CONNECTION... 134
FIGURE4-42: PHASE2AMC-1 COMPARED TO A COMPOSITE OF MC-1 + LARGE DEFLECTION CONNECTION ... 135
FIGURE4-43: PHASE2AMC-2 COMPARED TO A COMPOSITE OF MC-2 + LARGE DEFLECTION CONNECTION ... 135
FIGURE4-44: PHASE2B:PLAN DIAGRAM OF UVM ELEMENTS AND ASSOCIATED TEST PHASES ...................... 136
FIGURE4-45: PLAN DIAGRAM OF ACOUSTIC CONCEPT CONNECTIONS MEASURED IN PHASE 2B ..................... 137
FIGURE4-46: PLAN DIAGRAM OF STAGGERED PLATE SPECIMEN ................................................................ 137
FIGURE4-47: ELEVATION VIEW OF PHASE 2B TEST SPECIMENS IN FILLER WALL
(LEFT – PARALLEL, RIGHT – STAGGERED) ............................................................................ 138
FIGURE4-48: PHASE2BCLASSA SPECIMEN CONSTRUCTION OF SOLID PLATE TESTS ..................................... 139
FIGURE4-49: PHASE2BCLASSB SPECIMEN CONSTRUCTION OF SOLID PLATE TESTS ..................................... 140
FIGURE4-50: TL OF “PARALLEL PLATES” WITH 6” AND 4” AIR SPACES ....................................................... 146
FIGURE4-51: PLAN DRAWING OF CONNECTION ELEMENT TL13-612 (LEFT) AND UNITIZED MULLION
TL13-311 (RIGHT) ......................................................................................................... 146
FIG URE4-52: PHASE2BALUMINUM AND GYPSUM BOARD PLATE RESULTS ................................................ 147
FIGURE4-53: PHASE2BCLASSA: ALUMINUM OR GYPSUM BOARD PLATE WITH AN MLV DAMPING LAYER .... 149
FIGURE4-54: PHASE2BCLASSB,MLV COMPOSITE WITH ALUMINUM OR GYPSUM BOARD PLATE ................ 154
FIGURE4-55: PHASE2BCLASSB:MLV COMPOSITE WITH ALUMINUM OR GYPSUM BOARD PLATE ................ 155
FIGURE4-56: MULLIONMATE®(LEFT) AND MULL IT OVER® (RIGHT) ....................................................... 157
FIGURE4-57: MULLIONMATE(LEFT) AND MULL IT OVER (RIGHT) ........................................................... 157
FIGURE4-58: MULLIONMATE(LEFT) AND MULL IT OVER (RIGHT) ........................................................... 158
FIGURE4-59: TRANSMISSIONLOSS RESULTS OF THE MULL IT OVER PRODUCT FROM WEAL ......................... 159
FIG URE4-60: PHASE3:PLAN DIAGRAM OF UVM ELEMENTS AND ASSOCIATED TEST PHASES ........................ 161
FIGURE4-61: PHASE1CONFIGURATIONS USED AS THE CENTER VERTICAL MULLION IN PHASE 3 .................... 162
13
FIGURE4-62: COUPLED SEMI-ANECHOIC CHAMBERS 3S AND 3R SHOWN IN PLAN WITHIN THE MAIN
WEAL CHAMBERS. ......................................................................................................... 163
FIGURE4-63: CURTAIN WALL BAY INSERTED INTO THE APERTURE OF THE FILLER WALL ................................. 164
FIGURE4-64: CURTAIN WALL PLACED AND CENTERED IN THE FILLER WALL.................................................. 164
FIGURE4-65: CHAMBER3S(BUILT WITHIN THE WEAL SOURCE CHAMBER) ............................................... 164
FIGURE4-66: CHAMBER3R(BUILT WITHIN THE WEAL RECEIVING CHAMBER) ........................................... 164
FIGURE4-67: CHAMBER3S AND 3R PLAN DRAWINGS AND ACOUSTIC DETAILS ........................................... 165
FIGURE4-68: WEALCHAMBER[SECTION] ........................................................................................... 166
FIGURE4-69: WEALCHAMBER[ELEVATION] ........................................................................................ 166
FIGURE4-70: CHAMBER3S AND 3R PLAN DRAWINGS AND ACOUSTIC DETAILS ........................................... 168
FIGURE4-71: SUMMARY OF STC RANGES AT EACH UVM TEST PHASE ...................................................... 169
FIGURE5-1: DRAWING OF MULLION SECTION WITH IDENTIFIED FACE AREAS OF EACH CURTAIN WALL
ELEMENT: TRANSOM (0.16 SF), GLASS (0.7 SF), MULLION (2.42 SF), AND SILL (0.16 SF). ...... 174
FIGURE 5-2: TL13-311 (UNCONNECTED MULLI ON) ................................................................................ 175
FIGURE 5-3: TL14-170 (CONNECTED MULLION) .................................................................................... 175
FIGURE5-4: NOISEREDUCTIONSPECTRA BETWEEN TL13-311 AND TL14-170 ....................................... 175
FIGURE 5-5: TL13-316 (UNCONNECTED) .............................................................................................. 177
FIGURE 5-6: TL14-168 (CONNECTED) ................................................................................................... 177
FIG URE5-7: NOISEREDUCTIONSPECTRA BETWEEN TL13-316 AND TL14-168 ....................................... 177
FIGURE 5-8: TL13-323 (UNCONNECTED) .............................................................................................. 179
FIGURE 5-9: TL14-167 (CONNECTED) ................................................................................................... 179
FIG URE5-10: NOISEREDUCTIONSPECTRA BETWEEN TL13-323 AND TL14-167 ....................................... 179
FIGURE5-11: NOISEREDUCTIONSPECTRA BETWEEN TL13-311 AND TL13-323 ....................................... 181
FIGURE5-12: NOISEREDUCTIONSPECTRA BETWEEN TL14-170 AND TL14-167 ....................................... 182
FIGURE5-13: UNCONNECTEDMULLION WITH GYPSUM BOARD OVERCLAD, TL13-325, STC 42 5/8” .......... 184
FIGURE5-14: TRANSMISSIONLOSSEXTRAPOLATION OF TL13-325 IF TESTED WITH THE CURTAIN WALL
SYSTEM ........................................................................................................................ 185
FIGURE5-15: IDENTIFICATION OF ELEMENTS USED FOR THE COMPOSITE CALCULATIONS AND ASSOCIATED
SURFACE AREAS .............................................................................................................. 186
FIG URE5-16: TL13-311,STC 36 ........................................................................................................ 187
FIG URE5-17: TL13-323,STC 52 ........................................................................................................ 187
FIG URE5-18: TL13-621,STC37,STAGGEREDPLATE WITHOUT SEALS .................................................... 187
FIGURE5-19: TL13-622,STC51, STAGGEREDPLATE WITHOUT SEALS.................................................... 187
FIGURE5-20: TL14-170,STC37PLAN DRAWING OF UVM HOLLOW AND EXPOSED ................................. 188
FIGURE5-21: TL14-167,STC42PLAN DRAWING OF UVM FILLED AND OVERCLAD .................................. 188
FIGURE5-22: NRC-CNRC LABORATORY TEST TL-93-302 PERFORMANCE IN OCTAVE BAND CENTER
FREQUENCIES ................................................................................................................. 189
FIG URE5-23: COMPOSITEPREDICTION OF A LOW PERFORMING CURTAIN WALL JUNCTION ........................ 190
FIGURE5-24: COMPOSITEPREDICTION OF A LOW PERFORMING CURTAIN WALL SYSTEM WITH GLASS ......... 191
FIGURE5-25: COMPOSITEPREDICTION OF A HIGH PERFORMING CURTAIN WALL JUNCTION ....................... 192
14
FIGURE5-26: COMPOSITEPREDICTION OF A HIGH PERFORMING CURTAIN WALL SYSTEM WITH GLASS ........ 193
FIGURE5-27: STC SUMMARY ACROSS ALL LABORATORY TESTS ................................................................. 196
FIGURE5-28: PRELIMINARYRANKING OF UVM TEST ELEMENTS .............................................................. 198
FIGURE6-1: LIMITED SPECTRUM WITH CURTAIN WALL DESIGN ............................................................... 204
FIGURE6-2: LIMITED SPECTRUM WITH CURTAIN WALL DESIGN ............................................................... 206
FIGURE A- 1: NON-UNITIZED (STICK) SYSTEM ....................................................................................... 219
FIGURE A- 2: UNITIZED SYSTEM ........................................................................................................... 220
FIGURE A- 3: FLANKING PATH DIAGRAM (MOMMERTZ & MULLER, 2008) .............................................. 222
FIGURE A- 4: APPROXIMATE SOUND SPECTRA OF MALE AND FEMALE SPEECH (LOG-TERM AVERAGE) ........... 222
FIGURE A-5: SECTION OF LABORATORY TRANSMISSION LOSS TESTING CHAMBER ....................................... 227
FIGURE A-6: FIELD CONDITION FOR IN-SITU TRANSMISSION LOSS TESTING. ............................................... 228
FIG URE A-7: DIFFERENCE BETWEEN LABORATORY AND FIELD TRANSMISSION LOSS TESTS. ........................... 228
FIG URE C-1: PHASE 1-A TRANSMISSION LOSS CURVES .......................................................................... 314
FIGURE C-2: 4” AIRSPACE COMPARISON, MULLION AND PLATE CONNECTION ........................................... 315
FIGURE C-3 3” AIRSPACE COMPARISON, MULLION AND PLATE CONNECTION ........................................... 316
FIGURE C-4: 3” AIRSPACE AND BATT INFILL COMPARISON, MULLION AND PLATE CONNECTION ................... 317
FIGURE C-5: 3” AIRSPACE AND GYPSUM OVERCLAD, MULLION AND PLATE CONNECTION .......................... 318
FIGURE C-6: PHASE 2B PARALLEL PLATE TESTS WITH 3” AIR CAVITIES ...................................................... 319
FIGURE C-7: SOUND TRANSMISSION LOSS OF ALUMINUM PLATES ONLY WITH 3”, 4” OR 6” AIR CAVITY ........ 320
FIG URE C-8: TL SPECTRA OF PH3-MC1 WITH PH1-MC1 AND PH3-MC2 WITH PH1-MC2 ....................... 321
FIGURE C-9: TRANSMISSION LOSS COMPARISON BETWEEN CONTROL MULLIONS MC1 AND MC2 .............. 322
FIGURE C- 10: PHASE 1-MC1 (TL13-311) AND PHASE 3-MC1 (TL14-170) ............................................. 322
FIGURE C- 11: PHASE 1-M1A (TL13-316) AND PHASE 3-MC1A (TL14-168) ........................................... 323
FIGURE C- 12: PHASE 1-MC2 (TL13-323) AND PHASE 3-MC2 (TL14-167) ........................................ 323
FIGURE C- 13: TL PLOTS OF PHASE 1 (MC1/1A) AND PHASE 3 (MC1/1A) ................................................ 324
FIGURE C- 14: TL PLOTS OF COMPARISON MC1 AND MC1A .................................................................... 325
FIGURE D-1: TYPICAL LOCATION OF THE ACCELEROMETERS AT THE SOURCE AND RECEIVING CHAMBERS ........ 326
FIGURE D-2: PLAN DRAWING OF WEAL TEST CHAMBERS AND PHASE 3 TEST RIG 3S/3R CHAMBERS ............... 328
FIGURE D-3: ELEVATION A: VIEW OF ACCELEROMETER LOCATIONS ON CHAMBERS 3S AND 3R FOR EV2 ...... 328
FIGURE D-4: AXIS DESIGNATION FOR VIBRATION MEASUREMENTS ........................................................... 330
FIGURE D-5: EV1[AV]CURTAINWALLELEVATION WITH 3R CHAMBER ACCELEROMETER
LOCATIONS 02, 08, 03 ................................................................................................... 335
FIGURE D-6: EV2[BV]CURTAINWALLELEVATION WITH 3R CHAMBER ACCELEROMETER
LOCATIONS 02, 05, 03 ................................................................................................... 336
FIGURE D- 7: EV1 – MEASURED ACCELERATION LEVELS (dB) ................................................................... 339
FIGURE D-8: EV2–MEASUREDACCELERATIONLEVELS (dB) ................................................................. 339
15
FIGURE D- 9: EV1 – CALCULATED ACCELERATION (M/S2) ........................................................................ 340
FIGURE D- 10: EV2 – CALCULATED ACCELERATION (M/S2) ..................................................................... 340
FIGURE D- 11: EV1 – CALCULATED VELOCITY (M/S) ................................................................................ 340
FIGURE D- 12: EV2 – CALCULATED VELOCITY (M/S) ............................................................................... 340
FIGURE D- 13: EV1 – CALCULATED SOUND PRESSURE LEVEL (dB) ............................................................. 341
FIG URE D- 14: EV2 – CALCULATED SOUND PRESSURE LEVEL (dB) ............................................................. 341
FIG URE D- 15: EV1 – CALCULATED SOUND PO WER LEVEL (dB) ................................................................ 341
FIGURE D- 16: EV2 – CALCULATED SOUND POWER LEVEL (dB) ................................................................ 341
FIGURE D- 17: EV1, PREDICTED AND MEASURED SOUND PRESSURE LEVELS ................................................. 342
FIGURE D- 18: EV2, PREDICTED AND MEASURED SOUND PRESSURE LEVELS ................................................. 343
FIGURE D- 19: SUMMARY OF SOUND POWER AT EACH UVM MEASURED DURING PHASE 3 ........................... 343
16
ABSTRACT
“Cohesion is a measure of how well the parts of a component ‘belong
together.’ Cohesion is strong if all parts are needed for the functioning of
other parts. Strong cohesion promotes maintainability and adaptability
by limiting the scope of changes to a small number of components.”
1
The glass curtain wall is a cohesive design of glazing, aluminum framing, structural silicone, and neoprene
gasketing. These components for unitized and non-unitized systems are technologically sophisticated and
work together as a complex dynamic system. The intricate design accounts for the architecture of
structural deflections, thermal properties, acoustic performance, moisture control, fire and smoke
protection, amongst others. Therefore, the design and installation of these of these components should
work cohesively together to provide an effective building performance, including provisions for good
sound isolation. The sound isolation between occupied adjacencies located at the façade is highly
influenced by the architectural composite of the various elements. Additionally certain parts within the
composite can transfer sound energy more efficiently than others via sound flanking transmission paths.
Sound flanking transmission that exists at the façade curtain wall and interconnecting partition presents
a design challenge that reduces acoustic privacy and sound isolation design targets. This intersection can
create an acoustic weakness within the curtain wall assembly and where it fastens to the building.
Three architectural elements commonly contributing to this weakness are the curtain wall infill glazing,
the aluminum mullion extrusions, and the partition connection joining the mullion to an interconnecting
partition. The behavior of sound flanking transmission paths across each of these curtain wall elements is
currently not well understood for all systems. These architectural mechanisms can create lateral sound
paths and degrade the overall sound isolation integrity of the composite architecture. This is especially an
issue when a high sound isolation performance between adjacent spaces is expected from an acoustically
rated partition.
Research on sound flanking paths in curtain wall systems has been carried out in theoretical statistical
energy analysis (SEA) models, sound isolation prediction simulations, and with physical measurements on
laboratory and field installations. However, most of these studies are composite and do not necessarily
investigate the specific behavior of the separate system components. Typically the sound isolation
performance of mullions measured in a laboratory or field may not clearly identify which curtain wall
element most significantly contributes to overall performance. Acoustic products for curtain wall systems
are emerging with the intent to improve overall sound isolation performance, which is an indication that
this problem impacts the architectural practice.
In order to improve the acoustic performance of the curtain wall system, the critical components
attributed to the sound flanking transmission paths must be better understood, particularly at the
mullion. Three elements associated with the architecture of the curtain wall system were selected and
studied through a series of laboratory test measurements and sound isolation prediction calculations to
determine potential improvement: the partition connection, the vertical mullion, and the curtain wall
glazing. The testing method proposed is in accordance with ASTM E90, an acoustic testing procedure that
1
Man Lin, “Chapter 6 Architectural Design, Computer Science CSci485,” 2012,
http://cse.stfx.ca/~mlin/cs485/lectures/archdesign.ppt.
17
measures the transmission loss (TL) of a building specimen between two reverberant test chambers. The
single figure STC classification per ASTM E336 is obtained from these measurements. The STC is a
commonly used amongst architects in practice to identify the sound level resistance of walls and floors.
Approximately 80 acoustic laboratory tests were performed on select curtain wall elements and modified
to identify the highest practicable acoustic performance that may be achieved. Additionally, an auxiliary
set of vibration measurements were conducted at one of the test stages in order to examine the acoustic
energy injection at mechanically connected elements of the curtain wall system.
Comparisons between the independent test elements were examined in order to understand construction
and performance benefits associated with achievable performance. The sound transmission loss data
obtained from the laboratory test procedure is analytically calculated with the performance of an interior
partition assembly to understand composite effects. Results from this testing method indicate how the
performance of individual components influences a composite system and identifies elements that limit
the achievable sound isolation. Although global variations of curtain wall designs exist in practice, the
conclusions developed from the proposed experiment method are relevant to the specific curtain wall
specimen typology measured and have relevance to similar systems.
The research aims to enhance façade tectonic cohesion specific to acoustic design integration and to
inform building engineering design and performance decisions.
18
TERMS & ABBREVIATIONS
ACOUSTIC TERMS/TEST
See Appendix A for acoustic definition of terms and test procedures.
ACOUSTIC ABBREVIATIONS
ATI Architectural Testing, Inc.
Institute for testing and certifying building products and constructions
ASTM American Society for Testing & Materials
CW Curtain Wall
UVM Unitized Vertical Mullion
Refers to the UVM test methodology; the test method considers two associated mullion elements:
glass and partition connection.
CVM Center Vertical Mullion
Used in reference to the curtain wall test specimens in Phase 3. Modifications in this phase were
made to the center vertical mullion; vertical mullions were located at the outer edge of the bays.
IGU Insulated Glazing Unit
L
p
Sound Pressure Level (dB)
L
w
Sound Power Level (dB)
L
eq
Equivalent Sound Level, over a period of time
MC Mullion Constant
Mullions defined in Chapter 4 as MC1 and MC2. MC1 defines the lowest performing baseline and
MC2 defines the highest performing.
MLV Mass Limp Vinyl
NR Noise Reduction
R Acoustic Index - Sound Reduction Index, laboratory measurement
R
w
Acoustic Index - Sound Reduction Index, laboratory measurement, weighted
R’ Acoustic Index – Apparent Sound Reduction Index, field measurement
R’
w
Acoustic Index – Apparent Sound Reduction Index, field measurement, weighted
SEA Statistical Energy Analysis
STC Acoustic Index - Sound Transmission Class
TL Sound Transmission Loss
19
VEC Vertical Edge Connection
The vertical edge connection (VEC) is referenced in Chapter 4 during Phase 2 to define the various
test conditions at the vertical edge of a building specimen, i.e. resilient acoustic seal, partition
connection.
WEAL Western Electro-Acoustic Laboratory
The laboratory in Santa Clarita, CA where all specimens of the UVM test were measured.
20
CHAPTER 1 INTRODUCTION TO CURTAIN WALL SOUND FLANKING
TRANSMISSION
1.1 ACOUSTIC RELEVANCE TO GLASS CURTAIN WALL DESIGN
Lateral and vertical sound transmission across a façade limits the potential sound isolation between
spaces and is rarely considered in the design of glass curtain wall construction. Acoustic concerns typically
concentrate on sound transmission from environmental noise through the facade. However these sound
flanking paths across the façade compromise the interior sound isolation and transmit acoustic energy by
way of structural paths connected to the curtain wall, such as at the vertical mullion, horizontal mullion,
glass infill, partition connection, floor slab connection, etc. Interior sound isolation is degraded as a result
of these acoustic weaknesses. Flanking transmission paths can be both lateral and vertical (Figure 1-1).
FIGURE 1-1: DIAGRAM OF SOUND TRANSMISSION PATHS AT THE CURTAIN WALL FACADE
Since structural paths at the curtain wall are numerous, the scope of this research is limited to the lateral
sound transmission across three select façade elements to study limitations of their inherent structure:
(1) the vertical mullion, (2) its connection condition to an interior partition, and (3) a composite of the
curtain wall glazing and horizontal mullions. Sound transmission loss (TL) measurements from these
component investigations are analytically studied to evaluate frequency regimes, trends, and correlations
with respect to performance and material assemblies.
Although this acoustic relevance to the curtain wall system is important, its design is engineered to
perform other multi-disciplinary functions that cohesively integrate tectonic interactions including
structural deflection, thermal resistance, fire and smoke protection, and architectural detailing.
This integrated engineering has various design impacts relevant to the acoustic performance and
specification of the curtain wall (Figure 1-2).
Vertical Sound Flanking
Transmission Paths
Lateral Sound
Flanking Transmission
Paths
Vertical Mullion
Horizontal Mullion
(aka Transom)
21
FIGURE 1-2: DESIGN CONSIDERATIONS REQUIRED AT THE CURTAIN WALL SYSTEM INCLUDING ACOUSTICS
Part of the quality assurance in architecture is meeting the acoustic performance targets as practically
and cost effectively as possible. Therefore, methods to meet noise reduction criteria (e.g. STC, NIC, etc.)
with the potentially limiting connections at the façade and building interior must be considered.
Sound isolation performance indices in the USA typically assign sound transmission class (STC) ratings
between noise sensitive adjacencies. The STC is a single figure rating that defines how effectively a
building element (e.g. wall or floor) resists airborne sound transmission. This type of performance criteria
is commonly used to reduce disturbance between adjacent spaces during simultaneous activities and is
an index regularly recognized by architects. The STC performance is evaluated in an acoustic laboratory
where the building test specimen is mounted between source and receiving chambers and measured for
transmission loss. The testing procedure required to obtain the STC rating takes into account the overall
radiating surface area of the building specimen, which includes acoustic mechanisms influencing sound
flanking transmission. The STC measurement and testing standard procedure is defined in Chapter 2 and
Appendix A.
High-rise residential design requires demising partitions with high STC target ratings between dwellings.
The interior partition plus its edge connections at the façade, floor, and ceiling will contribute to meeting
this criteria to target good levels of acoustic privacy. Occupants expect robust sound isolation from
adjacent neighbors in a multi-unit high-rise tower. “The design and construction of multifamily dwellings
must include consideration of privacy, which in many cases is legally mandated, even if it is not controlled
by a building code or property line ordinance, it nevertheless forms part of the basis of the home buyer’s
or occupant’s reasonable expectation of quality.”
2
Often the curtain wall design will require some form of acoustic intervention where high sound isolation
is required. One example of this is at the hotel and residential tower at the LA Live development located
in Los Angeles, CA (FIGURE 1-3). Concern for vertical sound flanking transmission at a double story height
mullion system was brought to the attention of the curtain wall designers, Enclos Corp. This initiated a
2
Marshall Long, Architectural Acoustics (Elsevier, 2006), 509.
22
series of unique acoustic amelioration tests by the design team. The inherent condition impacted the
design process, building specification and detailing, and building material cost.
FIGURE 1-3: L.A. LIVE TOWER AND RESIDENCES, LOS ANGELES, CA (IMAGE COURTESY OF © 2015 ENCLOS CORP)
Acoustic performance implications associated with façade sound flanking at high-rise occupancies in the
commercial, retail, healthcare, education, or residential sectors include
Client expectation and criteria targets for the use and quality of the space.
Speech privacy and confidentiality degraded between adjacent spaces laterally and vertically.
Simultaneous activity between adjacencies is compromised.
Issues relating to cost effectiveness. The sound isolation performance of robust partitions is
devalued.
Future building market forecast for an increased demand in residential buildings. Therefore
design requirements and space planning issues will be more onerous.
Healthcare, residential, and building code violations.
Health risk including loss of sleep, reduced healing environment, stress, and loss of productivity.
These emphasize the importance of maintaining sound isolation that is consistent with various elements.
23
1.2 PROBLEM STATEMENT AND HYPOTHESIS
Sound flanking transmission paths reduce the achievable sound isolation between adjacent spaces
located at the façade. A demising partition may be built to a STC 55 specification, however once attached
to the glass curtain wall system, the composite of the all parts may reduce the performance in excess of
10 – 20 STC points (FIGURE 1-4). Much of the performance reduction is attributed to sound traveling across
the path of least resistance, e.g. light-weight building components. A glass curtain wall is considered a
light-weight building component as opposed to a concrete floor slab which is significantly heavier.
FIGURE 1-4: PLAN DIAGRAM, LATERAL SOUND FLANKING PATH AT THE INTERFACE BETWEEN A CURTAIN WALL SYSTEM
AND INTERIOR DEMISING PARTITION
This limitation is well known to acoustical consultants. Recent publications identify the design and
development of acoustic detailing at high-rise structures and discuss the common sound flanking issue at
the intersection of demising partitions and the curtain wall (for example LoVerde and Dong, 2008).
3
3
John LoVerde and Wayland Dong, “Methods for Reducing Flanking Airborne Noise Transmission through Mullions of Curtain
Wall Systems,” The Journal of the Acoustical Society of America 124, no. 4 (2008): 2463.
24
Industry manufacturers are also addressing the sound flanking issue. Products for curtain wall systems are
being designed to mitigate sound transmission by applying appendages to the mullion. Manufactured
product options are listed in Chapter 2.
Additionally, curtain wall specifications may include acoustic clauses to address sound flanking concerns,
but the language used may be difficult to enforce or lack responsible entities, such as in this example:
“Sound flanking transmission at demising walls and floors must be avoided through correct design and
detailing.”
The acoustic weakness of this interface is attributed to lightweight building components that are designed
to resiliently connect the curtain wall mullions to the building (FIGURE 1-4). Although design resolutions
have been identified in architectural acoustic practice, it still remains unclear what components of the
system are primarily impacting the overall STC performance, both individually and on the assemblage.
These components become mechanisms for lateral sound paths and degrade the integrity of the overall
sound isolation of the curtain wall and interconnecting partition.
Hypothesis:
Sound transmission loss testing of individual and composite architectural elements comprised of and
associated with the intersection of the unitized vertical mullion reveals sound flanking path
mechanisms controlling the overall sound isolation performance.
FIGURE 1-5: PLAN DIAGRAM OF A UNITIZED MULLION SYSTEM CONNECTED TO AN INTERIOR WALL PARTITION; THREE
SOUND PATHS IDENTIFIED.
Although many sound paths occur at the unitized mullion, three lateral sound paths are under
consideration. Each of these paths is associated with one of the following curtain wall elements: interior
wall connection, curtain wall mullion, and curtain wall glass. These sound paths are typical at most unitized
curtain wall mullion systems.
1. The sound path at the partition connection is located between the mullion and demising wall.
Unitized Vertical Mullion:
Direct Sound Path through mullion
Curtain Wall Glass:
Flanking Sound Path at the glazing assembly
Partition Connection:
Flanking Sound Path at the connection
between the mullion and interior wall
Interior Demising Wall
25
2. A direct sound path is located at the curtain wall mullion extrusion.
3. The sound flanking path located at the curtain wall glass occurs due to flexural waves from acoustic
energy excited from a source room and transmitting to a receiving room.
Improvements to each building element should not be judged in isolation, and sound flanking at the
curtain wall glazing is not typically accounted for as a target for mitigation. The commercial designs for
mullion enhancements may show high TL values, but those may be reduced in a real installation by other
flanking paths, including across the curtain wall glazing.
1.2.1 BUILDING SOUND FLANKING PATHS
Sound flanking transmission is inherent to building design, and their paths occur wherever building
elements join. It is not exclusive to the curtain wall system. Sound flanking may be defined as the
transmission of acoustic energy around a primary sound isolation barrier; this is also known as an indirect
sound path. An acoustic intersection detail may indicate the vertical mullion as the primary barrier and
the interior wall connection and curtain wall glazing are flanking paths (Figure 1-5). At a larger scale, the
primary barrier may be the interior demising wall (Figure 1-6).
FIGURE 1-6: DIAGRAM SHOWING FOUR JUNCTIONS AT A DEMISING WALL
The overall sound isolation performance between the two spaces is dependent on the resistance of sound
energy through the demising wall and its edge flanking. There are four edge flanking conditions at a
1
4
2
3
1. Indirect path at the façade and wall
2. Indirect path at the ceiling and wall
3. Indirect path at the corridor partition
and wall
4. Indirect path at the floor and wall
26
demising wall: at the facade, at the floor, at the adjacent interior wall (often at a corridor), and at the
ceiling where sound flanking or acoustical leaks may occur (Figure 1-6). Sound flanking is defined as sound
transmission through building components; an acoustic leak is defined as sound transmission through air
gaps or holes where they occur in the building construction.
The sound paths at the façade and interior wall will vary based on the context of the building design,
construction, and deflection requirements for wind and/or seismic loads. “The effect of flanking sound is
to lower the achieved sound insulation between adjacent areas below that which would be expected from
the known performance of the identified dividing barriers. Because flanking sound is always present
(other than within the ideal confines of an acoustic laboratory) practical site performance between non-
isolated’ constructions will be limited….”
4
1.2.2 COMPOSITE TRANSMISSION LOSS
“The sound isolation between rooms is dependent mainly on the mass of the separating wall and
composites like doors or windows and the degree to which they are sealed airtight.”
5
The overall sound
isolation performance between the two spaces will be the cumulative effect of the direct sound through
the demising partition and the many indirect paths through the coupled intersection at its edges (Figure
1-6). The composite sound transmission loss performance will consist of the TL for each individual element
and its relative area.
It can be challenging to identify the individual architectural element that most significantly controls the
resultant TL rating. It has not yet been determined in practice which of the three elements identified in
Figure 1-5 limits the overall sound isolation rating of the composite system shown.
A further discussion of composite TL is provided in Chapter 2.
4
Association of Interior Specialists, “Building Acoustics - Terminology,” AIS Association of Interior Specialists, accessed April 6,
2014, http://www.bre.co.uk/page.jsp?id=1146.
5
B. J. Smith, R. J. Peters, and S. Owen, Acoustics and Noise Control, 2 Sub (Addison Wesley Longman, 1996), 67.
27
1.3 OVERVIEW OF RESEARCH
This research investigates the lateral sound flanking at the connection between the demising partition
and the façade at the curtain wall.
This is important to note especially in the review of sound transmission loss test reports for curtain wall
mullions. A composite performance is typically shown with a demising partition but without the glass
infill. Therefore it is difficult to determine what the composite TL will be in an actual project because
one cannot identify the TL of the individual components or their interaction. The sound isolation
behavior of a singular component is therefore difficult to identify.
FIGURE 1-7: DIAGRAM OF RESEARCH DESIGN
The structure of the research and method will begin with the objectives which support the hypothesis,
including a course to set up the proposed empirical test method used to support an analysis method
(FIGURE 1-7). Vertical sound flanking transmission across curtain wall elements also occurs in practice but
will not form part of this research study.
28
1.3.1 RESEARCH OBJECTIVES
There are four research objectives. Objective 1 reviews practices to modify curtain wall elements and
procedures to measure them. Objective 2 develops the experiment design to measure unitized vertical
mullions. The results from the empirical testing conducted in Objective 2 will be applied to two different
analytical analysis methods to meet Objectives 3 and 4.
1. Identify curtain wall mullion design practices and procedures. Investigate methods used in
practice to modify curtain wall elements and review current sound isolation metrics and
measurement methods to identify uncertainties in current design and test procedures.
2. Develop an experiment methodology for unitized vertical mullion measurements. Develop a
test method to measure the unitized vertical mullion and associated connections individually in
accordance with ASTM E90 and without the influence of a composite demising wall. The approach
will define connected versus unconnected mullion conditions. Objective 2 is applied to two
different analytical analysis methods to meet Objectives 3 and 4.
3. Identify controlling sound paths at the unitized vertical mullion. Evaluate the sound transmission
loss of connected and unconnected mullion conditions to identify controlling frequency regimes,
trends and correlations between curtain wall elements.
4. Determine the acoustic relationship between vertical mullion and interconnecting walls.
Determine impacts to the demising partition using the composite transmission loss prediction
method. This will include the performance and areas of the curtain wall elements and demising
wall partition to provide information where diminishing returns occur between acoustic
performance and material construction.
1.3.2 RESEARCH METHOD
1. Identify existing methods of measuring transmission loss in curtain wall systems.
2. Identify current architectural interventions to improve the sound isolation performance.
3. Expand on precedent research to develop a test experiment to measure individual components
of the curtain wall system.
4. Design a test experiment to measure the three elements identified in the scope of the research
study and develop modifications to measure for each element. These elements include the
vertical mullion, the partition connection and the curtain wall glazing.
5. Profile the one-third octave band transmission loss, frequency regimes, and acoustic ratings from
the empirical measurements of each element.
6. Isolate the sound isolation performance of the connection between the mullion and the interior
wall including tolerances required for structural and thermal façade performances, for example
structural deflection and thermal expansions.
7. Measure the sound transmission loss of the unitized vertical mullion including the addition of the
curtain wall glazing in a system that simulates an outdoor condition to measure only sound energy
across the curtain wall and remove the influence of sound energy passing through the curtain
wall.
29
8. Compare the composite sound transmission loss predictions with a higher performing demising
wall to evaluate limitations.
1.3.3 APPROACH TO MEASURING THE TEST SPECIMENS
The curtain wall system used is detailed in Chapter 3 and consists of a unitized vertical mullion connected
to a glass curtain wall bay on either side with glass-infill and horizontal mullions. The experiment is
designed to acoustically test specimens of mullions both connected and unconnected to the curtain wall
system. Additionally, concept partition connections are tested with and without the vertical mullion. All
test measurements are in accordance with the ASTM E90 Test Method for Laboratory Measurement of
Airborne Sound Transmission Loss of Building Partitions and Elements.
As a means for relative performance comparisons, each building specimen is measured with modifications
typically seen in practice based on findings from Objective 1. Modifications to the building specimens are
made with the intent to identify the highest practicable sound transmission loss achievable and as a means
for relative comparison.
This experiment is divided into three test phases based on three lateral sound paths:
Phase 1 Sound path through the unconnected aluminum mullion extrusion
Phase 2 Sound path at the partition connection between the mullion and interior partition
A. Partition connection with the unconnected mullion
B. Partition connection without the unconnected mullion
Phase 3 Sound Path through acoustic vibration transmission at the composite curtain wall glazing
FIGURE 1-8: PLAN DIAGRAM OF THE EXPERIMENT TEST PHASES DEFINED BY THREE CURTAIN WALL SPECIMENS
30
The test method designed for measuring the lateral sound transmission of this curtain wall typology and
its associated parts under relative laboratory setup conditions is unprecedented.
The data obtained from the empirical test data will be analytically analyzed in Chapter 5 to identify
controlling architectural mechanisms and the influence these specimens may have on the demising wall
system.
1.4 DISSERTATION OUTLINE
The research study is organized into 7 chapters.
Chapter 1
Introduction to Curtain Wall Sound
Flanking Transmission
The sound flanking transmission at curtain wall systems is
defined. The research objective and architectural acoustic
test methods are described specifically for lateral sound
flanking at unitized curtain wall mullions.
Chapter 2
Façade Background Review and
Professional Applications in
Acoustics
Background of the current testing and research approach to
reduce sound flanking is explained. This includes global test
methods, manufactured products, and precedent building
measurements.
Chapter 3
Unitized Vertical Mullion Test
Method
This describes the test experiment procedure and phases (1,
2A, 2B, 3) for the Unitized Vertical Mullion (UVM) test
method. This includes the proposed specimens and required
test chamber conditions.
Chapter 4
Test Results and Analysis of the
Unitized Vertical Mullion
Measurements
This includes the sound transmission class (STC) results of 80+
acoustic laboratory tests. These are organized by test phase.
The details of the laboratory chamber conditions and final
specimens tested at each phase are reported.
Chapter 5
Analysis of Controlling Mechanisms
and Composite Transmission Loss
Sound transmission loss (TL) of frequency ranges are
compared and analyzed between phases. Results from the
UVM test method are applied to composite TL predictions.
Chapter 6
Conclusion
A summary of the objective conclusions and contributions is
presented.
Chapter 7
Future Work
Test measurements and analytical studies based on findings
from the UVM test method are proposed for future work.
Appendix A
Terminology
Acoustic terminology relevant to the research work is
defined.
Appendix B
UVM Laboratory Test Results
One-third octave band results for all laboratory tests
conducted at WEAL are given.
31
Appendix C
Ancillary Sound Analysis
Additional WEAL test results comparison and correlations are
shown.
Appendix D
Ancillary Vibration Analysis
Vibration measurement test results conducted during the
Phase 3 curtain wall bay test and analysis are provided.
TABLE1-1: DISSERTATION OUTLINE
32
CHAPTER 2 FAÇADE BACKGROUND REVIEW AND PROFESSIONAL
APPLICATIONS IN ACOUSTICS
This chapter provides background to the architecture of the glass curtain wall and acoustically relevant
characteristics. The approach to acoustic design practices and procedures are also discussed specifically
to sound isolation testing standards and current methods used in practice to improve acoustic
performance. In addition, precedent research and case studies were investigated to provide an
understanding of performance data analyzed in the design practice. The chapter is divided into four
sections:
1. Acoustic Detailing at the Curtain Wall Mullion
2. Acoustic Mitigation Practices and Products for Mullions
3. Sound Isolation Test Measurement Metrics and Methods
4. Acoustic Precedent Research Studies of Curtain Wall Systems
2.1 ACOUSTIC DETAILING AT CURTAIN WALL MULLIONS
This section investigates the anatomy of the unitized curtain wall system and how the design leads to
sound path weaknesses. The acoustic detailing of this architectural system is therefore important to
reduce sound flanking transmission.
2.1.1 THE ANATOMY OF THE GLASS CURTAIN WALL
Studying the anatomy of the unitized glass curtain wall systems can help identify mechanisms where
sound flanking paths occur. A glass curtain wall system is defined as an aluminum framed wall grid
containing glass infill panels or opaque infill panels of metal or thin stone.
6
Curtain wall system selection
drives a large part of the design process, performance, construction administration, building aesthetic,
and cost. Systems are supplied by manufacturers as off-the-shelf or custom designed solutions.
A unitized curtain wall system is modular, as opposed to a non-unitized or stick system. Unitized systems
are the focus of this research study. They are built and prefabricated in a factory as modular units
composed of a 4-sided perimeter of half mullions fastened to an insulated glazing unit (IGU) infill. The
prefabricated units are shipped to site and connected directly to the building. An anchoring system is
needed to connect the units to the floors and to wrap the units around the building as they interlock.
A stick system differs in that it requires that the framing and glass are pieced together in situ.
7
The vertical mullions of a glass curtain wall system are defined as structural elements that divide adjacent
infill glazing units.
8
Their purpose in a curtain wall system is to provide a rigid support to the infill glazing
of the window. When used to support glazing they are joined with horizontal mullions. Horizontal mullions
6
Nik Vigener and Mark Brown, “Building Envelope Design Guide - Curtain Walls,” Whole Building Design Guide: A Program of the
National Institute of Building Sciences, October 20, 2011, http://www.wbdg.org/design/env_fenestration_cw.php.
7
Ibid.
8
W. Müller and G. Vogel, Atlante Di Architettura (Milan: Hoepli, 1992), http://en.wikipedia.org/wiki/Mullion.
33
are also known as transoms; however the term horizontal mullion is more commonly used in practice
today.
Sound paths are created both laterally and vertically at the glass curtain wall façade (Figure 2-1). These
sound paths radiate acoustic energy inside a building where the vertical and horizontal mullions connect
at interior walls and floors. These connection points can reduce the sound isolation performance between
spaces, especially where no acoustic treatment has been considered.
FIGURE 2-1: SOUND TRANSMISSION PATHS ALONG A CURTAIN WALL FACADE
The aluminum mullion extrusions are considered lightweight in acoustic terms when compared to the
mass of other materials typically found in a building, such as concrete or steel.
A typical unitized curtain wall mullion section in plan has several points of connection (Figure 2-2). Labels
(a) through (d) indicate transitions where the mullion is connected to the glass infill:
(a) illustrates a mullion connected to the curtain wall glazing, this assembly condition defines the
connected mullion in the UVM test method;
(b) the mullion is shown without the glazing;
(c) identifies connected extrusions components of the unitized aluminum mullion; and
(d) illustrates the mullion independent from the curtain wall glazing and horizontal mullions, this
assembly condition defines the unconnected mullion in the UVM test method.
Vertical Sound Flanking
Transmission Paths
Lateral Sound
Flanking Transmission
Paths
Vertical Mullion
Horizontal Mullion
(aka Transom)
Stack Joint
34
FIGURE 2-2: UNITIZED VERTICAL MULLION EXTRUSION DISASSEMBLED FROM THE GLASS INFILL: THE CONNECTED
MULLION (A) AND UNCONNECTED MULLION (D)
The connected and unconnected vertical mullion assembly conditions are critical to understand for
application to the laboratory test methodology described in Chapter 3 (Figure 2-2 a and d). The mullion at
label (d) illustrates that the air cavity is divided in two by the interstitial “leg” extrusion, which joins at a
neoprene gasket (Figure 2-2).
A plastic mockup version of a unitized mullion extrusion identifies two metal anti-buckling clips fastening
both sides of the plastic mullion (Figure 2-3). Anti-buckling clips are used at all unitized systems to brace
the two halves of the mullion together. These clips are necessary to provide lateral stability of the system
and to maintain good weatherability by ensuring a positive pressure on the primary gasket.
FIGURE 2-3: SECTION OF PLASTIC MULLION MOCKUP WITH ANTI-BUCKLING CLIPS- COURTESY ENCLOS CORP
a b c d
ANTIBUCKLING
CLIPS
PRIMARY
NEOPRENE
GASKET
35
The clips are typically 2" long, are placed every 24" on center, and adhered in place by silicone dots.
Continuous clips may also be employed for other types of curtain wall systems. In the proposed laboratory
tests described in Chapter 3, the specimen module used two antibuckling clips at either end of the 5-foot
mullion specimen and was adhered by masking tape in the absence of a silicone dot. Images of this can
be found in Chapter 4.
The stack joint of a unitized system is located where two vertical mullions intersect with the horizontal
mullion. Normally the system is designed so that there is a continuous hollow cavity. The continuous air
cavity and frame enables a path for airborne and structure-borne sound to travel.
Due to economic demands and the goal of reducing material costs, unitized curtain wall systems often
span double story heights instead of traditional single story heights.
9
The vertical mullion is therefore
continuous between two floors and no stack joint occurs where a horizontal mullion intersects. This
potentially means that activities can be heard in a space from floors above or below the receiver floor.
An example of a double-span glass curtain wall system is at the Marriott-Ritz Carleton tower at LA Live,
designed by Enclos Corp. Acoustic studies were conducted to evaluate vertical sound flanking. This is
discussed as a case study later in this chapter.
2.1.1.1 LOSS DUE TO SOUND FLANKING
The total amount of deterioration in sound blocking due to the junction detail of an interior wall or floor
at the glass curtain wall can vary.
Acoustic consultants are familiar with dB loss estimates in the STC values due to sound flanking for typical
heavyweight junctions, for example, at a gypsum wall and concrete floor slab. Sources such as British
Gypsum
10
publish estimates of 4 – 5 dB loss where normal wall head meets a slab due to poor sealed
junction detailing.
Other sources such as from NRC-CNRC indicate up to a 15dB reduction from a continuous subfloor below
a partition.
11
Laboratory tests were conducted to measure the amount of noise reduction at the curtain wall. These
results are provided in Chapter 4.
9
Mic Patterson, “Structural Glass Facades,” 2011.
10
British Gypsum, “Education Sector Guide - 7 - Flanking Sound Transmission” (British Gypsum, Saint-Gobain, March 2014),
http://www.british-gypsum.com/~/media/Files/British-Gypsum/WHITE-BOOK-Sector-Guides/WBES/WBES-7-Flanking-Sound-
Transmission-04.pdf.
11
A.C.C Wamock, T. R. T. Nightingale, and M.R. Atif, “Estimation of Sound Transmission Class and Impact Insulation Class Rating
for Steel Framed Assemblies” (American Iron and Steel Institute / Steel Framing Alliance, 2008).
36
2.1.1.2 MULLION SOUND PATHS
Chapter 1 introduced the notion that sound can travel horizontally and laterally across a façade via paths
through the glass, mullions, and associated connections. There are several additional paths where sound
energy may transmit from one side of an internal wall to the other:
a. Glazing path is created when sound energy from the source room strikes the curtain wall glass
and transmits the sound energy as vibration to the lightweight mullion and subsequently to the
curtain wall glass of the receiver room. The receiver glazing becomes a diaphragm for transmitting
the structure-borne sound back into the air (Figure 2-4 Acoustic Key #1). This lateral path is
typically not considered in practice for methods of sound mitigation as obviously as the mullions.
b. Mullion path is when the structure-borne sound path is directly transmitted through the vertical
lightweight mullion extrusion of the curtain wall system (Figure 2-4 Acoustic Key #2).
c. Horizontal mullion path is the indirect sound transmission through the horizontal mullion
(transom) of the curtain wall system (Figure 2-4 Acoustic Key #3).
d. Connection path is the airborne sound path through air gaps/leaks where the mullion connects
to the demising partition. (Figure 2-4 Acoustic Key #4).
e. Direct wall path is the structure-borne sound path from direct transmission through the demising
partition (Figure 2-4 Acoustic Key #5).
Architectural Key:
a. Aluminum mullion extrusion
b. Glass infill of curtain wall framing
c. Aluminum horizontal mullion of the
curtain wall system
d. Indicative Internal wall partition
Acoustic Key:
1. Flanking sound transmission path across
glass -mullion -glass elements
2. Direct sound transmission path through
the vertical mullion extrusion
3. Flanking sound transmission path
through the horizontal mullion element
4. Flanking sound transmission path at the
resilient partition connection
5. Direct sound transmission path across
the interior wall partition
FIGURE2-4: PLAN DIAGRAM OF FIVE INDICATIVE SOUND PATHS AT THE GLASS CURTAIN WALL
37
2.1.1.3 COMPOSITE TRANSMISSION LOSS PERFORMANCE
Critical to understanding sound flanking transmission at the curtain wall is the significant influence to the
composite sound isolation performance.
The interior wall partition has a high sound isolation performance rating (Figure 2-5). This wall
construction is typically designed in buildings where high levels of acoustic privacy are required. The wall
consists of a double row of steel studs, two layers of gypsum wallboard at both sides, and batt insulation
in the air cavity.
The indicative STC rating for this double stud wall is degraded by 10 dB STC points due to the significantly
lower performing mullion. This amount of loss is significant because 10dB is perceived by the human ear
as twice the level of loudness, the doubling or halving of loudness level.
1213
FIGURE2-5: INDICATIVE DIAGRAMS SHOWING THE COMPOSITE PERFORMANCE BETWEEN A HIGH STC WALL RATING AND
LOW STC RATING OF A MULLION.
“Where a [demising] partition has a low isolation value of 35dB or less, flanking transmission is of little
consequence, but when partition values of 50 dB are reached, further improvement is limited by the
indirect sound paths.”
14
Good architectural detailing for coupling a heavyweight wall with a lightweight
mullion opens opportunities to improve acoustic performance ratings and validate the cost of building
materials.
Composite TL can be predicted with any element with an individually known STC rating.
12
Eckard Mommertz, Acoustics and Sound Insulation: Principles, Planning, Examples (Birkhäuser, 2009).
13
David A. Bies and Colin H. Hansen, Engineering Noise Control: Theory and Practice (Spon Press, 2003).
14
Smith, Peters, and Owen, Acoustics and Noise Control, 67.
38
2.1.1.4 ACOUSTIC BEHAVIOR OF SLITS AND GAPS
Another common reason why sound flanking occurs at the curtain wall system is due to the small air leaks
created by slits or gaps from the designed resiliency of the system or common field assembly conditions.
“When the gap size is larger than the wavelength, the wave passes through the gap and does not spread
out much on the other side. When the gap size is equal to the wavelength, maximum diffraction occurs
and the waves spread out greatly – the wave fronts are almost semicircular.”
15
Slits and gaps generally radiate only high frequency sound. This condition can occur at path #4 shown in
Figure 2-4 where the curtain wall mullion connects to the interior wall.
15
Trevor Cox, “Diffraction through a Single Slit,” Wave Diffraction, accessed July 18, 2014,
http://www.acoustics.salford.ac.uk/feschools/waves/diffract3.php.
39
2.1.2 MULTIDISCIPLINARY DESIGN PROVISIONS
There are several multidisciplinary considerations driving the design of details for a cohesive curtain wall
assemblage. These include resistance and control of fire, air and water infiltration, odor control, thermal
resistance, structural strength, durability, and control of sound and vibration
16
(Figure 2-6).
FIGURE2-6: SECTION THROUGH A CURTAIN WALL TRANSOM WHERE IT IS CONNECTED AT THE STRUCTURAL SLAB.
(MULLION SKETCH COURTESY OF MATT WILLIAMS, ARUP FACADES)
The mullion is resiliently fixed to the base building to tolerate façade deflections. Sound flanking paths
occur at lightweight and resilient connections at the perimeter of the walls and floors slabs.
There are several challenges associated with modifying the connections to the curtain wall, such as adding
mass to the system, maintaining resiliency, insisting on good workmanship, and improving code
enforcement. The cross-disciplinary requirements have significant implications on the acoustic
performance.
16
Chris Makepeace et al., Glass and Metal Curtain Walls: Best Practice Guide Building Technology (Public Works and Government
Services Canada: Canada Mortgage and Housing Corporation, n.d.), http://www.tboake.com/guides/curtain.pdf.
40
1. Adding Mass to the Curtain Wall System
Acoustic Issue Impact
Mass added to building materials generally
improves the sound isolation performance.
Sound flanking occurs at the edge condition
where a slab or wall meets the curtain wall
façade because the aluminum extrusions are
lighter in mass.
Adding mass to the curtain wall system will add
mass to the overall dead load of the façade and
would need to be considered structurally.
2. Maintaining Resiliency at the Curtain Wall System
Acoustic Issue Impact
Airtight seals are not necessarily required on
the inboard side between the curtain wall and
demising wall and can cause acoustic leaks.
Acoustic improvements result from full seals
and closure. Modifying/reducing the thermal
and structural resiliency afforded in a mullion
connection creates a challenge for acoustic
improvements.
Curtain walls require movement to expand and
contract as the building heats and cools, wind load
conditions, or movements from seismic events.
Modifications for acoustics can compromise the
resiliency required of the mullion connection.
3. Curtain Wall System Construction Administration
Acoustic Issue Impact
Poor workmanship of curtain wall system
constructions may compromise the acoustic
detail and degrade overall acoustic
performance. The field performance of the
acoustics is highly dependent on workmanship.
High construction quality must be carefully
implemented to avoid short circuiting or
bridging of building components.
Unitized systems (pre-assembled) versus stick
systems (field-assembled) can have significant
differences between the construction and
workmanship of the curtain wall system that can
compromise acoustic detailing and overall
acoustic performance. Some building construction
trades are not typically trained with acoustic
material techniques or installation process. Poor
construction practice can lead to misplacement of
sealants or even gaps between the elements. Field
inspections by qualified individuals are necessary
to insure good quality construction.
4. Sound Isolation Code Enforcement
Acoustic Issue Impact
Dated legislative code standards and
standardized US testing methods should be
revisited to clearly enforce specific
performances limits relevant to current
building technology.
Many sound isolation regulation requirements in
the United States are low compared to some
international standards. This limits the impetus
for design innovations to improve the acoustic
performance of sound flanking (Table 2-5).
41
2.1.3 ACOUSTIC DETAIL CONSIDERATIONS
The architectural design of the curtain wall system, its connections to the building interior, and required
provision for multidisciplinary design all contribute to possible sound flanking paths. The design’s details
and construction management should take into the consideration the following architectural conditions
to reduce sound flanking transmission, as proposed by LoVerde (2008)
17
with respect to primarily to
concrete structures:
- Mullions/windows
- Curtain wall connections to the slab
- Floor conditions
- Ceiling (slab) conditions
- Interior intersection details
- Penetrations
- Curtain wall intersection at wall
In a forum discussion amongst acoustic colleagues in the industry, it is noted that detail resolutions at
the curtain wall vary and can be challenging:
Larry Tedford, an Associate Principal at Arup in San Francisco, says, “There is no elegant solution
with the expectation of reasonably high acoustic separation performance at curtain wall mullions.
Typically an airtight separation with mass and a resilient disconnect is acoustically optimal and
aesthetically detrimental. So, the compromise point is what needs to be worked out.”
18
Kym Burgemeister, an Associate Principal at Arup in Melbourne, states that the “detailing of the
curtain wall is a challenge, because the façade is a living, breathing building element, that is
designed to move and expand/contract as the building heats and cools. Every connection to the
façade and mullion needs to be resilient and non-damaging and wrapped around the mullion but
not fixed to or through it.”
19
No one singular element of the curtain wall design is solely responsible for influencing of sound isolation
between two adjacent spaces. The dynamic façade system influences acoustic performance as a
composite, although certain elements within the composite may transmit more sound energy than
others.
The next section describes design considerations to improve the curtain wall systems.
17
LoVerde and Dong, “Methods for Reducing Flanking Airborne Noise Transmission through Mullions of Curtain Wall Systems.”
18
Larry Tedford, “Sound Flanking at Curtain Wall Mullions,” Arup Acoustics General Forum: Glazing and Facades, 2009.
19
Kym Burgemeister, “Sound Flanking at Curtain Wall Mullions,” Arup Acoustics General Forum: Glazing and Facades, 2009.
42
2.2 ACOUSTIC MITIGATION PRACTICES AND PRODUCTS FOR MULLIONS
Methods for improving the sound isolation performance between spaces have been considered by
acoustic consultants in the profession and by manufacturers in the industry. This often poses challenges
due to the multidisciplinary design provisions associated with the holistic design.
Opportunities for modification are often focused at the vertical mullion. Although this element has a
significant contribution to the performance and solutions offered in practice today tend to focus on it,
other components at the façade system, for example glass infill and connections, should be examined as
well.
2.2.1 PRACTICE-BASED MULLION MODIFICATIONS
Common practice resolutions to reduce sound flanking between adjacent spaces and to improve the
acoustic integrity of demising barriers are illustrated in Figure 2-7.
These acoustic concepts, considered by acoustic engineers and architects, are typically categorized by
adding mass and/or acoustic damping material in the mullion air cavity or as an overclad, i.e. enclosing
the mullion.
FIGURE2-7: CATEGORIES OF MULLION DETAIL MODIFICATIONS CONSIDERED TO IMPROVE SOUND ISOLATION
43
Alternatively, gypsum board wall partitions extend the drywall layers to enclose the mullion. Double
mullion systems with a center spandrel panel are used to entirely decouple acoustic paths between
spaces. These acoustic details only convey conceptual resolutions and are described in five categories;
window wall systems, air cavity fill inside the mullion, mullion overclad, double mullion and spandrel, and
unitized mullion without structural interstitial bridging. The compromise between cost and aesthetic
interventions with the amount of acoustic benefits also needs to be considered.
Window Wall Systems
Extending internal walls and floors to penetrate through the façade diaphragm is the most effective way
of reducing sound flanking transmission. This system is typically called a ‘window wall’ instead of a curtain
wall. The vibrational sound energy transmitting through the glazing and lightweight mullion path
terminates at the internal wall partition and does not pass through to the adjacent space. However,
penetrating the exterior glazing at these demising walls has significant architectural design implications.
Air Cavity Fill within the Mullion Cavity
Packing the hollow metal cavity of the mullion extrusion can add mass and damping to the element and
therefore improves the sound isolating performance. Materials seen in practice are cement board or steel
plate lining glued inside and gravel, sand or fiberglass fill. This is more often implemented in European
countries than in the U.S. Damping materials such as vinyl sheets are also an option to reduce structure
borne vibration. Other solutions involve “insulating the mullions by filling them with expanding foam,
sand, non-shrinking mortar, caulk or lightweight cement.”
20
Mullion Overclad
Encasing the mullions and sills both vertically and horizontally with layers of gypsum wallboard or
sheathing will improve the sound isolation of the lightweight mullion construction because mass and
damping are added. This modification to the mullion limits the sound across the mullion path, but only
addresses one of many paths for sound to travel.
In practice, mullion overclads can be a challenge to enforce in situ for the following reasons:
1. In the event that the demising partition is relocated (for example, in a retrofit), the overclad
attachment (screwed) or adherence (glued) will deface the mullion.
2. When MechoShades® (or similar product) are integrated as part of the architectural design,
the fitted dimension can be compromised.
3. Gypsum board cladding may not allow sufficient mullion deflection movement for certain
installations. Curtain wall mullions require allowable deflection movement ±1” toward
and/or away from the building at mid-span. This is dependent on the structural
requirements custom to each project.)
Double Mullion and Spandrel
Using a double mullion plus spandrel where walls and floors meet will reduce the sound flanking
transmission because this decouples the sound transmission paths. The double mullion arrangement
20
Dave Barista, “Glass Curtain Wall: Plenty of Light, but Is It Soundtight?,” Building Design + Construction, May 2, 2006,
http://www.bdcnetwork.com/glass-curtain-wall-plenty-light-it-soundtight?page=1&quicktabs_1=1.
44
allows a resilient disconnection between the mullions, but would need to be carefully detailed for relevant
design disciplines.
Unitized Mullion without Structural Interstitial Bridging
Isolating each structural member or “splitting” the mullion is a very effective way of reducing the sound
flanking path, but has to be carefully detailed for acoustic, structural, moisture and thermal integrity.
Assuming a design where the mullions running vertically and horizontally are seamless, this solution will
break the continuity requiring elements to be tied back to the structure separately.
Even when acoustic concept details are properly specified for a building, acoustic performance can still be
limited because installation is contingent on workmanship. Construction quality must be carefully
implemented to avoid acoustically short circuiting, i.e. mechanical bridging of building components. Many
building construction trades people are not trained properly with regard to acoustic material techniques
or installation processes.
45
2.2.2 PRODUCT-BASED MULLION MODIFICATION
Sound flanking transmission at the curtain wall is becoming a known issue in the industry, and some
manufacturers are patenting designs that aim to improve sound insulation performance between spaces.
This includes techniques such as the use of isolation clips, infill materials, and others (Table 2-1).
No. Product Description
1
Image from ©2012 PAC International, Inc. RSIC®-
AMI window mullion
This is a US based proprietary product
called Resilient Sound Isolation Clip. It is a
neoprene button that decouples both sides
of the curtain wall mullion from the
aluminum tube overclad.
The figure shows that the overclad is filled
with an MLV (mass limp vinyl) interlayer
and fiberglass insulation.
The composite performance per ©PAC
International, Inc. achieves STC 58.
2
Image from © Siderise Mullion / Transom Acoustic
Inserts
Siderise Mullion & Transom Inserts is a UK
based product. The mullion is filled with a
proprietary infill material to reduce the
vertical and horizontal sound transmission
between adjacent spaces.
The performance per © Siderise is “up to
41 dB R
w
‘through’ frame on a 50mm
mullion.”
46
No. Product Description
3
Image from © Siderise Acoustic Barrier Overlay
The Siderise Acoustic Barrier Overlay® is a
UK based product. It consists of a flexible
composite mass overlay to improve floor
to floor sound isolation.
4
Image from © Siderise Acoustic Void Barrier
The Siderise Acoustic Void Barrier® is a UK
based product.
The performance of the void barrier can
achieve over 51dB R
w
per the
manufacturer.
5
Image from © 2011 Gordon Incorporated, Mullion
Mate
Mullion mate® is a product by Gordon
Interior Specialties Division in the US. Its
main function is to close the gap between
a window mullion and partition wall with a
spring loaded device that snaps into place.
© Gordon Incorporated indicates an
attenuation performance of STC 38.
47
No. Product Description
6
Image from ©2014 Mull It Over Trim Cap
The Mull-It-Over Trim Cap is panel that is
capped to either side of a curtain wall
mullion. The panel leaf consists of a
composite of aluminum, foam, and a
damping interlayer.
The mullion performance per © Mull-It-
Over is an increase to STC 57.
7
Image from Michael Rizza Company
TM
, A Division
of Balco Inc. PCS Partition Closure Seal © 2014
Arcat, Inc.
The Partition Closure Seal® is a US product.
It is a silicone face seal join that provides a
closure between glass or mullion and an
interior wall.
8
Image from © 1998-2015 by EMSEAL Joint
Systems, Ltd.
EMSEAL® is a US based product consisting
of a mass-loaded acoustic seal that it used
to close the end of a partition to a glass
window or mullion.
The performance per © EMSEAL is STC 53
with one seal layer and STC 72 with two
layers.
TABLE2-1: MULLION PRODUCTS TO REDUCE SOUND FLANKING TRANSMISSION AT THE CURTAIN WALL
48
These products provide a mullion overclad, mullion infill, or mullion connection option to increase mass
and improve resilience of the mullion, but they do not address the flanking transmission across the glass
or necessarily at building connection components. Most of the acoustically tested products do not
necessarily include the glass infill in the test procedure.
Additionally, most of these products also require drilling holes to attach the product to the mullion. This
can be problematic especially if the partition is relocated or the building is repurposed.
In the proposed laboratory measurement procedure described in Chapter 3, three of the products
described here will be tested within the following phases:
Phase 2A: Partition Closure Seal®
Phase 2B: Mull-it-Over ® and Mullion Mate®
2.2.3 SUMMARY OF PRACTICE AND PRODUCT SOLUTIONS
Resolutions to improve the sound insulation performance of the mullion have been attempted by custom
engineered design methods or by proprietary products. However, both methods often have significant
impacts on the curtain wall systems design and construction including that they are
Usually not being cost effective,
Not aesthetically pleasing,
May not be part of the owner’s project requirements and criteria,
Not practical to construct,
Dependent on contractor’s ability to build it,
Not as acoustically effective as they should be,
Not able to resolve the flanking path through the glass.
49
2.3 SOUND ISOLATION METRICS AND MEASUREMENT METHODS
The performance due to sound flanking transmission is obtained with laboratory or field test
measurement for sound isolation. The US and Canada use test procedures based on ASTM standards and
Europe, the UK, Australasia and other parts of the world are based on ISO standards. Background
information for both standards will be referenced in this section.
Although the method for measuring and calculating the sound isolation performance of building
elements is generally similar, the results can significantly vary based on the applied acoustic indices. The
ASTM E90 standard for Sound Transmission Loss measurements in an acoustic laboratory will be used in
the test methodology proposed in Chapter 3.
The following section describes ASTM standardized test methods, rating procedures, and broad relevant
comparisons to ISO standards. A higher single number value indicates a higher sound isolating
performance of a specimen.
2.3.1 ASTM STANDARD TEST METHODS AND RATING PROCEDURES
The common laboratory test procedure for sound isolation in the US is defined by the ASTM E90-09
standard called “Standard Method for Laboratory Measurement of Airborne Sound Transmission Loss of
Building Partitions”
21
. Sound isolation performance is tested between two reverberant chambers in an
acoustic laboratory where a sound source is emitted in the source chamber and measured in the
adjacent receiving chamber. The difference between the sound emitted and sound received provides
the overall sound transmission loss (TL) at one-third octave band center frequencies. The frequency
range is defined from 125 Hz to 4000 Hz in this standard. Equation used to calculate Transmission Loss
(TL) is shown below:
= (
)+ ,
Where
is the average sound pressure level in the source room, dB
is average sound pressure level in the receiving room, dB
is the surface area of the partition (ft
2
)
is the absorption in sabins in the receiving room.
The term . is the normalizing factor. This needs to
be adjusted or normalized so that the Transmission Loss values
from different testing laboratories may be compared. It is used to
adjust for the different size of test specimens tested in each
laboratory and the amount of sabin absorption in each receiving
room.
EQUATION 2-1
22
21
E33 Committee, ASTM E90 - 09 Test Method for Laboratory Measurement of Airborne Sound Transmission Loss of Building
Partitions and Elements (ASTM International, 2009), http://www.astm.org/Standards/E90.htm.
22
E33 Committee, “ASTM E90 - 09 Test Method for Laboratory Measurement of Airborne Sound Transmission Loss of Building
Partitions and Elements” (ASTM International, 2009), 90, http://www.astm.org/Standards/E90.htm.
50
A single-figure numerical rating may be classified from the laboratory transmission loss called sound
transmission class (STC). The STC classification is based on the ASTM 413 standard and is one of the most
common indices used in the US to rate the sound isolation performance of all types of architectural
barriers (e.g. walls, floors, doors, windows).
23
This rating is only assigned to laboratory tested specimens.
It is possible that two different barrier assemblies perform with identical STC ratings, but may have
significantly divergent frequency regimes. Therefore the TL per octave band frequency is currently the
best way to understand the acoustic characteristics of a barrier than solely relying on the single figure
STC rating.
The STC rating is derived by weighing a reference contour based on the ASTM E413 standard to the
laboratory (standard noise reduction curve) measured one-third octave band TL values. The standard
was created to provide a single number rating for interior building partitions that are subjected to noises
from speech, television, radio, office equipment, and other mid to high frequency noise sources
24
.
An example of a STC reference contour adjustment to a TL performance is shown in FIGURE 2-8. The final
STC rating of the given test specimen is defined where the value at 500 Hz intersects at the defined
reference contour, in this example the TL is 32 dB at 500Hz, thus the specimen is rated STC 32.
FIGURE 2-8: STC REFERENCE CONTOUR AGAINST THE TRANSMISSION LOSS OF A TESTED SPECIMEN
25
23
E33 Committee, ASTM E413 -10 Classification for Rating Sound Insulation (ASTM International, 2010),
http://www.astm.org/Standards/E413.htm.
24
Architectural Testing, “Architectural Testing - Acoustical Performance Testing”, n.d., http://www.archtest.com/testing.
25
James E. Ambrose and Jeffrey Ollswang, Simplified Design for Building Sound Control (John Wiley & Sons, 1995).
51
The acoustic index that corresponds with a sound transmission measurement tested in the field is called
the field sound transmission class (FSTC). FSTC ratings can typically range 5– 10 dB less than the STC rating
for the same specimen. This is because building specimens tested in the laboratory are devoid of sound
flanking paths that reduce the achievable acoustic performance. The standardized procedure to measure
the transmission loss of a building element in the field is defined by ASTM E336
26
and rated per ASTM
E413.
Noise Isolation Class (NIC) is another valuable rating used to classify the sound isolation performance of
building specimens in the field. The rating is derived from Noise Reduction (NR) performance at one-third
octave band center frequencies. The NR is simply the arithmetic difference between the sound pressure
levels in the source and receiving rooms.
= ( )
Where
is the average sound pressure level in the source room, dB
is average sound pressure level in the receiving room, dB
EQUATION 2-2
27
Similar to the STC rating procedure, the single-figure NIC rating can be defined from NR values by
comparing the measured data to the standard reference contour per the ASTM E413 standard. The NIC
rating cannot be used in place of FSTC since it is only specific to the context in which it was measured: the
partition type, partition area, and amount of absorption present in the receiving room at the time of the
measurement
28
.
The NIC rating is different that the ASTM procedure for STC and FSTC because “no correction to the
measured [NR] data is made to account for partition size, receiving room absorption or sound flanking.
There are no widely used standards using the NIC rating, however the NIC rating is often used in lieu of
STC and FSTC ratings. NIC is used to assess the sound isolation performance of in situ partition
constructions, especially complicated ones that involve multiple sound transmission paths that are not
suited for laboratory testing.“
29
The ASTM standards corresponding to the sound isolation test rating are AST E90, ASM E336, and ASTM
E413.
30
ASTM E90 provides the measurement procedure to obtain transmission loss (TL) per one-third
octave band frequencies in an acoustic laboratory.
ASTM E336 provides the measurement procedure to obtain noise reduction (NR) and field
transmission loss (FTL) per one-third octave band frequencies in the field.
ASTM E413 provides the classification procedure to define the single number rating for sound
transmission class (STC), field sound transmission class (FSTC), and noise isolation class (NIC).
26
E33 Committee, “ASTM E336 - 11 Standard Test Method for Measurement of Airborne Sound Attenuation between Rooms in
Buildings” (ASTM International, 2011), http://www.astm.org/Standards/E336.htm.
27
Ibid., 33.
28
Marshall Long, Architectural Acoustics (Elsevier, 2006).
29
Malcolm J. Crocker, Handbook of Acoustics (John Wiley & Sons, 1998).
30
E33 Committee, “ASTM E413 -10 Classification for Rating Sound Insulation” (ASTM International, 2010), 413,
http://www.astm.org/Standards/E413.htm.
52
2.3.2 ISO STANDARD TEST METHODS AND RATING PROCEDURES
The ISO standards for sound isolation measurements range from 100- 3150 Hz. This differs slightly from
the ASTM standards that range from 125 – 4000 Hz. There are many ISO standards for testing the sound
isolation performance of specimens with and without flanking for various surface areas and beyond the
scope of this work. However relevant corresponding ISO indices that may be used as a comparison to
the ASTM standards are described in this section to inform case study comparisons later in this chapter.
The ISO single figure index R
w
is approximately comparable to the ASTM value STC. Similarly, the single
figure number D
w
is comparable to the ASTM value NIC. The final value of the comparable ISO and ASTM
indices will vary slightly. General equivalent performance descriptions between standards are
summarized from ASTM E90, ASTM E339, ASTM E413, ISO 10848, ISO EN 12354 and ISO EN 140 (Table
2-2).
ASTM
1
Index
ISO
2
Index
Measurement
Type
Value (dB) Description
TL R
Laboratory
(one-third octave)
TL = L1 – L2 + 10 log (S/A2)
R = L1 – L2 + 10 log (S/A2)
[TL] Transmission Loss
[R] Sound Reduction Index
STC Rw
Laboratory
(single figure)
Classified per TL (ASTM) or
R (ISO)
[STC] Sound Isolation Class
[Rw] Weighted Sound Reduction Index
NR D
Field
(one-third octave)
NR = L1 – L2
D = L1 – L2
[NR] Noise Reduction
[D] Level Difference
NIC Dw
Field
(single figure)
Classified per NR (ASTM) or
D (ISO)
[NIC] Noise Isolation Class
[Dw] Weighted Level Difference
-- Dn
Field
(one-third octave)
Dn = D – 10log (A/A0) [Dn] Normalized Level Difference
NNR DnT
Field
(one-third octave)
NNR = L1 – L2 + 10 log (T/T0)
DnT = L1 – L2 + 10 log (T/T0)
[NNR] Normalized Noise Reduction
[DnT] Standardized Level Difference
NNIC DnTw
Field
(single figure)
NIC
DnTw = D + 10log (T/T0)
[NNIC] Normalized Noise Isolation Class
[DnTw] Weighted Standardized Level Difference
TABLE2-2: CORRELATION BETWEEN ASTM AND ISO SOUND ISOLATION INDICES
1
The Transmission Loss frequency range in the ISO standard is from 100 – 3150 Hz.
2
The Transmission Loss frequency range in the ASTM standards is from 125 – 4000Hz.
Legend:
S Testing area of the specimen (ft
2
, m
2
)
A equivalent absorption area (ft
2
, m
2
)
A
0
reference absorption area (10m
2
)
T Reverberation Time (seconds)
T
0
Reverberation Time (0.5 seconds)
V Volume of receiving room (ft
3
, m
3
)
L
1
Average sound pressure level in the source room (dB)
53
L
2
Average sound pressure level in the receiving room (dB)
Sound isolation indices may be characterized as normalized, standardized, or weighted. This is a function
of the measurement conditions including sound flanking, room volumes, and sound absorption. The
indices are defined below:
Weighted: to establish a single figure rating descriptor, normalized or standardized levels are
compared to the Reference Curves published in BS EN ISO 717 or ASTM 413 for airborne noise
transmission.
Normalized: adding the Sabine equation (10 log (S/A) (metric) to the receiving room so that room to
room variation in the field will not influence the results. This is due to the variation of sound
absorbing materials encountered in the field.
Standardized: standardizing the sound pressure levels to a reverberation time of T = 0.5 sec is
equivalent to standardizing the equivalent area absorption of A
0
= 0.32 V (metric) if the
reverberation times differ.
There are many relevant ISO standards that are currently used to predict and test flanking sound
transmission (Table 2-3).
International Organization for Standardization
ISO 140-1 Acoustics -- Measurement of sound insulation in buildings and of building
elements -- Part 1: Requirements for laboratory test facilities with suppressed
flanking transmission
ISO 140-2 Acoustics -- Measurement of sound insulation in buildings and of building
elements -- Part 2: Determination, verification and application of precision data
ISO 140-3 Acoustics -- Measurement of sound insulation in buildings and of building
elements -- Part 3: Laboratory measurements of airborne sound insulation of
building elements
ISO 140-4 Acoustics -- Measurement of sound insulation in buildings and of building
elements -- Part 4: Field measurements of airborne sound insulation between
rooms
ISO 140-5 Acoustics -- Measurement of sound insulation in buildings and of building
elements -- Part 5: Field measurements of airborne sound insulation of façade
elements and façades
ISO 10848-1 Acoustics; Laboratory measurement of the flanking transmission of airborne
and impact sound between adjoining rooms - Part 1: Frame Document
ISO 10848-2 Acoustics; Laboratory measurement of the flanking transmission of airborne
and impact sound between adjoining rooms - Part 2: Application to light
elements when the junction has a small influence
ISO 10848-3 Acoustics -- Laboratory measurement of the flanking transmission of airborne
and impact sound between adjoining rooms -- Part 3: Application to light
elements when the junction has a substantial influence
ISO 10848-4 Acoustics -- Laboratory measurement of the flanking transmission of airborne
and impact sound between adjoining rooms -- Part 4: Application to junctions
with at least one heavy element
ISO 717-1:
A1:2006
Acoustics; Rating of sound insulation in building and of building elements - Part
1: Airborne sound insulation.
54
International Organization for Standardization
ISO 12354-1 Building Acoustics – Estimation of acoustic performance of building from the
performance of elements – Part 1: Airborne sound insulation between rooms
TABLE2-3: INTERNATIONALSTANDARDS FOR SOUND ISOLATION AND FLANKING TRANSMISSION
There are several relevant ASTM standards that address sound isolation and measure flanking sound
transmission (Table 2-4).
American Society for Testing and Materials
ASTM E90 Test Method for Laboratory Measurement of Airborne Sound
Transmission Loss of Building Partitions and Elements
ASTM E336
Annex A1 and Annex
A2
Standard Test Method for Measurement of Airborne Sound Attenuation
between Rooms in Buildings
ASTM E413 Classification for Rating Sound Insulation
TABLE2-4: AMERICANSTANDARDS FOR SOUND ISOLATION AND FLANKING TRANSMISSION
The ISO 12354 standard provides a prediction methodology for sound flanking transmission. Four paths
identified in the standard used for analytical calculation are defined (FIGURE 2-9). Designations for the
flanking paths and separating elements are identified between the source and receiving chambers. The F
is designated for the flanking element and D for a separating element in the source room and f and d for
respective flanking and separating elements at the receiving room.
31
FIGURE 2-9: DEFINITION OF SOUND TRANSMISSION PATHS BETWEEN ROOMS PER ISO 12354 (IMAGE: ISO 12354)
There are currently no ASTM standards for prediction methodologies.
31
ISO 12354-1:2000 Building Acoustics. Estimation of Acoustic Performance in Buildings from the Performance of Elements.
Airborne Sound Insulation between Rooms (BSI, July 15, 20000).
55
2.3.3 SOUND ISOLATION CRITERIA
The minimum sound isolation criteria regulated by building legislation varies between countries, both in
performance and acoustic index, i.e. ASTM and ISO. A summary of select international building legislation
primarily for residential building code is normalized to the same acoustic metric, sound transmission class
(STC) (Antonio, 2008) (Table 2-5). This is an indicator of what building code dictates as the absolute
minimum performance that can be built to meet acoustic privacy needs for health, safety, and
disturbance.
STC Building Legislation
69
1
Draft Nordic Standard ‘very good sound conditions’
65
2
German Standard (Highest Class Acoustical Comfort)
64
1
Draft Nordic Standard ‘satisfactory sound conditions’
61
1
Draft Nordic Standard ‘acceptable sound conditions’
61
3
UK ‘Quiet Homes’ minimum recommendation
60
4
Minimum Australian requirement: Bathrooms and Kitchens to habitable rooms
59
2
Minimum German Standard ‘lowest class acoustical comfort’
56
5
Minimum UK requirement
55
1
Draft Nordic Standard ‘less satisfactory sound conditions’
55
4
Minimum Australian requirement: Sole occupancy units
50
6
Minimum 2013 California Building Code (CBC), Residential
40
6
Minimum 2013 California Building Code (CBC), Non-residential
Note: The STC comparisons include assumptions and broad estimates of equivalence between
different acoustic indices. Accordingly, the table should not be used as an absolute justification
of criteria, but an indicator of approximate comparisons.
32
1. Draft Nordic Standard INSTA 122:1997 Sound Classification of Buildings
2. German Standard DIN 4109: Sound Insulation in Buildings
3. Quiet Homes: a guide to good practice and reducing the risk of poor sound insulation
between dwellings, Building Research Establishment
4. City of Sydney DCP, 1996
5. Building Regulations, Approved Document E, 2002
6. Minimum requirement 2013 California Building Code Section 1208
TABLE2-5: APPROXIMATE COMPARISON OF SOUND ISOLATION CRITERIA FOR INTERNATIONAL RESIDENTIAL BUILDING
CODE, STC
Sound isolation regulation requirements of the United States have not been developed to the same extent
as the international standards. This is due to the history of the federal noise legislation in the United
32
Nick Antonio, “Residential Sound Insulation and Building Code” (Acoustic Society of America, LA Chapter Meeting, Los Angeles,
January 15, 2008).
56
States. After federal noise regulation responsibilities transferred from national to state and local
governments, further research in building acoustics (e.g. sound flanking transmission) was curtailed.
Congress ended funding of the federal noise control program ONAC (Office of Noise Abatement and
Control) in 1981. The ONAC was originally established by the EPA (Environmental Protection Agency).
Before funding ended, the EPA established regulations and programs which were salient to the
development of many state and local government noise control laws across the United States
33
. However,
after ONAC closed, the responsibility of noise and abatement was transferred to State and local
governments, which truncated the development of further national regulations.
It is assumed that this may be an indication of the status of United States building research in acoustics
trailing those of other countries. For example, there are no testing standards or acoustic laboratories in
the United States which test sound flanking conditions like the Building Research Establishment in the
UK
34
, the NRC-IRC Flanking Sound Transmission Facility in Canada
35
or the Sound Flanking Transmission
Lab at the ift Rosenheim
36
in Germany.
In the United States, the EPA retains authority to conduct research and publish information on noise and
its effects on the public
37
.
Since building science research can often be driven by regulations, this is potentially why sound isolation
is at a minimum in certain state jurisdictions. The California Building Code (CBC) requirements are lower
than most international standards. CBC requires a minimum STC 40 interior sound transmission
performance between separating non-residential spaces and STC 50 at separating residential spaces.
38
.
Design incentives to improve the sound isolation performance in the USA may be limited because an
increase in code requirements can translate to added cost of construction.
The American measurement standard, which includes sound flanking transmission, is ASTM E336, but this
standard does not necessarily target the sound flanking weaknesses and instead is aimed at an overall
composite demising assembly performance. In the context of a demising wall joined at a curtain wall, it
would be difficult to extract the specific transmission loss contributions from the curtain wall.
Although ISO Standards have both a calculation prediction method (ISO 12354-1) and a testing method
(ISO 140-4) specific to sound flanking transmission, they do not apply to lightweight elements like a curtain
wall mullion. The computation developed in the ISO EN 12354 methodology for the apparent sound
insulation of building assemblies is for wall and floor elements that are assumed to be heavy, monolithic,
homogeneous and moderately damped
39
.
33
US EPA, “Noise Pollution | Air and Radiation | US EPA”, 1981, http://www.epa.gov/air/noise.html.
34
“BRE Group: Acoustics Laboratory”, n.d., http://www.bre.co.uk/page.jsp?id=1146.
35
F. King T. Estabrooks, “NRC-IRC Flanking Sound Transmission Facility,” National Research Council Canada, no. NRCC-51390
(October 2009): 3.
36
ift Rosenheim, “Laboratory Building Acoustics,” accessed May 14, 2015, https://www.ift-rosenheim.de/en/labor-bauakustik.
37
US EPA, “Noise Pollution | Air and Radiation | US EPA.”
38
2013 California Green Building Standards Code California Code of Regulations,Title 24, Part 11 (CALGreen) (International Code
Council, ICC, 2013).
39
T. R. T. Nightingale, “On Using Multiple Kij’s in the EN12354 Acoustics Prediction Model to Represent Excess Attenuation in
Flanking Surfaces” (presented at the Proceedings of the 17th International Congress on Acoustics, Rome, Italy, 2001).
57
The standardization of prediction and measurement methodologies of sound flanking transmission at
glass curtain wall remains under development for both ASTM and ISO standards.
Another consideration that standardized tests do not take into account is the potential room modes that
can be excited due to the location of the vertical mullion element at the corner of the room. A corner
location can amplify the sound level in the receiving room by exciting modes, emulating how a
loudspeaker at the corner of a room can provide this acoustic excitation. This acoustic effect should be
noted but is generally assimilated with the overall sound isolation rating.
The modal theory of sound indicates that by “aiming the loudspeaker into a corner of the room (especially
in smaller rooms), all resonant modes are excited, because all modes terminate in the corners.”
40
Modes
can excite and amplify frequencies based on room dimension and room shape. The aluminum mullion,
sill, and transom of a curtain wall are located at the corners of the room and sound that leaks through
these elements may be amplified based on their corner location in a room.
40
F. Alton Everest and Ken C. Pohlmann, Master Handbook of Acoustics (McGraw-Hill Professional, 2009).
58
2.4 ACOUSTIC PRECEDENT RESEARCH STUDIES
The following section describes academic and professional entities that have conducted research and
development in the field of sound flanking transmission relevant to facades.
Acoustic test laboratories that investigate façade sound isolation for acoustic consulting or design
fabrication are also identified, including test methods to evaluate lightweight building elements.
Although outside the scope of this research study, models for predicting sound flanking mechanism are
described based on the ISO 12354-1 method. These research studies conducted by others on predictive
models inform methods to compare physical experiment transmission loss (or sound reduction) data.
In addition, physical testing case studies specific to glass curtain wall elements are described. These test
experiments conducted by others inform the development of the UVM test procedure described in
Chapter 3.
2.4.1 LABORATORY ENTITIES AND INVESTIGATIONS OF FAÇADE SOUND ISOLATION
A proposal for a sound flanking laboratory to conduct research on lightweight construction was published
as early as 1974 by Nagy at the Technical University in Budapest.
41
The intent of the laboratory was for
students to conduct simultaneous airborne sound reduction measurements within a multi-room room
facility, which included decoupled volumes to measure curtain wall sound flanking transmission. At this
time, very few multi-chamber laboratories existed, except for those that investigated sound flanking
transmission across ceiling plenums. The research conducted by Nagy confirmed the complexity of paths
influencing sound flanking sound transmission for lightweight versus heavy weight specimens and the
variation of responses dependent upon partition quality.
Flanking transmission along a façade became a concern in countries where construction technology
connected lightweight aluminum façade assemblies to heavy weight concrete floor slabs. Experiments
and prediction of potential flanking at the junction was examined by Martin with an apartment building
in the Netherlands.
42
Later the in situ measurements were reported once the apartment building was
complete.
43
The result of the in situ measurements validated the prior laboratory and prediction study
conducted by the team. It was noted that the resilient coupling between the façade and the floor slab was
a challenge to predict using methods per ISO 12354 and measured vibration reduction index were
required to support the calculation. Once the in situ measurements were conducted the team validated
that the prediction results showed a dominant sound transmission via the façade, further validating the
importance of sound paths at façade connections.
Another example of experimental testing to investigate sound flanking at the façade was conducted by
Ando and Koga at the Technical Research Institute of Okumura Corp. and the Kajima Technical Research
41
J.P. Nagy, “Laboratory for Flanking Sound Transmission of Lightweight Constructions,” Civil Engineering 18, no. 3 (1974): 169–
78.
42
H.J. Martin, M.A.E. Schoffelen, and W.M. Siebesma, “Flanking Transmission along an Aluminium Façade – Experiments vs
Prediction-,” in Proceedings 17th International Congress on Acoustics Rome, vol. 3 (ICA, Rome, Italy, 2001).
43
H.J. Martin, M.A.E. Schoffelen, and W.M. Siebesma, “Flanking Transmission along an Aluminium Façade – Experiments vs
Prediction-,” in 18th Proceedings International Congress on Acoustics (ICA, Kyoto, 2004).
59
Institute.
44
In this case, the façade specimen consisted of lightweight concrete connected to a double stud
gypsum demising wall. The test setup used a semi-anechoic chamber to represent the exterior of the
façade. The experiment measurements were based on methods from EN 12354-1 2000. Vibration
measurements were also conducted. The test setup provided data for sound and vibration characteristics
of the flanking path of various excited and radiating areas.
In 2009 the National Research Council Canada (NRC-IRC) introduced an unprecedented flanking sound
transmission facility consisting of eight-rooms.
45
The facility would be used to characterize airborne and
impact sound transmission paths between rooms both laterally and vertically and support the
development of designs in accordance with building code requirements. The facility is most predominately
used for wood framed constructions. The NRC-IRC has extensive publications relevant to sound flanking
transmission prediction models using semi-empirical, statistical, and analytic methods, involving
collaborators from different countries.
The Building Research Establishment (BRE) in the UK provides UKAS accredited sound testing at their
sound flanking laboratory. The facility is used to compare sound isolation performance data with building
regulations in accordance with ISO measurement standards.
46
Curtain wall manufacturers, such as Permasteelisa Group and Schüco have also built research laboratories
to conduct acoustic test measurements in order to improve design and installation. Permasteelisa has a
Laboratory for Acoustic Research on Glass and Large Envelopes (L.A.R.G.E) located in Italy. Schüco has a
research facility called the Technology Center that includes a four room laboratory independent from one
another to conduct various acoustic testing.
47
Schüco has conducted sound flanking research studies on glass curtain wall facades in accordance with
DIN 52210 Part 7: Airborne and impact sound insulation, calculation of insulation against noise
transmission” at the ift Rosenheim laboratory in Germany. One research study was conducted in 2000 on
a non-unitized system where modifications were made to the mullion and transom profiles. Information
in this test report is proprietary.
48
In 2004 Schüco conducted another study on sound flanking transmission at the ift Rosenheim lab on the
currently known ‘USC 65’
49
mullion, which at the time was called the ‘Skyline S 65F’. This research study
tested the sound flanking transmission of a full scale ‘Skyline S’ curtain wall rig in accordance with ISO
10848. The frame profiles were modified with mass and damping materials to measure the flanking
transmission with reduced influence at certain areas of the curtain wall assembly.
50
Additionally,
predictive calculations were used per ISO 12354 to estimate normalized flanking level differences, D
n,f,w
(reference Table 2-2).
44
Kei Andow and Takashi Koga, “Experimental Study on Effect of Lining for Flanking Transmission of Building Facade,” in RBA-04
- SOUND INSULATION OF MULTI-FAMILY DWELLINGS, RBA-04 - SOUND INSULATION OF MULTI-FAMILY DWELLINGS (Forum
Acusticum Sevilla 2002, Sevilla, Spain, 2002), 6, http://www.sea-acustica.es/Sevilla02/rba04002.pdf.
45
F. King T. Estabrooks, “NRC-IRC Flanking Sound Transmission Facility,” National Research Council Canada, no. NRCC-51390
(October 2009): 3, doi:irc_id:20490.
46
“BRE Group: Acoustics Laboratory,” accessed March 24, 2012, http://www.bre.co.uk/page.jsp?id=1146.
47
“Sound Insulation of the Original Component” (Schüco, n.d.), http://www.schueco.com/web2/de-
en/investors/technology_center/specialist_areas.
48
Laing, “Systems for Facades, aluminum/PVC-U Windows and Doors - Fittings,” Acoustic Laboratory Test, Insulation against
Noise Transmission for Facades (Germany: Institute for Window Technology (ift) Rosenheim, November 13, 2000).
49
Schüco, “Overview of Profiles for Schüco Facade USC 65,” accessed May 19, 2014,
http://schilloh.2netmedia.de/pdf/produkte/fassaden_daecher/172256.pdf.
50
Bernd SaB and Ulrich Sieberath, “Classification Report, Sound Reduction and Flanking Transmission Loss of Building Elements,”
Acoustic Laboratory Test, Schuco Skyline S 65 F (Germany: ift Rosenheim, August 25, 2004).
60
Research studies and testing conducted by curtain wall manufacturers such as Schüco is often proprietary
and therefore may routinely conduct these type of laboratory measurements for a designated client
entities.
2.4.2 SOUNDFLANKINGPREDICTIONMETHODS
Although the prediction of sound flanking transmission is outside the scope of this research study, it is
noted that there continues to be research investigations to predict sound flanking paths at lightweight
building elements. Further development in this field of research may be applied to the results of the
proposed UVM test measurements described in Chapter 3.
The only method to analytically predict sound flanking in building elements is found in the international
standard ISO 12354-1. The method is typically applied to heavyweight homogeneous building elements
and therefore is not entirely applicable to lightweight elements such as curtain wall mullions. The
apparent sound reduction index of building elements relies in comparing the results of the measurements
obtained according to ISO 140-6.
51
An experimental study was conducted by researchers at Lund University in Sweden that studied flanking
transmission in lightweight buildings that used the EN 12354-1 calculation prediction standard. They
found that the standard used in lightweight applications predicted the transmission loss to be lower than
other tested methods. “The EN 12354-1 standard overestimates the transmission in lightweight buildings
by 1 to 8 dB. The orientation of the floor beams is important for transmission in the low frequency range.
Continuous floor plate as connector transmits the same amount of energy at high frequency independent
of floor beams orientation. This connector transmits the same amount of energy at high frequency
independent of floor beams orientation. This connector also has the best agreement with the EN 12354-
1 standard.”
52
The ISO standard EN 12354 for predicting flanking transmission, it is understood that bending waves are
primarily considered because this wave type has an out-of-plane displacement normal to the surface of
the building element, typically the dominant motion for acceptance and radiation of sound by a surface.
Further, building elements are usually weakly coupled at the junction and only the resonant wave
component, due to free bending waves, is transmitted structurally from one element to the other.
53
The
prediction method in EN 12354 is intended for homogeneous heavyweight monolithic structures and is
therefore not valid for lightweight hybrid curtain wall mullion assemblies.
Carl Hopkins, who is currently a professor at the University of Liverpool and previously with the Building
Research Establishment published a book which comprehensively discusses the theory of sound and
vibration in buildings called Sound Insulation.
54
The book includes sound isolation measurement and
prediction methods for application to the design and construction of buildings. Sound flanking prediction
methods are discussed with respect to statistical energy analysis.
51
B. Szudrowicz and A. Izewwska, “Empirical Verification of the Prediction Model Designed to Estimate the Flanking Transmission
in Buildings,” in Proceedings 17th International Congress on Acoustics Rome, vol. 3 (ICA, Rome, Italy, 2001).
52
Lars-Göran Sjökvist, “Flanking Transmission in Lightweight Buildings” (Lund University, Sweden: Department of Engineering
Acoustics, 2004), http://lup.lub.lu.se/record/929711.
53
T. R. T. Nightingale, “On Using Multiple Kij’s in the EN12354 Acoustics Prediction Model to Represent Excess Attenuation in
Flanking Surfaces” (Proceedings of the 17th International Congress on Acoustics, Rome, Italy, 2001).
54
Carl Hopkins, Sound Insulation (Elsevier / Butterworth-Heinemann, 2007).
61
Currently there are various software prediction programs understood to estimate sound transmission
including BASTIAN®, ENC®, WinFLAG®, Sound of Numbers®, Insul®, COMSOL® and Abaqus®. Based on
research conducted at Chalmers University of Technology, BASTIAN® can provide a reliable prediction of
sound flanking; however it is necessary for the user to have a theoretical understanding of the calculation
factors in order to make appropriate adjustments.
55
The research also states that BASTIAN® is a reliable
prediction program with the exception of low frequency discrepancies, which can be expected since
measurement uncertainties typically occur in this range.
2.4.3 PRECEDENT ACOUSTIC TEST MEASUREMENTS OF CURTAIN WALL ELEMENTS
Lateral sound transmission measurements across glass curtain wall elements has been studied and tested
by acoustic consultants and façade engineers for projects in the US and abroad. Available information
from the measurement reports can be limited since commissions are often proprietary and arranged by
client or owner entities for performance based designs.
A case study consisting of in situ lateral sound transmission measurements at a curtain wall mullion was
presented by Louwers
56
in 2012 at the Inter-Noise conference. The intent of the measurement study was
to reduce sound flanking transmission at the vertical curtain wall mullion by modifying the profile (filling
the façade stud or applying panel damping). The results indicated significant improvement with fill and
cladding additions to the mullion. However the best performance was achieved with a split mullion (Figure
2-10).
FIGURE2-10: PLAN DRAWING OF “SPLIT” MULLION (LEFT) AND D
NF
PERFORMANCE (RIGHT) (IMAGE: LOUWERS, 2012)
57
55
Jason Esan Cambridge, “An Evaluation of Various Sound Insulation Programs and Their Use in the Design of Silent Rooms”
(Chalmers University of Technology, 2006).
56
Marc Louwers, “Improvement of Acoustical Flanking Transmission through Light-Weight Façades,” in INTER-NOISE and NOISE-
CON Congress and Conference Proceedings, vol. 2012, 10 (Institute of Noise Control Engineering, 2012), 1998–2003.
57
Ibid.
62
A split mullion is one where the stiffening connections on either side of the internal mullion cavity do not
mechanically connect. Instead they are joined by a resilient rubber gasketing connection.
The mullions measured in the study were of a different type where the internal cavity did not have
interstitial stiffening connections. This mullion was modified and measured on site, initially with a mineral
wool fill and then an overclad with a damping material and 2mm steel plate. The single figure rating
difference between the filled mullion and the overclad mullion with fill was 3dB D
nf,w
.
All field tests in the Louwer study included a higher performing demising wall, therefore it was noted that
the Df and Fd paths would be dominated by the Ff path (see FIGURE 2-9 for reference acronyms).
It was also noted in the Louwer study that limited data was available to compare the mullion profile
modified with and without mineral wool fill. Chapter 4 includes this type of comparison as well as other
modifications to the same mullion profile.
Professional consulting firms such as Veneklasen Associates (VA) in the US have conducted an extensive
amount of field testing to better understand inter-spatial sound flanking transmission. Associates from
the firm John LoVerde and Wayland Dong presented historical detailing and modification to intersection
at curtain systems in 2008 and correlated the test data with various constructions.
58
This work identified
the extent of architectural interventions required in order to create robust sound isolation commensurate
with the demising partitions. It provided an indication of mullion modifications required to improve the
NIC rating. The performances ranged from approximately 30dB to 65dB where the higher performing
façade assemblies included two mullions separated by a wide spandrel pane. All field tests included a high
performing demising partition that varied per test.
The research and development team at Enclos Corporation, a leading firm of façade engineering and
curtain wall designers, has investigated various sound isolation issues associated with curtain wall building
facades, including sound flanking transmission vertically and horizontally. They collaborated on an
acoustic study with Veneklasen Associates on the LA Live project in Los Angeles, CA. The Enclos Corp team
summarized their work in a report titled Inter-Story Acoustical Evaluation of Unitized Curtain Wall
Systems
59
in 2008 that set a precedent for the proposed work in this research study. The focus of the
Enclos work was on sound transmission between vertical adjacencies at residential and hotel dwellings of
the LA Live Ritz Marriott building. This acoustic issue is especially important where the hollow internal
cavity of the vertical mullions spanned double-story heights between dwellings.
They evaluated, “the effect of [the] continuous pathway for airborne sound through the vertical mullion
on inter-story acoustical performance.” They proposed, “If found to be significant, identify strategies that
can be employed to mitigate the effect.”
60
The vertical transmission path was tested with a test rig that
included vertical mullion members. Two issues were investigated: the continuity of the aluminum
members without a stack joint across the double height and the void in the center of the mullion as
continuous to allow a conduit for sound. The study enabled the team to improve the sound transmission
performance of the vertical mullion and identify modifications providing the most value acoustically and
economically.
58
LoVerde and Dong, “Methods for Reducing Flanking Airborne Noise Transmission through Mullions of Curtain Wall Systems.”
59
TJ Dehghanyar et al., “Inter-Story Acoustical Evaluation of Unitized Curtain Wall Systems” (Culver City, CA: Enclos Corp, July
2008).
60
Ibid.
63
The tests were conducted at the Western Electro-Acoustic Laboratory (WEAL) in Valencia, CA. An image
of the lab test setup is shown where the aluminum curtain wall element has been separated from the
curtain wall system and located in an aperture in the filler wall. Testing was conducted per ASTM E90.
FIGURE2-11: SIMULATED FLOOR SLAB CONDITION AND TEST CHAMBER CONFIGURATION (LEFT) AND PHYSICAL TEST SETUP
(RIGHT), (IMAGES: ©ENCLOS CORP)
61
The conclusions from these tests indicated that significant acoustic improvements can be made to the
mullion condition by either capping the ends where the mullion discontinues, using interior finishes to
close the air path at the mullion termination, or by adding an insulation plug at either end of the vertical
mullion length. It was noted that structure sound transmission significantly contributed to the overall
sound levels.
Subsequent to the inter-story investigation, Enclos continued to do conduct studies at WEAL on
horizontal mullion acoustic performance and summarized findings the following year in a presentation
titled Partition Mullions: Curtain Wall Acoustical Enhancements.
62
Two mullion systems were studied
and compared to an unmodified mullion condition (System A) an overclad consisting of an MLV layer
and aluminum plate and (System B) an aluminum tube overclad filled with MLV pillows and attached to
the mullion with Pac-International RSIC® clips.
Images of the latter modification and test rig set up are shown in FIGURE 2-12.
61
Ibid.
62
TJ Dehghanyar, “Partition Mullions: Curtain Wall Acoustical Enhancements” (Enclos Studio, January 29, 2009).
64
FIGURE2-12: TEST SPECIMEN OF AN OVERCLAD MULLION ELEMENT AND LOCATION IN THE FILLER WALL APERTURE
(IMAGES:©ENCLOSCORP)
Laboratory Sound Transmission Loss results from the test are plotted (FIGURE 2-13).
FIGURE2-13: TL OF THE MODIFIED MULLIONS (IMAGES: © ENCLOS CORP)
65
Modifications in both System A and B significantly improved the base mullion performance and it is
noted that a dominant resonance frequency occurs at 400 Hz at the base condition.
66
2.5 SUMMARY OF PROFESSIONAL BACKGROUND AND PRECEDENT RESEARCH
This chapter provided background research about acoustic conditions and performance evaluation of the
curtain wall mullions to support Objective 1. The architecture of the vertical mullion was acoustically
evaluated for architectural mechanisms that enable sound flanking transmission paths. Methods for
improving sound isolation performance by modifying the curtain wall design was identified by methods
conducted by design consultants and by manufactured products. The proposed architectural
interventions are typically applied to vertical mullions and do not necessarily take into account other
architectural connections or sound flanking across the glass infill.
New building design projects are required to comply with codes and standards respective to project
location. Code dictates the absolute minimum requirements for health, safety, and/or disturbance
limitations. Legislative requirements for residential sound isolation in the United States are low compared
to certain international regulations. Therefore incentives for improvement are limited in the US, making
it difficult to convince owner or developer entities to approve a higher performance standards, especially
when the acoustic testing and enhanced design adds cost to a project.
Acoustic test methods per ASTM and ISO standards for façade sound transmission were identified and
broadly compared. Field and laboratory performance specific to flanking sound transmission is generally
limited with ASTM standards, whereas ISO standards provide more developed procedures. There is no
existing sound flanking prediction method within ASTM standards and predictive sound flanking methods
provided by ISO standards are limited since typically applied to heavyweight monolithic homogenous
barriers. Currently, techniques to improve the prediction of sound flanking methods for lightweight
systems are in development.
Precedent research on sound flanking transmission was reviewed for known laboratories conducting
measurements on curtain walls as well as analytical prediction methods to calculate sound flanking.
Laboratory and field test methodologies for lateral sound transmission at the curtain wall are often limited
since the elements are part of a composite, making it a challenge to identify which component contributes
to the dominating sound flanking path. Lateral sound transmission measurements for curtain wall systems
are more often tested as a composite and not necessarily per their individual parts. Although there are a
few investigations conducted by professional curtain wall manufacturers who are testing individual
elements disassociated from the curtain wall glazing.
There are various calculation and physical measurement techniques to evaluate and rate sound flanking
elements, such as with ISO 10848 and ISO EN 12354 methods. In addition to this, acoustic software has
been developed using statistical energy analysis and finite element analysis models to predict
transmission loss.
Additionally curtain wall case studies were evaluated to inform the test procedures and methods
proposed in this research study. Specifically the acoustic studies conducted by Enclos Corp
63
provided
valuable results to spearhead the test experiment proposed in Chapter 3. Beyond the precedent research
described in this chapter, it is not currently known if curtain wall test measurements have been conducted
by others in the US on elements associated with the curtain wall façade both independently and as a
composite to identify and compare acoustic characteristics of the architectural elements, in the absence
of an interconnecting partition.
63
Dehghanyar et al., “Inter-Story Acoustical Evaluation of Unitized Curtain Wall Systems.”
67
CHAPTER 3 UNITIZED VERTICAL MULLION RESEARCH METHODOLOGY
3.1 INTRODUCTION
This chapter describes the methodology of the proposed research experiment and analytical analysis
procedure. An overview of the methodology to support the research objectives is outlined below.
1. Review of Mullion Practice and Procedures
Identify sound flanking paths in a unitized curtain wall system
Identify existing methods of measuring transmission loss for
curtain wall systems
Identify current architectural interventions to improve the
sound isolation performance at mullions
Chapter 2
2. Mullion Test Experiment
Create a test method to measure the transmission loss of
individual and composite parts of a unitized glass curtain wall
specimen
Chapter 4
3. Performance Evaluation and Composite TL Analysis
Profile the STC performance of each test specimens
Compare/correlate noise reduction (NR) and transmission
loss (TL) levels and frequency regimes
Apply transmission loss levels to predictions using the
composite TL equation
Chapter 5
TABLE3-1: OVERVIEW OF APPROACH TO RESEARCH STUDY
As part of the first research objective, the background review revealed the anatomy of the unitized curtain
wall system and how its design inherently leads to sound flanking path weaknesses. Mitigation of these
weaknesses has been studied and measured by others in the industry. The investigation reveals limitations
with these approaches and that individual parts of the curtain wall system have not been evaluated in a
uniformly systemic method.
The background review supports the experimental test proposed to support Objective 2. This involves
measuring the transmission loss performances of individual and composite components of the curtain
wall specimen in a laboratory setting.
Results from test experiment will be applied to two different analytical analysis methods to meet research
Objectives 3 and Objective 4. An evaluation of the performance results will be compared between and
within test phases and the applied to a Composite TL calculation method.
68
3.2 REVIEW OF ACOUSTIC PRACTICES AND PROCEDURES AT CURTAIN WALL FACADES
The background review in Chapter 2 broadly identified laboratory test procedures for sound isolation and
current methods of design interventions for curtain wall mullions. Standards for measuring and predicting
lightweight systems, eg. glass curtain walls, are under further development for ISO and ASTM
methodologies. The common US standards for airborne sound isolation testing of building specimens in a
laboratory is the ASTM E90 method.
Sound transmission tests for curtain wall systems have been conducted by laboratory institutions in the
past; some known studies for sound flanking were conducted per Schüco. The review of the laboratory
test results demonstrated a challenge to distinguish which building element of the system contributes
most significantly to the overall sound transmission loss performance.
The empirical test method proposed in this research study is unique in a laboratory that conducts sound
transmission measurements per the ASTM E90 procedure. The method obtains the sound isolation
performance of individual elements that comprise the curtain wall system and provides a relative means
for comparison between modified elements. All boundary conditions are uniform by maintaining a
structural break between the test element and laboratory.
3.3 LABORATORY TEST PROCEDURE
The test method will be designated the UVM (Unitized Vertical Mullion) Method. The UVM experiment
will measure individual and composite parts of a curtain wall specimen. The approach consists of physical
laboratory tests so that the acoustic limitations of architectural interventions used in practice can be
quantified and relative comparisons between tests may be made.
The method primarily investigates the acoustic performance of three specific elements associated with
construction mechanisms supporting sound paths across the curtain wall system:
1. the vertical mullion extrusion,
2. the building connection element between the mullion and interior partition, and
3. the glass infill and aluminum framing.
In order to measure the independent transmission loss of the first two building specimens, they must be
decoupled from the curtain wall system.
69
Components of the glass curtain wall system to be measured are identified, each component or specimen
defines a test phase, and an approach to measure the lateral sound transmission loss (TL) across the
specimens is also described (Table 3-2).
TEST
PHASE
MEASURED SPECIMENS
(ISOLATED AND COMPOSITE)
MEASUREMENT APPROACH
Mullion Connection
Glass
Curtain Wall
PHASE 1
Individual unitized mullion measured with and
without architectural modifications at the
external face and/or internal air cavity
PHASE 2A
Individual mullion measured with various
resilient demising wall connections
PHASE 2B
Resilient demising wall connections measured in
the absence of a mullion
PHASE 3
Center mullion between two glass curtain wall
bays measured as a whole
TABLE3-2: UVMMETHODDESCRIPTION
The three phases in the UVM method are directly associated with the three sound flanking transmission
paths (Figure 3-1).
FIGURE3-1: PLAN DIAGRAM OF BUILDING ELEMENTS AND TEST PHASES ASSOCIATED WITH THE CURTAIN WALL ASSEMBLY
70
All measurements will be conducted at the Western Electro-Acoustic Laboratory (WEAL), an acoustic
laboratory in Santa Clarita, California. The testing will be in accordance with ASTM E90-09 Standard
Method for Laboratory Measurement of Airborne Sound Transmission Loss of Building Partitions.
The laboratory tests will be limited to lateral sound transmission loss. Vertical sound transmission loss at
curtain wall systems is not conducted as part of the laboratory set-up although many mechanisms
influencing sound flanking are not mutually exclusive vertically or laterally.
Phase 1 UVM
M
Modifications to these building elements will also be tested during each phase to identify the highest
practicable STC that may be achieved and acoustic findings relevant to performance. These will create
subcategories in each phase, for example in Phase 1 there are three main subcategories:
Class A – Unmodified mullion constant tests
Class B – Filled mullion tests
Class C – Overclad mullion tests with cladding and/or a combination of fill material
Phase 2 UVM
C
During Phase 2, the connection tests are measured with and without a mullion present. This will be
separated into Phase 2a and 2b respectively.
Phase 2 UVM
G
The third phase of the UVM lab tests focuses on the curtain wall glazing. The glass infill will be supported
by the perimeter aluminum extrusions of the transom, sill, and vertical mullion. The transmission loss of
this system will be compared with the initial Phase 1 UVM
M
Transmission Loss. All transmission loss results
will be analyzed for critical sound level and frequency correlations as well as composite calculations to
understand the influence to holistic system design.
3.3.1 TEST CONSTANTS AND VARIABLES
The physical elemental constant used in the experiement is the unmodified unitized vertical mullion, i.e.
hollow and exposed. The TL value in dB of this physical element is compared with mullions that were
modified with variable construction materials.
The variable materials are applied to the inside and/or outside of the physical mullion which remains
contant throughout most test phases.
3.3.2 CURTAIN WALL SYSTEM SPECIMEN
The curtain wall system was provided by Enclos Corp and decoupled to independently test the vertical
mullion separately. Shop drawings of the specimen created by Enclos are shown (FIGURE 3-2 and FIGURE
3-3).
71
FIGURE3-2: UNITIZEDCURTAINWALLSYSTEM SHOP DRAWINGS (COURTESY OF ENCLOS CORP), PLAN AND ELEVATION
72
FIGURE 3-3: UNITIZED CURTAIN WALL SYSTEM SHOP DRAWINGS (COURTESY OF ENCLOS CORP), DETAIL AT THE JOINT
(DIM. IS DIMENSION, REF. IS REFERENCE)
This curtain wall specimen is a unitized system composed of an aluminum perimeter frame extrusion and
insulating glazing unit (IGU) infill (FIGURE 3-2 and FIGURE 3-3).
Two bays of the unitized curtain wall specimen are shown in plan and elevation (FIGURE 3-2). The weight
of a single bay weighs approximately 335 pounds. The total height of the curtain wall system is 5’-0” and
total length of both bays when connected is 9’-10 ¾”. The total surface area of the glass infill is 21.5 ft
2
at each bay.
The vertical mullion connected to the glass infill is shown in plan (FIGURE 3-3). Its dimensions are 3” wide
x 6-3/4” deep x 60” high. The IGU assembly consists of ½” laminated glass - 1” air space – 3/8”
monolithic glass.
Notes regarding the specimen per phase:
UVM Phases 1 and 2 employed the vertical mullion only. Destructive and non-destructive tests
were designed to modify the performance of the mullion (Phase 1) and connection (Phase 2A
and 2B).
UVM Phase 3 included the entire assembly including both glass bays (FIGURE 3-2). The center
vertical mullion was modified in this phase only. No connection component was used.
½” laminate glass
1” air space
3/8” monolithic glass
73
3.3.3 TEST SPECIMEN COMPONENTS
The curtain wall elements used for in the UVM test procedure are individually described below for each
phase.
UVM Phase 1
(Mullion Element)
UVM Phase 2a and 2B
(Connection Element)
UVM Phase 3
(Glass System)
FIGURE 3-4: PLAN DIAGRAM OF THE
MULLION USED IN PHASE 1.
FIGURE 3-5: PLAN DIAGRAM OF A
CONCEPT CONNECTION USED PHASE 2.
FIGURE 3-6: PLAN DIAGRAM OF GLASS
CURTAIN WALL USED IN PHASE 3.
The vertical mullion extrusion is
shown decoupled from the
curtain wall glass, transom and
sill system.
The weight of the aluminum
mullion element is 26 lbs (11.8
kg) for a 5’0” length.
The connection consists of two
parallel aluminum plates which
would join a mullion at one end,
and an interior wall at the other.
Various other acoustic concept
connections are measured.
Phase 2a measured concept
connections with a mullion.
Phase 2b measured concept
connections without a mullion.
The largest specimen will be
used in Phase 3 where the
center vertical mullion is
coupled between the two
curtain wall bays. This includes
the glass infill.
All test specimens were placed in an aperture at the filler wall located between two reverberant test
chambers (FIGURE 3-7).
FIGURE 3-7: BASIC SETUP OF THE TEST SPECIMEN LOCATED IN THE TEST CHAMBER FILLER WALL.
74
3.3.3.1 TEST BASE CASE, CONTROLS AND VARIABLES
The laboratory experiment includes the following conditions so that the results of the test specimens may
be analogously compared between phases.
Test Base Case
Two base case performances were obtained with the physical mullion element that remained constant
throughout the testing experiment. The base cases are identified in Test Phase 1 and described in Chapter
4.
The lowest performance base case is defined as Mullion Constant 1 (MC1) and the highest performance
base case is defined as Mullion Constant 2 (MC2). These base cases are compared to curtain wall building
elements tested in Phases 2 and 3. The physical mullion shape and dimension remained constant at all
test phases although building mass and damping infill, overclad and connections materials varied.
Test Controls
The boundary condition was controlled at each test phase. This generally required all test specimens to
be placed in the filler wall aperture with a minimum ¼” perimeter air gap. The edge condition created by
the gap was sealed with an acoustically resilient material, putty or caulking.
The acoustic influence of this boundary conditions is potentially changed by the length of the linear
perimeter of a test specimen.
Test Variables
Test specimens were modified with the following variables depending on the test phase.
Test mullion infill materials: sand, ¼” diameter pea-gravel, damping materials, mineral wool
Test mullion overclad materials: gypsum board, steel sheet metal, damping materials
Test specimen edge condition: foam, silicone, rubber gasketing
Test specimen structural supports: wood battens
75
3.3.4 LABORATORY TEST CHAMBERS
The laboratory transmission loss test chambers at WEAL include a reverberant sound source chamber
decoupled from the reverberant receiving chamber (FIGURE 3-8). A high sound isolating filler wall
separates the two chambers. The loudspeaker in the source chamber excites acoustic energy in the
room. The resistance to this acoustic energy incident on the test specimen was measured in the
receiving chamber. The microphone located in the receiving chamber measured the residual acoustic
energy transmitted through the test specimen.
Greater detail for customized test rig preparation and setups are shown in Chapter 4 at each phase.
FIGURE 3-8: DIAGRAM PLAN DRAWINGS OF THE TRANSMISSION LOSS CHAMBER AT WEAL AND THE TEST ELEMENTS
FROM THE CURTAIN WALL SYSTEM.
The Receiving Chamber dimensions are 6.30 m (20.67 ft) x 4.53 m (14.88 ft) x 5.18 m (17.00 ft), and
the volume is 148.0 m
3
(5226.1 ft
3
).
The Source Chamber dimensions are 6.55 m (21.50 ft) x 5.09 m (16.71 ft) x 6.10 m (20.00 ft), and the
volume is 203.4 m
3
(7184.6 ft
3
).
3.3.4.1 FILLER WALL AND TEST SPECIMEN APERTURE
The highlighted wall (chamber filler wall) between the source and receiving chambers represents the
extents of the laboratory filler wall at WEAL (FIGURE 3-8). This is the designated wall area typically used
76
to insert specimen modules for doors, wall, and façade assemblies. The filler wall fills-out the remaining
area around a given specimen module size. The sound isolation performance of the filler wall must be
high in order to obtain the correct transmission loss value of the test specimen. The filler wall assembly
consists of a double stud wall with four layers of 5/8” type ‘X’ gypsum wall board on the source side and
three layers at the receiving side. The wall air cavity is filled with 9” R-30 batt insulation, and the overall
width of the wall is approximately 13-1/2”. The aperture within the filler wall will be sized to fit the
specimens with a perimeter air gap of ¼” so there is no contact between specimen and filler wall. The
single figure transmission loss rating performance for chamber filler wall is STC 74 in accordance ASTM
E90.
The total face area of the test specimens varies for every phase.
The test aperture is framed with a set of wood studs consisting of 2”x 6” stud at the receiving side and
2”x 8” stud at the source side. The double wood studs are separated by a ¼” to ½” air space (FIGURE 3-9).
FIGURE 3-9: CHAMBER FILLER WALL CONSTRUCTION (DETAIL COURTESY OF WEAL)
Test specimen was consistently centered unless otherwise noted in the detail descriptions in this
section.
Standardized location of the mullion is centered in the chamber.
Comparison examples of a hollow and filled aperture are shown (Figure 3-10 and Figure 3-11).
77
FIGURE3-10: APERTURE IN CHAMBER FILLER WALL IS
OPEN WITHOUT A TEST SPECIMEN
FIGURE3-11: APERTURE IN CHAMBER FILLER WALL IS
FILLED WITH A TEST SPECIMEN
(SPECIFICALLY FILLED WITH A MULLION AND
SILICONE CONNECTION USED IN PHASE 2A)
3.3.4.2 FILLER WALL TL PERFORMANCE
The transmission loss performance of the filler wall with no aperture is measurement TL13-232 and an
open aperture is measurement TL13-331.
Weal Test STC Description
TL13-331 0 Filler wall with an aperture (opening of 7-1/4" x 60-1/2")
TL13-232 74 Filler Wall Data (with no aperture)
TABLE3-3: FILLER WALL TESTS, PERFORMANCES AND DESCRIPTION S
Open aperture in
chamber filler wall
Filled aperture in
chamber filler wall
78
FIGURE3-12: SOUND TRANSMISSION PLOTS OF FILLER WALL
3.3.4.3 PHASE 3 TEST CHAMBER RIG
A special test rig setup is required for Phase 3. Measuring transmission loss laterally across the curtain
wall bay will require the specimen to sit perpendicular to the filler wall. The vertical center mullion
between the bays would sit in the filler wall aperture.
Structural reinforcements are required to hold the curtain wall bays on either side of the filler wall.
Acoustic detailing to limit the passage of sound at the exterior face of the curtain wall system will be
detailed to simulate an outdoor condition. This will limit the sound transmission through the glass that
somehow re-enters the area on the other side of demising partitions in practice. This sound transfer will
be limited by creating auxiliary semi-anechoic chambers at the outboard side of the curtain wall bay and
at either side of the filler wall (FIGURE 3-13). These smaller chambers will be filled with batt insulation
and named Chamber 3R at the receiving room and Chamber 3S at the source room.
-20
0
20
40
60
80
100
63 80 100 125 160 200 250 315 400 500 630 800 1000 1250 1600 2000 2500 3150 4000 5000
Transmission Loss (dB)
One-third Octave Band Center Frequency (Hz)
Filler Wall Transmission Loss
TL13-331, STC 0 TL13-232, STC 74
79
FIGURE3-13: PLAN DRAWING OF AUXILIARY SEMI-ANECHOIC CHAMBERS 3S AND 3R CUSTOM BUILT FOR THE PHASE 3 TEST MEASUREMENTS
80
FIGURE3-14: INTERSECTIONDETAIL AT THE WEAL FILLER WALL AND THE CURTAIN WALL BAY
Chamber 3S represents the semi-anechoic enclosure inside the WEAL Source Chamber (FIGURE 3-15 and
FIGURE 3-16).
FIGURE 3-15: SEMI-ANECHOIC CHAMBER 3S AT THE SOURCE SIDE FILLED WITH BATT INSULATION
CHAMBER 3S
81
FIGURE 3-16: SEMI-ANECHOIC CHAMBER 3S AT THE SOURCE SIDE
Chamber 3R represents the semi-anechoic enclosure inside the WEAL Receiving Chamber (FIGURE 3-17).
FIGURE 3-17: SEMI-ANECHOIC CHAMBER 3R AT THE SOURCE SIDE
CHAMBER 3S
CHAMBER 3R
82
3.3.5 TEST EXPERIMENT FACTORS
1. Flanking paths at the sill and transom have intentionally been removed from the primary testing
phases 1 and 2 in order to focus on the behavior of the vertical mullion and connection
elements. In practice, all parts of the composite assembly are important; however the
measurement results from the separated elements are to be compared with the composite glass
curtain wall system in Phase 3.
2. Vertical flanking paths are not analyzed. Although similar principles regarding the acoustic
treatment of the junction between the slab and curtain wall apply, no laboratory tests or
analysis are conducted.
3. Materials used in these test measurements are selected based on standard assemblies used in
practice today, i.e. material used at the mullion infill or overclad.
4. The IGU assembly in Phase 3 includes a laminated pane; the PVB interlayer is between surface
#3 and #4. Flexural vibration through the laminated pane may differ to a monolithic pane and
influence the transmission loss performance.
5. Choices available to set the specimen flush with one side of filler wall or at the center were
considered. It was decided to set the mullions at the center of the filler wall for two reasons: to
continue from the Enclos Corp precedent testing as discussed in Chapter 2 and to simulate a
field condition where the mullion is typically centered at a demising wall.
6. Putty and/or wet seal caulking are used to seal the perimeter edges of the specimens during all
phases.
7. Pink noise is used at the source chamber.
3.4 LABORATORY RESULT ANALYSIS
Test measurement results obtained through Objective 2 will be used to support Objectives 3 and 4. The
laboratory performances will be analytically compared for significant finding between phases and applied
to transmission loss predictions for composite conditions with an interconnecting wall.
3.4.1 SOUND TRANSMISSION CORRELATIONS AND COMPARISONS
The test data acquired from the UVM test phases will be analyzed in one-third octave bands to extract
significant contributions related the noise reduction (NR) and/or transmission loss. (TL). Sound
transmission class (STC) ratings will be identified to categorize the highest and lowest performing
assemblies.
3.4.2 COMPOSITETRANSMISSIONLOSSPREDICTIONS
The test data acquired from the UVM test procedure will be applied to composite TL calculation
predictions. This composite will include an interior demising wall assembly.
The transmission loss for select curtain wall elements (e.g. mullion, partition connection, glass) will be
acquired from testing the specimens at WEAL.
83
The laboratory transmission loss data for the high performing demising wall partition is obtained from the
National Research Council Canada
64
, one of the independent acoustical laboratories that catalog the TL of
building materials. As a note, there are other acoustic laboratory institutions around the world that
catalogue all types of partition constructions.
The composite transmission loss for the overall building system may be analytically predicted once all
Transmission Loss performances are collected.
Composite TL Equation
The composite transmission loss of a non-homogeneous wall may be estimated with the following
equation.
=10 10
=
=
Equation 3-1: Composite
Transmission Loss
6566
64
R.E. Halliwell et al., “NRC-CNRC Gypsum Board Walls: Transmission Loss Data,” Internal Report (Canada: Institute for Research
in Construction, March 1998), http://archive.nrc-cnrc.gc.ca/obj/irc/doc/pubs/ir/ir761/ir761.pdf.
65
David A. Bies and Colin H. Hansen, Engineering Noise Control: Theory and Practice (Taylor & Francis, 2009).
66
Peter Hubert Parkin, Henry Robert Humphreys, and J.R. Cowell, Acoustics, Noise, and Buildings, Fourth Edition (Faber and Faber,
Boston MA, 1979).
84
CHAPTER 4 TEST RESULTS AND ANALYSIS OF THE UNITIZED VERTICAL
MULLION MEASUREMENT PHASES
4.1 INTRODUCTION
This chapter provides the laboratory measurement results of the unitized vertical mullion (UVM) test
method conducted between May 2013 and March 2014 at four different stages:
1. UVM test phase 1: May 2013
2. UVM test phase 2:
A. July 2013
B. October 2013
3. UVM test phase 3: March 2014
The outline of each section will include the following information (Table 4-1):
1. Phase specific laboratory test set up description
2. Test specimen description, acoustic modifications of materials, and configuration
3. Tabulated STC results for each test specimen per phase (one-third octave band sound
transmission loss spectrum results can be found in Appendix B UVM Laboratory Test )
4. Transmission loss overlays of significant test specimen comparisons
5. Summaries of notable observations at each test phase, field notes and considerations for future
explorations
TEST
PHASE
CHAPTER
SECTION
MEASURED SPECIMENS
(CONNECTED AND UNCONNECTED) ACOUSTIC METRIC
Mullion Connection Glass Infill
PHASE 1 Ch 4: 4.2
Sound transmission class (STC)
ratings
One-third octave band sound
transmission loss (per center at
frequency range 63 Hz –5000 Hz
PHASE 2A Ch 4: 4.3
PHASE 2B Ch 4: 4.4
PHASE 3 Ch 4: 4.5
PHASE 3 Ch 6
Vibration acceleration level
(dB re 10
-6
G) measurements
TABLE4-1: TEST PHASE OUTLINE OF DATE, UVM, AND RESULTS
85
4.2 PHASE 1 – MULLION ELEMENT (ISOLATED)
This laboratory test phase measures the TL performance of the vertical mullion (Figure 4-1). The mullion
is modified with test variables including infill and overclad materials and measured in the absence of a
partition connection or glazing element. Specific configuration descriptions are provided in the following
sections. A total of 22 laboratory measurements were conducted.
Date May 28, 30, 31, 2013
Laboratory: WEAL
Specimen: Vertical unitized
mullion element; connection
and glazing element not
included
Dimensions: 3”x 6-3/4” x
60” (76 mm x 171mm x
152mm)
Surface Area: 2.71 ft
2
(0.25m
2
)
Transmission: Horizontal
Procedure: ASTM E90-09
Total Tests: 22
FIGURE4-1: PHASE1:PLAN DIAGRAM OF UVM ELEMENTS AND ASSOCIATED TEST PHASES
4.2.1 PHASE 1 SPECIMEN AND TEST CHAMBER DESCRIPTION
The aluminum mullion profile illustrates how the hollow air cavity is divided in plan by the interstitial leg
stiffener connections (Figure 4-2). The various fill materials used for mullion modifications were placed in
the larger air cavity. The fill materials of various densities included sand, mineral wool, pea gravel, and
damping materials. The materials used to overclad the mullion also consisted of various densities such as
gypsum wall board, steel plates, and vinyl damping materials. These were adhered or appended to the 6-
3/4” face of the mullion.
86
FIGURE4-2: PLAN DRAWING OF UNITIZED VERTICAL MULLION PROFILE
A photograph of the vertical mullion profile taken at the laboratory is shown (Figure 4-3).
The silicone gasket is present between the interstitial mullion leg stiffeners. A wooden spacer was placed
at either end of the mullion cavity to maintain an overall width of 3” as would be the case in situ.
FIGURE4-3: IMAGE OF HOLLOW EXPOSED MULLION PROFILE WITH ANTI-BUCKLING CLIPS, WOOD SPACES AND SILICONE
GASKET
Anti-
buckling
clips
Wood
Spacer
Primary
silicone
gasket
Large air cavity
(Modified with fill materials)
Small air cavity
(Not modified for testing)
Interstitial “legs” interlock at a
silicon gasket (not shown here)
87
All specimens sat on a neoprene spacer and the perimeter edge gap was sealed with backer rod and putty
(Figure 4-4, FIGURE 4-5, and FIGURE 4-6).
FIGURE4-4: IMAGE OF HOLLOW EXPOSED MULLION PROFILE WITH ANTI-BUCKLING CLIPS, WOOD SPACES AND SILICONE
GASKET
FIGURE 4-5: A ¼” GAP BETWEEN MULLION PERIMETER
AND FILLER WALL
FIGURE 4-6: PUTTY APPLIED TO SPECIMEN PERIMETER
EDGE TO SEAL ACOUSTIC LEAKS
88
The test specimen was placed into an aperture in the chamber filler wall with a face area dimension of
60-1/2” x 6-1/2" (Figure 4-7). This allowed a ¼” perimeter airspace between the specimen and the filler
wall to avoid direct contact with each other.
FIGURE4-7: ELEVATION OF THE MULLION IN THE FILLER WALL WITH SEALED PUTTY PERIMETER
4.2.2 TESTING CLASSIFICATIONS
The modifications to the unitized vertical mullion were categorized into three measurement classes:
Mullion Class A – Exposed and hollow mullion (test constant)
Mullion Class B – Cavity filled mullion tests
Mullion Class C – Overclad mullion tests (with a combination of fill material) The Class C test series was
further subdivided based on the overclad material type:
C1 aluminum + MLV layer,
C2 gypsum board + MLV layer,
C3 gypsum board, and
C4 aluminum tubes.
All the overclad materials were screwed to the mullion, not glued.
89
4.2.3 MATERIAL DESCRIPTIONS
Mullion Cavity Fill Materials
Material
Description
mineral wool
2 pcf
sand
Filled in small plastic bags and laid in the
mullion cavity
pea gravel
Approximately ¼” diameter, small plastic
bags were filled with pea gravel and laid in
the mullion cavity
Mass-loaded vinyl (MLV)
pillows
2 layers of MLV material (3/16”) thick,
arched side to side in the mullion cavity,
with mineral wool packed in the remaining
air gap
Overclad Materials:
Material
Description
Aluminum tube
1-1/2” (16 mm) aluminum tube (1/8” thick
with 1-1/4” airspace)
¼” RSIC isolator
Gypsum Wall Board
5/8” thick
Steel Plate
1/8” thick
MLV 3/16” layer
No tests were conducted where a damping compound was attached to the inboard walls of the mullion.
Compounds were attached only to the outboard walls.
90
FIGURE 4-8: CLASS A – MULLION (HOLLOW AND EXPOSED)
FIGURE 4-9: CLASS B – FILLED MULLIONS
FIGURE4-10: CLASSC–OVERCLADMULLIONS
MINERAL WOOL
91
4.2.4 PHASE 1 CLASS A TEST SEQUENCE
Class A mullion test results (HOLLOW AND EXPOSED MULLION) (Table 4-2).
WEAL Test No. STC Material Layers [description] Element Drawing [Plan]
(1) TL13-309 37
[1] 1/8” (3mm) aluminum mullion
[2] 2-3/4” (70mm) air space
[3] 1/8” (3mm) aluminum mullion
[mullion is flush with source room side]
[silicone gasket is not included]
(2) TL13-310 34
[1] 1/8” (3mm) aluminum mullion
[2] 2-3/4” (70mm) air space
[3] 1/8” (3mm) aluminum mullion
[mullion is centered in the filler wall]
[silicone gasket is not included]
(3) TL13-311 36
[1] 1/8” (3mm) aluminum mullion
[2] 2-3/4” (70mm) air space
[3] 1/8” (3mm) aluminum mullion
[silicone gasket is included]
[mullion test constant]
(4) TL13-312 47
[1] 1/8” (3mm) aluminum mullion
[2] 3-3/4” (95 mm) air space
[3] 1/8” (3mm) aluminum mullion
[mullion leaves disconnected by 1” (25mm)]
TABLE4-2: PHASE1CLASSA,STCRESULTS, AND SPECIMEN DESCRIPTION
The goal of this mullion test series was to determine the effect of the position on the transmission loss. It
was decided that all mullion positions should be located at the center of the filler wall to simulate a
centered condition in situ with a demising wall. Other laboratories test specimens that are flush to one
side of the chamber wall, but that was not the choice for this case.
92
4.2.4.1 PHASE 1-A DEDUCTIONS AND OBSERVATIONS
Test specimen TL13-311 is identified as one of the base cases and applied to variable conditions in
subsequent test measurements. It is considered the minimum base case performance and used for
measurement comparisons in other phases.
TL13-312 was conducted to understand the acoustic impact of a 2-leaf system without interconnections.
This mullion does not have a practical use when completely separated in this way because the stability of
the system is compromised. Further development of this is conducted in Phase 2B.
The TL spectra of the mullions in Phase 1 Class A include
Lowest STC: (2) TL13-310, STC 34
Highest STC: (4) TL13-312, STC 47
Mullion specimen TL13-309 is located in the filler wall flush against the source side of the room and has a
resonance at 160 Hz. TL13-311 (test constant) is centered in the filler wall and has a resonance at 400 Hz
(Figure 4-11).
TL13-309 TL13-310 TL13-311
TL13-312
FIGURE4-11: PHASE1-ATRANSMISSIONLOSSCURVES
*UNTREATEDMULLIONBASECASE,MC1
0
20
40
60
80
100
63
80
100
125
160
200
250
315
400
500
630
800
1000
1250
1600
2000
2500
3150
4000
5000
Transmission Loss (dB)
One-Third Octave Band Center Frequency (Hz)
Sound Transmission Loss of Phase 1-A Test Sequence
Hollow and Exposed Mullions
(1) TL13-309, STC 37
(2) TL13-310, STC 34
*(3) TL13-311, STC 36
(4) TL13-312, STC 47
93
4.2.5 PHASE 1 CLASS B TEST SEQUENCE
Class B mullion test results (MULLION CAVITY FILLED) (Table 4-3).
WEAL Test No. STC Material Layers [description] Element Drawing [Plan]
(5) TL13-313 39
[1] 1/8” (3mm) aluminum mullion
[2] 2-3/4” (70mm) pea gravel
[3] 1/8” (3mm) aluminum mullion
[specimen weight is 49.5 lbs (22.5 kg)]
(6) TL13-314 38
[1] 1/8” (3mm) aluminum mullion
[2] 2-3/4” (70mm) sand
[3] 1/8” (3mm) aluminum mullion
[specimen weight is 45.5 lbs (20.6 kg)]
(7) TL13-315 36
[1] 1/8” (3mm) aluminum mullion
[2] 2-3/4” (70mm) mineral fiber 2.5 pcf
[3] 1/8” (3mm) aluminum mullion
[mineral fiber 2.5 pcf (40 kg/m
3
)]
[mineral fiber laid in, not ram packed]
(8) TL13-316 38
[1] 1/8” (3mm) aluminum mullion
[2] 2-3/4” (70mm) MLV pillow
[3] 1/8” (3mm) aluminum mullion
[specimen weight is 34.5 lbs (20.6 kg)]
TABLE4-3: PHASE1CLASSB,STCRESULTS, AND SPECIMEN DESCRIPTION
94
FIGURE4-12: CLASSBFILLEDMULLION(A) PEA GRAVEL (B) SAND (C) MINERAL WOOL (D) MLV PILLOWS
a b d
c
95
4.2.6.1 PHASE 1-B DEDUCTIONS AND OBSERVATIONS
The highest performing mullions in Class B is TL13-313, filled with pea gravel and TL13-316, filled with
MLV pillows. These fill materials are used for measurements in the next sequence of measurements, Class
C. Additionally these fill configurations are understood to be cost effective and practical.
The TL spectra of mullions in Phase 1 Class B are shown include (Figure 4-13):
Lowest STC: (7) TL13-315, STC 36
Highest STC: (5) TL13-313, STC 39
FIGURE4-13: PHASE1-BTRANSMISSIONLOSSCURVES
*UNTREATEDMULLIONBASECASE,MC1
There are several observations that can be drawn with reference to Figure 4-13 and Table 4-4:
Mullion infill provides up to a +6 STC dB point average increase compared to TL13-311.
There is a +3dB variation between material fill variations when comparing the maximum and
minimum TL values.
The standard deviation between Class B tests #5 - #8 indicate a 1 – 2 STC dB point change
depending on the mullion fill.
There is a 3 to 5 STC dB standard deviation between Class B tests #5 -# 8 and TL13-311 between
the 800 Hz to 5 kHz frequency range.
The mullion cavity infill provides significant TL improvement above 800Hz when compared to
TL13-311.
The transmission loss results of Class B show a significant improvement at 630 Hz to 5000 Hz in
comparison to the MC-1 frequency spectrum.
0
20
40
60
80
100
63
80
100
125
160
200
250
315
400
500
630
800
1000
1250
1600
2000
2500
3150
4000
5000
Transmission Loss (dB)
One-Third Octave Band Center Frequency (Hz)
Sound Transmission Loss of Phase 1-B Test Sequence
Filled Mullion Cavity
(5) TL13-313, STC 39
(6) TL13-314, STC 38
(7) TL13-315, STC 36
(8) TL13-316, STC 38
*(3) TL13-311, STC 36
96
Class B Standard Deviation
One-third Octave Band Center Frequency (Hz)
63 80 100 125 160 200 250 315 400 500 630 800 1k 1.25k 1.6k 2000 2.5k 3.15k 4k 5k
Class B tests 0.4 0.5 0.6 0.6 0.6 1.9 2.0 1.2 1.0 0.5 0.8 1.7 1.8 1.5 2.0 1.6 1.7 1.8 2.2 2.0
Class B tests
and TL13-311
0.4 0.4 0.6 0.7 0.5 1.7 2.0 1.3 0.9 0.6 1.2 4.5 5.0 3.5 5.1 3.8 2.9 3.4 4.6 2.9
TABLE4-4: CLASSB STANDARD DEVIATION
97
The following images are indicative of the overclad typologies used at each Class C variation:
FIGURE4-14: CLASSCOVERCLADMULLIONIMAGES FROM TEST SERIES C1, C2, C3, AND C4
C1 1/8” Aluminum plate and 3/16”
MLV
C2 5/8” gypsum board and 3/16” MLV
C4 1-1/2” aluminum tubes resiliently attached
C3 5/8” gypsum board
98
4.2.7 PHASE 1 CLASS C1 TEST SEQUENCE
Class C1 mullion test results [1/8” ALUMINUM PLATE PLUS 3/16” MASS LIMP VINYL LAYER OVERCLAD] (Table 4-5).
WEAL Test
No.
STC Material Layers [description] Element Drawing [Plan]
(9) TL13-317 46
[1] 1/8” (3mm) aluminum plate
[2] 3/16” (5 mm) MLV layer
[3] 1/8” (3mm) aluminum mullion
[4] 2-3/4” (70mm) MLV pillow
[5] 1/8” (3mm) aluminum mullion
[6] 3/16” (5 mm) MLV layer
[7] 1/8” (3mm) aluminum plate
[specimen weight is 54 lbs (24.5 kg)]
(10) TL13-318 48
[1] 1/8” (3mm) aluminum plate
[2] 3/16” (5 mm) MLV layer
[3] 1/8” (3mm) aluminum mullion
[4] 2-3/4” (70mm) air space
[5] 1/8” (3mm) aluminum mullion
[6] 3/16” (5 mm) MLV layer
[7] 1/8” (3mm) aluminum plate
[specimen weight is 45.25 lbs (20.5 kg)]
(12) TL13-320 46
[1] 1/8” (3mm) aluminum plate
[2] 3/16” (5 mm) MLV layer
[3] 1/8” (3mm) aluminum mullion
[4] 2-3/4” (70mm) pea gravel
[5] 1/8” (3mm) aluminum mullion
[6] 3/16” (5 mm) MLV layer
[7] 1/8” (3mm) aluminum plate
[specimen weight is 69 lbs (31 kg)]
TABLE4-5: PHASE1CLASSC1,STCTESTRESULTS AND SPECIMEN DESCRIPTION
99
4.2.7.1 PHASE 1-C1 DEDUCTIONS AND OBSERVATIONS
The TL spectra of mullions in Phase 1 Class C1 (Figure 4-15) include
Lowest performing STC: (9) TL13-317, STC 46
(12) TL13-320, STC 46
Highest performing STC: (10) TL13-318, STC 48
(11) TL13-319, STC 48
FIGURE4-15: PHASE1-BTRANSMISSIONLOSSCURVES
*UNTREATEDMULLIONBASECASE,MC1
There are several observations that can be drawn with reference to Figure 4-15 and Table 4-6:
TL13-318 performed 2dB STC points higher than TL13-317 although the mullion cavity is hollow,
which reduced the overall mass of the specimen.
The overclad in this series (aluminum and MLV) provides up to a 14dB STC increase compared to
a hollow exposed mullion.
Composite variations of mullion cavity infill with the overclad provide a 2dB STC improvement
The overclad provides significant improvement above 250 Hz.
TL13-318 (hollow cavity) has more than a 5 dB TL reduction at 800Hz and 1 kHz compared to TL13-
317 and TL13-320 which have filled mullion cavities.
Class C1 test measurements improved over 10dB TL across the frequency ranges above 250 Hz
compared to TL13-311 mullion constant.
The standard deviation within the Class C1 measurement tests 9-12 is 1 dB – 3 dB across all
frequencies.
The standard deviation between Class C1 (tests 9-12) and TL13-311 is 4 dB -8 dB at the 250 Hz to
5 kHz frequency range.
0
20
40
60
80
100
Transmission Loss (dB)
One-Third Octave Band Center Frequency (Hz)
Sound Transmission Loss of Phase 1-C1 Test Sequence
1/8" Thick Aluminum Plate + MLV Overclad
(9) TL13-317, STC 46
(10) TL13-318, STC 48
(12) TL13-320, STC 46
*(3) TL13-311, STC 36
100
Class C1 Standard Deviation
One-third Octave Band Center Frequency (Hz)
63 80 100 125 160 200 250 315 400 500 630 800 1k 1.25k 1.6k 2000 2.5k 3.15k 4k 5k
Class C1 tests 0.9 0.5 0.8 0.4 0.5 0.6 0.3 0.5 0.9 0.9 1.7 3.2 2.8 0.9 0.4 0.2 0.2 0.1 0.4 1.3
Class C1 tests
and TL13-311
0.9 0.4 0.7 0.5 0.8 1.5 4.3 4.7 4.7 4.7 5.9 8.4 8.3 6.5 7.9 7.0 6.3 6.9 7.1 5.1
TABLE4-6: CLASSC1 STANDARD DEVIATION
101
4.2.8 PHASE 1 CLASS C2 TEST SEQUENCE
Class C2 mullion test results [5/8” GYPSUM BOARD PLATE PLUS 3/16” MASS LIMP VINYL PLATE OVERCLAD] (Table
4-7).
WEAL Test
No.
STC Material Layers [description] Element Drawing (Plan)
(13) TL13-321 50
[1] 5/8” (16 mm) gypsum board plate
[2] 3/16” (5 mm) MLV layer
[3] 1/8” (3mm) aluminum mullion
[4] 2-3/4” (70mm) air space
[5] 1/8” (3mm) aluminum mullion
[6] 3/16” (5 mm) MLV layer
[7] 5/8” (16 mm) gypsum board plate
[specimen weight is 49.5 lbs (22.5 kg)]
(14) TL13-322 47
[1] 5/8” (16 mm) gypsum board plate
[2] 3/16” (5 mm) MLV layer
[3] 1/8” (3mm) aluminum mullion
[4] 2-3/4” (70mm) pea gravel
[5] 1/8” (3mm) aluminum mullion
[6] 3/16” (5 mm) MLV layer
[7] 5/8” (16 mm) gypsum board plate
[specimen weight is 72.5 lbs (33 kg)]
(15) TL13-323 52
[1] 5/8” (16 mm) gypsum board plate
[2] 3/16” (5 mm) MLV layer
[3] 1/8” (3mm) aluminum mullion
[4] 2-3/4” (70mm) MLV pillows
[5] 1/8” (3mm) aluminum mullion
[6] 3/16” (5 mm) MLV layer
[7] 5/8” (16 mm) gypsum board plate
[specimen weight is 57.5 lbs (26 kg)]
TABLE4-7: PHASE1CLASSC2,STCTESTRESULTS AND SPECIMEN DESCRIPTION
4.2.8.1 PHASE 1-C2 DEDUCTIONS AND OBSERVATIONS
The TL spectra of mullions in Phase 1 Class C2 (Figure 4-16) include
Lowest performing STC: (14) TL13-322, STC 47
Highest performing STC: (15) TL13-323, STC 52
TL13-323 is the highest performing test in the Phase 1 series and is used as the second test constant for
subsequent testing Phases.
102
FIGURE4-16: PHASE1-C2TRANSMISSIONLOSSCURVES
*UNTREATEDMULLIONBASECASE,MC1
There are several observations that can be drawn with reference to Figure 4-16 and Table 4-8:
It is unclear why TL13-322 (including a pea gravel mullion fill) performs lower than TL13-321
(hollow mullion cavity) even though the former includes additional mass.
TL13-321 in this test series performs 2 dB STC points higher than TL13-318 from the last test series
C1. The difference may be attributed to the difference in overclad mass as both mullion have
hollow cavities. The mass of the gypsum board overclad is heavier than the aluminum plate
overclad in the last series.
In general, the overclad of gypsum board + MLV provides up to a 15dB STC increase when
compared to TL13-311 (hollow and exposed mullion).
Class C2 tests provide a 10 dB improvement across the 250 Hz to 5 kHz frequency region.
Class C2 Standard Deviation
One-third Octave Band Center Frequency (Hz)
63 80 100 125 160 200 250 315 400 500 630 800 1k 1.25k 1.6k 2000 2.5k 3.15k 4k 5k
Class C2 tests 0.3 0.3 0.4 0.2 0.5 0.7 0.8 1.0 1.4 3.6 3.9 1.3 1.5 2.4 2.1 0.9 0.8 0.7 0.5 0.8
Class C2 tests
and TL13-311
0.5 0.3 0.3 0.3 1.1 2.3 6.0 6.8 6.8 7.0 7.9 9.3 9.3 6.7 8.6 8.3 7.4 8.2 8.6 6.2
TABLE4-8: CLASSC2 STANDARD DEVIATION
The standard deviation within the Class C2 tests 13-15 is 1 dB - 3dB across one-third octave band
center frequencies, with the exception of 500 Hz and 630 Hz.
0
20
40
60
80
100
Transmission Loss (dB)
One-Third Octave Band Center Frequency (Hz)
Sound Transmission Loss of Phase 1-C2 Test Sequence
Gypboard +MLV Overclad
(13) TL13-321, STC 50
(14) TL13-322, STC 47
(15) TL13-323, STC 52
*(3) TL13-311, STC 36
103
The standard deviation between Class C2 tests 13-15 and TL13-311 is 6 dB – 9 dB from the 250
Hz to 5 kHz frequency region.
104
4.2.9 PHASE 1 CLASS C3 TEST SEQUENCE
Class C3 mullion test results [5/8” GYPSUM BOARD PLATE OVERCLAD] (Table 4-9).
WEAL Test
No.
STC Material Layers [description] Element Drawing (Plan)
(16) TL13-324 47
[1] 5/8” (16 mm) gypsum board plate
[2] 1/8” (3mm) aluminum mullion
[3] 2-3/4” (70mm) MLV pillows
[4] 1/8” (3mm) aluminum mullion
[5] 5/8” (16 mm) gypsum board plate
[specimen weight is 47 lbs (21 kg)]
(17) TL13-325 42
[1] 5/8” (16 mm) gypsum board plate
[2] 1/8” (3mm) aluminum mullion
[3] 2-3/4” (70mm) air space
[4] 1/8” (3mm) aluminum mullion
[5] 5/8” (16 mm) gypsum board plate
(18) TL13-326 45
[1] 5/8” (16 mm) gypsum board plate
[2] 1/8” (3mm) aluminum mullion
[3] 2-3/4” (70mm) pea gravel
[4] 1/8” (3mm) aluminum mullion
[5] 5/8” (16 mm) gypsum board plate
[specimen weight is 61.5 lbs (28 kg)]
TABLE4-9: PHASE1CLASSC3,STCTESTRESULTS, AND SPECIMEN DESCRIPTION
105
4.2.9.1 PHASE 1-C3 DEDUCTIONS AND OBSERVATIONS
The TL spectra of mullions in Phase 1 Class C3 (Figure 4-17) include
Lowest performing STC: (17) TL13-325, STC 42
Highest performing STC: (16) TL13-324, STC 47
FIGURE4-17: PHASE1-C3TRANSMISSIONLOSSCURVES
*UNTREATEDMULLIONBASECASE,MC1
There are several observations that can be drawn with reference to Figure 4-17 and Table 4-10:
The gypsum board overclad provides up to an 11dB STC increase compared to a hollow exposed
mullion, TL13-311.
There is 3dB STC between material variations in the mullion cavity
The performance of the Class C3 test sequence is greater than 5 dB compared to TL13-311 from
the 250 Hz – 5 kHz frequency region.
The standard deviation within Class C3 tests 16-18 is 1 dB - 3dB across all frequencies.
The standard deviation within Class C3 tests 16-18 and TL13-311 is 5 dB – 8 dB from the 250 Hz to
5 kHz frequency region.
Class C3 Standard Deviation
One-third Octave Band Center Frequency (Hz)
63 80 100 125 160 200 250 315 400 500 630 800 1k 1.25k 1.6k 2000 2.5k 3.15k 4k 5k
Class C3 tests
(16-18)
0.4 0.4 0.5 0.1 1.3 1.4 0.8 1.9 1.2 2.4 2.8 2.5 2.8 2.3 2.4 1.2 0.8 1.5 1.9 2.1
Class C3 tests
and TL13-311
0.3 0.3 0.5 0.3 1.3 2.1 4.6 5.3 4.7 4.2 4.1 6.4 7.0 4.8 6.9 7.1 6.6 7.1 7.9 5.8
TABLE4-10: CLASSC3 STANDARD DEVIATION
0
20
40
60
80
100
63
80
100
125
160
200
250
315
400
500
630
800
1000
1250
1600
2000
2500
3150
4000
5000
Transmission Loss (dB)
One-Third Octave Band Center Frequency (Hz)
Sound Transmission Loss of Phase 1-C3 Test Sequence
Gypsum Board Overclad
(16) TL13-324, STC 47
(17) TL13-325, STC 42
(18) TL13-326, STC 45
*(3) TL13-311, STC 36
106
4.2.10 PHASE 1 CLASS C4 TEST SEQUENCE
Class C4 mullion test results (1-1/2” HOLLOW ALUMINUM TUBE OVERCLAD WITH RESILIENT CONNECTION] (Table
4-11).
WEAL Test
No.
STC Material Layers [description] Element Drawing (Plan)
(19) TL13-327 31
[1] 1-1/2” (16 mm) hollow aluminum tube
[2] 1/4” (6mm) airspace RSIC isolator
[3] 1/8” (3mm) aluminum mullion
[4] 2-3/4” (70mm) air space
[5] 1/8” (3mm) aluminum mullion
[6] 1/4” (6mm) airspace RSIC isolator
[7] 1-1/2” (16 mm) hollow aluminum tube
[specimen weight is approximately 47 lbs]
(20) TL13-328 38
[1] 1-1/2” (16 mm) MLV + aluminum tube
[2] 1/4” (6mm) airspace RSIC isolator
[3] 1/8” (3mm) aluminum mullion
[4] 2-3/4” (70mm) air space
[5] 1/8” (3mm) aluminum mullion
[6] 1/4” (6mm) airspace RSIC isolator
[7] 1-1/2” (16 mm) MLV+ aluminum tube
[specimen weight is 56.75 lbs (21 kg)]
(21) TL13-329 48
[1] 1-1/2” (16 mm) MLV pillow+ alum tube
[2] 1/4” (6mm) airspace RSIC isolator
[3] 1/8” (3mm) aluminum mullion
[4] 2-3/4” (70mm) air space
[5] 1/8” (3mm) aluminum mullion
[6] 1/4” (6mm) airspace RSIC isolator
[7] 1-1/2” (16 mm) MLV pillow+ alum tube
[specimen weight is 58 lbs (26 kg)]
(22) TL13-330 48
[1] 1-1/2” (16 mm) MLV pillow+ alum tube
[2] 1/4” (6mm) airspace MLV buttons
[3] 1/8” (3mm) aluminum mullion
[4] 2-3/4” (70mm) air space
[5] 1/8” (3mm) aluminum mullion
[6] 1/4” (6mm) airspace MLV buttons
[7] 1-1/2” (16 mm) MLV pillow+ alum tube
[specimen weight is 58.5 lbs (26.5 kg)]
TABLE4-11: PHASE1CLASSC4,STCTESTRESULTS AND SPECIMEN DESCRIPTION
107
4.2.10.1 PHASE 1-C4 DEDUCTIONS AND OBSERVATIONS
The TL spectra of mullions in Phase 1 Class C4 (Figure 4-18) include:
Lowest performing STC: (19) TL13-327, STC 31
Highest performing STC: (21) TL13-329, STC 48
(22) TL13-330, STC 48
FIGURE4-18: PHASE1-C4TRANSMISSIONLOSS
*UNTREATEDMULLIONBASECASE,MC1
There are several observations that can be drawn with reference to Figure 4-20 and Table 4-12:
A significant resonance at 630 Hz is present in all measurements of this C4 test sequence. The
acoustic excitation frequency may be the same as the natural frequency of aluminum tube.
With the exception of 630 Hz, class C4 tests, there is a general TL improvement to the TL13-311
mullion from 250 Hz – 5 kHz.
The difference between test specimens TL13-327 and TL13-330 is the resilient isolation
connection to the aluminum tube overclad, the Pac-International RSIC isolators and MLV buttons
respectively. The performance difference between the two tests is negligible.
The aluminum tube overclad provides up to a 9dB STC increase compared to the TL13-311 hollow
exposed mullion. This is a significant improvement, however not as high as previous overclad
systems due to the resonance seen at 630Hz that lowers the overall STC rating.
Infill variations within mullion cavity provide a 3dB STC improvement.
There is 3dB STC between material variations Class C4 Investigation.
Standard deviation between Class C4 tests 19-22 is up to 9dB across the frequency spectrum.
Standard deviation between Class C4 tests 19-22 and TL13-311 is 5 dB – 10 dB at 250 Hz to 5 kHz.
0
10
20
30
40
50
60
70
80
63
80
100
125
160
200
250
315
400
500
630
800
1000
1250
1600
2000
2500
3150
4000
5000
Transmission Loss (dB)
One-Third Octave Band Center Frequency (Hz)
Sound Transmission Loss of Phase 1-C4 Test Sequence
Aluminum Tube Overclad
(19) TL13-327, STC 31
(20) TL13-328, STC 38
(21) TL13-329, STC 48
(22) TL13-330, STC 48
*(3) TL13-311, STC 36
108
Class C4 Standard Deviation
Standard
Deviation
One-third Octave Band Center Frequency (Hz)
63 80 100 125 160 200 250 315 400 500 630 800 1k 1.25k 1.6k 2000 2.5k 3.15k 4k 5k
Class C4 tests
(19-22)
0.2 0.8 0.3 0.4 1.0 1.4 2.1 2.8 3.8 5.4 8.5 9.0 5.8 2.6 2.1 1.1 1.2 3.1 2.0 1.0
Class C4 tests
and baseline
(19-22 & 3)
0.3 0.7 0.3 0.4 1.5 2.5 5.4 6.4 7.4 5.9 7.5 10.1 9.7 7.8 8.7 8.0 7.1 7.7 7.9 6.2
TABLE4-12: CLASSC4 STANDARD DEVIATION
PAC International® RSIC Clips are compared to C4 tests 19 – 22 (FIGURE 4-19).
FIGURE4-19: TRANSMISSIONLOSSOVERLAY OF PHASE 1-C4 AND PAC INTERNATIONAL® RSIC SPECIMEN
FIGURE4-20: WEALTL13-329STC48,LABFILLER
WALLSTC74 FILLER WALL
FIGURE4-21: RALTL05-167,STC58, COMPOSITE WALL
PARTITION STC 64 © PAC INTERNATIONAL®
Results from the PAC International® test specimen does not indicate the same resonance as the Class C4
specimens. It should be noted that the test specimens are not measured under the same laboratory
0
20
40
60
80
100
63 100 160 250 400 630 1000 1600 2500 4000
Transmission Loss (dB)
One-Third Octave Band Center Frequency (Hz)
Class C4 Mullions and Pac International RSIC-1® Mullion Transmission Loss
(19) TL13-327, STC 31
(20) TL13-328, STC 38
(21) TL13-329, STC 48
(22) TL13-330, STC 48
RSIC-1 , RAL-TL05-167, STC 58
109
conditions. The Phase 1 C4 tests are measured in the absence of a composite partition, and the PAC
International® includes an STC 64 wall partition.
110
4.2.11 PHASE 1 SUMMARY
The transmission loss of all Phase 1 mullions are plotted (FIGURE 4-22).
FIGURE4-22: TRANSMISSIONLOSS SPECTRA OF ALL PHASE 1 E90 LABORATORY TESTS
*UNTREATED MULLION BASE CASE, MC1
There is a trend of resonant frequencies between 400 – 630 Hz, common in all Phase 1 test
measurements (FIGURE 4-22). Mullion specimens MC 1 (mullion control 1, TL13-311, Figure 4-23) and MC
2 (mullion control 2, TL13-323, Figure 4-24) are identified as the lowest and highest performing mullions
to be applied to subsequent phases as a means of comparison.
0
10
20
30
40
50
60
70
80
90
100
63 80 100 125 160 200 250 315 400 500 630 800 1000 1250 1600 2000 2500 3150 4000 5000
Transmission Loss (dB)
1/3 Octave Band Center Frequency (Hz)
Test Phase 1: Mullion
(1) TL13-309, STC 37 (2) TL13-310, STC 34 *(3) TL13-311, STC 36 (4) TL13-312, STC 47
(5) TL13-313, STC 39 (6) TL13-314, STC 38 (7) TL13-315, STC 36 (8) TL13-316, STC 38
(9) TL13-317, STC 46 (10) TL13-318, STC 48 (12) TL13-320, STC 46 (13) TL13-321, STC 50
(14) TL13-322, STC 47 (15) TL13-323, STC 52 (16) TL13-324, STC 47 (17) TL13-325, STC 42
(18) TL13-326, STC 45 (19) TL13-327, STC 31 (20) TL13-328, STC 38 (21) TL13-329, STC 48
(22) TL13-330, STC 48
MC2
TL13-323
MC1
TL13-311
111
FIGURE4-23: MULLIONCONSTANT1(MC1) SHOWN IN PLAN (LEFT) AND SPECIMEN PHOTO (RIGHT)
FIGURE4-24: MULLIONCONSTANT2(MC2) SHOWN IN PLAN (LEFT) AND SPECIMEN PHOTO (RIGHT)
The lowest performing mullion in Phase 1 was the TL13-327 (STC 31) test specimen. This is not selected
as the mullion constant since it includes an atypical modification (aluminum tube overclad) with an
irregular spectrum. The standard deviation between Test #3, and #5 through #22 ranges from 0.4 to 6.6
(Table 4-13).
The table below provides the standard deviation between Test #3, and #5 through #22.
Phase 1 Standard deviation
Tests
3,
5 - 22
One-third Octave Band Center Frequency (Hz)
63 80 100 125 160 200 250 315 400 500 630 800 1k 1.25k 1.6k 2k 2.5k 3.15k 4k 5k
0.6 0.5 0.6 0.4 1.3 2.1 4.3 5.2 5.8 5.1 6.6 6.2 5.7 4.9 5.0 4.9 4.8 5.1 4.7 4.1
TABLE4-13: STANDARD DEVIATION BETWEEN ALL PHASE 1 TEST SPECIMENS
The dB improvements ranged from 6 to 15 dB above the TL13-311 mullion constant (Table 4-14).
112
Phase Mullion Specimen Description dB Improvement
Ph1-A
Infill: none
Overclad: none
(TL13-311, STC 36 is used to
compare other phases)
Ph1-B
Infill: Varied
Overclad: None
6 dB improvement at frequencies
above 800 Hz
Ph1-C1
Infill: Pea Gravel or MLV Pillows
Overclad: 1/8” Alum + 3/16” Mass Limp Vinyl
14 dB improvement at
frequencies above 250 Hz
Ph1-C2
Infill: Pea Gravel or MLV Pillows
Overclad: 5/8” Gypsum + 3/16” Mass Limp Vinyl
15 dB improvement at
frequencies above 250 Hz
Ph1-C3
Infill: Pea Gravel or MLV Pillows
Overclad: 5/8” Gypsum
11 dB improvement at
frequencies above 250 Hz
Ph1-C4
Infill: Pea Gravel or MLV Pillows
Overclad: Aluminum Tube with Resilient Connection
9 dB improvement at frequencies
above 250 Hz
TABLE4-14: PHASE1COMPARISONS WITH MC-1 (TL13-311)
Overall observations:
The greatest improvement influence with filled mullions occurred at mid to high frequencies.
Overclad generally outperformed filling the mullion cavity.
Modifying a mullion with both an overclad and mullion cavity infill is significantly effective to
improve a hollow and bare mullion.
113
4.3 PHASE 2A – COMPOSITE SEAL AND CONNECTION ELEMENTS (WITH MULLION)
This laboratory test phase measures the TL performance of the vertical mullion and a partition connection
or partition seal, therefore specimens in this phase consist of a composite of both elements. The selected
vertical mullions used in this phase are the MC-1 (TL13-311) mullion and the MC-2 (TL13-323) mullion.
The glazing element is not included in this phase of testing. Specific configuration descriptions are
provided in the following sections. A total of 25 laboratory measurements were conducted in this phase
(Figure 4-25).
Date July 17 – 20, 2013
Laboratory: WEAL
Specimen: Connection
Element with mullion , no
glazing
Dimensions: Connection
depth varies
Surface Area: varies, 2.81 ft
2
to 3.75 ft
2
Transmission: Horizontal
Procedure: ASTM E90-09
Total Tests: 25
FIGURE4-25: PHASE2A:PLAN DIAGRAM OF UVM ELEMENTS AND ASSOCIATED TEST PHASES
Mullion partition connections and seals often vary in material and width and are dependent on the in situ
condition. They are often used to seal the deflection gap required between partition and the curtain wall
to accommodate wind, seismic or thermal loads. These connection products typically consists of a
material capable of static deflections that accommodate these requirements, such as foam, rubber,
silicone, etc.
Some of the tests in this phase do not necessarily target sound flanking performance but potential
acoustic leaks instead, so that standard approaches seen in practice may be compared.
114
4.3.1 PHASE 2A TEST SPECIMEN DESCRIPTION
In this phase, the filler wall was modified for a two aperture sizes to accommodate various test specimens.
The narrow aperture dimension is 3.26 ft
2
(7-3/4” x 60-1/2”), 0.3 m
2
(Figure 4-26). The dimension allows
measurements with the vertical mullion and an edge seal condition of ½” to ¾” gaps.
FIGURE4-26: HEAD OF NARROW APERTURE WIDTH OF 7-3/4”
SHOWN WITH A MULLION AND ½” BACKER ROD (ALSO SHOWN IS 4” REMOVABLE SECTION OF THE FILLER WALL)
The wide aperture dimension is 3.89 ft
2
(9-1/4” width x 60-1/2” height), 0.36 m
2
(FIGURE 4-27).
This allowed mullions to be tested with a 2-½” deep connection element. However, the typical depth was
8-3/4” to provide ¼” compression at the silicone element between the mullion and filler wall. The ¼”
compression is per the Silicone Compression Seal® product specification.
FIGURE4-27: HEAD OF WIDE APERTURE WIDTH OF 8-1/4”
MULLION AND SILICONE CLOSURE SHOWN WITH PUTTY AT PERIMETER (4” SECTION IS REMOVED)
Narrow edge seal
condition
Wide edge seal condition
7-3/4”
9-1/4”
115
All mullions were inserted into the aperture first followed by the fitting of the vertical edge seal or
partition connector element. Similar to phase 1, the vertical mullion had no direct contact with the filler
wall. However the proposed resilient connection typically was compressed between the mullion and filler
wall.
The resilient connection materials and dimensions tested for the Phase 2A series (TABLE 4-17 and TABLE
4-18) including the following:
¼” – ¾” backer rod and wet seal
½” Armacell® foam
PCS-1 Silicone Compression Seal® product by Michael Rizza Company
TM
: (2) 10' strips, width 2"
min to 2-1/2" width
PCS-1 Silicone Compression Seal® product and modified with an overclad of aluminum or gypsum
board plates
Also tested in Phase 2A was the Mull-it-Over
TM
(see TABLE 4-19) product, which is not defined in this
research as a “connection element” but will be included in the evaluation for transmission loss
performance of products.
FIGURE4-28: IMAGE OF THE PCS-1 SILICONE
COMPRESSION SEAL® PROFILE;
UNCOMPRESSED PROFILE DIMENSION 2-
1/2” X 7/8”
FIGURE4-29: ISOMETRIC DRAWING OF THE PCS-1
SILICONE COMPRESSION SEAL®
INSTALLATION (©2011 BALCO USA, INC.)
(Note: Image shows the compression seal
terminating at glass, not the mullion.)
This product has been selected to simulate a resilient seal connection between a demising wall and
mullion. It should be noted that there are other means and providing this type of resilient in practice.
116
4.3.2 PHASE 2A TESTING CLASSIFICATIONS
Tested configurations are categorized and tabulated (TABLE 4-15 TO TABLE 4-19).
TESTED SPECIMENS Phase 2A Class A: 1/2” Vertical gap both sides of mullion
TL13-398, STC 42
TL13-399, STC 36
TL13-400, STC 42
TL13-401, STC 46
TL13-402, STC 49
TL13-404, STC 35
A 1/2" vertical edge gap is at all vertical edges. Vertical gaps in Phase 1 were
1/4" and used acoustic putty for sealing.
Class A configurations follow from Phase 1 (isolated mullion tests) to assess the
influence of the open area of the test aperture. The horizontal width of the filler
wall aperture was 7-1/4” during Phase 1. The width is 7-3/4” in this Phase 2A.
This allows a ½” gap on either side of the mullion instead of ¼’ when the mullion
specimen is centered in the aperture.
TABLE 4-15: PHASE 2A CLASS A TEST CONFIGURATION DESCRIPTION, PLAN DRAWING
TESTED SPECIMENS Phase 2A Class B: 1/2" - 3/4" Foam or Backer Rod Tests with Mullion
TL13-405, STC 44
TL13-406, STC 52
TL13-407, STC 34
TL13-408, STC 38
TL13-409, STC 49
TL13-410, STC 49
TL13-411, STC 49
A 3/4” vertical edge gap with a resilient connection and wet seal is at one side
of the mullion.
A 1/4” vertical edge gap is at opposite side and sealed with acoustic putty.
The width of these connections are the smallest tested in Phase 2A and are
considered the minimum allowable façade deflection in practice. The 3/4” gap
is filled with backer rod and caulking wet seal. The connection for Test TL13-
405 consists of ½” compressed Armacell® in lieu of EMSEAL.
TABLE 4-16: PHASE 2A CLASS B TEST CONFIGURATION DESCRIPTION, PLAN DRAWING
117
TESTED SPECIMENS Phase 2A Class C: 2-1/4" Silicone Product Tests with Heavy Mullion
TL13-412, 31/29
TL13-413, 4One-third7
TL13-414, 32/30
TL13-415, 36/31
TL13-416, 34/32
Mullion configuration from TL13-323 is connected to a 2-1/4” Rizza Silicone
partition enclosure product. A 1/4” gap is at the opposite vertical edge.
The intent of these tests is to assess the influence of the Michael Rizza Silicone
Partition Closure® product on the best performing mullion specimen, MC-2
(TL13-323).
TABLE4-17: PHASE2ACLASSCTESTCONFIGURATIONDESCRIPTION,PLANDRAWING
TESTED SPECIMENS Phase 2A Class D 2-1/4" Silicone Product Tests with Hollow/ Exposed mullion
TL13-417, STC 30
TL13-418, STC 35
TL13-419, STC 28
TL13-420, STC 31
TL13-421, STC 34
TL13-422, STC 22
Mullion configuration from TL13-311 is connected to a 2-1/4” Rizza Silicone
partition enclosure product. A 1/4” gap is at the opposite vertical edge.
The intent of these tests is to assess the influence of the Michael Rizza Silicone
Partition Closure on the MC1 (TL13-311) the hollow and exposed mullion.
TABLE 4-18: PHASE 2A CLASS D TEST CONFIGURATION DESCRIPTION, PLAN DRAWING
118
TESTED SPECIMENS Phase 2A Class E Product Test (Mull-it-Over
TM
)
TL13-423, STC 46 A single test using the Mull-It-Over product and hollow mullion was measured.
Chapter 2 contains further description of the Mull-it-Over
TM
product.
TABLE 4-19: PHASE 2A CLASS E TEST CONFIGURATION DESCRIPTION, PLAN DRAWING
Results in test configurations [A] through [D] tabulated below specifically call out the vertical edge
condition (VEC) for each test assembly.
In all cases the PCS-1 Silicone Compression Seal® is compressed at least ¼” in all installations between the
mullion and filler wall.
119
4.3.3 PHASE 2A CLASS A TEST SEQUENCE
Results from the Class A testing sequence are summarized (Table 4-20).
WEAL Test
No.
STC Material Layers [description] Element Drawing [Plan]
(25)TL13-398 42
[1] 1-1/2” (16 mm) MLV pillow+ alum tube
[2] 1/4” (6mm) airspace MLV buttons
[3] 1/8” (3mm) aluminum mullion
[4] 2-3/4” (70mm) air space
[5] 1/8” (3mm) aluminum mullion
[6] 1/4” (6mm) airspace MLV buttons
[7] 1-1/2” (16 mm) MLV pillow+ alum tube
[7/8” edge gap sealed with putty]
[TL13-329, 48 / 42]
(26) TL13-399 36
[1] 1/8” (3mm) aluminum mullion
[2] 2-3/4” (70mm) air space
[3] 1/8” (3mm) aluminum mullion
[1/2” edge gap sealed with putty]
[TL13-311, 36 / 33]
(27) TL13-402 49
[1] 5/8” (16 mm) gypsum board plate
[2] 3/16” (5 mm) MLV layer
[3] 1/8” (3mm) aluminum mullion
[4] 2-3/4” (70mm) MLV pillows
[5] 1/8” (3mm) aluminum mullion
[6] 3/16” (5 mm) MLV layer
[7] 5/8” (16 mm) gypsum board plate
[1/2” edge gap sealed with putty]
[TL13-323, 52 / 43]
(28) TL13-404 35
[1] 1/8” (3mm) aluminum mullion
[2] 2-3/4” (70mm) air space
[3] 1/8” (3mm) aluminum mullion
[1/2” edge gap sealed with putty]
[TL13-315, 36 / 34]
TABLE4-20: PHASE2ACLASSA,STCRESULTS AND SPECIMEN DESCRIPTION
120
4.3.3.1 PHASE 2A-A DEDUCTIONS AND OBSERVATIONS
The following Class A observations are compared with similar mullion composition in Phase 1:
TL13-398 (STC 42) compared to Phase 1 TL13-329 (STC 48)
The test performance in Phase 2a is significantly lower than the previous Phase 1 and not as tonal
however shares a similar coincidence dip and overall reduction.
TL13-399 (STC 36) compared to Phase 1 TL13-311 (STC 36)
The test in this phase has a similar performance as the previous Phase 1 but has a reduced TL at low
frequencies. This may be attributed to the increased perimeter gap, i.e. from ¼” to ½” on each side.
TL13-402 (STC 49) is compared to Phase 1 TL13-323 (STC 52)
The test performance is less than the test in Phase 1. This may be due to the increased perimeter gap.
TL13-404 (STC 35) is compared to TL13-315 (STC 36)
The test performance is almost identical. This indicates that ram-packing mineral wool versus laying
in the fill material does not make a significant difference.
In general, all the retested mullions performed lower than the similar Phase 1 mullions. The air slot created
between the mullion and the chamber filler wall may be adversely influencing the lower TL performance.
The gap dimension is typically ½” wide x 60-1/2” tall x 3” deep.
TL13-398 TL13-399 TL13-402 TL13-404
FIGURE4-30: PHASE2A-A,TL OF MULLIONS WITH ½” VERTICAL SEAL
0
10
20
30
40
50
60
70
80
90
100
Transmission Loss (dB)
One-third Octave Band Center Frequency (Hz)
Phase 2A Class A Transmission Loss
Mullions with 1/2" seals
(25) TL13-398, STC 42
(26) TL13-399, STC 36
(27) TL13-402, STC 49
(28) TL13-404, STC 35
121
4.3.4 PHASE 2A CLASS B TEST SEQUENCE
Results from the Class B testing sequence are summarized (Table 4-21).
WEAL Test
No.
STC Material Layers [description] Element Drawing [Plan]
(29) TL13-405 44
[1] 5/8” (16 mm) gypsum board plate
[2] 3/16” (5 mm) MLV layer
[3] 1/8” (3mm) aluminum mullion
[4] 2-3/4” (70mm) MLV pillows
[5] 1/8” (3mm) aluminum mullion
[6] 3/16” (5 mm) MLV layer
[7] 5/8” (16 mm) gypsum board plate
VEC: ½” Armacell ®
[3 edges 1/4" putty, 1/4" wood shim for compression on 1/2"
Armacell®]
[TL13-323, 52 / 43]
(30) TL13-406 52
[1] 5/8” (16 mm) gypsum board plate
[2] 3/16” (5 mm) MLV layer
[3] 1/8” (3mm) aluminum mullion
[4] 2-3/4” (70mm) MLV pillows
[5] 1/8” (3mm) aluminum mullion
[6] 3/16” (5 mm) MLV layer
[7] 5/8” (16 mm) gypsum board plate
VEC: 3/4” foam backer rod +wet seal
[3 edges 1/4" putty, 1/4" neoprene shim]
[TL13-323, 52 / 43]
(31) TL13-407 34
[1] 1/8” (3mm) aluminum mullion
[2] 2-3/4” (70mm) air space
[3] 1/8” (3mm) aluminum mullion
VEC: 3/4” foam backer rod +wet seal
[3 edges 1/4" wet seal, 1/4" neoprene shim]
[TL13-311, 36 / 33; TL13-399, 36 / 33]
(32) TL13-408 38
[1] 1/8” (3mm) aluminum mullion
[2] 2-3/4” (70mm) air space
[3] 1/8” (3mm) aluminum mullion
VEC: 3/4” foam backer rod +wet seal
[3 edges wet seal + masking tape + putty, neoprene shim]
[TL13-311, 36 / 33; TL13-399, 36 / 33; TL13-407, 34/32]
122
WEAL Test
No.
STC Material Layers [description] Element Drawing [Plan]
(33) TL13-411 49
[1] 1-1/2” (16 mm) MLV pillow+ alum tube
[2] 1/4” (6mm) airspace MLV buttons
[3] 1/8” (3mm) aluminum mullion
[4] 2-3/4” (70mm) air space
[5] 1/8” (3mm) aluminum mullion
[6] 1/4” (6mm) airspace MLV buttons
[7] 1-1/2” (16 mm) MLV pillow+ alum tube
VEC: 3/4” foam backer rod +wet seal
[3 edges 1/4" wet seal, 1/4" neoprene shim]
[TL13-329/30, 48 / 42; TL13-398, 42 / 37]
TABLE4-21: PHASE2ACLASSB,STCRESULTS AND SPECIMEN DESCRIPTION
4.3.4.1 PHASE 2A-B DEDUCTIONS AND OBSERVATIONS
The following Class B observations are compared with similar mullion composition in Phase 1:
TL13-406 has the same rating as the same mullion tested in Phase 1, TL13-323, despite the wider edge
condition. This may imply opportunities for design and/or savings.
Putty did not perform as well at wet seal for the hollow aluminum tube overclad tests
TL13-405 (STC 44) compared to Phase 1 TL13-323 (STC 52) and Phase 2A TL13-402 (STC 49):
The test performance of TL13-405 is significantly lower than the previous 323 and 402. The Armacell®
foam material is porous and may indicate the acoustic leak occurring above 500 Hz.
TL13-406 (STC 52) compared to Phase 1 TL13-323 (STC 52) and Phase 2A TL13-402 (STC 49):
The test performances between conditions are similar – the seal condition of TL13-406 includes a ¾”
gap at one edge with wet seal + backer rod, whereas the seal condition for TL13-323 and TL14-402
were applied with a dense putty on a narrower edge gap.
This may indicate that a “connection” gap of ¾” for lateral facades deflection may not adversely
influence the overall Transmission Loss performance.
TL13-407 (STC 34) compared to TL13-399 (STC 36) and Phase 1 TL13-311 (STC 36):
Generally, the performance is similar. The greater gap of ¾” in test 407 dips at 400Hz.
TL13-408 (STC 38) compared to -399, -311, -407
This test is identical to TL13– 407 with the exception of adding masking tape and putty to 3 sides of
the mullion to isolate the ¾” wet seal edge. There is a noted improvement where the entire curve
shifts up 4dB. The improvement may be attributed to the improved edge seal condition further
reducing sound leaks and the increased damping from the putty impeding additional vibration.
TL13-411 (STC 49) is compared to TL13-330 (STC 48), TL 13-398 (STC 42)
The mullion in Phase 1 test -330 was filled with MLV pillows and had a ¼” air space on either side. The
configuration in test -411 used had a 1-1/4” west seal edge on one side and ½” edge seal on the other.
123
Test TL13-398 performed significantly poorer than TL13-411. The main difference between the two
setups was the edge seal condition. Putty covered the 7/8” edge gap in test TL13-398 and a wet seal
caulk was used in test TL13-411. This seems to imply there is no significant difference between using
wet seal versus putty and that the wider edge condition (i.e. 1-1/4” versus ¼”) does not adversely
impact the overall performance as long as the seal is airtight. Further there are diminishing returns
with adding mass to the mullion in the form of fill and overclad material. In this test, there was no
significant difference between a filled mullion and non-filled mullion. The hollow tube overclad filled
with a damping layer was identical for both tests.
No significant difference between putty and wet seal.
The air slot created on either side of the mullions and the chamber filler wall varies: one side is
generally ¼” wide x 60-1/2” tall x 3” deep and the other side is general ¾” wide and filled with to strips
of backer rod and wet sealed. These gaps are not adversely influencing the performance of the
mullion and are performing similar to the Phase 1 tests.
Transmission loss plots TL13-408 -399, -311, -407 should be compared in more detail for trends based
on the different sealed conditions. This analysis is not included as part of the research study.
Possible resonant frequency for mullion is at 400Hz. Mullion resonant frequency should be tested.
TL13-405 (STC 44) TL13-406 (STC 52) TL13-407 (STC 34) TL13-408 (STC 38) TL13-411 (STC 49)
FIGURE4-31: PHASE2A-B,TL OF MULLIONS WITH 1/2”AND 3/4” SEALS
0
10
20
30
40
50
60
70
80
90
100
63
80
100
125
160
200
250
315
400
500
630
800
1000
1250
1600
2000
2500
3150
4000
5000
Transmission Loss (dB)
One-third Octave Band Center Frequency (Hz)
Phase 2 Class B Trasnmission Loss
1/2" foam and 3/4" backer rod + caulking
(29) TL13-405, STC 44
(30) TL13-406, STC 52
(31) TL13-407, STC 34
(32) TL13-408, STC 38
(33) TL13-411, STC 49
124
4.3.5 PHASE 2A CLASS C TEST SEQUENCE
Results from the Class C testing sequence are summarized (Table 4-22):
WEAL Test
No.
STC Material Layers [description] Element Drawing [Plan]
(34) TL13-412 31
[1] 5/8” (16 mm) gypsum board plate
[2] 3/16” (5 mm) MLV layer
[3] 1/8” (3mm) aluminum mullion
[4] 2-3/4” (70mm) MLV pillows
[5] 1/8” (3mm) aluminum mullion
[6] 3/16” (5 mm) MLV layer
[7] 5/8” (16 mm) gypsum board plate
VEC: (1) 2-1/2” Silicone Partition Closure ®
at source side
[wet seal top and bottom of silicone, 3 edges 1/4" putty, 1/4"
neoprene shim]
(35) TL13-413 41
[1] 5/8” (16 mm) gypsum board plate
[2] 3/16” (5 mm) MLV layer
[3] 1/8” (3mm) aluminum mullion
[4] 2-3/4” (70mm) MLV pillows
[5] 1/8” (3mm) aluminum mullion
[6] 3/16” (5 mm) MLV layer
[7] 5/8” (16 mm) gypsum board plate
VEC: (2) 2-1/2” Silicone Partition Closure ®
at both source and receiver sides
[wet seal top and bottom of silicone, 3 edges 1/4" putty, 1/4"
neoprene shim]
(36) TL13-414 32
[1] 5/8” (16 mm) gypsum board plate
[2] 3/16” (5 mm) MLV layer
[3] 1/8” (3mm) aluminum mullion
[4] 2-3/4” (70mm) MLV pillows
[5] 1/8” (3mm) aluminum mullion
[6] 3/16” (5 mm) MLV layer
[7] 5/8” (16 mm) gypsum board plate
VEC: (1) 2-1/2” Silicone Partition Closure ®
at receiver side
[wet seal top and bottom of silicone, 3 edges 1/4" putty, 1/4"
neoprene shim]
125
WEAL Test
No.
STC Material Layers [description] Element Drawing [Plan]
(37) TL13-415 36
[1] 5/8” (16 mm) gypsum board plate
[2] 3/16” (5 mm) MLV layer
[3] 1/8” (3mm) aluminum mullion
[4] 2-3/4” (70mm) MLV pillows
[5] 1/8” (3mm) aluminum mullion
[6] 3/16” (5 mm) MLV layer
[7] 5/8” (16 mm) gypsum board plate
VEC: (1) 2-1/2” Silicone Partition Closure ®
at source side + 1/8” aluminum overclad
[wet seal top and bottom of silicone, 3 edges 1/4" putty, 1/4"
neoprene shim, aluminum plate is wet sealed at vertical filler
wall edge and adhered to mullion with masking tape]
(38) TL13-416 34
[1] 5/8” (16 mm) gypsum board plate
[2] 3/16” (5 mm) MLV layer
[3] 1/8” (3mm) aluminum mullion
[4] 2-3/4” (70mm) MLV pillows
[5] 1/8” (3mm) aluminum mullion
[6] 3/16” (5 mm) MLV layer
[7] 5/8” (16 mm) gypsum board plate
VEC: (2) 2-1/2” Silicone Partition Closure ®
at both sides + 1/8” aluminum overclad
[wet seal top and bottom of silicone, 3 edges 1/4" putty, 1/4"
neoprene shim, aluminum plate is wet sealed at vertical filler
wall edge and adhered to mullion with masking tape]
TABLE4-22: PHASE2ACLASSC,STCRESULTS AND SPECIMEN DESCRIPTION
*VEC (vertical edge condition)
4.3.5.1 PHASE 2A-C OBSERVATIONS
The following are based on the results from Phase 2A Class C test specimens:
TL13-412 (STC 31) – one silicone connection, obvious tone 2 kHz, leaks at the top of the silicone
connection.
TL13-413 (STC 41) – two silicone connections; tone is not audible.
TL13-414 (STC 32) – subjectively heard the same tone at 2 kHz, similar to TL13 -412. No difference in
silicone placement at source or receiver side.
TL13-415 (STC 36) – one silicone connection covered by a 1/8” metal plate; significant improvement
at the mid to high frequencies (no 2 kHz tone), but reduced performance below 315 Hz.
TL13-416 (STC 34) - same configuration as TL13-415 but with 2 silicone connections; improvement
over entire spectrum at low and high frequencies with the exception of 400Hz. STC is irrelevant here since
discriminating at 400Hz.
126
The transmission loss of MC-2 is plotted against versions of the composite MC-2 with a silicone connection
(FIGURE 4-32).
FIGURE4-32: PHASE2A-C,TL OF HIGHER PERFORMING MULLION WITH SILICONE PARTITION CLOSURE®
*TREATED MULLION BASE CASE, MC2
The following two specimens are set up identically in the filler wall with the exception of 1/8” aluminum
plates enclosing the silicone partition connection at TL13-416 (FIGURE 4-33):
TL13-413 STC 41 (No overclad)
TL13-416 STC 34 (Added 1/8” aluminum overclad)
TL13-413, STC 41
TL13-416, STC 34
FIGURE4-33: PLANDRAWING OF TL13-413 (LEFT) AND TL 416 (RIGHT)
0
10
20
30
40
50
60
70
80
90
100
63
80
100
125
160
200
250
315
400
500
630
800
1000
1250
1600
2000
2500
3150
4000
5000
Transmission Loss (dB)
One-third Octave Band Center Frequency (Hz)
Phase 2 Class C Trasnmission Loss
High Performing Mullion (MC-2) with 2-1/4" Silicone Connection
(34) TL13-412, STC 31
(35) TL13-413, STC 41
(36) TL13-414, STC 32
(37) TL13-415, STC 36
(38) TL13-416, STC 34
*(15) TL13-323, STC 52
127
The TL13-416 configuration performs 7dB lower than the TL13-413 specimen without the aluminum
enclosure. This is due to a resonance created by the aluminum plate at 400 Hz (FIGURE 4-34).
FIGURE4-34: SOUNDTRANSMISSIONLOSS CURVES OF TL13-413 AND TL13-416
0
10
20
30
40
50
60
70
80
90
100
Transmission Loss (dB)
One-third Octave Band Center Frequency (Hz)
Sound Transmission Loss
TL13-413 and TL13-416
TL13-413
TL13-416
128
4.3.6 PHASE 2A CLASS D TEST SEQUENCE
Results from the Class C testing sequence are summarized (Table 4-31).
WEAL Test
No.
STC Material Layers [description] Element Drawing [Plan]
(39) TL13-417 30
[1] 1/8” (3mm) aluminum mullion
[2] 2-3/4” (70mm) air space
[3] 1/8” (3mm) aluminum mullion
VEC: (1) 2-1/2” Silicone Partition Closure ®
at source side
[wet seal top and bottom of silicone, 3 edges 1/4" putty, 1/4"
neoprene shim]
(40) TL13-418 35
[1] 1/8” (3mm) aluminum mullion
[2] 2-3/4” (70mm) air space
[3] 1/8” (3mm) aluminum mullion
VEC: (2) 2-1/2” Silicone Partition Closure ®
at both source and receiver sides
[wet seal top and bottom of silicone, 3 edges 1/4" putty, 1/4"
neoprene shim]
(41) TL13-419 28
Data not used, incorrect surface area TL13-420
(42) TL13-420 31
[1] 1/8” (3mm) aluminum mullion
[2] 2-3/4” (70mm) air space
[3] 1/8” (3mm) aluminum mullion
VEC: (2) 2-1/2” Silicone Partition Closure ®
at both sides + 1/8” aluminum overclad
[wet seal top and bottom of silicone, 3 edges 1/4" putty, 1/4"
neoprene shim; wet seal at vertical edge of aluminum plate
and filler wall; adhered plate to mullion with masking tape]
(43) TL13-421 34
[1] 1/8” (3mm) aluminum mullion
[2] 2-3/4” (70mm) air space
[3] 1/8” (3mm) aluminum mullion
VEC: (1) 2-1/2” Silicone Partition Closure ®
at source side + 1/8” aluminum overclad
[wet seal top and bottom of silicone, 3 edges 1/4" putty, 1/4"
neoprene shim, wet seal at vertical edge of aluminum plate
and filler wall; adhered plate to mullion with masking tape]
129
WEAL Test
No.
STC Material Layers [description] Element Drawing [Plan]
(44) TL13-422 22
[1] 1/8” (3mm) aluminum mullion
[2] 2-3/4” (70mm) air space
[3] 1/8” (3mm) aluminum mullion
VEC: 1/8” aluminum plate overclad
[3 edges 1/4" putty, 1/4" neoprene shim; wet seal at vertical
edge of aluminum plate and filler wall; adhered plate to
mullion with masking tape]
TABLE4-23: PHASE2ACLASSD,STCRESULTS AND SPECIMEN DESCRIPTION
4.3.6.1 PHASE 2A-D OBSERVATIONS
The following are based on the results from Phase 2A Class D test specimens:
TL13-417 (STC 30) – performs lower than TL13-311 and performs similarly to TL13-412
10dB improvement between TL13-412 and TL13-413
5dB improvement between TL13-417 and TL13-418
The TL of MC-1 is plotted against versions of the composite MC-1 with a silicone connection (FIGURE 4-35).
FIGURE4-35: TL OF HOLLOW/EXPOSED MULLION CONNECTED TO RIZZA SILICONE PARTITION CLOSURE®
*UNTREATED MULLION BASE CASE, MC1
0
10
20
30
40
50
60
70
80
90
100
63
80
100
125
160
200
250
315
400
500
630
800
1000
1250
1600
2000
2500
3150
4000
5000
Transmission Loss (dB)
One-third Octave Band Center Frequency (Hz)
Phase 2 Class D Trasnmission Loss
Low Performing Mullion (MC-1) with 2-1/4" Silicone Connection
(39) TL13-417, STC 30
(40) TL13-418, STC 35
(42) TL13-420, STC 31
(43) TL13-421, STC 34
(44) TL13-422, STC 22
*(3) TL13-311, STC 36
130
The following two specimens are set up identically in the filler wall with the exception of 1/8” aluminum
plates enclosing the silicone partition connection at TL13-420 (FIGURE 4-36):
TL13-418 STC 35 (No overclad)
TL13-420 STC 31 (Added 1/8” aluminum overclad)
TL13-418, STC 35
TL13-420, STC 31
FIGURE4-36: PLANDRAWING OF TL13-418 (LEFT) AND TL 420 (RIGHT)
The TL13-420 configuration performs 4dB lower than the TL13-418 specimen without the aluminum
enclosure. This is due to a resonance at 500 Hz attributed to the aluminum plates (FIGURE 4-37). This is not
an outstanding difference; however it indicates that a wider partition connection has less of an influence
on a low performing mullion.
FIGURE4-37: SOUNDTRANSMISSIONLOSS CURVES OF TL13-418 AND TL13-420
0
10
20
30
40
50
60
70
80
90
100
Transmission Loss (dB)
One-third Octave Band Center Frequency (Hz)
Sound Transmission Loss
TL13-418 and TL13-420
TL13-418
TL13-420
131
4.3.7 PHASE 2A CLASS E TEST SEQUENCE
The Class E sequence was limited to one test shown (TABLE 4-24).
WEAL Test
No.
STC Material Layers [description] Element Drawing [Plan]
(45) TL13-423 46
[1] Mull-It-Over ® (alum+vinyl+foam)
[2] 2-1/4” (57mm) air space
[3] 1/8” (3mm) aluminum mullion
[4] 2-3/4” (70mm) air space
[5] 1/8” (3mm) aluminum mullion
[6] 1-1/2” (38mm) air space
[7] Mull-It-Over ® (alum+vinyl+foam)
[overall width 8-1/4” (210mm) ]
[mullion: putty on 3 sides, 1/2" backer rod +wet seal on one
side; mull-it-over®: putty on top and bottom, wet seal on
screws, compression seal at the ¾” vertical continuous wood
spacer]
TABLE 4-24: PHASE 2A CLASS E, STC RESULTS, AND SPECIMEN DESCRIPTION
132
4.3.8 PHASE 2A CONCLUSION
The mullions tested in Phase 2A contained vertical edge conditions (VEC) including either a narrow seal or
wide connection (FIGURE 4-38).
FIGURE4-38: TLSPECTRA OF ALL PHASE 2A TESTS
There are no obvious trends in this test series as the specimen modifications varied significantly.
0
10
20
30
40
50
60
70
80
90
100
63
80
100
125
160
200
250
315
400
500
630
800
1000
1250
1600
2000
2500
3150
4000
5000
Transmission Loss (dB)
One-third Octave Band Center Frequency (Hz)
Phase 2A All Tests
(25) TL13-398, 42/37
(26) TL13-399, 36/33
(27) TL13-402, 49/41
(28) TL13-404, 35/32
(29) TL13-405, 44/38
(30) TL13-406, 52/41
(31) TL13-407, 34/32
(32) TL13-408, 38/35
(33) TL13-411, 49/40
(34) TL13-412, 31/29
(35) TL13-413, 41/37
(36) TL13-414, 32/30
(37) TL13-415, 36/31
(38) TL13-416, 34/32
(39) TL13-417, 30/28
(40) TL13-418, 35/32
(42) TL13-420, 31/29
(43) TL13-421, 34/30
(44) TL13-422, 22/22
(45) TL13-423, 46/41
133
The lowest performing test in Phase 2A is TL13-422 (STC 22). The highest performing is TL13-406 (STC 52),
which is nearly identical to TL13-323 (MC-2) in Phase 1. The lowest performing with a silicone closure
connection is TL13-417 (STC 30). The highest performing with a silicone closure connection is TL 413 (STC
41).
TL13-413, STC 41
TL13-417, STC 30 TL13-422, STC 22
FIGURE4-39: PLANDRAWING OF TL13-413 (LEFT), TL13-417 (MIDDLE) AND TL 422 (RIGHT)
Table 4-25 summarizes the single figure STC range of the mullions with narrow or vertical edge condition
(VEC).
PHASE 2A
NARROW VEC (SEAL) WIDE VEC (CONNECTION)
Range of STC results Range of STC results
Mullion Constant 1
(TL13-311 STC36)
34 - 38 22 - 35
Mullion Constant 2
(TL13-323 STC 52)
49 – 52 31 – 41
TABLE4-25: RESULT SUMMARY OF STC RANGE FOR SMALL AND LARGE VEC
*The Armacell® foam connection test with a heavy mullion is STC 44 and was not included in the summary table
above since it was only measured with MC-2 and not MC-1.
As would be expected, a narrow connection does not significantly reduce the performance of the
respective mullion constants (e.g. TL13-311 and TL13-323) tested in Phase 1. Whereas a wider connection
significantly impacts the mullion performance as seen with specimen TL13-413, STC 41. This is an 11dB
STC reduction from the mullion constant TL13-323, STC 52 tested in Phase 1.
This confirms the importance of potential acoustic leaks that can occur in practice and provides a standard
for comparison against the mullions in Phase 1.
4.3.8.1 TL SUMMARY GRAPHS
The following transmission loss tables (FIGURE 4-40 AND Figure 4-44) are shown below for archival
purposes. They include the transmission loss test results in this phase with overlays of the mullion
constants as means for comparison.
134
MC-1 with a Narrow Vertical Edge Seal
The measurement condition in this configuration considers a representative deflection less than ¾” in one
axis with Mullion Constant 1 (non-modified unitized mullion). Standard deviation of the transmission loss
curves shown indicate that a small edge condition (< 3/4") has a greater influence on a lighter mullion
(FIGURE 4-40).
FIGURE4-40: PHASE2AMC-1 COMPARED TO A COMPOSITE OF MC-1 + SMALL DEFLECTION CONNECTION
*UNTREATED MULLION BASE CASE, MC1
MC-2 with a Narrow Vertical Edge Seal
The measurement condition in this configuration considers a representative deflection less than ¾” in one
axis with Mullion Constant 2 (modified unitized mullion, high performing). Standard deviation of the
transmission loss curves shown indicate that a small edge condition ( < 3/4") has less influence on a
heavier mullion (FIGURE 4-41).
FIGURE4-41: PHASE2AMC-2 COMPARED TO A COMPOSITE OF MC-2 + SMALL DEFLECTION CONNECTION
0
20
40
60
80
100
63 80 100 125 160 200 250 315 400 500 630 800 10001250160020002500315040005000
Transmission Loss (dB)
One-third Octave Band Center Frequency (Hz)
Ph2a-A/B: 1/2" - 3/4" Foam or Backer Rod Tests
(Edge Condition using Mullion Constant 1 )
Ph2a-B STDEVA w/ TL13-311
(26) TL13-399, 36
(31) TL13-407, 34
(32) TL13-408, 38
MC1 (3) TL13-311, 36
0
20
40
60
80
100
63 80 100 125 160 200 250 315 400 500 630 800 10001250160020002500315040005000
Transmission Loss (dB)
One-third Octave Band Center Frequency (Hz)
Ph2a-A/B: 1/2" - 3/4" Foam or Backer Rod Tests
(Edge Condition using Mullion Constant 2)
Ph2a-B STDEVA w/ TL13-323
(30) TL13-406, 52
(27) TL13-402, 49
MC2 (15) TL13-323, 52
135
MC-1 with a Wide Vertical Edge Seal
The measurement condition in this configuration considers a representative deflection greater than ¾” in
one axis with Mullion Constant 1. Standard deviation of the transmission loss curves indicate that a large
edge condition (1" - 2") has a smaller influence on a lighter mullion (FIGURE 4-42).
FIGURE4-42: PHASE2AMC-1 COMPARED TO A COMPOSITE OF MC-1 + LARGE DEFLECTION CONNECTION
*UNTREATED MULLION BASE CASE, MC1
MC-2 with a Wide Vertical Edge Seal
The measurement condition in this configuration considers a representative deflection greater than ¾” in
one axis with Mullion Constant 2 (modified unitized mullion, high performing). Standard deviation of the
transmission loss curves shown indicate that a large edge condition (1" - 2") has a greater influence on a
heavier mullion (FIGURE 4-43).
FIGURE4-43: PHASE2AMC-2 COMPARED TO A COMPOSITE OF MC-2 + LARGE DEFLECTION CONNECTION
0
20
40
60
80
100
63 80 100 125 160 200 250 315 400 500 630 800 10001250160020002500315040005000
Transmission Loss (dB)
One-third Octave Band Center Frequency (Hz)
Ph2a-D: 2-1/4" Silicone Tests with Mullion Constant 1 (Light Mullion)
*Ph2a-D STDEVA w/ TL13-311
*Ph2a-D STDEVA only
(40) TL13-418, 35
(42) TL13-420, 31
(43) TL13-421, 34
MC1 (3) TL13-311, 36
0
20
40
60
80
100
63 80 100 125 160 200 250 315 400 500 630 800 10001250160020002500315040005000
Transmission Loss (dB)
One-third Octave Band Center Frequency (Hz)
Ph2a-C: 2-1/4" Silicone Tests with Mullion Constant 2 (Heavy Mullion)
*Ph2a-C STDEVA w/ TL13-323
*Ph2a-C STDEVA only
(35) TL13-413, 41
(37) TL13-415, 36
(38) TL13-416, 34
MC2 (15) TL13-323, 52
136
4.4 PHASE 2B –CONNECTION ELEMENT (WITHOUT MULLION)
The following laboratory test phase measures the TL performance of acoustic concept connections used
between a demising partition and a mullion. All partition connections are measured in the absence of the
unitized vertical mullion and glazing elements. The intent of these measurements is to assess the influence
of varying mass, airspace cavity and acoustic seal conditions. The TL results will be applied to composite
TL predictions to support objective 4. Detail configuration descriptions are provided in the following
sections. A total of 29 laboratory measurements were conducted in this phase (Figure 4-44).
Date October 8 – 11, 2013
Laboratory: WEAL
Specimen: Connection
Element (no mullion)
Dimensions: Connection
depth varies
Surface Area: 2.73 ft
2
(0.25m
2
), varies for product
tests.
Transmission: Horizontal
Procedure: ASTM E90-09
Total Tests: 29
FIGURE4-44: PHASE2B:PLAN DIAGRAM OF UVM ELEMENTS AND ASSOCIATED TEST PHASES
The acoustic concepts connections designed for this test sequence consider either a parallel plate or
staggered plate configuration (Figure 4-45). These diagram concepts were developed subsequent to
discussions with curtain wall designers and architects.
137
FIGURE4-45: PLAN DIAGRAM OF ACOUSTIC CONCEPT CONNECTIONS MEASURED IN PHASE 2B
The acoustic concept details are representative of building materials used in practice to either conceal
gaps between a partition and a mullion (i.e. parallel plates) or to create a configuration that enables façade
deflections (i.e. staggered plates) ( Figure 4-45 ).
The staggered plates are specifically indicative of resilient male-female connections which occur in
practice to satisfy deflection conditions in at least 2-axis: Horizontal deflections normal to the façade (See
A in Figure 4-46) and minimal off axis deflections accommodated by the ¼” airspace between plates (See
B in Figure 4-46).
FIGURE4-46: PLAN DIAGRAM OF STAGGERED PLATE SPECIMEN
PARTITION SIDE
MULLION SIDE
A
B
138
4.4.1 PHASE 2B TEST SPECIMEN DESCRIPTION
The primary connection elements used during this test phase were placed into a 6-1/2” x 60-1/2” aperture
in the chamber filler wall. Figure 4-47 shows an elevation view of the test specimen inserted into the filler
wall.
FIGURE4-47: ELEVATION VIEW OF PHASE 2B TEST SPECIMENS IN FILLER WALL (LEFT – PARALLEL, RIGHT – STAGGERED)
Both test specimens in the figures above show configurations with 1/8” thick aluminum plates (60” x 6”)
materials. Other materials and seal elements used in this test phase are listed below and are shown in
Figure 4-48 and Figure 4-49.
1/8” aluminum plates – 60” x 6” and 60”x4”
5/8” gypsum wall board – 60” x 6”, 60”x4.5” and 60”x3.5”
3/16” Mass Loaded Vinyl (MLV) – 60” x 6”
Mullion Mate Product: (60”height, width 2-7/8" minimum to 3-15/16" maximum opening)
Mull-it-Over Product: (2) 5' height, 4-1/2" wide
Neoprene bulb seals
¾” x 1” wood furring strips
Bulb seals were used on the plate conditions to simulate a resilient seal to the glass or mullion. The furring
strips were used, especially with the staggered plate condition to structurally attach the strips to the filler
wall with neoprene pads. As with all UVM test phases, the horizontal perimeter gap is maintained at ¼”
around all test elements so that no mechanical connection occurs between the specimen and the chamber
filler wall. The dimensions of the filler wall aperture were modified for select mullion products tested at
the end of the phase.
Laboratory
filler wall
Test
specimens
139
“Solid Plate” specimen construction
FIGURE4-48: PHASE2BCLASSA SPECIMEN CONSTRUCTION OF SOLID PLATE TESTS
140
“Staggered Plate” specimen construction
FIGURE4-49: PHASE2BCLASSB SPECIMEN CONSTRUCTION OF SOLID PLATE TESTS
141
4.4.3 PHASE 2B TESTING CLASSIFICATIONS
Tested configurations are categorized and tabulated (TABLE 4-26 TO TABLE 4-28). All assemblies shown are
drawn in plan.
TESTED SPECIMENS Phase 2B Class A: SOLID PLATE
TL13-605 through
TL13-619
The parallel plate specimens tested were placed on either side of a 6”, 4” or 3”
airspace. The plates used in this series consisted of 1/8” aluminum and 5/8”
gypsum board. Variations included a resilient layer of 3/16” mass loaded vinyl
and/or batt insulation in the cavity. Each plate was cut to dimensions of 6”x60”.
TABLE 4-26: PHASE 2B CLASS A SOLID PLATE DESCRIPTION AND PLAN DRAWING
TESTED SPECIMENS Phase 2B Class B: STAGGERED PLATE
TL13-620 through
TL13-629
The set up and variations to the staggered plate test specimens were similar to
the solid plate series. The exception is that the plate materials were broken to
allow a 2” overlap between plates offset by ¼” airspace.
TABLE 4-27: PHASE 2B CLASS A STAGGERED PLATE DESCRIPTION AND PLAN DRAWING
142
TESTED SPECIMENS Phase 2B Class C: Product Tests (Mullion Mate ® and Mull-it-over®)
TL13-630 through
TL13-633
Pre-manufactured products were tested for Transmission Loss performance.
These tests were measured in isolation and as a composite with the unitized
vertical curtain wall mullion.
The Mullion Mate® product is tested with and without a mullion.
The Mull-it-over® product is tested in Phase 2A with a mullion, the product is
measured again in Phase 2B without the mullion.
TABLE 4-28: PHASE 2B CLASS C PRODUCT TEST DESCRIPTION AND PLAN DRAWING
The majority of test configurations measured in Phase 2B assume a 3” air cavity so their performance may
be directly compared with the mullions tested in Phase 1 and Phase 2a. Parallel plate configurations are
measured with 6”, 4” and 3” airspaces to be used for different means of comparison.
The staggered plate configurations are measured to assess the influence of a labyrinthine sound path from
one side to another.
143
4.4.4 PHASE 2B CLASS A TEST SEQUENCE
Results from the Class A testing sequence are summarized (Table 4-29).
WEAL
Test No.
STC Material Layers [description] Element Drawing [Plan]
TL13-607 51
[1] 1/8” (3mm) aluminum plate
[2] 6” (150mm) air space
[3] 1/8” (3mm) aluminum plate
[¼” Bulb Seal at assumed vertical mullion edge, the other
3 edges are sealed with putty]
TL13-608 51
[1] 1/8” (3mm) aluminum plate
[2] 6” (150mm) batt insulation
[3] 1/8” (3mm) aluminum plate
[¼” Bulb Seal at assumed vertical mullion edge, the other
3 edges are sealed with putty]
TL13-610 47
[1] 1/8” (3mm) aluminum plate
[2] 4” (100mm) air space
[3] 1/8” (3mm) aluminum plate
[¼” Bulb Seal at assumed vertical mullion edge, the other
3 edges are sealed with putty]
TL13-611 49
[1] 1/8” (3mm) aluminum plate
[2] 4” (100mm) batt insulation
[3] 1/8” (3mm) aluminum plate
[¼” Bulb Seal at assumed vertical mullion edge, the other
3 edges are sealed with putty]
TL13-612 44
[1] 1/8” (3mm) aluminum plate
[2] 3” (75mm) air space
[3] 1/8” (3mm) aluminum plate
[¼” Bulb Seal at assumed vertical mullion edge, the other
3 edges are sealed with putty]
144
WEAL
Test No.
STC Material Layers [description] Element Drawing [Plan]
TL13-613 47
[1] 1/8” (3mm) aluminum plate
[2] 3” (75mm) batt insulation
[3] 1/8” (3mm) aluminum plate
[¼” Bulb Seal at assumed vertical mullion edge, the other
3 edges are sealed with putty]
TL13-614 46
[1] 1/8” (3mm) aluminum plate
[2] 3/16” (5 mm) MLV layer
[3] 3” (75mm) air space
[4] 3/16” (5 mm) MLV layer
[5] 1/8” (3mm) aluminum plate
[¼” Bulb Seal at assumed vertical mullion edge, the other
3 edges are sealed with putty]
TL13-615 48
[1] 1/8” (3mm) aluminum plate
[2] 3/16” (5 mm) MLV layer
[3] 3” (75mm) batt insulation
[4] 3/16” (5 mm) MLV layer
[5] 1/8” (3mm) aluminum plate
[¼” Bulb Seal at assumed vertical mullion edge, the other
3 edges are sealed with putty]
TL13-616 45
[1] 5/8” (16 mm) gypsum board plate
[2] 3” (75mm) air space
[3] 5/8” (16 mm) gypsum board plate
[¼” Bulb Seal at assumed vertical mullion edge, the other
3 edges are sealed with putty]
TL13-617 48
[1] 5/8” (16 mm) gypsum board plate
[2] 3” (75mm) batt insulation
[3] 5/8” (16 mm) gypsum board plate
[¼” Bulb Seal at assumed vertical mullion edge, the other
3 edges are sealed with putty]
145
WEAL
Test No.
STC Material Layers [description] Element Drawing [Plan]
TL13-618 50
[1] 5/8” (16 mm) gypsum board plate
[2] 3/16” (5 mm) MLV layer
[3] 3” (75mm) batt insulation
[4] 3/16” (5 mm) MLV layer
[5] 5/8” (16 mm) gypsum board plate
[¼” Bulb Seal at assumed vertical mullion edge, the other
3 edges are sealed with putty]
TL13-619 47
[1] 5/8” (16 mm) gypsum board plate
[2] 3/16” (5 mm) MLV layer
[3] 3” (75mm) air space
[4] 3/16” (5 mm) MLV layer
[5] 5/8” (16 mm) gypsum board plate
[¼” Bulb Seal at assumed vertical mullion edge, the other
3 edges are sealed with putty]
TABLE4-29: PHASE2BCLASSA,STCTESTRESULTS, AND SPECIMEN DESCRIPTION
146
4.4.4.1 TL OF ALUMINUM PLATES, 6” AND 3” AIR SPACE
The TL of aluminum plates tested with 6” and 3” airspaces (Figure 4-50).
FIGURE4-50: TL OF “PARALLEL PLATES” WITH 6” AND 4” AIR SPACES
*UNTREATED MULLION BASE CASE, MC1
The measured unitized vertical mullion is 3” wide (TL13-311, Figure 4-51) and is compared with TL13-612,
also 3” wide.
TL13-612, STC 44 TL13-311 (MC1), STC 36
FIGURE4-51: PLAN DRAWING OF CONNECTION ELEMENT TL13-612 (LEFT) AND UNITIZED MULLION TL13-311 (RIGHT)
The TL spectrum of TL13-612 is significantly higher than MC1 (TL13-311) (Figure 4-50). This indicates that
the structural bridging at the internal “leg” connection within the cavity of the unitized mullion is
significantly reducing the TL performance. The improved TL performance of specimen TL13-612 may be
Interstitial “Leg”
Connection
3” WIDTH 3” WIDTH
147
influenced by the acoustically separated aluminum plates (i.e. non-bridged) and the plate stiffness from
the wood battens.
4.4.4.2 TL OF ALUMINUM AND GYPSUM BOARD PLATES, 3” AIR SPACE
The TL performance of the aluminum and gypsum board plates are plotted with Phase 1 mullions(MC-1)
TL13-311, (28) TL13-404 and (17) TL13-325 as building materials are common between the test sequences.
(53) TL13-612
(54) TL13-613 (57) TL13-616 (58) TL13-617
(MC-1) TL13-311
(28) TL13-404 (17) TL13-325
FIGURE4-52: PHASE2BALUMINUM AND GYPSUM BOARD PLATE RESULTS
*UNTREATED MULLION BASE CASE, MC1
0
10
20
30
40
50
60
70
80
90
100
Transmission Loss (dB)
One-third Octave Band Center Frequency (Hz)
Phase 2b: Plate Connection with a 3" airspace
(1/8" Aluminum and 5/16" Gypsum)
(53) TL13-612 - 44
(54) TL13-613 - 47
(57) TL13-616 - 45
(58) TL13-617 - 48
*(3) TL13-311 - 36
(28) TL13-404 - 35
(17) TL13-325 - 42
148
Several observations can be drawn with reference to Figure 4-52:
The TL performance of all parallel plate connections shown perform higher than mullions. This
indicates that the connection element is not necessarily controlling the overall transmission loss
of a curtain wall. The lower mullion performance may be influenced by the factors discussed in
the section above, i.e. interstitial leg connections in the mullion cavity as well as structural
stiffness.
The average parallel plate performance is 11 dB STC more than MC-1. The average performance
of parallel plate configurations filled with mineral fiber is 9 dB STC above the phase 1 Mullion test
TL13-404.
The mullion specimens from earlier phases perform better at lower frequencies (below 250Hz)
and the parallel plate specimens generally outperform the mullions above 250 Hz.
Gypsum board plates perform higher than aluminum plates. This is likely due to the heavier
weight of the gypsum board.
149
4.4.4.3 TL OF ALUMINUM AND GYPSUM BOARD PLATE WITH MLV DAMPING LAYER, 3” AIR SPACE
The TL performance of select Phase 2B connections are plotted with Phase 1 mullions (10) TL13-318 and
(13) TL13-321 as building materials are common between the two test phases (Figure 4-53).
(55) TL13-614
(56) TL13-615
(60) TL13-619
(59) TL13-618
(10) TL13-318 (13) TL13-321
FIGURE4-53: PHASE2BCLASSA: ALUMINUM OR GYPSUM BOARD PLATE WITH AN MLV DAMPING LAYER
Several observations can be drawn with reference to Figure 4-53:
The performance of the mullions and plates are closer in performance. This indicates that the MLV
damping layer has a significant influence on the performance of the mullion. The Phase 1 mullions
are within 1 - 2 dB STC points of the connection elements.
0
10
20
30
40
50
60
70
80
90
100
Transmission Loss (dB)
One-third Octave Band Center Frequency (Hz)
Phase 2b: Plate Connection with a 3" airspace
(1/8" Aluminum+MLV vs 5/8" Gypsum+MLV)
(55) TL13-614 - 46
(56) TL13-615 - 48
(60) TL13-619 - 47
(59) TL13-618 - 50
(53) TL13-612 - 44
(10) TL13-318 - 48
(13) TL13-321 - 50
150
At frequencies below 400Hz the mullions outperform the plate specimens. This result is
unexplained; however it may be an indication that the difference in structural stiffness between
mullions and plate specimens should be examined further.
Similar to the previous section, the gypsum board plates performed slightly higher than the
aluminum plates; however the performance between materials is much closer with the
introduction of batt within the parallel plate cavity.
151
4.4.5 PHASE 2B CLASS B TEST SEQUENCE
Results from the Class B testing sequence are summarized (Table 4-30).
WEAL
Test No.
STC Material Layers [description] Element Drawing [Plan]
TL13-620 22
[1] 5/8” (16 mm) staggered gyp bd plates
[2] 3” (75mm) air space
[3] 5/8” (16 mm) staggered gyp bd plates
[Putty seal at all filler wall edges]
[1/4” air space between staggered plates]
TL13-621 37
[1] 5/8” (16 mm) staggered gyp bd plates
[2] 3” (75mm) batt insulation
[3] 5/8” (16 mm) staggered gyp bd plates
[Putty seal at all filler wall edges]
[1/4” air space between staggered plates]
TL13-622 51
[1] 5/8” (16 mm) staggered gyp bd plates
[2] 3” (75mm) batt insulation
[3] 5/8” (16 mm) staggered gyp bd plates
[Putty seal at all filler wall edges]
[1/4” bulb seal between staggered plates, closing air gap]
TL13-623 44
[1] 5/8” (16 mm) staggered gyp bd plates
[2] 3” (75mm) air space
[3] 5/8” (16 mm) staggered gyp bd plates
[Putty seal at all filler wall edges]
[1/4” bulb seal between staggered plates, closing air gap]
152
WEAL
Test No.
STC Material Layers [description] Element Drawing [Plan]
TL13-624 47
[1] 5/8” (16 mm) staggered gyp bd plates
[2] 3” (75mm) air space
[3] 5/8” (16 mm) staggered gyp bd plates
[Putty seal at all filler wall edges]
[1/4” bulb seal between staggered plates, closing air gap;
2 receiver/1 source]
TL13-625 20
[1] 1/8” (3mm) aluminum plate
[2] 3” (75mm) air space
[3] 1/8” (3mm) aluminum plate
[Putty seal at all filler wall edges]
[1/4” air space between staggered plates]
TL13-626 31
[1] 1/8” (3mm) aluminum plate
[2] 3” (75mm) batt insulation
[3] 1/8” (3mm) aluminum plate
[Putty seal at all filler wall edges]
[1/4” air space between staggered plates]
TL13-627 49
[1] 1/8” (3mm) aluminum plate
[2] 3” (75mm) batt insulation
[3] 1/8” (3mm) aluminum plate
[Putty seal at all filler wall edges]
[1/4” bulb seal between staggered plates, closing air gap]
TL13-628 47
[1] 1/8” (3mm) aluminum plate
[2] 3” (75mm) air space
[3] 1/8” (3mm) aluminum plate
[Putty seal at all filler wall edges]
[1/4” bulb seal between staggered plates, closing air gap]
153
WEAL
Test No.
STC Material Layers [description] Element Drawing [Plan]
TL13-629 48
[1] 1/8” (3mm) aluminum plate
[2] 3” (75mm) air space
[3] 1/8” (3mm) aluminum plate
[Putty seal at all filler wall edges]
[1/4” bulb seal between staggered plates, closing air gap;
2 receiver/1 source]
TABLE4-30: PHASE2BCLASSB,STCTESTRESULTS AND SPECIMEN DESCRIPTION
154
4.4.5.1 TL OF STAGGERED PLATE - ALUMINUM
The TL performance of the aluminum staggered plates are plotted in Figure 4-54.
(66) TL13-625
(67) TL13-626 (69) TL13-628 (68) TL13-627
FIGURE4-54: PHASE2BCLASSB,MLV COMPOSITE WITH ALUMINUM OR GYPSUM BOARD PLATE
0
10
20
30
40
50
60
70
80
90
100
Transmission Loss (dB)
One-third Octave Band Center Frequency (Hz)
Phase 2b: Staggered 1/8" Aluminum Plate - 3" air space
(with and without silicone bead seal/batt insulation)
(66) TL13-625 - 20
(67) TL13-626 - 31
(68) TL13-627 - 49
(69) TL13-628 - 47
(70) TL13-629 - 48
155
4.4.5.2 TL OF STAGGERED PLATE - GYPSUM BOARD
The TL performance of the aluminum staggered plates are plotted in Figure 4-55.
(61) TL13-620
(62) TL13-621 (64) TL13-623 (63) TL13-622
FIGURE4-55: PHASE2BCLASSB:MLV COMPOSITE WITH ALUMINUM OR GYPSUM BOARD PLATE
TL13-622 is the highest performing connection element.
0
10
20
30
40
50
60
70
80
90
100
Transmission Loss (dB)
One-third Octave Band Center Frequency (Hz)
Phase 2b: Staggered 5/8" Gypsum Board Plate - 3" air space
(with and without silicone bead seal/batt insulation)
(61) TL13-620 - 22
(62) TL13-621 - 37
(63) TL13-622 - 51
(64) TL13-623 - 44
(65) TL13-624 - 47
156
4.4.6 PHASE 2B CLASS C TEST SEQUENCE
Results from the Class C testing sequence are summarized (Table 4-31).
WEAL
Test No.
STC Material Layers [description] Element Drawing [Plan]
TL13-630 23
[1] 1/16” (1.5mm) aluminum plate
[2] 2-3/16” (55.5mm) batt insul
[3] 1/16” (1.5mm) aluminum plate
[Mullion Mate ® assembly isolated]
[wet seal at all vertical edges]
TL13-631 30
[1] 1/16” (1.5mm) aluminum plate
[2] 3” (75mm) air space
[3] 1/16” (1.5mm) aluminum plate
VEC: 3-1/2” Mullion Mate ®
[wet seal at all vertical edges]
[TL13-311 mullion used]
TL13-632 31
[1] 1/16” (1.5mm) aluminum plate
[2] 3” (75mm) air space
[4] 1/16” (1.5mm) aluminum plate
VEC: 3-1/2” Mullion Mate ®
[wet seal at all vertical edges]
[TL13-323 mullion used]
TL13-633 50
[1] Mull-It-Over ® (alum+vinyl+foam)
[2] 6” (150mm) air space
[3] Mull-It-Over ® (alum+vinyl+foam)
[overall width 7-1/2” (190mm) ]
TABLE4-31: PHASE2BCLASSC,STCTESTRESULTS AND SPECIMEN DESCRIPTION
*VEC (vertical edge condition)
157
4.4.6.1 MANUFACTURED PRODUCT RESULTS
FIGURE4-56: MULLIONMATE®(LEFT) AND MULL IT OVER® (RIGHT)
FIGURE4-57: MULLIONMATE(LEFT) AND MULL IT OVER (RIGHT)
158
4.4.6.2 TL OF MULLION MATE ® PERFORMANCE BELOW
The Mullion Mate® product is tested in an isolated condition and then coupled with the lowest and highest
performing mullion base cases from phase 1 (Figure 4-58).
FIGURE4-58: MULLIONMATE(LEFT) AND MULL IT OVER (RIGHT)
*UNTREATED MULLION BASE CASE, MC1
The graph in Figure 4-58 includes the TL overlay of the mullion base case TL13-311. The test
measurements with the product perform significantly less than the untreated (hollow and exposed)
mullion base case. There is also an unexplained resonance at 400Hz with the product tests.
Similar to test measurements conducted in Phase 1 Class C4, the discrete resonance seen with the product
tests reduces the overall STC rating and requires further investigation.
0
20
40
60
80
100
63
80
100
125
160
200
250
315
400
500
630
800
1000
1250
1600
2000
2500
3150
4000
5000
Transmission Loss (dB)
One-third Octave Band Center Frequency (Hz)
Mullion Mate Product
(with and without mullions)
(71) TL13-630 - 23 (72) TL13-631 - 30
(73) TL13-632 - 31 *(3) TL13-311, STC 36
159
4.4.6.3 MULL-IT-OVER® PERFORMANCE
All the Mull-it-Over Transmission Loss plots for assemblies tested at WEAL in phase 2A and 2B are
summarized in Figure 4-59. The tested product in phase 2A (TL13-623) structurally bridges either side of
the filler wall with a wood stud and includes a mullion. The product tested during phase 2B (TL13-633)
does not include a mullion and is not bridged.
FIGURE4-59: TRANSMISSIONLOSS RESULTS OF THE MULL IT OVER PRODUCT FROM WEAL
0
20
40
60
80
100
63
80
100
125
160
200
250
315
400
500
630
800
1000
1250
1600
2000
2500
3150
4000
5000
Transmission Loss (dB)
One-third Octave Band Center Frequency (Hz)
Phase 2: Mull-It-Over Tests (WEAL)
(45) TL13-423, 46 (74) TL13-633 - 50
160
4.4.7 PHASE 2B OBSERVATIONS AND CONCLUSION
The following are general observations and field notes based on the TL of the parallel and staggered
plates:
Adding silicone bead seals to create airtight conditions with the staggered plate test sequence (eg.
TL13-623) significantly increases the transmission loss better than adding mineral fiber to the cavity
(eg. TL13-621).
Adding batt insulation provides a greater TL improvement with narrow airspaces (3”) versus wider
airspaces (6”)
The elements with a wider airspace and/or greater mass achieved higher transmission loss
performances.
o (48) TL13-607 and (49) TL13-608 highest performing due to 6” airspace.
o Gypsum board elements performed higher than similar configuration setups with aluminum
plates. This indicates that the denser material property of gypsum board is influencing
performance.
Test (53) TL13-612 resulted in STC 44 performs better than a 3” wide aluminum mullion extrusion
(TL13-311) since there is no bridging between each side.
Highest performing test is (63) TL13-622 with STC 51.
Batt insulation and seals are significant TL improvement factors; this will influence detailing in
practice.
Lowest performing shows 2 major elements
o Low mass - indicated by dips at lower frequencies 160 – 630
o Holes/leaks - indicated by dips at high frequencies, 2.5kHz
The following are conclusions from the product measured in this phase:
Products that are lighter or narrower than the connected mullion will degrade the overall
performance, i.e. Mullion Mate®.
Products that act like an extension of the demising wall can provide good transmission loss
performance; appropriate field assembly construction is critical to avoid leaks or degradation of the
overall transmission loss.
In the next phase of testing, no connection detail is added to the curtain wall bay system so that any
variable influence will not impact the Transmission loss results.
Analytical studies to be investigated further include:
Single panel TL comparisons- Mass dependent - single panels generally increase 6 dB per octave.
o If mass is increased on both sides, the curve shifts up.
o The specimen starts to act like a single panel (i.e. mullion) when more connections bridge
both sides.
Double Panel TL comparisons - generally increase 10 db per octave.
Larger hole or leak - the frequency dip in the curve shifts toward low frequency.
The influence of low frequency wavelengths on small apertures, i.e. 6” aperture
¼” perimeter slits may have an influence on the frequency dips.
161
4.5 PHASE 3 – GLAZING ASSEMBLY ELEMENT
The Phase 3 laboratory tests are the final UVM sound transmission loss measurement sequence (Figure
4-60). Unlike the previous test phases, most significant in this phase is the introduction of the following
new elements:
Insulated glazing unit infill (IGU)
Upper horizontal mullion (i.e. transom)
Lower horizontal mullion (i.e. sill)
Date March 17 - 20, 2014
Laboratory: WEAL
Specimen: Glazing Element
including Mullion and
Transom
Dimensions: 8-3/4” depth
Surface Area: 2.92 ft
2
(0.27m
2
)
Transmission: Horizontal
Procedure: ASTM E90-09
Total Tests: 5 airborne, 4
vibration sets
FIGURE4-60: PHASE3:PLAN DIAGRAM OF UVM ELEMENTS AND ASSOCIATED TEST PHASES
The composite unitized curtain wall bay assembly includes the unitized vertical and horizontal mullions
and the insulated glazing unit (IGU) infill for two adjacent bays. A partition connection element from Phase
2a and 2b was not included in this test phase so that any variable influence at this path is eliminated. A
total of five sound transmission loss tests were conducted.
Four sets of vibration tests were also conducted during this phase as part of a linear sequence to study
the acoustic energy injection across the multiple elements that comprise the curtain wall. The
measurement detail and results are included in Appendix D. Vibration measurements were not conducted
in other test phases because the airborne sound transmission loss was discrete for all other elements,
whereas the architecture of the curtain wall assembly is inherently composite and vibration
measurements give an indication of acoustic energy transmitted to surfaces at the receiving chamber.
162
4.5.1 PHASE 3 TEST SPECIMEN DESCRIPTION
The unitized glass curtain wall system was tested with three modifications to the center vertical mullion
which align directly with select tests in Phase 1 (Figure 4-61):
TL13-323, STC 52 (MC-2)
2
TL13-316, STC 38 TL13-311, STC 36 (MC-1)
1
FIGURE4-61: PHASE1CONFIGURATIONS USED AS THE CENTER VERTICAL MULLION IN PHASE 3
1
MC1 (Mullion Control 1): the hollow and exposed mullion tested in isolation during Phase 1 with an
STC 36 performance.
2
MC2 (Mullion Control 2): the MLV pillow filled mullion with gypsum plus MLV overclad. The highest
performing mullion tested in Phase 1 with an STC 52 performance.
The construction of the testing rig is labor intensive due to the weight of the large scale specimen and
design of the semi-anechoic chambers. Therefore the number of tests conducted in this phase were
limited to the highest (MC-2) and lowest (MC-1) performing mullions. An opportunity to test a third
variation (TL13-316) was possible based on ease of disassembling the MC-2 mullion.
The vibration measurements procedure and results are described in Appendix D. The vibration
measurements were conducted to explore the energy radiation passing from the source to receiving
chambers at three curtain wall surfaces:
[1] The insulated glazing unit (z-axis),
[2] The lower horizontal mullion (y-axis), and
[3] The center vertical mullion (x-axis).
163
4.5.2 PHASE 3 TEST CHAMBER CONSTRUCTION
Two gypsum board wall enclosures (Chamber 3S and 3R in Figure 4-62) were built within the WEAL source
and receiving chambers in order to conduct the airborne and vibration test series. These smaller semi-
anechoic chambers, 3S chamber at the source side and 3R chamber at the receiving side are both semi-
anechoic. The envelope of the chambers was designed to enclose the outboard side of the curtain wall
system as seen in Figure 4-62. The intent of this integral chamber was to simulate an in situ installation
where direct sound transmitting directly through the IGU would be absorbed by the atmosphere.
FIGURE4-62: COUPLED SEMI-ANECHOIC CHAMBERS 3S AND 3R SHOWN IN PLAN WITHIN THE MAIN WEAL CHAMBERS.
The follow is a description of Chambers 3S and 3R and the curtain wall specimen (Figure 4-63 and Figure
4-66).
3S Cubic Volume: 210 ft
3
3R Cubic Volume: 138 ft
3
Absorption: 3S/3R chambers filled with mineral wool and fiber glass (approximately 15 batt
layers per chamber )
Specimen Weight: 335 lbs one curtain wall bay (without mullion fill or overclad modifications)
Filler wall aperture: 7" x 60" (includes depth of mullion and backer rod/wet seal, not IGU depth)
164
FIGURE4-63: CURTAIN WALL BAY INSERTED INTO THE
APERTURE OF THE FILLER WALL
FIGURE4-64: CURTAIN WALL PLACED AND CENTERED IN
THE FILLER WALL.
FIGURE4-65: CHAMBER3S(BUILT WITHIN THE WEAL
SOURCE CHAMBER)
FIGURE4-66: CHAMBER3R(BUILT WITHIN THE WEAL
RECEIVING CHAMBER)
165
The perimeter edge of the curtain wall was spaced by ¼” thick neoprene and sealed with backer rod and acoustic caulking (This is the perimeter
seal condition uniform at all test phases.)
FIGURE4-67: CHAMBER3S AND 3R PLAN DRAWINGS AND ACOUSTIC DETAILS
166
The fiberglass insulation filled the 3S and 3R chambers from floor to ceiling. The curtain wall system sat
on a wood bulkhead platform and mounted on neoprene (Figure 4-68 and Figure 4-69).
FIGURE4-68: WEALCHAMBER[SECTION]
FIGURE4-69: WEALCHAMBER[ELEVATION]
167
4.5.3 PHASE 3 TEST SEQUENCE
Summarized below are single figure STC ratings for the Phase 3 configurations (Table 4-32).
WEAL Test
No.
STC Material Layers [description] Element Drawing [Plan]
Ph3 [MC2]
TL14-167
42
Applying center vertical mullion
configuration: TL13-323 (STC 52)
[overclad (gyp+MLV) and filled
(MLV pillows)]
Ph3 [MC1
A
]
TL14-168
37
Applying center vertical mullion
configuration: TL13-316 (STC 38)
[filled MLV pillows]
[measurement taken after
overclad was removed]
Ph3 [MC1]
TL14-170
32
Applying center vertical mullion
configuration: TL13-311 (STC 36)
TL14-171 32
TL13-311 STC 36
[Receiving chamber (3S) removed]
TABLE4-32: PHASE3STCTESTRESULTS
168
The transmission loss calculation per ASTM E90 did not use modified cubic volumes at the source and
receiving chambers of the WEAL facility. The TL and STC estimates ignore the cubic volume of the 3S and
3R chambers.
For test specimen TL14-167, the MLV pillow infill extended the entire length of the center vertical mullion,
i.e. past the depth of the horizontal mullion elements. However, the overclad only extended to the
exposed sides of the center vertical mullion and terminated where the horizontal mullion interested.
4.5.4 PHASE 3 SUMMARY
Three primary configurations were measured in phase 3:
STC 32 is the performance of the curtain wall system with the MC 1 configuration at the center
vertical mullion.
STC 42 is the performance of the curtain wall system with MC 2 configuration at the center
vertical mullion.
STC 38 is the performance of the curtain wall system with a center vertical mullion filled with
MLV pillows.
FIGURE4-70: CHAMBER3S AND 3R PLAN DRAWINGS AND ACOUSTIC DETAILS
Subjective observations when evaluating acoustic leaks with the rigid stethoscope at the receiving side
of the specimen:
Audible energy at the IGU significantly lower than energy at the mullion and horizontal mullions.
Significantly greater acoustic energy at the horizontal mullions (upper and lower) than any other
structural part of the specimen.
Audible low frequency energy at the overclad clad mullion (TL14-167).
Audible acoustic energy through the 1/2" silicone gasket between the glass and mullion.
0
10
20
30
40
50
60
70
80
90
100
63 80 100 125 160 200 250 315 400 500 630 800 1000 1250 1600 2000 2500 3150 4000 5000
Transmission Loss (dB)
One-Third Octave Band Center Frequency (Hz)
Phase 3 Transmission Loss Results of Curtain Wall System
(TL14-167, STC 42) (TL14-168, STC 37) (TL14-170, STC 32)
TL14-170, STC 32 TL14-168, STC 37 TL14-167, STC 42
169
4.6 UVM MEASUREMENT SUMMARY
Approximately 80 laboratory tests were conducted in four test phases in accordance with ASTM E90. All
measurements were tested in the same type of laboratory conditions so that they may be compared to
each other. Test measurements are not intended to simulate an in situ condition but instead inform
relative changes based on the architectural modifications to each element.
The results from these modifications provide information on what architectural mechanism of the curtain
wall architecture are controlling the overall sound isolation performance (as discussed in Chapter 5) and
additionally may be applied in practice to improve curtain wall mullions.
Publications similar to the California Catalog of STC and IIC Ratings for Wall and Floor/Ceiling Assemblies
67
or the NRC-CNRC labs Gypsum Board Walls: Transmission Loss Data
68
provide laboratory tests data for
batteries of various partition typologies in order to guide designers with performance metrics and relative
improvements between wall or floor specimens. The potential value with the measurements conducted
in the UVM testing sequence is to make relative comparisons and not necessarily use the face value
performance rating. The many modifications to test elements reveal relative changes that can be taken
from the laboratory and applied in practice.
FIGURE 4-71 summarizes the range of STC results for each phase.
FIGURE 4-71: SUMMARY OF STC RANGES AT EACH UVM TEST PHASE
Transmission loss curves were observed after each testing phase in this chapter. Addition notable
observations not mentioned previously include:
Mullions in Phase 1 and Partition connections in Phase 2B are capable of achieving commensurate
TL performances.
67
Russell B. DuPree and California. Dept. of Health Services. Office of Noise Control, Catalog of STC and IIC Ratings for Wall and
Floor/ceiling Assemblies (Berkeley, California 94704: Office of Noise Control, California Department of Health Services, 1980).
68
Halliwell et al., “NRC-CNRC Gypsum Board Walls: Transmission Loss Data.”
52
41
50
42
36
22
25
36
0
10
20
30
40
50
60
70
80
90
100
Sound Transmission Class (STC)
Test Phases
STC Result Ranges at Each Test Phase
PHASE 1
PHASE 2A
PHASE 2B
PHASE 3
170
Discrete resonances occurred at various phases and were observed to be an indication of
material resonance, structural stiffness, or possibly the size of the aperture.
171
CHAPTER 5 ANALYSIS OF CONTROLLING MECHANISMS AND COMPOSITE
TRANSMISSION LOSS
5.1 INTRODUCTION
This chapter describes the analysis of the one-third octave sound transmission loss tests characteristics of
the individual elements tested in the Unitized Vertical Mullion (UVM) test method. Critical comparisons
are conducted between connected and unconnected mullion conditions in order to characterize
mechanisms controlling the overall sound isolation performance. Transmission loss results from the UVM
test method are applied to composite predictions including theoretical demising wall performance to
establish where diminishing returns may occur between acoustic performance and material construction.
Unconnected mullions are those that that were tested independent from the curtain wall system glass,
transom, and sills. These were tested in Phase 1.
Connected mullions are those that were tested while mechanically connected to the curtain wall system.
These were tested during Phase 3.
Noise reduction (NR) ratings are also evaluated and provide the direct source level difference between
the source and receiving chambers. This differs from transmission loss (TL) because it does not include
normalizing factors of the specimen size and receiving room absorption.
5.2 CONNECTED VERSUS UNCONNECTED MULLION CONDITIONS
The STC rating of the connected mullions in Phase 3 performed lower than the respective unconnected
mullion conditions in Phase 1. The single-figure STC and NR ratings and the one-third octave band sound
transmission loss between these two test phases are compared below.
The unconnected mullion conditions specifically studied here are
Phase 1 Test TL13-311 is named PH1-MC1 for Mullion Constant 1 and is the independent exposed
and hollow mullion condition.
Phase 1 Test TL 14-323 is named PH3-MC2 for the Mullion Control 1 and is overclad with 5/8”
thick gypsum wall board and a 3/16” thick limp mass vinyl later including Mall Limp Vinyl (MLV)
pillows filled with mineral fiber in the mullion air cavity.
The connected mullion conditions specifically studied here are
Phase 3 Test TL14-170 is named PH3-MC1 and is the exposed and hollow mullion condition which
forms the center vertical mullion connected to the curtain wall bay test rig with glass.
Phase 3 Test TL14-167 is named PH3-MC2 and is the overclad and filled mullion which forms the
center vertical mullion connected the curtain wall bay test rig with glass
172
5.2.1 COMPARISON: SINGLE FIGURE RATING
Sound transmission class (STC) and noise reduction (NR) ratings are tabulated for lightweight and heavy
weight mullions measured with or without the curtain wall glass (Table 5-1).
Lab Test STC NIC
Center Vertical Mullion
Unconnected
(without glass)
Connected
(with glass)
Exposed/
Hollow
Filled Overclad/
Filled
TL13-311 36 49
TL14-170 32 47
TL13-316 38 52
TL14-168 37 50
TL13-323 52 64
TL14-167 42 55
TABLE5-1: STC AND NR RATING COMPARISON OF SELECT MULLION TESTS FROM PHASE 1 AND PHASE 3
All specimens tested in Phase 3 have a lower STC performance than the respective mullion assemblies
tested in Phase 1. The following performance deltas were observed:
The STC 36 hollow/exposed mullion (TL13-311) performs 4 dB STC points lower when connected
to the curtain wall glass and aluminum frame at STC 32 (TL14-170).
The STC 38 filled mullion (TL13-316) performs 1 dB STC point lower when connected to the curtain
wall at STC 37 (TL14-168).
The STC 52 overclad/filled mullion (TL13-323) performs 10 dB STC points lower when connected
to the curtain wall glass system at STC 42 (TL14-167).
It is evident from these results that the addition of the curtain wall glass, sill, and transom to the highest
rated specimen (TL13-323) resulted in the greatest reduction in STC and NIC ratings. The conclusion is that
one reaches the point of diminishing returns with improving the STC of a mullion once it is part of a
composite curtain wall system. The mullion is not the weakest link in this case and the performance is
dependent on other components in the system.
Observations between test specimens in Phase1 and Phase 3 are compared in Table 5-2.
173
PHASE 3 TEST AND DESCRIPTION OBSERVATIONS
TL14-167 Plan Drawing of Test Rig TL14-167 Observations
The heavy center mullion had the greatest performance
reduction when connected to curtain wall system than
the other lighter mullion assemblies in Phase 3, e.g.
TL14-168 and TL14-1470.
The significant performance reduction may indicate
there are diminishing returns between material
assembly and acoustic performance.
It is clear that flanking paths within the test rig are at
the transom, sill, and glass. It is unclear which of these
three elements is contributing to the reduction more
significantly, the glass infill or horizontal aluminum
members.
Future Studies:
Future test studies should include modifications to the
horizontal mullions to explore STC performance
improvement and the influence of the glass connection.
TL14-168 Plan Drawing of Test Rig TL14-168 Observations
Based on the sound flanking path observations from
TL14-167, it appears that
sound flanking within the TL14-168 specimen did
not significantly reduce the performance of the
center vertical mullion assembly (i.e. 1dB STC
difference between Phase 1 and 3)
The performance between the unconnected version of
this mullion and connected version does not vary
significantly. This indicates a possible balance between
the sound energy through the mullion compared to the
curtain wall assembly.
TL14-170 Plan Drawing of Test Rig TL14-170 Observations
Based on the flanking paths within the TL14-167 test
specimen (i.e. at the transom, sill, and glass), it appears
that these same paths have contributed to the further
reduction of the center vertical mullion in this test rig.
Future Studies
It is unclear if the center vertical mullion in this specific
test rig is controlling the overall STC performance. Two
deductions may be considered:
174
PHASE 3 TEST AND DESCRIPTION OBSERVATIONS
The vertical mullion is not controlling the STC; the
glass and horizontal mullions are reducing the TL of
the exposed and hollow mullion.
The vertical mullion is controlling the STC; the
difference between the mullion alone and the
mullion + curtain wall system is less with the
previous test, TL14-168 (where the center mullion
is almost balanced with the curtain wall system)
TABLE5-2: OBSERVATIONS OF THREE COMPARABLE TEST SPECIMENS FROM PHASE 1 AND PHASE 3
The other sound flanking paths within the fully connected curtain wall system identified in Phase 3 are
glass infill and horizontal mullions (FIGURE 5-1). These paths limit the achievable performance of the
connected center vertical mullion.
FIGURE 5-1: DRAWING OF MULLION SECTION WITH IDENTIFIED FACE AREAS OF EACH CURTAIN WALL ELEMENT:
TRANSOM(0.16 SF), GLASS (0.7 SF), MULLION (2.42 SF), AND SILL (0.16 SF).
5.2.2 NOISE REDUCTION COMPARISON: PHASE 1 AND PHASE 3
The Noise Reduction (NR) spectra between the comparable Phase 1 and Phase 3 unitized vertical mullion
test rigs are compared in this section. This intent is to identify performance limitations at specific
frequency regions.
[PATH 1]
TOP HORIZONTAL MULLION
0.16 sf
[PATH 4]
BOTTOM HORIZONTAL MULLION
0.16 sf
[PATH 3]
CENTER VERTICAL MULLION
2.42 sf
[PATH 2]
GLASS INFILL
0.70 sf
175
5.2.2.1 UNCONNECTED VS CONNECTED EXPOSED/HOLLOW UVM
The NR curves for Ph1-MC1 (TL 13-311) and Ph3-MC1 (TL14-170) are compared below.
FIGURE 5-2: TL13-311 (UNCONNECTED MULLION) FIGURE 5-3: TL14-170 (CONNECTED MULLION)
FIGURE 5-4: NOISE REDUCTION SPECTRA BETWEEN TL13-311 AND TL14-170
Trends Observed
It is unclear what is causing the resonance at 250 Hz that is bringing down the performance Ph3-MC1
(TL14-170). This resonance is lower in NR level and frequency than the unconnected vertical mullion from
Phase 1, i.e. 400Hz. An examination of this resonance could be a valuable opportunity for future study.
0
10
20
30
40
50
60
70
80
90
100
63 80 100 125 160 200 250 315 400 500 630 800 1k 1.25k 1.6k 2k 2.5k 3.15k 4k 5k
Noise Reduction (dB)
One-Third Octave Band Center Frequency (Hz)
NR COMPARISON
UNCONNCECTED (TL13-311) VS CONNECTED (TL14-170)
HOLLOW/EXPOSED CENTER VERTICAL MULLION
TL13-311 (STC 36, NIC 49) TL14-170 (STC 32, NIC 47)
TL13-311
Unconnected
TL14-170
Connected
176
It is observed that in the low frequency region between 63 Hz and 200 Hz that the vibrational waves in
the glass and horizontal mullions of TL14-170 are transmitted into the mullion and then into the curtain
wall system in the receiving room.
In the frequency region between 630 Hz to 1.6k Hz, the glass and horizontal mullions of TL14-170 may be
structurally stiffening (clamping) the center vertical mullion and therefore reducing vibration into the
curtain wall bay at the receiving room.
At the high frequency region above 2 kHz, the delta between the curves is small. This may indicate that
both the glass and horizontal mullions are loosely coupled to the mullion in this region.
177
5.2.2.2 UNCONNECTED VS CONNECTED FILLED UVM
The curtain wall specimen Ph3-MC1a (TL14-168) performs 1 dB STC points lower than the mullion
specimen of the same weight and assembly, Ph1-MC1a (TL 13-316).
FIGURE 5-5: TL13-316 (UNCONNECTED) FIGURE 5-6: TL14-168 (CONNECTED)
FIGURE 5-7: NOISE REDUCTION SPECTRA BETWEEN TL13-316 AND TL14-168
0
10
20
30
40
50
60
70
80
90
100
63 80 100 125 160 200 250 315 400 500 630 800 1k 1.25k 1.6k 2k 2.5k 3.15k 4k 5k
Noise Reduction (dB)
One-Third Octave Band Center Frequency (Hz)
NR COMPARISON
UNCONNCECTED (TL13-316) VS CONNECTED (TL14-168)
FILLED CENTER VERTICAL MULLION
TL13-316 (STC 38, NIC 52) TL14-168 (STC 37, NIC 50)
Unconnected
Connected
178
Trends Observed
Although the TL of the connected mullion in TL14-168 is not as high as the one in the previous test TL14-
170, the same trends as the previous comparison occur.
At frequencies below 315 Hz, it is deduced that the connected mullion is performing lower due to the
sound energy vibrating through the glass and horizontal mullions. A resonance at 250 Hz is still present.
The noise reduction spectrum from the previous test specimen TL14-170 has a higher performance at low
frequencies than the TL14-168 specimen shown here. However at mid-frequencies, TL14-168 performs
higher than TL14-170. This is an indication that the added mass and dampening inside the mullion
improves the performance.
179
5.2.2.3 UNCONNECTED VS CONNECTED OVERCLAD/FILLED UVM
The curtain wall specimen Ph3-MC2 (TL14-167) performs 10 dB STC points lower than the mullion
specimen of the same weight and assembly, Ph1-MC2 (TL 13-323). The spectral plots of both test curves
are overlaid (FIGURE 5-10).
FIGURE 5-8: TL13-323 (UNCONNECTED) FIGURE 5-9: TL14-167 (CONNECTED)
FIGURE5-10: NOISEREDUCTIONSPECTRA BETWEEN TL13-323 AND TL14-167
0
10
20
30
40
50
60
70
80
90
100
63 80 100 125 160 200 250 315 400 500 630 800 1k 1.25k 1.6k 2k 2.5k 3.15k 4k 5k
Noise Reduction (dB)
One-Third Octave Band Center Frequency (Hz)
NR COMPARISON
UNCONNCECTED (TL13-323) VS CONNECTED (TL14-167)
OVERCLAD/FILLED CENTER VERTICAL MULLION
TL13-323 (STC 52, NIC 64) TL14-167 (STC 42, NIC 55)
Unconnected
Connected
180
Trends Observed
The connected mullion has lower noise reduction throughout the entire frequency region, as much as 12
dB at low frequencies and 20 dB at high. The previous comparisons indicated certain mid to high frequency
regions where the connected mullion performed slightly higher than the unconnected mullion. In this
particular case however, the connected mullion that was mass loaded (with infill and an overclad) is
significantly lower than its respective unconnected mullion assembly.
This reveals that the glass and horizontal mullions are reducing the potential sound isolation of the overall
Phase 3-MC2 system. These three elements (glass, upper, and lower horizontal mullion) are flanking paths
and weaken almost all frequency domains.
The resonance seen in the previous comparisons at 250Hz does not exist with PH3-MC2 (TL14-167). This
may be an indication that the added mass and damping at the vertical mullion improve this resonance.
The NR curves indicate that the vertical mullion is closely coupled to the glass and horizontal mullions at
630 Hz.
No overclad could be placed on the connected center vertical mullion where the horizontal mullion joins,
i.e. the stack joint. The overall weight of the center vertical mullion is therefore slightly less than the
unconnected condition since the overclad was cut back at top and bottom ends of the vertical mullion.
181
5.2.3 MECHANISMS LIMITING SOUND ISOLATION PERFORMANCE
5.2.3.1 HORIZONTAL MULLIONS AND GLASS
The Noise Reduction graph below compares select unconnected mullion constants from Phase 1:
- TL13-311: The unconnected vertical mullion, exposed and hollow
- TL13-323: The unconnected vertical mullion, overclad and filled
FIGURE5-11: NOISEREDUCTIONSPECTRA BETWEEN TL13-311 AND TL13-323
There is an average 16 dB noise reduction improvement above 250Hz from TL13-311 and TL13-323.
A significant difference in noise reduction is expected based on the known mass and damping
modifications. The delta in performance occurs across most of the frequency region with the exception
at low frequencies.
0
10
20
30
40
50
60
70
80
90
100
63 80 100 125 160 200 250 315 400 500 630 800 1k 1.25k 1.6k 2k 2.5k 3.15k 4k 5k
Noise Reduction (dB)
One-Third Octave Band Center Frequency (Hz)
NR COMPARISON
UNCONNCECTED CENTER VERTICAL MULLION
TL13-311 (EXPOSED/HOLLOW) AND TL13-323(OVERCLAD/FILLED)
TL13-311 (STC 36, NIC 49) TL13-323 (STC 52, NIC 64)
Exposed/Hollow
Overclad/Filled
182
Similarly, the Noise Reduction graph below compares the connected mullion constants from Phase 3:
- TL14-170: Connected vertical mullion, exposed and hollow
- TL13-168: Connected vertical mullion, filled
- TL14-167: Connected vertical mullion, overclad and filled
FIGURE5-12: NOISEREDUCTIONSPECTRA BETWEEN TL14-170 AND TL14-167
However the difference between the two curves TL14-167 and TL14-170 in Figure 5-12 is not as significant
as those seen in FIGURE 5-11.
It is not clear why there is not a consistent and significant difference throughout the entire frequency
range when the mass and damping modification between the two vertical mullion conditions is disparate
or why the curves flatten between 1000 Hz to 4000 Hz.
These limitations in improvement and minimal change in the frequency regime strongly indicates that
mechanism limiting the sound isolation performance is at the horizontal mullions and glass.
0
10
20
30
40
50
60
70
80
90
100
63 80 100 125 160 200 250 315 400 500 630 800 1k 1.25k 1.6k 2k 2.5k 3.15k 4k 5k
Noise Reduction (dB)
One-Third Octave Band Center Frequency (Hz)
NR COMPARISON
CONNCECTED CENTER VERTICAL MULLION
TL14-170 (EXPOSED/HOLLOW) TL14-168 (FILLED) TL14-167 (OVERCLAD/FILLED)
TL14-170 (STC 32, NIC 47) TL14-168 (STC 37, NIC 50) TL14-167 (STC 42, NIC 55)
Overclad/Filled
Filled
Exposed/Hollow
183
5.2.3.2 VERTICAL MULLION
Although the NR comparison seen in FIGURE 5-11 and FIGURE 5-12 provide an indication of performance
limitations, it should be noted that the vertical mullion still has a significant contribution to the sound
isolation despite the degradation of flanking at the horizontal mullions and glass. .
As noted in Table 5-1, a 10 dB STC difference exists between lower (TL14-170, STC 32) and higher (TL14-
167, STC 42) connected mullions in Phase 3. A 10dB difference is acoustically significant, however based
on the horizontal flanking limitations understood from the previous section, it is questioned if the amount
of materials used in the TL14-167 test (i.e. overclad and fill material) has diminishing returns.
Performance observations from Phase 1 and Phase 2B indicate that high performances can be achieved
with minimal materials, i.e. overclad mullions generally outperform filled mullions. Therefore the center
mullion used in TL14-167 may have performed equally as well with less materials, for example, without
the infill of MLV pillows.
A balance between materials and performance may be considered with TL13-316 and TL14-168 (Table
5-3). There is a 1 dB STC difference in performance between the modified mullion without (TL13-316) and
with (TL14-168) the glass curtain wall and horizontal mullions. This indicates that the physical
modifications made to the unconnected mullion are appropriate to the achievable performance obtained
when it is connected to the curtain wall system.
Unconnected Mullion Connected Mullion
PERFORMANCE
DIFFERENCE
(STC)
TL13-316
STC 38
TL14-168
STC 37
- 1 dB
Table 5-3: TL13-316 and TL14-168
An overclad-only system (without a mullion infill) was not tested during Phase 3. However it may be
possible to extrapolate the performance of the system analytically based on the performance of an
unconnected overclad-only mullion (FIGURE 5-13) and the discrete frequency trends seen in FIGURE 5-12.
This analytical extrapolation may inform if the overclad-only version of an unconnected mullion has a
commensurate lateral sound isolation performance if it were connected to the overall curtain wall system.
The unconnected mullion specimen selected for extrapolation is TL13-325 from Phase 1 Class C2 (FIGURE
5-13). The specimen assembly consists of the unitized vertical mullion and an overclad of 5/8” gypsum
wall board. The specimen performance obtained in Phase 1 is STC 42.
184
FIGURE5-13: UNCONNECTEDMULLION WITH GYPSUM BOARD OVERCLAD, TL13-325, STC 42 5/8”
The extrapolated adjustments to the TL13-325 TL curve are plotted (FIGURE 5-14). The adjustments and
assumptions to create the extrapolated STC 41 curve are described below:
A. From 63 Hz to 160 Hz the TL from TL14-168 was directly applied.
In this frequency range the performance between TL13-325 and TL13-316 are almost identical.
Therefore the TL values from TL14-168 were used based on the assumption that TL13-325 will
behave similar to TL13-316 if it were connected to the curtain wall.
B. From 200 Hz – 800 Hz the TL from TL14-168 was increased by the TL difference between TL13-325
and TL13-316.
In this frequency range the TL13-325 measurement has a higher Transmission Loss than the TL13-
316 between 200 Hz and 800 Hz. Therefore the TL extrapolation applied the difference between
these two spectra and adds it to the performance of TL14-168.
C. From 1000 Hz – 5000 Hz the TL from TL14-167 was increased based on an average between TL14-167
and TL14-168.
All phase 3 tests had limited transmission loss performances in this frequency range. The Phase 3
test TL14-167 with the heavy mullion performed lower than the Phase 3 test TL14-168, with a
filled cavity (no overclad). Therefore the TL extrapolation used was the median between the two
curves.
Unitized Mullion Description
Unconnected
Mullion
Connected
Mullion
PERFORMANCE
DIFFERENCE (STC)
(Overclad/Filled Mullion)
TL13-323
STC 52
TL14-167
STC 42
- 10 dB
(Overclad Mullion)
TL13-325
STC 42
(Extrapolated)
STC 41
(Extrapolated)
- 1 dB
TABLE5-4: SUMMARY OF STC EXTRAPOLATION BETWEEN UNCONNECTED AND CONNECTED MULLIONS
The extrapolated transmission loss of TL13-325 is predicted at STC 41, if it were connected to the curtain
wall (FIGURE 5-14).
185
FIGURE5-14: TRANSMISSIONLOSSEXTRAPOLATION OF TL13-325 IF TESTED WITH THE CURTAIN WALL SYSTEM
0
10
20
30
40
50
60
70
80
90
100
63 80 100 125 160 200 250 315 400 500 630 800 1000 1250 1600 2000 2500 3150 4000 5000
Transmission Loss (dB)
One-Third Octave Band Center Frequency (Hz)
Trasmission Loss of an Unconnected Overclad Mullion (TL13-325 )and
a Connected Version of an Overclad Mullion (Extrapolated )
Unconnected Overclad Mullion (TL13-325, STC 42) Connected Overclad Mullion (Extrapolated STC 41)
186
5.3 COMPOSITE TRANSMISSION LOSS PREDICTIONS
This section uses the UVM (Unitized Vertical Mullion) test results in composite sound isolation predictions
with an internal wall partitions. This analysis will evaluate where diminishing returns occur between
material construction and acoustic performance. Additionally, the analysis will inform how the mullion
influences the overall STC rating. The composite transmission loss predictions apply TL results from UVM
Phases 1, 2B, and 3.
Two primary prediction combinations were conducted with the area of a high performing wall and area
of a UVM element:
1. Low STC performing UVM element with a high performing demising wall
2. High STC performing UVM elements with a high performing demising wall
The composite STC from these combinations will indicate the dB reduction that may be expected when a
curtain wall element is attached to a robust demising partition.
5.3.1 CALCULATION VARIABLES AND DESCRIPTIONS
Figure 5-15 diagrams the location of elements used in the composite calculations with respective surface
areas. Composite calculations assume the surface area in elevation facing the wall. The greatest surface
area is the demising wall, and each of the curtain wall elements is significantly smaller.
FIGURE5-15: IDENTIFICATION OF ELEMENTS USED FOR THE COMPOSITE CALCULATIONS AND ASSOCIATED SURFACE AREAS
[1] mullion:: 0.26 m
2
(2.8ft
2
)
[2] connection:: 0.25m
2
(2.7 ft
2
)
[4] demising partition:: 12.5 m
2
(135 ft
2
)
[3] glass and curtain wall frame:: 0.27 m
2
(2.92 ft
2
)
187
[1] Mullion
The TL performance used for the mullion element is taken from the Phase 1 testing specimens MC1 (TL13-
311) and MC2 (TL13-323). The surface area assumed for the face of the vertical extrusion is 0.26 m
2
(2.8ft
2
).
[1] MULLION: PH1-MC1
Low Performing
[1] MULLION: PH1-MC2
High Performing
FIGURE5-16: TL13-311,STC36 FIGURE5-17: TL13-323,STC52
[2] Connection
Two connection elements were selected from Phase 2B, TL13-621 STC 37 and TL13-622 STC 51. These
specimens were selected because they are comparable in performance to mullions MC1 and MC2
respectively.
The face area of the connection elements is 0.25 m
2
(2.7 ft
2
).
[2] CONNECTION: PH2B
Low Performing
[2] CONNECTION: PH2B
High Performing
FIGURE5-18: TL13-621,STC37,STAGGERED
PLATE WITHOUT SEALS
FIGURE5-19: TL13-622,STC51, STAGGEREDPLATE
WITHOUTSEALS
[3] Glass
The glass element is a contruction composite of curtain wall glazing and the aluminum perimeter mullion
frame. The transmission loss performances used in the composite calculations are taken from the Phase
3 testing, specifically TL14-167 and TL14-170. These represent the highest and lowest performances in
188
Phase 3. Mullions from Phase 1 are not combined with Phase 3 tests to predict a composite transmission
loss.
The surface area assumed is 0.3 m
2
(2.92 ft
2
) including the mullion face area and thickness of the glass.
[3] GLASS: PH3-MC1
Low Performing
[1] GLASS: PH3-MC2
High Performing
FIGURE5-20: TL14-170,STC37PLAN DRAWING OF
UVM HOLLOW AND EXPOSED
FIGURE5-21: TL14-167,STC42PLAN DRAWING
OF UVM FILLED AND OVERCLAD
[4] Partition
The performance of the demising partition is taken from the laboratory tests at the NRC Institute for
Research in Construction in Canada
69
, TL93-302 (STC 64). This is considered a high performing wall. The
wall face area assumed is 12.5 m
2
(135 ft
2
). The wall assembly consists of two layers of 16mm (5/8”)
gypsum board at either side of a double row of steels studs each 65mm (2-1/2”) wide, two layers of batt
insulation in the air cavity, and 16mm clearance between studs.
69
Ibid.
189
FIGURE5-22: NRC-CNRC LABORATORY TEST TL-93-302 PERFORMANCE IN OCTAVE BAND CENTER FREQUENCIES
5.3.2 COMPOSITE TL WITH LOW PERFORMING UVM ELEMENTS (WITHOUT GLASS)
The composite transmission Loss for item 2 in Table 5-5, where low performing curtain wall elements are
applied to the high performing wall is described (Figure 5-23). No curtain wall glass or frame is included
with this prediction.
190
FIGURE5-23: COMPOSITEPREDICTION OF A LOW PERFORMING CURTAIN WALL JUNCTION
These results indicate the ineffectiveness of attempting to terminate a double stud partition, common in
residential design into a common vertical mullion with no overclad or fill.
0
20
40
60
80
100
63
80
100
125
160
200
250
315
400
500
630
800
1k
1.25k
1.6k
2k
2.5k
3.15k
4k
5k
Transmission Loss, dB
1/3 Octave Band Center Frequencies, Hz
Composite Transmission Loss
Low Performing UVM with High Performing Partition (no curtain wall)
STC 50 Composite
(Mullion+Connection+Wall)
STC 64 Wall, NRCC TL93-302
191
5.3.3 COMPOSITE TL WITH LOW PERFORMING UVM ELEMENTS (WITH GLASS)
The composite transmission loss for item 4 in Table 5-5, where the low performing curtain wall system
(including the glass) is applied to the high performing wall (Figure 5-24).
FIGURE5-24: COMPOSITEPREDICTION OF A LOW PERFORMING CURTAIN WALL SYSTEM WITH GLASS
The composite transmission loss performance of the two elements results in STC 48, a 16 dB reduction
from the highest achievable partition element.
.
0
20
40
60
80
100
63
80
100
125
160
200
250
315
400
500
630
800
1k
1.25k
1.6k
2k
2.5k
3.15k
4k
5k
Transmission Loss, dB
1/3 Octave Band Center Frequencies, Hz
Composite Transmission Loss
Low Performing UVM System with High Performing Partition (with glass)
STC 48 Composite (CWglass
system+Connection+Wall)
STC 64 Wall, NRCC TL93-302
192
5.3.4 COMPOSITE TL WITH HIGH PERFORMING UVM ELEMENTS (WITHOUT GLASS)
The composite transmission loss for item 2 in Table 5-6, where high performing curtain wall elements are
applied to the high performing wall is described (Figure 5-25). No curtain wall glass or frame is included
with this prediction.
FIGURE5-25: COMPOSITEPREDICTION OF A HIGH PERFORMING CURTAIN WALL JUNCTION
The composite Transmission Loss performance of the three elements results in STC 631, a 3dB reduction
from the highest achievable partition element.
There is very little change when the highest achievable elements tested from the UVM method are added
to a high performing wall.
In the subsequent composite prediction the influence of the glass and horizontal mullions will be included.
0
20
40
60
80
100
63
80
100
125
160
200
250
315
400
500
630
800
1k
1.25k
1.6k
2k
2.5k
3.15k
4k
5k
Transmission Loss, dB
1/3 Octave Band Center Frequencies, Hz
Composite Transmission Loss
High Performing UVM with High Performing Partition (no curtain wall)
STC 61 Composite
(Mullion+Connection+Wall)
STC 64 Wall, NRCC TL93-302
193
5.3.5 COMPOSITE TL WITH HIGH PERFORMING UVM ELEMENTS (WITH GLASS)
The composite transmission loss for item 4 in Table 5-6, where the high performing curtain wall system
(including the glass) is applied to the high performing wall is described (Figure 5-26).
FIGURE5-26: COMPOSITEPREDICTION OF A HIGH PERFORMING CURTAIN WALL SYSTEM WITH GLASS
The composite transmission loss performance of the two elements (i.e. connected mullion [3] and concept
connection [2] ) with the high performing wall results in STC 57, a 7 dB reduction from the highest
achievable partition element.
This indicates value in acoustically modifying the vertical mullion and connection.
0
20
40
60
80
100
000 000000 0000000 0000
Transmission Loss, dB
1/3 Octave Band Center Frequencies, Hz
Composite Transmission Loss
High Performing UVM System with High Performing Partition (with glass)
STC 57 Composite (CWglass
system+Connection+Wall)
STC 64 Wall, NRCC TL93-302
194
5.3.6 SUMMARY OF COMPOSITE TL ANALYSIS
Table 5-5 and Table 5-6 summarize the overall composite STC predictions.
The composite STC reduction compared to the wall STC is identified in the last column of each table.
Low Performing UVM System Summary
LOW PERFORMANCE UVM
[4]
WALL COMPOSITE STC
#
[1]
MULLION
[2]
CONNECTION
[3]
GLASS MC1
REDUCTION
FROM [4]
1
STC 36 -- -- STC 64 STC 53
-11
2
STC 36 STC 37 -- STC 64 STC 50
-14
3
-- -- STC 32 STC 64 STC 49
-15
4
-- STC 37 STC 32 STC 64 STC 48
-16
TABLE5-5: SUMMARY OF PREDICTED COMPOSITE TRANSMISSION LOSS
There is a 2dB difference between the composite STC of low performing UVM elements and a high
performing wall, with or without glass, i.e. STC 50 and STC 48. This is not considered a significant change,
and therefore the curtain wall glazing has less of an impact on poor performing curtain wall elements.
The composite STC results are 11dB to 16dB points less than the STC performance of a high performing
demising wall. This indicates significant privacy reductions when attaching a lightweight curtain wall
system to a heavy partition.
High Performing UVM System Summary
HIGH PERFORMANCE UVM
[4]
WALL COMPOSITE STC
#
[1]
MULLION
[2]
CONNECTION
[3]
GLASS MC1
REDUCTION
FROM [4]
1
STC 52 -- -- STC 64 STC 62
-2
2
STC 52 STC 51
--
STC 64 STC 61
-3
3
-- -- STC 42 STC 64 STC 57
-7
4
-- STC 51 STC 42 STC 64 STC 57
-7
TABLE5-6: SUMMARY OF PREDICTED COMPOSITE TRANSMISSION LOSS
There is a 4dB difference between the composite STC of high performing UVM elements and a high
performing wall, with or without glass, i.e. STC 61 and STC 57. This is a significant change and therefore
the curtain wall glazing has more of an impact on high performing curtain wall elements.
195
The composite STC results are 2dB to 7dB STC points less than the STC performance of a high performing
demising wall. This indicates that despite the significant modifications to the curtain wall system, sound
flanking paths possibly at the sill and transom reduce the sound isolation performance.
When applying the curtain wall performances to a composite, the achievable TL or NR between adjacent
spaces will be limited by the glass infill and horizontal mullions.
196
5.4 RANKING RELATIVE PERFORMANCE
Many modifications to test elements during the UVM test phases revealed relative changes in sound
isolation performance that can be taken from the laboratory and applied in practice.
In Phase 1, significant building modifications were made to the exposed face and internal cavity
of the unconnected vertical mullion.
In Phase 2, acoustic concept connections between a mullion and interior demising wall were
designed to represent possible façade deflection and seal conditions in practice. Various mass and
damping materials were used to create acoustically sealed and unsealed test specimens.
In Phase 3, modifications were strictly applied to the center vertical mullion.
FIGURE5-27: STC SUMMARY ACROSS ALL LABORATORY TESTS
The table below provides initial thoughts for future development to rank indicative mullion modifications
relative to a baseline. The work may be further developed to create a foundation to characterize and
further analyze physical variables associated with the mullion design and construction.
0
20
40
60
80
100
0 10 20 30 40 50 60 70
Sound Transmission Class (STC) Rating
UVM Test Number
STC Summary of Unitized Vertical Mullion (UVM) Test Method
PHASE 1
PHASE 2B
PHASE 2A PHASE 3
197
198
FIGURE5-28: PRELIMINARYRANKING OF UVM TEST ELEMENTS
199
5.5 ANALYSIS SUMMARY
Comparisons between Phases
Direct comparisons of tests between phases 1 and 2 show that
At the low frequency region, generally below 250 Hz, the curtain wall has a greater performance
degradation than the unconnected vertical mullion due to the sound energy vibrating the glass
and horizontal mullions that transfers to the receiving room.
The vertical mullion is not the weak point (see 3 graphs) with connected mullion conditions. Based
on the observed trends and frequency correlations, there is an indication that the glass infill and
horizontal mullions are the weak links.
Composite Transmission Loss Predictions
When applying the curtain wall performances to a composite, the achievable TL or NR between adjacent
spaces will be limited by the glass infill and horizontal mullions.
The connection elements from Phase 2b do not control the sound performance rating. These connection
elements can be controlled and tuned to perform as well as the mullion.
Objectives of the Hypothesis
Sound transmission loss testing of individual and composite architectural elements comprised of and
associated with the intersection of the unitized vertical mullion reveals sound flanking path
mechanisms controlling the overall sound isolation performance.
This work was designed to reveal the sound flanking path mechanisms controlling the overall sound
isolation performance; this objective has been satisfied by this analysis. The glass infill and horizontal
mullions have impacted specific regions of the frequency regime of different UVM test specimens and
therefore reduce the overall performance of the sound isolation rating.
Sound paths at the glass and horizontal mullions at the source room transmit sound energy into the
connected mullion and this subsequently transfers to the receiving room.
This generates an interesting future study to overclad the horizontal mullions and dampen the glazing at
the source room for a laboratory measurement test. Enclosing the horizontal mullions and glazing would
limit the acoustic energy incident on the specimen to the mullion. Reradiated energy contributions from
the glass and horizontal mullions would be limited at the receiving chamber. Therefore the amount of
residual energy in the receiving room would primarily be a result of the exposed vertical mullion
contribution.
200
CHAPTER 6 CONCLUSION
HYPOTHESIS:
Sound transmission loss testing of individual and composite architectural elements comprised of and
associated with the intersection of the unitized vertical mullion reveals sound flanking path mechanisms
controlling the overall sound isolation performance.
6.1 INTRODUCTION
The lateral transmission loss performance of connected and unconnected curtain wall mullions was
investigated through acoustic laboratory tests called the Unitized Vertical Mullion (UVM) test method.
The impetus for this investigation relates to sound flanking transmission at glass curtain wall façade
systems that currently influence construction and design building practices.
Lateral sound flanking paths occurring at the curtain wall system and partition interconnections were
identified and the sound isolation reduction at high STC rated demising partitions was investigated. The
composite architectural components of the curtain wall façade work dynamically together to influence
the lateral sound isolation performance between adjacencies, although certain elements of the composite
may transfer sound paths more efficiently than others. The research investigation aimed to understand
the independent performance of select curtain wall elements associated with defined sound paths and
identify architectural mechanisms influencing the overall sound isolation performance.
Four unique laboratory test phases were conducted to measure the lateral sound transmission at the
vertical mullion and associated architectural components. The research objectives were designed to
support the sound flanking investigation and construction mechanisms controlling the overall sound
isolation performance between spaces sharing a common mullion. Conclusions for each of the following
research objective are defined in this chapter:
1. Identify curtain wall mullion practices and procedures.
2. Develop a test experiment designed to measure the unitized vertical mullion and associated
components.
3. Identify controlling sound paths at the unitized vertical mullion from the measurement results.
4. Apply the measurement results to predictive composite transmission loss calculations and
determine impacts between the vertical mullion and interconnecting walls.
Methods to support the research objectives led to the following final conclusions:
It was possible to remove the influence of a demising partition to isolate the dominant horizontal
sound transmission path of the test elements. The test method revealed the sound isolation
performance of individual (flanking) elements.
201
Some design solutions were substantially more effective to improve sound isolation performance,
e.g. overcladding mullions versus filling mullion cavities.
The primary acoustic mechanism of energy transfer is the vibrational excitation of the horizontal
mullion and glazing by the common unitized vertical mullion. The interaction of these two
elements is important and will be the subject of additional and future work. The dynamics of the
glass and mullion are coupled; sound incident on the glass displaces as a membrane which applies
bending at the boundary condition (mullion) and excites the membrane at the opposite side.
The test measurements and data show that the curtain wall glazing and horizontal mullions are
the controlling the sound paths as demonstrated through the analysis. The glass is a dominant
source due to the larger radiating surface area at the receiving room.
The one-third octave band sound transmission analysis indicated that the lateral sound paths (i.e.
at the glazing and horizontal mullions) limit the overall sound flanking isolation of the curtain wall
system at specific frequency regions.
The composite TL analysis indicated significant value in modifying the unitized vertical mullion,
although the overall performance is limited by the glazing and horizontal mullions.
This indicates that the vertical mullion is a mechanism which highly controls the passage of
airborne sound transmission across the curtain wall system and can significantly influence the
sound isolation rating.
Details of these conclusions are described in the following sections of this chapter.
6.2 PRACTICES AND PROCEDURES
The first objective investigated various areas relevant to sound flanking transmission in research and
practice: global test methods, design and manufactured methods, and precedent investigations and case
studies.
Conclusions from the background research reveals that the laboratory and field test methodologies are
limited with regards to identifying dominating paths for sound flanking transmission. Laboratory testing
procedures for sound flanking transmission is not common in the US. The ASTM E90 procedure for
obtaining a STC performance accounts for the overall sound radiating surface area of a building element
which includes acoustic mechanisms influencing sound flanking transmission. Similarly field performance
ratings (NIC or FSTC) for sound isolation in accordance with ASTM E336 accounts for the radiating surfaces
including the composite performance of an interconnecting wall. Since most of the measurement
conditions include composite elements such as wall partitions, this also limits how to approach the
improvement for the curtain wall design.
Manufacturer solutions are generally limited to product resolutions at the vertical mullion. They do not
take into account other defined sound paths at the curtain wall system, such as at the partitions
connections, glass, and horizontal mullion.
202
Precedent case studies of laboratory or field measurements on curtain walls are typically conducted as
composites with the exception of the LA LIVE
70
case study project by Enclos Corp. The Unitized Vertical
Mullion (UVM) testing methodology expands upon the Enclos precedent to support the second objective
of this research study.
6.3 TEST METHODOLOGY FOR UNITIZED MULLIONS
The second objective was achieved by designing a laboratory test procedure (Unitized Vertical Mullion -
UVM method). This method was developed to first measure the TL of individual components (vertical
mullion and connectors) independently and separately from the curtain wall. Measurements were
conducted in the absence of a demising wall assembly. The WEAL filler wall is rated STC 74. The high filler
wall rating removes the influence of a composite interconnecting wall so that the UVM unconnected and
connected elements may be measured independently. The unique test procedure included a consistent
and controlled approach to measure the unitized vertical mullion with and without the glazing and
horizontal mullion elements.
The results of these tests provide sound transmission loss data for individual mullion modifications, and
thus they may be compared to the sound flanking measurement of the curtain wall.
The sound flanking curtain wall measurement included a unique test set up designed specifically to
enclose the full size curtain wall bays. This test chamber assembly in Phase 3 was an effective way to
target the transmission loss performance in the absence of other sound flanking variables that normally
exist in a building.
Overall significant conclusions and contributions based on the experiment design of the UVM test method
are listed:
Test Method
The experiment design is a unique method to measure the individual elements of a curtain wall
system.
The lateral sound flanking transmission loss measurement of a full-scale curtain wall specimen is
unprecedented for a two-chamber laboratory that includes the construction of semi-anechoic
chambers to enclose the curtain wall bays.
All test specimens were measured in the absence of a composite wall and perimeter seal and
mounting conditions were uniform to develop relative comparisons.
Phase 1 (Unconnected Mullions)
A broad range of mullion performances (with variations of mass and damping) have been
collected for relative comparison:
The highest unconnected mullion performance achieved was STC 52.
The lowest unconnected mullion performance achieved was STC 36.
70
Dehghanyar et al., “Inter-Story Acoustical Evaluation of Unitized Curtain Wall Systems.”
203
An overclad-only modification at the mullion significantly improves performance over a mullion
cavity infill with no overclad.
Gypsum board overcladding is more effective than aluminum overcladding.
Phase 2a and 2b (Partition Connections)
The highest performing connection without a mullion was STC 51.
The highest performing partition connections or seals are comparable to or outperform
unconnected mullions. Therefore the intersection between the demising partition and mullion is
not necessarily the component controlling the sound isolation performance of the system.
The acoustic detailing of edge seals (e.g. material and gap size) can significantly influence
performance.
Bead seals and mineral wool fill can significantly influence the transmission loss performance of
the connection elements (with no mullion). Sealed air tight conditions without batt infill
performed higher than conditions with batt-filled cavities and no bead seals.
Parallel aluminum plate conditions reveal that the profile of the aluminum mullion extrusion may
influence the achievable isolation provided by the mullion even though interstitial leg connections
are connected by a resilient gasket.
The parallel aluminum plate condition performed higher than the hollow and exposed mullion.
This indicates that the interstitial leg connections that exists within the mullion profile is
influencing the achievable sound isolation.
Gypsum board connection configurations are more effective than those with aluminum.
Phase 3 (Connected Mullions)
Sound paths at the curtain wall glazing and the horizontal mullions are significantly impacting the
overall sound isolation performance.
Modifications at the mullion significantly improve performance; however overall performance is
limited by the horizontal mullion and glazing.
The greatest depreciation in performance due to sound flanking at the center vertical mullion was
10dB STC points: STC 52 (unconnected mullion) to STC 42 (connected mullion condition).
The highest connected mullion performance achieved STC 42.
The highest connected mullion performance achieved STC 32.
The results from the modifications at each test phase provide information on what architectural
mechanisms of the curtain wall are controlling the overall sound isolation performance.
An additional value from the test measurement series is its potential application in the profession to
improve curtain wall mullions. The test series enables designers to make comparisons between different
modifications and not necessarily take the face value performance rating. The various architectural
enhancements conducted in the empirical testing reveal relative changes that can be taken from the
laboratory and applied in practice to guide designers of relative improvements.
The data collected to satisfy objective 2 of this research study is applied to two different analysis methods
to meet objectives 3 and 4.
204
6.4 CONTROLLING ELEMENTS AT THE UNITIZED VERTICAL MULLION
The method to support the third objective was to evaluate the transmission loss and noise reduction
results of the connected and unconnected conditions of the UVM test method. The summary of STC
performance ranges for each UVM test phase are listed below:
Phase 1 Unconnected Mullion (without glazing) STC 36 – STC 52
Phase 2a Unconnected Mullion + Partition Connection (without glazing) STC 22 – STC 41
Phase 2b Partition Connection (without glazing) STC 25 – STC 51
Phase 3 Connected Mullion (with glazing) STC 36 – STC 42
The sound transmission loss performance is limited to STC 42 where the unitized mullion is connected to
the curtain wall system. It may be concluded that the highest performing elements tested in Phase 1 and
Phase 2 are not controlling the overall STC performance in the main curtain wall system because they are
capable of achieving such high performances in their respective phases.
Additionally, comparison of the Phase 3 frequency spectra revealed a limit to the transmission loss
improvement between 1000 Hz – 3150Hz (FIGURE 6-1).
FIGURE 6-1: LIMITED SPECTRUM WITH CURTAIN WALL DESIGN
Based on the comparative analysis of Phase 3, glass infill may not be the only limiting factor, as previously
believed. Sound paths created at the horizontal mullion are critical variables influencing the transmission
loss.
This opens opportunities for future test experiments to determine performance characteristics of the
upper and lower horizontal mullions.
0
20
40
60
80
100
Transmission Loss (dB)
One-third Octave Band Center Frequency (Hz)
Phase 3 Transmission Loss Results of Curtain Wall System
(TL14-167, STC 42) (TL14-168, STC 37) (TL14-170, STC 32)
TL14-167 TL14-168 TL14-170
Limited improvement
in this spectral region
205
Based on the overlay of noise reduction curves in FIGURE 6-2 the curtain wall system significantly changed
the results of the unconnected mullion tests. Therefore the introduction of the glazing and horizontal
mullions have a significant impact. The behavior of some of the frequency regions are not yet explained,
and further study on these specific regions is necessary.
0
20
40
60
80
100
Noise Reduction (dB)
One-Third Octave Band Center Frequency (Hz)
NR Comparison
Unconnected and Connected Center Vertical Mullion
Highest and Lowest Performing
TL14-167 (STC 42, NR 55) TL14-170 (STC 32, NR 47)
TL13-311 (STC 36, NR 49) TL13-323 (STC 52, NR 64)
Without Glass
With Glass
206
FIGURE 6-2: LIMITED SPECTRUM WITH CURTAIN WALL DESIGN
6.4.1 IMPROVEMENTS AT THE UNCONNECTED UNITIZED MULLION
It was concluded that overclad design modifications to the vertical mullion perform better that filling the
internal air cavity. Relative improvements for the unconnected unitized mullion condition are tabulated
(Table 6-1).
MULLION MULLION DESCRIPTION STC
RELATIVE STC
IMPROVEMENT FROM
BASELINE
MC1: (3) TL13-311 Exposed and Hollow, no modification 36 BASELINE
MC1a: (7) TL13-315 Filled with mineral fiber 36 +0
MC1b: (17) TL13-325 Overclad with 5/8” gypsum board 42 +6
MC1c: (10) TL13-318 Overclad with 1/8” Alum + 3/16” MLV 48 +12
MC1d: (13) TL13-321 Overclad with 5/8” gypsum + 3/16” MLV 50 +14
MC-1 MC-1a MC-1b MC-1c MC-1d
Table 6-1: Relative STC improvements for various mullion retrofit options
207
6.4.2 IMPROVEMENTS AT THE CONNECTED UNITIZED MULLION
Relative improvements for the connected unitized mullion condition are tabulated (Table 6-2).
MULLION VERTICAL MULLION DESCRIPTION STC
RELATIVE STC
IMPROVEMENT FROM
BASELINE
Typical Curtain wall system
(TL14-170)
Exposed and Hollow, no
modification
32
BASELINE
Curtain wall system with a
filled mullion (TL14-168)
Filled with MLV Pillows (Limp Mass
Vinyl/Mineral Fiber)
37
+5
Curtain wall system with a
filled and overclad mullion
(TL14-167)
Overclad with 5/8” gypsum board +
limp mass vinyl plate and filled with
MLV Pillows
42
+10
TABLE6-2: RELATIVESTC IMPROVEMENTS FOR VARIOUS MULLION RETROFIT OPTIONS
6.5 COMPOSITE TRANSMISSION LOSS PERFORMANCE
Methods used to fulfill the fourth objective used the composite TL equation to determine the impact of
a demising partition. This objective returns to the issue in architectural acoustics where an acoustically
weak link impacts a high performing element when joined as part of a composite system. In this case,
the method targets the performances obtained from objective 2 and applied to a demising partition.
Table 6-3 provides the summary of impacts when elements from the UVM test method are joined with a
high performing demising wall.
UVM PERFORMANCE
WALL
RATING COMPOSITE RATING
# MULLION CONNECTION
CURTAIN WALL
(GLASS)
1
DELTA FROM
WALL STC
1
STC 36 STC 37 -- STC 64 STC 50
14 dB
2
STC 37 STC 32 STC 64 STC 48
16 dB
3
STC 52 STC 51 STC 64 STC 61
3 dB
4
STC 51 STC 42 STC 64 STC 57
7 dB
TABLE6-3: SUMMARY OF PREDICTED COMPOSITE TL PERFORMANCES
1
Sound flanking present
208
The composite rating based on the assembly from line #1 and #2 in Table 6-3 shows a 14 dB STC – 16 dB
STC point depreciation to the STC 64 partition. This confirms the ineffectiveness of terminating high
performing partition into a common vertical mullion with no modification (i.e. overclad or fill).
The composite rating based on the assembly from line #3 and #4 in Table 6-3 shows a 3dB – 7 dB STC
point depreciation to the STC 64 partition. This lower range in performance reduction is similar to what
is typically measured as an NIC the field and therefore is in line with in situ performance expectations of
a high performance partition.
The deltas listed in Table 6-3 are indicative for a high performing interconnecting wall and will be different
with a demising partition of a lower performance, ie. STC 50 partition.
A conclusion from the composite TL analysis reveals that there is a benefit to modifying the center vertical
mullion, but the overall performance of the system is limited by the sound paths at the glass and horizontal
mullions.
6.6 LIMITATIONS
Laboratory tests were conducted during Phase 1 and Phase 2 with an objective to identify the highest
performing specimen based on acoustic concept designs used in practice.
More tests were conducted in these specific phases than anticipated since it was achievable to manipulate
the scale of the test rig over modest periods of time.
Phase 3 was limited to 3 test configurations. Each of these test rigs in this phase required significant
manpower, machinery (e.g. forklift), time, and construction; therefore investigations were limited to
rating the highest and lowest performing mullions. Ideally more tests would have been conducted with
modifications to the horizontal mullions; however this may be conducted for the future laboratory test
investigations to better understand limitations at the glass and vertical mullion.
209
CHAPTER 7 FUTURE WORK
7.1 FUTURE TESTING AND DESIGN INVESTIGATIONS
Unexpected and esoteric findings from the Unitized Vertical Mullion (UVM) test method have led to ideas
for future investigations for sound and vibration test measurements, ideas for mullion design concepts,
and further analytical studies on some of the research findings. These are categorized in Table 7-1.
Study 1
Airborne Sound
Measurement
Advance the UVM Test Method by modifying the horizontal mullions
in additional to the vertical mullion to evaluate the glazing path.
Study 2
Airborne Sound
Measurement
Advance the UVM Test method by modifying the inboard glass lite of
the IGU assembly to compare the lateral transmission loss of
laminated versus monolithic panes.
Study 3
Vibration
Measurements
Refine and develop the initial vibration analysis conducted in
Appendix D.
Study 4
Intensity
Measurements
Conduct intensity measurements to evaluate sound energy at the
curtain wall glazing, vertical mullions, and horizontal mullions.
Study 5 Design Concepts
Develop and explore concept designs at the horizontal and vertical
mullion (stack joint) to resilient decouple elements.
Study 6 Design Concepts
Develop and explore concept design within the vertical mullion
cavity where the interstitial “leg” extrusions connection both sides
of the unitized parts.
Study 7 Analytic Study
Further develop analysis of the unexplained transmission loss
depreciation of Phase 3 specimens.
Study 8 Analytic Study
Further develop the initial ranking of relative architectural
modifications from the UVM test method (Section 5.4).
Study 9 Analytic Study
Study and identify mechanisms contributing to the discrete
resonances occurring at overclad mullion specimens and certain
gaps at the filler wall aperture.
Study 10 Analytic Study
Determine applications of the UVM test method to predictive
analysis in accordance with ISO EN standard definitions.
TABLE7-1: FUTUREWORK
7.1.1 FUTURELABORATORYTESTSMEASUREMENTS
Propose future studies #1- through #4 are measurements conducted with the laboratory test rig used in
Phase 3 of the UVM test method, including the semi-anechoic chamber enclosures to frame around the
curtain wall bays.
One of the primary findings from the UVM Phase 3 indicates that the horizontal mullion and glass are
significant sound paths. This was deduced by comparison trends seen in the plotted one-third octave band
210
frequency regimes that point to influences from the horizontal mullions or glazing. Study #1 would better
identify discrete contributions at the glazing by adding mass and damping to the horizontal and vertical
mullions. In addition to measuring influences from modifications at the aluminum frame, the glazing
should also be dampened or enclosed to isolate the influence of the vertical mullion.
The curtain wall test specimen in the UVM test method included laminate glass pane at the inboard side
of the IGU. The laminated pane will influence the flexural vibration transfer across the glass from the
source to receiving side. Study #2 would investigate the lateral transmission loss of a curtain wall system
with a monolithic insulated glazing unit.
Studies #3 would refine and improve upon the initial measurement study discussed in Appendix D. The
initial work infers that the glazing, as opposed to the vertical and horizontal mullion, is the dominating
path due its radiating surface area. Repeat measurements at the vertical and horizontal mullions and
glazing should be conducted to include an impulsive transfer function.
Study #4 would consider a sound intensity survey at the wall, glass, sill, and mullion to obtain sound
intensity levels at curtain wall elements at the receiving chamber. Both the vibration and intensity
measurements may provide better indications of the controlling elements.
7.1.2 FUTURE CURTAIN WALL DESIGN CONCEPT STUDIES
Studies #5 and #6 would further explore architectural design concepts at the stack joint and the internal
connection extrusions at the mullion. Part of the study will require background research into the global
models of curtain wall extrusion and connection typologies.
Significant sound transmission loss reductions occurred during the UVM Phase 3 testing that are believed
to be caused by the mechanical connections located at the vertical and horizontal mullion intersection
and at the “leg” extrusion that connects each side of a unitized mullion at a gasket. Exploring resilient
connection modifications may lead to an improved transmission loss across the façade.
Designs by Schüco are already developing interesting variations to the mullion connections. As an
example, the Schüco mullion Type USC 65
71
is designed to connect interstitial leg extrusions and a silicone
gasket. This seems to have promising performance characteristics, and it would be interesting to compare
this mullion element type with the mullion element used in the UVM tests.
7.1.3 FUTURE RESEARCH ANALYTICAL STUDIES
Unexplained frequency patterns were identified during the Phase 3 test measurement analysis from the
influence of the glazing and horizontal mullion elements. Study #7 would further analyze these frequency
regimes to determine the mechanisms contributing to discrete resonances and certain responses at low
frequencies.
Study #8 proposes to finalize some of the initial studies seen in Chapter 5 and create a matrix that classifies
and ranks the percentage improvement in modifying mullion elements so that designers can be informed
of the relative difference for sound isolation in practice.
71
Schüco, “Overview of Profiles for Schüco Facade USC 65.”
211
Discrete resonances were noted in all phases of the UVM test method. Deductions indicated critical
frequencies depended on natural frequency of the materials used or potentially the size of certain test
apertures. Study #9 is proposed to evaluate resonances for select specimens to reveal if corrections to
the frequency regimes are necessary. For example, a discrete resonance at 630Hz occurred at
unconnected mullion specimens with an aluminum tube overclad (Phase 1 Class C4). The tube consisted
of a 6" depth x 1-1/2" width x 60" height and an aluminum thickness of 1/8". These resonant
characteristics controlled much of the overall sound transmission performance of the specimen.
Therefore if the resonance could be corrected by means of structural stiffening, this would inform
resolutions to improve the overall sound isolation performance. Other notable resonances occurred in
Phase 2A where an aluminum overclad was placed over silicone connection elements, and the effect
reduced the overall sound transmission loss.
The boundary condition around the test specimens may have also influenced the performance of the
unconnected mullion tests. Air slots created on either side of the mullions and the chamber filler wall
were generally ¼” - ½” wide x 60-1/2” tall x 3” deep. The influence of these slits and gaps may be analyzed
by research and theories developed by Uris et al (2003)
72
and Gomperts and Kihlman (1967)
73
.
It is noted that discrete resonances occurred due to the small aperture size. This may be limiting the
overall TL of the elements. The future analytical study should include a calculation of the resonance
frequency based on the aperture dimensions at each phase. This should be compared to test specimens
with common resonance frequencies to see where this may analytically be corrected.
Study #10 would apply the ASTM E90 test measurement data in the UVM test method to standard sound
flanking prediction models in accordance with ISO 12354 and ISO 10848 so that normalized sound flanking
indices may be identified for broader applications. It would be informative to correlate the indices
between ISO and ASTM standards and normalize sound flanking transmission for the UVM test specimen.
7.2 CONCLUSION
The lateral transmission loss performance of connected and unconnected curtain wall mullions was
investigated using the Unitized Vertical Mullion (UVM) method. The original research was conducted over
approximately 80 acoustic laboratory tests to measure the performance of select curtain wall elements
measured independently and then modified to identify the highest practicable STC that may be achieved
and relatively compared. Although great progress was made in understanding how critical components
are responsible for sound flanking transmission paths, many potential future studies are possible that
would add to the field of architectural acoustics. Continuing research will enhance designers
understanding of façade tectonic cohesion specific to acoustic design integration and to inform building
engineering design and performance decisions.
72
Antonio Uris et al., “The In uence of Slits on Sound Transmission through a Lightweight Partition,” vol. 65 (Applied Acoustics,
Elsevier Ltd., 2003), 421–30.
73
M.C. Gomperts and T. Kihlman, “The Sound Transmission Loss of Circular and Slit-Shaped Apertures in Walls,” Acta Acustica
United with Acustica 18, no. 3 (1967): 144–50.
212
BIBLIOGRAPHY
2013 California Green Building Standards Code California Code of Regulations,Title 24, Part 11 (CALGreen).
International Code Council, ICC, 2013.
Andow, Kei, and Takashi Koga. “Experimental Study on Effect of Lining for Flanking Transmission of
Building Facade.” In RBA-04 - SOUND INSULATION OF MULTI-FAMILY DWELLINGS, 6. RBA-04 - SOUND
INSULATION OF MULTI-FAMILY DWELLINGS. Sevilla, Spain, 2002. http://www.sea-
acustica.es/Sevilla02/rba04002.pdf.
Antonio, Nick. “Residential Sound Insulation and Building Code.” presented at the Acoustic Society of
America, LA Chapter Meeting, Los Angeles, January 15, 2008.
Association of Interior Specialists. “Building Acoustics - Terminology.” AIS Association of Interior
Specialists. Accessed April 6, 2014. http://www.bre.co.uk/page.jsp?id=1146.
ASTM E33 Committee. “ASTM E90 - 09 Test Method for Laboratory Measurement of Airborne Sound
Transmission Loss of Building Partitions and Elements.” ASTM International, 2009.
http://www.astm.org/Standards/E90.htm.
ASTM E33 Committee.. “ASTM E336 - 11 Standard Test Method for Measurement of Airborne Sound
Attenuation between Rooms in Buildings.” ASTM International, 2011.
http://www.astm.org/Standards/E336.htm.
ASTM E33 Committee.. “ASTM E413 -10 Classification for Rating Sound Insulation.” ASTM International,
2010. http://www.astm.org/Standards/E413.htm.
Barista, Dave. “Glass Curtain Wall: Plenty of Light, but Is It Soundtight?” Building Design + Construction,
May 2, 2006. http://www.bdcnetwork.com/glass-curtain-wall-plenty-light-it-
soundtight?page=1&quicktabs_1=1.
Bies, David A., and Colin H. Hansen. Engineering Noise Control: Theory and Practice. Taylor & Francis,
2009.
“BRE Group: Acoustics Laboratory.” Accessed March 24, 2012. http://www.bre.co.uk/page.jsp?id=1146.
213
British Gypsum. “Education Sector Guide - 7 - Flanking Sound Transmission.” British Gypsum, Saint-
Gobain, March 2014. http://www.british-gypsum.com/~/media/Files/British-Gypsum/WHITE-BOOK-
Sector-Guides/WBES/WBES-7-Flanking-Sound-Transmission-04.pdf.
Cambridge, Jason Esan. “An Evaluation of Various Sound Insulation Programs and Their Use in the Design
of Silent Rooms.” Chalmers University of Technology, 2006.
Cox, Trevor. “Diffraction through a Single Slit.” Wave Diffraction. Accessed July 18, 2014.
http://www.acoustics.salford.ac.uk/feschools/waves/diffract3.php.
Crocker, Malcolm J. Handbook of Acoustics. John Wiley & Sons, 1998.
Dehghanyar, TJ. “Partition Mullions: Curtain Wall Acoustical Enhancements.” Enclos Studio, January 29,
2009.
Dehghanyar, TJ, Mic Patterson, Kevin Cole, and Michel Michno. “Inter-Story Acoustical Evaluation of
Unitized Curtain Wall Systems.” Culver City, CA: Enclos Corp, July 2008.
DuPree, Russell B., and California. Dept. of Health Services. Office of Noise Control. Catalog of STC and IIC
Ratings for Wall and Floor/ceiling Assemblies. Berkeley, California 94704: Office of Noise Control,
California Department of Health Services, 1980.
Everest, F. Alton, and Ken C. Pohlmann. Master Handbook of Acoustics. McGraw-Hill Professional, 2009.
Gomperts, M.C., and T. Kihlman. “The Sound Transmission Loss of Circular and Slit-Shaped Apertures in
Walls.” Acta Acustica United with Acustica 18, no. 3 (1967): 144–50.
Halliwell, R.E., T. R. T. Nightingale, A.C.C Warnock, and J.A. Birta. “NRC-CNRC Gypsum Board Walls:
Transmission Loss Data.” Internal Report. Canada: Institute for Research in Construction, March 1998.
http://archive.nrc-cnrc.gc.ca/obj/irc/doc/pubs/ir/ir761/ir761.pdf.
Hopkins, Carl. Sound Insulation. Elsevier / Butterworth-Heinemann, 2007.
ift Rosenheim. “Laboratory Building Acoustics.” Accessed May 14, 2015. https://www.ift-
rosenheim.de/en/labor-bauakustik.
214
ISO 12354-1:2000 Building Acoustics. Estimation of Acoustic Performance in Buildings from the
Performance of Elements. Airborne Sound Insulation between Rooms. BSI, July 15, 20000.
Kym Burgemeister. “Sound Flanking at Curtain Wall Mullions.” Arup Acoustics General Forum: Glazing and
Facades, 2009.
Laing. “Systems for Facades, aluminum/PVC-U Windows and Doors - Fittings.” Acoustic Laboratory Test.
Insulation against Noise Transmission for Facades. Germany: Institute for Window Technoogy (ift)
Rosenheim, November 13, 2000.
Lin, Man. “Chapter 6 Architectural Design, Computer Science CSci485.” 2012.
http://cse.stfx.ca/~mlin/cs485/lectures/archdesign.ppt.
Long, Marshall. Architectural Acoustics. Elsevier, 2006.
Louwers, Marc. “Improvement of Acoustical Flanking Transmission through Light-Weight Façades.” In
INTER-NOISE and NOISE-CON Congress and Conference Proceedings, 2012:1998–2003. 10. Institute of
Noise Control Engineering, 2012.
LoVerde, John, and Wayland Dong. “Methods for Reducing Flanking Airborne Noise Transmission through
Mullions of Curtain Wall Systems.” The Journal of the Acoustical Society of America 124, no. 4 (2008):
2463.
Makepeace, Chris, Brian Kyle, Anik Shooner, and Luis de Miguel. Glass and Metal Curtain Walls: Best
Practice Guide Building Technology. Public Works and Government Services Canada: Canada Mortage and
Housing Corporation, n.d. http://www.tboake.com/guides/curtain.pdf.
Martin, H.J., M.A.E. Schoffelen, and W.M. Siebesma. “Flanking Transmission along an Aluminium Façade
– Experiments vs Prediction-.” In Proceedings 17th International Congress on Acoustics Rome, Vol. 3.
Rome, Italy, 2001.
“Flanking Transmission along an Aluminium Façade – Experiments vs Prediction-.” In 18th Proceedings
International Congress on Acoustics. Kyoto, 2004.
Mehta, Madan, James Allison Johnson, and Jorge Rocafort. Architectural Acoustics: Principles and Design.
Prentice Hall, 1999.
Mommertz, Eckard. Acoustics and Sound Insulation: Principles, Planning, Examples. Birkhäuser, 2009.
215
Müller, W., and G. Vogel. Atlante Di Architettura. Milan: Hoepli, 1992.
http://en.wikipedia.org/wiki/Mullion.
Nagy, J.P. “Laboratory for Flanking Sound Transmission of Lightweight Constructions.” Civil Engineering
18, no. 3 (1974): 169–78.
Nightingale, T. R. T. “On Using Multiple Kij’s in the EN12354 Acoustics Prediction Model to Represent
Excess Attenuation in Flanking Surfaces.” Rome, Italy, 2001.
Parkin, Peter Hubert, Henry Robert Humphreys, and J.R. Cowell. Acoustics, Noise, and Buildings. Fourth
Edition. Faber and Faber, Boston MA, 1979.
Patterson, Mic. “Structural Glass Facades.” 2011.
Patterson, Mic. Structural glass facades and enclosures. John Wiley & Sons, 2011.
SaB, Bernd, and Ulrich Sieberath. “Classification Report, Sound Reduction and Flanking Transmission Loss
of Building Elements.” Acoustic Laboratory Test. Schuco Skyline S 65 F. Germany: ift Rosenheim, August
25, 2004.
Schüco. “Overview of Profiles for Schüco Facade USC 65.” Accessed May 19, 2014.
http://schilloh.2netmedia.de/pdf/produkte/fassaden_daecher/172256.pdf.
Sjökvist, Lars-Göran. “Flanking Transmission in Lightweight Buildings.” Lund University, Sweden:
Department of Engineering Acoustics, 2004. http://lup.lub.lu.se/record/929711.
Smith, B. J., R. J. Peters, and S. Owen. Acoustics and Noise Control. 2 Sub. Addison Wesley Longman, 1996.
“Sound Insulation of the Original Component.” Schüco, n.d. http://www.schueco.com/web2/de-
en/investors/technology_center/specialist_areas.
Szudrowicz, B., and A. Izewwska. “Empirical Verification of the Prediction Model Designed to Estimate the
Flanking Transmission in Buildings.” In Proceedings 17th International Congress on Acoustics Rome, Vol.
3. Rome, Italy, 2001.
Tedford, Larry. “Sound Flanking at Curtain Wall Mullions.” Arup Acoustics General Forum: Glazing and
Facades, 2009.
216
T. Estabrooks, F. King. “NRC-IRC Flanking Sound Transmission Facility.” National Research Council Canada,
no. NRCC-51390 (October 2009): 3. doi:irc_id:20490.
Uris, Antonio, JoseMaria Bravo, Jaime Llinares, and Hermelando Estelles. “The In uence of Slits on Sound
Transmission through a Lightweight Partition,” 65:421–30. Elsevier Ltd., 2003.
Vigener, Nik, and Mark Brown. “Building Envelope Design Guide - Curtain Walls.” Whole Building Design
Guide: A Program of the National Institute of Building Sciences, October 20, 2011.
http://www.wbdg.org/design/env_fenestration_cw.php.
Wamock, A.C.C, T. R. T. Nightingale, and M.R. Atif. “Estimation of Sound Transmission Class and Impact
Insulation Class Rating for Steel Framed Assemblies.” American Iron and Steel Institute / Steel Framing
Alliance, 2008.
217
APPENDIX A TERMINOLOGY
A.1 LABORATORY TEST STANDARDS
ASTM E90
ASTM E90 Test Method for Laboratory Measurement of Airborne Sound Transmission Loss of Building
Partitions and Elements
74
The laboratory test measurement procedure defines the airborne sound transmission loss of building
elements. The test element is mounted in a filler wall between two reverberant chambers which isolate
the sound source room from the sound receiving room. The test specimen is exposed to a diffuse sound
field so that the performance may be compared to other specimens in a similar sound field. The significant
path for sound transmission between the test chambers is through the specimen, which is mounted to
the chamber filler wall with the intent to remove sound flanking paths. The sound transmission loss is
based on one-third octave band center frequencies from 125 Hz to 4000 Hz. The TL is corrected for area
of the specimen size and the absorption in the receiving room. The test method is used to calculate the
single-figure STC per the ASTM 413 classification method.
ASTM E336
ASTM E336 Standard Test Method for Measurement of Airborne Sound Attenuation between Rooms in
Buildings
75
The field test measurement procedure defines the sound isolation between two spaces in a building. The
measurement includes the direct sound transmission through the separating building element and the
transmission of various sound flanking paths. The procedure measures noise reduction (NR), normalized
noise reduction (NNR) or apparent transmission loss (ATL). The corresponding single figure number rating
to these measurements is NIC, NNIC, and ASTC. One of the significant differences between the E90
laboratory method and the E336 field method is the presence of sound flanking paths.
ASTM E413
ASTM E413 Classification for Rating Sound Insulation
76
The classification method used to calculate the single-figure number ratings for laboratory or field
measurements of building elements in one-third octave bands. The calculation method covers the single
figure number classification of the following test measurement methods:
ASTM E90 laboratory test procedure for STC (Sound Transmission Class)
ASTM E336 field test procedure for field sound transmission class (FSTC), noise isolation class
(NIC), and normalized noise isolation class (NNIC).
74
E33 Committee, “ASTM E90 - 09 Test Method for Laboratory Measurement of Airborne Sound Transmission Loss of Building
Partitions and Elements,” 90.
75
E33 Committee, “ASTM E336 - 11 Standard Test Method for Measurement of Airborne Sound Attenuation between Rooms in
Buildings.”
76
E33 Committee, “ASTM E413 -10 Classification for Rating Sound Insulation.”
218
A.2 ACOUSTIC AND ARCHITECTURAL TERMS
ACOUSTICS
“Acoustics” is derived from the Greek word ‘akouein’, which means “to hear”, and is the branch of science
that deals with sound, including its generation, transmission, analysis and perception
77
.
ACOUSTIC LEAK
An acoustic leak occurs where a gap or hole in the construction occurs, whereas sound flanking is sound
transmission through building components. (See sound flanking.)
ANECHOIC CHAMBER
A room devoid of sound reflections; designed so that all sound reflections are completely absorbed.
Usually all six interior surfaces of the room are covered with special sound absorbing treatment.
ACOUSTIC PRIVACY
Obtaining acoustic privacy between adjacent spaces depends upon adequate sound isolation of the
demising partition and an appropriate level of background noise in the receiving space. The level of
acoustic privacy required for a space can be categorized by the used of the space. There are three major
components that define acoustic privacy:
the level of intrusive sound from the source
the sound isolation between source and receiver spaces
the background noise level at the receiver location
Background noise is the continuous HVAC system noise generated by fans and air velocities in ductwork
and vents which often serves as masking noise, especially to provide acoustic privacy.
AIRBORNE SOUND
Speaking or playing music in a room causes the enclosing components to vibrate. These oscillations
propagate within the construction (structure-borne noise) and are radiated in an adjacent room in the
form of airborne sound. The sound propagation takes palace not only via the separating component, but
also via the adjoining, so called flanking components
78
. Accordingly, building acoustics has to consider and
evaluate both the separating and the flanking components. Airborne and structure-borne sound
excitations often occur together. (Also see sound flanking.)
77
Mommertz, Acoustics and Sound Insulation.
78
Ibid.
219
CONNECTED MULLION
(See Unconnected Mullion) In the UVM testing phases, the “connected” mullion is described where the
center vertical mullion is connected to the curtain wall glazing and horizontal mullions.
CURTAIN WALL SYSTEMS
Non-Unitized Curtain Wall System
Non-unitized or ‘stick’ systems are curtain wall installations constructed from long vertical framing
members called mullions, or sticks, spanning across supporting floor slabs. The framing members are
shop fabricated, factory painted, and installed one piece at a time. The glass or other cladding panels
are then attached to the framing members. The system type is site labor intensive. Consequently,
stick systems have been replaced by unitized systems in many applications
79
.
FIGURE A-1: NON-UNITIZED (STICK) SYSTEM
80
Unitized Curtain Wall System
79
Mic Patterson, Structural Glass Facades and Enclosures (John Wiley & Sons, 2011), 36–37.
80
Helmut Kientz, “Shedding Light on Curtain Wall Systems,” HIXSON Architecture Engineering Interiors, n.d., http://www.hixson-
inc.com/_images/SheddingLight_Curtainwall_article0308.pdf.
220
Unitized systems are curtain wall systems where large framed units are built up under factory-controlled
conditions, shipped to the site, and the entire unit lifted and set into position. Multiple glazing panels are
typically incorporated within a single unit. Unitized systems strategically shift labor requirements from
the site to the factory, which potentially allows improved quality and greater economy, at least in areas
with high field labor rates.
81
Normally the system is designed so that there is a continuous hollow cavity within the unitized frame
system, both vertically and horizontally. This continuity creates a path for airborne and structure-borne
sound to travel.
FIGURE A- 2: UNITIZED SYSTEM
82
DECIBEL, dBA
The unit used for measuring A-weighted sound pressure level is dB(A). The A-weighting is based on the
frequency response of human hearing, where subjectively low frequency sounds do not seem as loud as
mid- or high- frequency sounds for a given sound pressure level. The A-weightings curves which plot the
81
Patterson, Structural Glass Facades and Enclosures, 38.
82
Helmut Kientz, “Shedding Light on Curtain Wall Systems,” HIXSON Architecture Engineering Interiors, n.d., http://www.hixson-
inc.com/_images/SheddingLight_Curtainwall_article0308.pdf.
221
frequency response of human hearing are incorporated into sound level meters for direct measurement
results measured in dB(A). Typical noise levels are given in the chart below that is widely used:
Noise Level dB(A) Example
130 Threshold of pain
120 Jet aircraft take-off at 100 m
110 Chain saw at 1 m
100 Inside disco
90 Heavy truck at 5 m
80 Curbside of busy street
70 Loud radio (in typical domestic room)
60 Office or restaurant
50 Domestic fan heater at 1m
40 Living room
30 Theatre
20 Remote countryside on still night
10 Sound insulated test chamber
0 Threshold of hearing
TABLE A- 1: DECIBELS
EQUIVALENT CONTINUOUS SOUND LEVEL (LEQ)
The equivalent noise level, L
eq
, is the sound pressure level of a steady sound that has, over a given period,
the same energy as the fluctuating sound in euqestion. It is an average and is measured in dB(A).
83
FLANKING SOUND TRANSMISSION
Sound transmitted through flanking paths propagates acoustic vibration through a continuous structural
component. In architecture, this typically occurs at a structural connection between two adjacent rooms
which is rigid enough to transmit sound energy. Examples of sound flanking paths at a partition separating
two spaces are at the ceiling, floor or intersecting walls if they are seamlessly connected between rooms.
Specific conduits of sound flanking are rigid connections found at ducts, plumbing piping, electrical
conduit, openings, structural elements, window mullions, etc.
Sound flanking paths reduce the sound transmission integrity of a partition because it circumvents a wall
or floor between two spaces by way of an independent structural path. As a result the sound isolation
performance of the demising wall is compromised.
83
Smith, Peters, and Owen, Acoustics and Noise Control.
222
FIGURE A- 3: FLANKING PATH DIAGRAM (MOMMERTZ & MULLER, 2008)
The diagram in Figure A- 3: shows common paths of flanking. The red arrows represent the
direction of sound energy propagating from the source to receiver rooms:
84
a. Location of the source room
b. Location of the receiving room
c. The demising partition between the source and receiving room
d. Partitions which are a continuous structural element between source and receiver rooms
FREQUENCY HZ
Frequency is cycles per second measured in Hertz (Hz), subjectively understood as pitch. The audible
frequency range for human hearing is typically 20Hz – 20,000Hz.
Speech Frequencies
Sound frequencies are a critical part of how humans perceive sound. Speech frequencies are the primary
sounds we hear between office spaces, classrooms, or residential units. Speech frequencies are shown
(Figure A-4). Speech frequencies are targeted for laboratory sound isolation tests for ASTM E90.
FIGURE A- 4: APPROXIMATE SOUND SPECTRA OF MALE AND FEMALE SPEECH (LOG-TERM AVERAGE)
85
(FROM MEHTA,
ARCHITECTURAL ACOUSTICS PRINCIPLES AND DESIGN)
84
Eckard Mommertz, Acoustics and Sound Insulation: Principles, Planning, Examples (Birkhäuser, 2009).
85
Madan Mehta, James Allison Johnson, and Jorge Rocafort, Architectural Acoustics: Principles and Design (Prentice Hall, 1999).
223
MULLION
The aluminum extrusion that frames curtain wall glazing and runs vertically or horizontally.
NOISE
Noise is unwanted sound and provokes disturbance to human comfort and productivity. In the health
industry this can have an adverse effect on healing process.
ONE-THIRD OCTAVE BANDS
Octave bands are referred to by their center frequency and are used to analyze acoustic measurements
and calculations. Acoustics assessment is usually within the frequency range of human hearing, typically
63Hz to 8 kHz.
However, building acoustics frequency ranges are assessed in one-third octave bands (three frequency
bands per octave) typically 100 to 3150Hz, for a more detailed frequency analysis. The extended frequency
range is defined from 50 and 5000Hz one third octave bands. This is because airborne sound isolation
performs better at high frequencies than at low frequencies, therefore airborne sound above 5000Hz is
usually not a problem. It is common to describe specific trends in analyzing the spectral content of building
Transmission Loss by defining low, medium and high frequency ranges.
86
Low frequency range: 500 – 200Hz
Mid frequency range: 250- 100Hz
High frequency range: 1250 – 5000Hz
OVERCLAD
The term used to enclose or wrap a mullion with a mass or damping building material.
PARTITION
The word “partition” in this standardized test methods includes all types of walls, floors, or any other
boundaries separating two spaces. The boundaries may be permanent, operable, or movable
87
.
PERCEPTION
Auditive perception
Physical acoustic descriptions Auditive human perception
Frequency Pitch
86
Hopkins, Sound Insulation.
87
E33 Committee, ASTM E336 - 11 Standard Test Method for Measurement of Airborne Sound Attenuation Between Rooms in
Buildings (ASTM International, 2011), http://www.astm.org/Standards/E336.htm.
224
Sound pressure level Loudness
Combination of frequencies Timbre
TABLE A- 2: PHYSICAL DESCRIPTIONS AND AUDITIVE PERCEPTION
88
Loudness
Human perception of loudness is subjective. The subjective effect of sound pressure level changes are
described (Table A-3). These ratings may be applied to relative differences between partition
performances when making indicative comparisons.
Difference in Levels, dB
(Increase or Decrease)
Apparent Loudness
(Subjective Ratings)
1 dB Just barely audible
3 dB Just audible
5 dB Clearly audible
10 dB Subjective doubling of loudness (Half or twice as loud)
20 dB Subjective four-fold increase in loudness (Much quieter or louder)
TABLE A-3: SUBJECTIVE EFFECT OF CHANGES IN SOUND PRESSURE LEVEL
8990
REVERBERATION
Reverberation is a space is dependent on the cubic volume and the amount of sound absorbing treatment
applied to surfaces of a room. Reverberant spaces with multiple noise sources will increase noise buildup
which can create a loud environment. Reverberant spaces can also reduce speech intelligibility.
REVERBERATION TIME (RT)
Reverberation time of an enclosed space is defined as the length of time taken in seconds for the sound
pressure level to decrease by 60dB after the source sound has stopped. The RT is dependent upon the
total sound absorbing surfaces and cubic volume of the space.
SOUND
Sound is pressure waves which occur through vibration travelling through a medium, either air or solid.
The human ear perceives sounds through the fluctuation of pressure change at the ear drum. Sound can
88
Ibid.
89
Eckard Mommertz, Acoustics and Sound Insulation: Principles, Planning, Examples (Birkhäuser, 2009).
90
David A. Bies and Colin H. Hansen, Engineering Noise Control: Theory and Practice (Spon Press, 2003).
225
be experienced through pressure changes in the air such as from a car horn or the structure borne from
a vibrating diaphragm such as a drum.
SOUND PRESSURE LEVEL (LP)
Sound is pressure waves. The human ear can accommodate an enormous range of pressures from the
threshold of hearing 20µPa to the threshold of pain 100,000,000 µPa. A logarithmic measurement scale
is used to accommodate sound pressures into levels using ratio of one to one million. The resulting
parameter is sound pressure level (L
p
) and the associated unit of sound measurement is decibel (dB), 0dB
(threshold of hearing) to 140dB (threshold of pain).
STRUCTURE-BORNE SOUND
Where walls or suspended floors are not excited by airborne sound, but instead are caused to vibrate by
way of direct mechanical actions, we speak of structure-borne sound. This is particularly the case when
walking across a floor (impact sound), or when moving chairs, but also when operating building services.
The sound transferred into components propagates through the construction as structure-borne sound
and is radiated in neighboring rooms in the form of (secondary) airborne sound. Airborne and structure-
borne sound excitations often occur together
91
.
SOUND ISOLATION
Sound isolation (sound insulation) is concerned with preventing sound propagation into a building and
within a building, in order to avoid the spread of disturbing noise. Sound isolation is based the amount of
noise transmitting through a wall or floor. All building partition elements and materials have an indicative
sound isolation performance, e.g. STC rating. This acoustic design of sound insulation entails the
arrangement of different functions within a building, and the design constructions and components of
building partitions
92
.
SOUND ISOLATION RATING
Sound insulation ratings provide an indication of how well sound is transmitted through a barrier. The
ratings are single figure numbers assigned to building elements such as, walls, floors, doors, windows, etc.
Various rating types are given to building element to identify specific characters of their sound isolating
properties. The rating descriptions vary between countries and are comparable, e.g. STC in the USA is the
near equivalent to R
w
used internationally.
STC(ASTM)
Sound Transmission Class (STC) is a single number rating used to describe the airborne sound
Transmission Loss performance of a partition. The number rating is derived by comparing
Transmission Loss values measured at 16 one-third octave bands (125 Hz – 4 kHz) to a reference curve.
91
Mommertz, Acoustics and Sound Insulation.
92
Ibid.
226
The STC value is obtained in accordance with ASTM E90 and ASTM E413. A higher STC rating indicates
a better sound isolation performance.
FSTC(ASTM)
Transmission loss data obtained in the field is reported as Field Sound Transmission Class (FSTC). The
FSTC value is obtained in accordance with ASTM E336.
DNT,W (ISO)
The sound isolation required between two spaces may be determined by the sound level difference
needed between them. A single figure descriptor, the standardized weighted sound level difference,
D
nT,w
is the index in the regulations (see BS EN ISO 717-1).
RW (ISO)
The Transmission Loss of a building element is a measure of the loss of sound through the barrier. It
is similar to STC in that the rating is a characteristic of the building component and not affected by
the common area between the rooms and the room acoustic of the receiving room as opposed to
D
nT,w
. The weighted sound reduction index R
w
is a single figure description of the sound reduction
index defined in BS EN ISO 717-1:1997.
SOUND LEVEL
The unit of measurement for sound levels is the decibel, (dB). The human threshold of hearing is 0 decibels
and the human threshold of pain is approximately 130dB. Normal conversational speech is approximately
50dB.
TRANSOM
Horizontal Mullion
TRANSMISSION LOSS, TL
Sound can reach an occupied room by propagating through the air or through vibration paths traveling
within the building structure. These two forms of sound propagation are referred to as airborne or
structure-borne sound.
Airborne sound isolation of a building element, such as a vertical mullion, depends upon some of the
following characteristics:
- The mass (lbs/ft
2
) of the mullion
- The depth of the air space between both sides of the mullion
- The structural connection mechanically fastening both sides of mullion together
- The amount of sound absorption in the air space of the mullion
- Transmission Loss is dependent on damping, mass and coincidence effect
227
Laboratory Transmission Loss Test
The transmission loss of a panel is measured in octave bands by comparing the level in the source room
with the level in the receiving room. The results are normalized for the area of the partition and the
absorption in the receiving room. The test is conducted in a reverberation chamber
FIGURE A-5: SECTION OF LABORATORY TRANSMISSION LOSS TESTING CHAMBER
93
Field Transmission Loss Test
While laboratory represent “idealized conditions” in-site conditions are different. Inevitably there is a
reduction in the apparent Transmission Loss performance of the partition due to flanking and sound
leakage. A 5 to 10 dB loss based on the lab tested insulation value is common.
93
M. David Egan, Architectural Acoustics (J. Ross Publishing, 2007).
228
FIGURE A-6: FIELD CONDITION FOR IN-SITU
TRANSMISSION LOSS TESTING
94
.
FIGURE A-7: DIFFERENCE BETWEEN LABORATORY AND FIELD
TRANSMISSION LOSS TESTS
95
.
UNCONNECTED MULLION
(See Connected Mullion) In the UVM testing phases, the “unconnected” mullion is described where the
center vertical mullion is separated from the curtain wall glazing and horizontal mullions.
VIBRATIONS
Vibrations are generally low-frequency structure-borne sound excitations (below about 63 Hz) which, for
example, are caused by trains, construction activities or industrial operations. If such vibrations could have
negative effects for people, historical buildings or sensitive laboratory apparatus, dynamic analyses are
usually required
96
.
94
Ibid.
95
Ibid.
96
Mommertz, Acoustics and Sound Insulation.
229
APPENDIX B UVM LABORATORY TEST RESULTS
B.1 INTRODUCTION
Laboratory transmission loss test reports of all tests in the UVM phase are provided. All tests are in
accordance with ASTM E90 Test Method for Laboratory Measurement of Airborne Sound Transmission
Loss of Building Partitions and Elements.
TEST
PHASE
MEASURED SPECIMENS
(ISOLATED AND COMPOSITE)
WEAL TEST NUMBER
Mullion Connection
Glass
Curtain Wall
PHASE 1
TL13-309 – TL13-330
PHASE 2A
TL13-398 – TL13-423
PHASE 2B
TL13-605 – TL13-633
PHASE 3
TL14-197 – TL14-171
TABLE B- 1: WEALTESTNUMBERS AT EACH PHASE
B.1.1 Test Measurement Standards
Test procedures is in accordance with ASTM E90-1990. STC ratings are in accordance with E413-1987.
230
B.2 PHASE 1 WEAL TEST RESULTS
Test Number STC Class Mullion Fill Cladding
TL13-309 37 A
bare mullion - flush src
room
none none
TL13-310 34 A
bare mullion - center
position
none none
TL13-311 36 A mullion with gasket none none
TL13-312 47 A
mullion separated by
1"
none none
TL13-313 39 B mullion with gasket bags of pea gravel none
TL13-314 38 B mullion with gasket bags of sand none
TL13-315 36 B mullion with gasket mineral fiber none
TL13-316 38 B mullion with gasket
mass loaded vinyl
pillows
none
TL13-317 46 C1 mullion with gasket
mass loaded vinyl
pillows
1/8" aluminum plate over
mass loaded vinyl
TL13-
318/319
48 C1 mullion with gasket none
1/8" aluminum plate over
mass loaded vinyl
TL13-320 46 C1 mullion with gasket pea gravel
1/8" aluminum plate over
mass loaded vinyl
TL13-321 50 C2 mullion with gasket none
5/8" gypsum board over mass
loaded vinyl
TL13-322 47 C2 mullion with gasket pea gravel
5/8" gypsum board over mass
loaded vinyl
TL13-323 52 C2 mullion with gasket
mass loaded vinyl
pillows
5/8" gypsum board over mass
loaded vinyl
TL13-324 47 C3 mullion with gasket
mass loaded vinyl
pillows
5/8" gypsum board
TL13-325 42 C3 mullion with gasket none 5/8" gypsum board
TL13-326 45 C3 mullion with gasket pea gravel 5/8" gypsum board
TL13-327 31 C4 mullion with gasket none
1-1/2" alum tube with PAC
isolators
TL13-328 38 C4 mullion with gasket MLV in tubes
1-1/2" alum tube with PAC
isolators
TL13-329 48 C4 mullion with gasket
MLV and mineral
fiber in tubes
1-1/2" alum tube with PAC
isolators
TL13-330 48 C4 mullion with gasket
MLV and mineral
fiber in tubes
1-1/2" alum tube with MLV
isolators
TABLE B- 2: PHASE 1 WEAL TEST NUMBERS, AREA AND DESCRIPTION
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
B.3 PHASE 2A WEAL TEST RESULTS
The area of the filler wall aperture varied for certain test specimens. Final transmission loss test results
were corrected after the phase was complete (Table B-2).
Test
Number
Area
(SF)
STC
mullion
fill
overclad connection edge seal
TL13-398 2.81 42 none
1-1/2" alum tube
with MLV isolators,
filled w MLV pillows
none 1/2" perimeter, putty
TL13-399 2.81 36 none none none 1/2" perimeter, putty
TL13-400 2.81 42
TL-402
air leaks
TL13-401 2.81 46
TL-402
air leaks
TL13-402 2.81 49
MLV
Pillows
5/8" gypsum board
+ MLV plate,
screwed to mullion
none 1/2" perimeter, putty
TL13-404 2.81 35
ram
packed
mineral
fiber
none none 1/2" perimeter, putty
TL13-405 3.02 44
MLV
Pillows
5/8" gypsum board
+ MLV plate,
screwed to mullion
1/2" Armacell,
3 edges 1/4" putty, 1/4" wood
shim for compression on 1/2"
armacell
TL13-406 3.13 52
MLV
Pillows
5/8" gypsum board
+ MLV plate,
screwed to mullion
3/4" backer rod
with caulking
3/4" wet seal,
3 edges 1/4" putty, 1/4"
neoprene shim
TL13-407 3.13 34 none none
3/4" backer rod
with caulking
all 4 edges wet seal,
incl 3/4" edge, 1/4" neoprene
shim
TL13-408 3.13 38 none none
3/4" backer rod
with caulking
3/4" wet seal, other 3 edges
wet seal + masking tape +
putty, neoprene shim
TL13-409 3.13 49
TL-411
flagged
TL13-410 3.13 49
TL-411
flagged
TL13-411 3.13 49
MLV
pillows
1-1/2" alum tube
with PAC isolators,
filled w MLV pillows
3/4" backer rod
with caulking
all 4 edge wet seal,
incl 3/4" edge, 1/4" neoprene
shim
TL13-412 3.75 31
MLV
Pillows
5/8" gypsum board
+ MLV plate,
screwed to mullion
source side, (1) 2-
1/2" width silicone
strip compressed
1/4"
wet seal top and bottom of
silicone,
3 edges 1/4" putty, 1/4"
neoprene shim
TL13-413 3.75 41
MLV
Pillows
5/8" gypsum board
+ MLV plate,
screwed to mullion
both sides, (2) 2-
1/2" width silicone
strip compressed
1/4"
wet seal top and bottom of
silicone,
3 edges 1/4" putty, 1/4"
neoprene shim
TL13-414 3.75 32
MLV
Pillows
5/8" gypsum board
+ MLV plate,
screwed to mullion
receiver side, (1) 2-
1/2" width silicone
strip compressed
1/4"
wet seal top and bottom of
silicone,
3 edges 1/4" putty, 1/4"
neoprene shim
253
Test
Number
Area
(SF)
STC
mullion
fill
overclad connection edge seal
TL13-415 3.75 36
MLV
Pillows
5/8" gypsum board
+ MLV plate,
screwed to mullion
source side, (1) 2-
1/2" width silicone
strip compressed
1/4", metal overclad
adhered with
masking tape
wet seal top and bottom of
silicone,
3 edges 1/4" putty, 1/4"
neoprene shim, wet seal on
one edge of metal plate
TL13-416 3.75 34
MLV
Pillows
5/8" gypsum board
+ MLV plate,
screwed to mullion
both sides (2) 2-
1/2" width silicone
strip compressed
1/4", metal overclad
adhered with
masking tape
wet seal top and bottom of
silicone,
3 edges 1/4" putty, 1/4"
neoprene shim, wet seal on
one edge of metal plate
TL13-417 3.75 30 none none
source side, (1) 2-
1/2" width silicone
strip compressed
1/4"
wet seal top and bottom of
silicone,
3 edges 1/4" putty, 1/4"
neoprene shim
TL13-418 3.75 35 none none
both sides, (2) 2-
1/2" width silicone
strip compressed
1/4"
wet seal top and bottom of
silicone,
3 edges 1/4" putty, 1/4"
neoprene shim
TL13-419 2.81 28 none none
both sides (2) 2-
1/2" width silicone
strip compressed
1/4", metal overclad
adhered with
masking tape
wet seal top and bottom of
silicone,
3 edges 1/4" putty, 1/4"
neoprene shim, wet seal on
one edge of metal plate
TL13-420 3.75 31 none none
both sides (2) 2-
1/2" width silicone
strip compressed
1/4", metal overclad
adhered with
masking tape
wet seal top and bottom of
silicone,
3 edges 1/4" putty, 1/4"
neoprene shim, wet seal on
one edge of metal plate
TL13-421 3.75 34 none none
source side, (1) 2-
1/2" width silicone
strip compressed
1/4", metal overclad
adhered with
masking tape
wet seal top and bottom of
silicone,
3 edges 1/4" putty, 1/4"
neoprene shim, wet seal on
one edge of metal plate
TL13-422 3.75 22 none none
none,metal
overclad adhered
with masking tape
3 edges 1/4" putty, 1/4"
neoprene shim, wet seal on
one edge of metal plate
TL13-423 3.125 46 none none mull-it-over
bare mullion: putty on 3 sides,
1/2" backer rod +wet seal on
one side
mullitover: putty on top and
bottom, wet seal on screws,
compression seal on one side
TABLE B- 3: PHASE 2A, WEAL TEST NUMBERS, AREA AND DESCRIPTION
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
B.4 PHASE 2B WEAL TEST RESULTS
WEAL TL STC
Area (SF)
print out
Specimen (Connection)
Silicone Bulb
Seal
Edge Seal
Plate/Airspace
Series
TL13-605 51 2.73
1/8" Alum. Plate
6" Cavity - NO Batt
Mullion edge
Bulb seal - mullion
edge
Putty - 3 sides
TL13-606 51 2.73 re-run
TL13-607 51 2.73 re-run
TL13-608 51 2.73
1/8" Alum. Plate
6" Cavity - FILLED batt insulation
Mullion edge
Bulb seal - mullion
edge
Putty - 3 sides
TL13-609 47 2.73
1/8" Alum. Plate
4" Cavity - NO Batt
(LEAK)
Mullion edge
Bulb seal - mullion
edge
Putty - 3 sides
TL13-610 47 re-run
TL13-611 49 2.73
1/8" Alum. Plate
4" Cavity - FILLED batt insulation
Mullion edge
Bulb seal - mullion
edge
Putty - 3 sides
TL13-612 44 2.73
1/8" Alum. Plate
3" Cavity - NO Batt
Mullion edge
Bulb seal - mullion
edge
Putty - 3 sides
TL13-613 47 2.73
1/8" Alum. Plate
3" Cavity - FILLED batt insulation
Mullion edge
Bulb seal - mullion
edge
Putty - 3 sides
TL13-614 46 2.73
1/8" Alum. Plate + MLV
3" Cavity - NO BATT
Mullion edge
Bulb seal - mullion
edge
Putty - 3 sides
TL13-615 48 2.73
1/8" Alum. Plate + MLV
3" Cavity - FILLED batt insulation
Mullion edge
Bulb seal - mullion
edge
Putty - 3 sides
TL13-616 45 2.73
5/8" gypsum board plates
3" Cavity - NO BATT
Mullion edge
Bulb seal - mullion
edge
Putty - 3 sides
TL13-617 48 2.73
5/8" gypsum board plates
3" Cavity - FILLED batt insulation
Mullion edge
Bulb seal - mullion
edge
Putty - 3 sides
TL13-618 50 2.73
5/8" gypsum board + MLV + Alum.
Plate
3" Cavity - FILLED batt insulation
Mullion edge
Bulb seal - mullion
edge
Putty - 3 sides
TL13-619 47 2.73
5/8" gypsum board + MLV + Alum.
Plate
3" cavity - NO Batt
Mullion edge
Bulb seal - mullion
edge
Putty - 3 sides
Staggered
Series
TL13-620 22 2.73
5/8" gypsum bd
2" overlap+ 1/4" air b/w plates
NO Batt
none
TL13-621 37 2.73
5/8" gypsum bd
2" overlap+ 1/4" air b/w plates
Cavity - filled batt insulation
none
277
WEAL TL STC
Area (SF)
print out
Specimen (Connection)
Silicone Bulb
Seal
Edge Seal
TL13-622 51 2.73
5/8" gypsum bd
2" overlap+ 1/4" air b/w plates
Cavity - filled batt insulation
1 - receiver side
1 - source side
TL13-623 44 2.73
5/8" gypsum bd
2" overlap+ 1/4" air b/w plates
NO Batt
1 - receiver side
1 - source side
TL13-624 47 2.73
5/8" gypsum bd
2" overlap+ 1/4" air b/w plates
NO Batt
2 - receiver side
1 - source side
TL13-625 20 2.73
1/8" aluminum plates (4'x60")
2" overlap+ 1/4" air b/w plates
NO Batt
none Putty Edge
TL13-626 31 2.73
1/8" aluminum plates (4'x60")
2" overlap+ 1/4" air b/w plates
Cavity - filled batt insulation
none Putty Edge
TL13-627 49 2.73
1/8" aluminum plates (4'x60")
2" overlap+ 1/4" air b/w plates
Cavity - filled batt insulation
1 - receiver side
1 - source side
Putty Edge
TL13-628 47 2.73
1/8" aluminum plates (4'x60")
2" overlap+ 1/4" air b/w plates
NO Batt
1 - receiver side
1 - source side
Putty Edge
TL13-629 48 2.73
1/8" aluminum plates (4'x60")
2" overlap+ 1/4" air b/w plates
NO Batt
2 - receiver side
1 - source side
Putty Edge
Products
TL13-630 23 1.41
Mullion Mate
ONLY
top/bottom - putty
vert sides - wet seal
2x6 on either side
TL13-631 30 4.22
Mullion Mate
BARE MULLION
Wet Seal
TL13-632 31 4.22
Mullion Mate
BEST MULLION
Wet Seal
TL13-633 50 2.81
Mull-it-over
isolated leaves
NO MULLION
Wet Seal
TABLE B- 4: PHASE 2B, WEAL TEST NUMBERS, AREA AND DESCRIPTION
Note:
WEAL tests TL13-605 throughTL13-629: The TL was reduced by 0.38 dB to adjust to the specimen area
(6" x 60") from the opening area (6.5" x 60.5") which was used in the original TL calculations.
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
B.5 PHASE 3 WEAL TEST RESULTS
WEAL TL
STC Area (SF) Specimen (Connection) Edge Seal Notes
TL14-167 42 2.92 sf
STC 52 Mullion - overclad
(gyp+MLV) and filled (MLV pillows)
backer rod + wet seal
overclad screwed into
mullion
TL14-168 37 2.92 sf STC 38 Mullion - filled MLV pillows backer rod + wet seal 3 non-puttied holes
TL14-169 32 2.92 sf STC 36 Mullion - bare and hollow Backer rod + wet seal 3 non-puttied holes
TL14-170 32 2.92 sf STC 36 Mullion - bare and hollow Backer rod + wet seal all holes puttied
TL14-171 32 2.92 sf STC 36 Mullion - bare and hollow Backer rod + wet seal
all holes puttied Receiver
chamber removed
TABLE B- 5: PHASE 3, WEAL TEST NUMBERS, AREA AND DESCRIPTION
308
309
310
311
312
313
APPENDIX C ANCILLARY SOUND ANALYSIS
C.1 INTRODUCTION
Additional transmission loss comparison overlays between tests phases are provided for archival
purposes.
314
C.2 PHASE 1 MULLION CONTROLS
MC-1, (3) TL13-311 - 36 MC-1a, (28) TL13-404 - 35 MC-1b, (17) TL13-325 - 42 MC-1c, (10) TL13-318 - 48 MC-1d, (13) TL13-321 - 50
FIGURE C-1: PHASE 1-A TRANSMISSION LOSS CURVES
0
10
20
30
40
50
60
70
80
90
100
63 80 100 125 160 200 250 315 400 500 630 800 1000 1250 1600 2000 2500 3150 4000 5000
Transmission Loss (dB)
One-third Octave Band Center Frequency (Hz)
Phase 2b: Mullion Controls 1, 1a, 1 b, 1c, 1d
MC-1, (3) TL13-311 - 36
MC-1a, (28) TL13-404 - 35
MC-1b, (17) TL13-325 - 42
MC-1c, (10) TL13-318 - 48
MC-1d, (13) TL13-321 - 50
315
C.3 UVM TEST ELEMENT COMPARISON WITH A 4” AIRSPACE
(4) TL13-312
Phase 1 – impractical mullion separated to create a
4” airspace
TL13-610
Phase 2b – parallel plates of aluminum
without an interconnection
FIGURE C- 2: 4” AIRSPACE COMPARISON, MULLION AND PLATE CONNECTION
0
10
20
30
40
50
60
70
80
90
100
Transmission Loss (dB)
One-third Octave Band Center Frequency (Hz)
Phase 2b: Plate Configurations - 4" airspace with mullion
(1/8” aluminum plate)
*(4) TL13-312, STC 47 (51) TL13-610 - 47
316
C.4 UVM TEST ELEMENT COMPARISON WITH A 3” AIRSPACE
MC-1, (3) TL13-311 - 36
TL13-612 – STC 44
FIGURE C- 3 3” AIRSPACE COMPARISON, MULLION AND PLATE CONNECTION
0
10
20
30
40
50
60
70
80
90
100
Transmission Loss (dB)
One-third Octave Band Center Frequency (Hz)
3" Airspace Comparison with Interconnections
[aluminium mullion versus aluminum plates]
MC-1, (3) TL13-311 - 36 (53) TL13-612 - 44
317
C.5 UVM TEST ELEMENT COMPARISON WITH A 3” AIRSPACE, BATT INFILL
MC-1, (3) TL13-311 - 36 MC-1a, (28) TL13-404 - 35 (54) TL13-613 - 47
FIGURE C- 4: 3” AIRSPACE AND BATT INFILL COMPARISON, MULLION AND PLATE CONNECTION
0
10
20
30
40
50
60
70
80
90
100
63 80 100 125 160 200 250 315 400 500 630 800 1000 1250 1600 2000 2500 3150 4000 5000
Transmission Loss (dB)
One-third Octave Band Center Frequency (Hz)
Comparison: 3" Mineral Wool Filled Airspace
[aluminium mullion versus aluminum plates]
MC-1, (3) TL13-311 - 36 MC-1a, (28) TL13-404 - 35 (54) TL13-613 - 47
318
C.6 UVM TEST ELEMENT COMPARISON WITH A 3” AIRSPACE, GYPSUM OVERCLAD
MC-1, (3) TL13-311 - 36 MC-1b, (17) TL13-325 - 42 (57) TL13-616 - 45
FIGURE C- 5: 3” AIRSPACE AND GYPSUM OVERCLAD, MULLION AND PLATE CONNECTION
0
10
20
30
40
50
60
70
80
90
100
63 80 100 125 160 200 250 315 400 500 630 800 1000 1250 1600 2000 2500 3150 4000 5000
Transmission Loss (dB)
1/3 Octave Band Center Frequency (Hz)
Mullion Control Comparison: MC-1b
with Phase 2b gypboard plates (57) TL13-616
MC-1, (3) TL13-311 - 36 MC-1b, (17) TL13-325 - 42 (57) TL13-616 - 45
319
C.7 PHASE 2B CLASS A SUMMARY GRAPHS
FIGURE C- 6: PHASE 2B PARALLEL PLATE TESTS WITH 3” AIR CAVITIES
Test Number Plate Assembly
(53) TL13-612 - 44/37 1/8" aluminum
(54) TL13-613 - 47/38 1/8" aluminum + batt
(55) TL13-614 - 46/39 1/8" aluminum + MLV
(56) TL13-615 - 48/39 1/8" aluminum + MLV + batt
(57) TL13-616 - 45/38 5/8" gypsum board plates
(58) TL13-617 - 48/39 5/8" gypsum board plates + batt
(60) TL13-619 - 47/39 5/8" gypsum board + MLV + Alum. Plate
(59) TL13-618 - 50/41 5/8" gypsum board + MLV + Alum. Plate+ batt
0
10
20
30
40
50
60
70
80
90
100
Transmission Loss (dB)
One-third Octave Band Center Frequency (Hz)
Phase 2B Class A Parallel Plates with 3" air space
(53) TL13-612 - 44
(54) TL13-613 - 47
(55) TL13-614 - 46
(56) TL13-615 - 48
(57) TL13-616 - 45
(58) TL13-617 - 48
(60) TL13-619 - 47
(59) TL13-618 - 50
320
C.8 PHASE 2B PARALLEL PLATE WITH VARIED AIR SPACE
FIGURE C- 7: SOUND TRANSMISSION LOSS OF ALUMINUM PLATES ONLY WITH 3”, 4” OR 6” AIR CAVITY
Test Number Description of Air Space between Aluminum Plates
(48) TL13-607 - 51/40 6” air cavity
(49) TL13-608 - 51/40 6” air cavity + batt
(51) TL13-610 - 47/39 4” air cavity
(52) TL13-611 - 49/39 4” air cavity + batt
(53) TL13-612 - 44/37 3” air cavity
(54) TL13-613 - 47/38 3” air cavity + batt
0
10
20
30
40
50
60
70
80
90
100
Transmission Loss (dB)
One-third Octave Band Center Frequency (Hz)
Phase 2b: Plate w 6", 4" and 3" air space
(1/8" Aluminum Sheets, 60"x6")
(48) TL13-607 - 51/40 (49) TL13-608 - 51/40 (51) TL13-610 - 47/39
(52) TL13-611 - 49/39 (53) TL13-612 - 44/37 (54) TL13-613 - 47/38
321
C.9 COMPARISONS BETWEEN PHASE 1 AND PHASE 3
Comparisons between Phase 1 mullions, TL13-311 (MC-1) and TL13-323 (MC-2), are compared with Phase
3 testing.
FIGURE C- 8: TL SPECTRA OF PH3-MC1 WITH PH1-MC1 AND PH3-MC2 WITH PH1-MC2
0
20
40
60
80
100
63 80 100 125 160 200 250 315 400 500 630 800 1000 1250 1600 2000 2500 3150 4000 5000
Transmission Loss (dB)
One-third Octave Band Center Frequency (Hz)
TL Plots of Highest and Lowest Performing
Phase 1 and Phase 3
PH3-MC1 (TL14-170 - 32) PH3-MC2 (TL14-167 - 42)
PH1-MC1 (TL13-311 - 36) PH1-MC2 (TL13-323 - 52)
LINEAR
AVERAGE (dB)
4 dB 6 dB
322
FIGURE C- 9: TRANSMISSION LOSS COMPARISON BETWEEN CONTROL MULLIONS MC1 AND MC2
FIGURE C-10: PHASE1-MC1(TL13-311) AND PHASE 3-MC1 (TL14-170)
0
20
40
60
80
100
63 80 100 125 160 200 250 315 400 500 630 800 1000 1250 1600 2000 2500 3150 4000 5000
Transmission Loss (dB)
One-third Octave Band Center Frequency (Hz)
Mullion Controls 1 and 2
PH1-MC1 (TL13-311 - 36) PH1-MC2 (TL13-323 - 52)
16 dB
LINEAR
AVERAGE dB
0
10
20
30
40
50
60
70
80
90
100
63 80 100 125 160 200 250 315 400 500 630 800 1000 1250 1600 2000 2500 3150 4000 5000
Transmission Loss (dB)
One-third Octave Band Center Frequency (Hz)
Transmission Loss Comparison
TL13-311 (MC1) and TL14-170 (Ph3-MC1)
STDEVA: TL13-311 & TL14-170 PH1-MC1 (TL13-311 - 36) PH3-MC1 (TL14-170 - 32)
323
FIGURE C-11: PHASE1-M1A(TL13-316) AND PHASE 3-MC1A (TL14-168)
FIGURE C-12: PHASE1-MC2(TL13-323) AND PHASE 3-MC2 (TL14-167)
0
10
20
30
40
50
60
70
80
90
100
63 80 100 125 160 200 250 315 400 500 630 800 1000 1250 1600 2000 2500 3150 4000 5000
Transmission Loss (dB)
1/3 Octave Band Center Frequency (Hz)
TRANSMISSION LOSS COMPARISON
TL13-316 (MC1A) AND TL14-168 (PH3-MC1A)
Ph3 STDEVA w/ TL13-316 PH1-MC1a (TL13-316 - 38) PH3-MC1a (TL14-168 - 37)
0
10
20
30
40
50
60
70
80
90
100
63 80 100 125 160 200 250 315 400 500 630 800 1000 1250 1600 2000 2500 3150 4000 5000
Transmission Loss (dB)
1/3 Octave Band Center Frequency (Hz)
Transmission Loss Comparison
TL13-323 (MC2) and TL14-167 (Ph3-MC2)
Ph3 STDEVA w/ TL13-323 PH1-MC2 (TL13-323 - 52) PH3-MC2 (TL14-167 - 42)
324
C.10 UNCONNECTED AND CONNECTED (HOLLOW AND FILLED)
PHASE 1-MC1 (TL13-311) PHASE 3-MC1 (TL14-170)
PHASE 1-MC1A (TL13-316) PHASE 3-MC1A (TL14-168)
FIGURE C-13: TLPLOTS OF PHASE 1 (MC1/1A) AND PHASE 3 (MC1/1A)
0
20
40
60
80
100
63 80 100 125 160 200 250 315 400 500 630 800 1000 1250 1600 2000 2500 3150 4000 5000
Transmission Loss (dB)
1/3 Octave Band Center Frequency (Hz)
COMPARISON 1 AND COMPARISON 1A
PH1-MC1 (TL13-311 - 36) PH3-MC1 (TL14-170 - 32)
PH1-MC1a (TL13-316 - 38) PH3-MC1a (TL14-168 - 37)
325
FIGURE C-14: TLPLOTS OF COMPARISON MC1 AND MC1A
0
20
40
60
80
100
63 80 100 125 160 200 250 315 400 500 630 800 1000 1250 1600 2000 2500 3150 4000 5000
Transmission Loss (dB)
1/3 Octave Band Center Frequency (Hz)
COMPARISON MC1 AND MC1A
PH1-MC1 (TL13-311 - 36) PH1-MC1a (TL13-316 - 38)
10 dB
LINEAR
AVERAGE dB
326
APPENDIX D ANCILLARY VIBRATION ANALYSIS
D.1 INTRODUCTION
Vibration measurements were conducted on the curtain wall assembly during Phase 3 to compare the
acoustic energy passing laterally from the source to receiving chamber at the glass, vertical mullion and
horizontal mullion. Description of the multi-chamber test setup including semi-anechoic enclosures is
described in Chapter 3.
Typical accelerometer measurement locations on the curtain wall at the source chamber was mirrored at
the receiving chamber so that measurements may be conducted simultaneously (Figure D-1).
FIGURE D-1: TYPICAL LOCATION OF THE ACCELEROMETERS AT THE SOURCE AND RECEIVING CHAMBERS
The vibration measurements are intended to provide a basis to compare
the acoustic energy loss from source to receiving side at mirrored accelerometer locations and
vibration levels between the vertical mullion, horizontal mullion, and glass at the receiving chamber.
The preliminary vibration measurement results and analysis is subject to future work. The initial
investigation discussed in Appendix D is provided for archival purposes.
GLASS
VERTCAL
MULLION
HORIZONTAL
MULLION
GLASS
VERTCAL
MULLION
HORIZONTAL
MULLION
327
D.2 VIBRATION MEASUREMENT SET UP
Four sets of vibration measurements were taken based on several configurations (Table D – 1).
Configuration
Number
UVM Assembly Used
(from Phase 3)
Chamber 3S (Semi-
anechoic enclosure)
Chamber 3R (Semi-
anechoic enclosure)
EV1 [A
v
] TL14-167 (STC 42) Fully Enclosed Fully Enclosed
EV2 [B
v
] TL14-170 (STC 32) Fully Enclosed Fully Enclosed
EV3a [C
v
] TL14-171 (STC 32) Fully Enclosed Removed
EV3b [D
v
] TL14-170 (STC 32) Removed Removed
TABLE D- 1: PHASE 3 CONFIGURATION OF VIBRATION MEASUREMENT
[A
v
] vibration measurement
The EV1 [A
v
] vibration measurement was conducted on the TL14-167 test specimen; the center
vertical mullion is overclad and filled.
The 3S and 3R chambers were fully enclosed.
[B
v
] vibration measurement
The EV2 [B
v
] vibration measurement was conducted on the TL14-170 test specimen; the center
vertical mullion is exposed and hollow.
The 3S and 3R chambers were fully enclosed.
[C
v
] vibration measurement
Vibration measurement was conducted on the TL14-171 test rig.
The 3R chamber was removed.
[D
v
] vibration measurement
Vibration test EV3b does not have a corresponding laboratory test.
Both the 3S and 3R chambers were removed.
The semi-anechoic chamber enclosures at configurations [C
v
] and [D
v
] were removed during the
measurement as indicated (Table D-1). This changes the structural stiffness of the system and may
influence the results.
D.2.1 Measurement and Chamber Set Up
The Phase 3 UVM Test chamber set up described in Chapter 3 was used to locate the accelerometers at
typical locations indicated on the curtain wall specimen (Figure D-2).
328
FIGURE D- 2: Plan drawing of weal test chambers and phase 3 test rig 3s/3r chambers
An elevation of the curtain wall bay specimen is shown with the accelerometer measurement locations at
the 3S and 3R chambers (Figure D-3).
FIGURE D- 3: Elevation A: View of Accelerometer locations on chambers 3S and 3R for EV2
329
The accelerometers numbered in Figure D-3 correspond to the axis and location description for the EV2
[B
v
] (Table D – 2).
Source ReceiverAxis Surface Material Location
1 1 X gypsum mid-span on gypsum board header
2 2 X aluminum mid-span on lower transom
3 3 Z glass mid-span center on glass bay
4 4 X aluminum on lower transom 2" from vertical mullion
5 5 Y aluminum on vertical mullion 2'-3" mid-span
6 6 Y aluminum on vertical mullion 2'-3" mid-span
TABLE D- 2: PHASE 3 CONFIGURATION OF VIBRATION MEASUREMENT
D.2.2 Equipment and Procedure
Several pieces of equipment were used for the measurements (Table D – 3).
Device Manufacturer Model Serial Number Comments
Accelerometer Endevco 7703A-1000 10125 Source
Accelerometer Endevco 7706-1000 AD71, 977.0 pC/g Receiver
Sound Level
Meter Input
Brüel & Kjær 2260 (VA Meter #7) Source
Sound Level
Meter Input
Brüel & Kjær 2260 (VA Meter #4) Receiver
Computer
Software
Analyzer
Brüel & Kjær
Evaluator Type 7820
Version 4.16.4
-- --
TABLE D- 3: VIBRATION MEASUREMENT EQUIPMENT
The settings on the Brüel & Kjær 2260 meter were
Range: 10dB - 100dB
One-third octave: Low Frequency
Statistical Measurements: FAST
measurements flat spectrum, L&L
The Brüel & Kjær Evaluator Type 7820 Version 4.16.5 software was used to analyze the data.
The two B&K 2260 meters were time synced to take simultaneous measurements.
The acoustic source of energy was from the loudspeaker in the source laboratory chambers which emitted
120 dB of pink noise.
330
D.3 VIBRATION MEASUREMENT RESULTS
Acceleration level (dB) L
EQ
measurements were taken over a period of 10 second (Figure D – 4).
FIGURE D- 4: Axis designation for vibration measurements
Each measurement was 10 seconds. A 5 second dB L,L
EQ
time period was used for analysis to omit ramp
up and down from the loudspeaker.
331
D.3.1 [A
v
] Vibration Measurement Results
The 5 sec dB L,L
eq
at the source and receiving chambers for the [A
v
] test rig are summarized (Table D-4).
Test
L,Leq
5 sec, dB
Axis
[Surface Material]
Accelerometer Location
[A
v
] S1 83
X
[gyp board] mid-span on gypsum
board header
[A
v
] R1 80
[A
v
] S2 100
X
[aluminum] mid-bay horizontal on
bottom horizontal mullion
[A
v
] R2 96
[A
v
] S3 102
Z
[glass] mid-bay horizontal, 12"
above bottom horizontal mullion
[A
v
] R3 89
[A
v
] S4 --
Z
[glass] same as S3/R3, (ambient)
[A
v
] R4,5 --
[A
v
] S5 103
X
[aluminum] at bottom transom, 2"
from center vertical mullion
[A
v
] R6 98
[A
v
] S6 100
Z
[glass] 12" above bottom transom
2" from center vertical mullion
[A
v
] R7 89
[A
v
] S7 100
Y
[gypsum+mlv] on mullion 12"
above bottom horizontal mullion
[A
v
] R8 93
[A
v
] S8 101
Y
[gypsum+mlv] repeat - taped
accelerometer from detaching
[A
v
] R9 94
TABLE D- 4: [A
V
] SOURCE AND RECEIVER RESULTS, 5 SEC DB L,L
EQ
332
D.3.2 [B
v
] Vibration Measurement Results
The 5 sec dB L,L
eq
at the source and receiving chambers for the [B
v
] test rig are summarized (Table D-5).
Test
L,Leq
5 sec, dB
Axis
[Surface Material]
Accelerometer Location
[B
v
] S1 92
X
[gyp board] mid-span on gypsum
board header
[B
v
] R1 81
[B
v
] S2 94
X
[aluminum] mid-span on lower
horizontal mullion
[B
v
] R2 92
[B
v
] S3 93
Z
[glass] mid-span center on glass
bay
[B
v
] R3 88
[B
v
] S4 94
X
[aluminum] on lower horizontal
mullion 2" from vertical mullion
[B
v
] R4 93
[B
v
] S5 93
Y
[aluminum] on vertical mullion 2'-
3" mid-span
[B
v
] R5 93
[B
v
] S6 --
Y
[aluminum] on vertical mullion 2'-
3" mid-span (ambient)
[B
v
] R6 39
TABLE D- 5: [B
V
] SOURCE AND RECEIVER RESULTS, 5 SEC DB L,L
EQ
333
D.3.3 [C
v
] Vibration Measurement Results
The 5 sec dB L,L
eq
at the source and receiving chambers for the [C
v
] test rig are summarized (Table D-6).
Test
L,Leq
5 sec, dB
Axis
[Surface Material]
Accelerometer Location
[C
v
] S1 101
Z
[glass] 6" above horizontal
mullion, 2" from vert mullion
[C
v
] R1 86
[C
v
] S2 102
Z
[glass] 2'-3" above horizontal
mullion, 2" from vertical mullion
[C
v
] R2 87
[C
v
] S3 102
Z
[glass] 6" above horizontal
mullion, 2'-3" from vertical
mullion mid-span center on glass
bay [C
v
] R3
87
[C
v
] S4 102 Z [glass] 2'-3" above horizontal
mullion, 2'-3" from vertical
mullion
[C
v
] R4 85
TABLE D- 6: [C
V
] SOURCE AND RECEIVER RESULTS, 5 SEC DB L,L
EQ
334
D.3.4 [D
v
] Ph3 - Vibration Measurement Results
The 5 sec dB L,L
eq
at the source and receiving chambers for the [D
v
] test rig are summarized (Table D-7).
Test
L,Leq
5 sec, dB
Axis
[Surface Material]
Accelerometer Location
[D
v
] S5 102
Z
[glass] 6" above horizontal
mullion, 2'-3" from vertical
mullion
[D
v
] R5 90
[D
v
] S6 102
Z
[glass] 2'-3" above horizontal
mullion, 2'-3" from vertical
mullion
[D
v
] R6 91
[D
v
] S7 101
Z
[glass] 6" above horizontal
mullion, 2" from vertical mullion
[D
v
] R7 90
[D
v
] S8 102
Z
[glass] 2'-3" above horizontal
mullion, 2'-3" from vertical
mullion
[D
v
] R8 92
[D
v
] S9 --
Z
Same as above
(ambient measurement) [D
v
] R9 --
[D
v
] S10 88*
Z
Same as above
(impulse measurement: tap on
glass mid bay center)
*L,L
max
[D
v
] R10 67*
TABLE D- 7: [D
V
] SOURCE AND RECEIVER RESULTS, 5 SEC DB L,L
EQ
335
D.4 VIBRATION ANALYSIS
A preliminary analysis investigation is summarized for the vibration measurement configurations EV1 [A
v
]
and EV2 [B
v
]. Configurations EV3a [C
v
] and EV3b [D
v
] are not included in this analysis.
The initial analysis includes the conversion of the measured vibration acceleration levels in dB (re 10
-6
m/s
2
) at curtain wall surfaces to sound pressure level in dB (re 20 µPa).
D.4.1 Measurement EV1 [A
v
] and EV2 [B
v
].
Measurements locations for the configurations used in the analysis are identified (Table D -8).
CURTAIN WALL
SURFACE
EV1 [A
v
] Figure D-5 Accelerometer
Location (Measurement)
EV2[B
v
] Figure D-5 Accelerometer
Location (Measurement)
LOWER HORIZONTAL
MULLION
2 (EV1_3R-02) 2 (EV2_3R-02)
VERTICAL MULLION 8 (EV1_3R-08) 5 (EV2_3R-05)
GLASS 3 (EV1_3R-03) 3 (EV2_3R-03)
TABLE D- 8: SELECTED VIBRATION MEASUREMENTS FROM TEST EV1 AND EV2
Below are elevations of the curtain wall bay that correspond with the accelerometer measurements
identified in Table D-8.
FIGURE D-5: EV1[A
V
] Curtain Wall Elevation with 3R Chamber Accelerometer Locations 02, 08, 03
336
FIGURE D-6: EV2[B
V
] Curtain Wall Elevation with 3R Chamber Accelerometer Locations 02, 05, 03
D.4.2 Calculation Procedure
Vibrational Acceleration Levels (dB) are converted to Sound Pressure Levels (dB) based on the following
calculation procedure described.
[1] Vibration Acceleration Levels (dB) measured at WEAL with the B&K 2260 meters (Table D-4 and
Table D-5).
[2] Data results in dB from [1] were converted to acceleration( ), in / using the following
equation:
= 10
Where,
=(9.8 10
) / EQUATIOND-1
337
[3] Acceleration is converted to velocity( ) in( / ) using angular frequency per One-third Octave
Band Center Frequency.
= / Where,
=2
= ( )
EQUATION D-2
[4] Velocity is then converted to pressure ( )in pascals using the following equation:
( ) = 407
EQUATION D-3
[5] Pressure is converted to sound pressure level ( ) in( ) using
( ) = 20LOG(
2 10
) EQUATION D-4
[6] Sound pressure level( ) is converted to sound power level( ) using
( ) = +10log( , )
EQUATION D-5
[7] Sound power level is logarithmically added for the 3 elements measured, at the sill, mullion and
glass:
( ) = 10log 10
+ 10log 10
+ 10log 10
EQUATION D-6
[8] Sound power level is converted to reverberation sound pressure level in the room: Reverberant
Sound Level (from RT) - L
p,rev
from L
w
( )
= 10log +10 = 10log +14
EQUATION D-7
Where,
is the room volume (m³)
is the reverberation time (s)
is the number of power sources (L
w
contributing to the reverberant
field)
338
D.4.2.1Acceleration Level (dB) to Acceleration (m/s
2
)
( ) = 20log 0
Where,
= (9.8 10
) / EQUATION D-8
339
D.4.3 Graphed Results of Conversions
Results based on Equations D-1 through Equation D-8 are graphed for systems EV1 and EV2.
EV1 TL14-167 (Heavy Mullion) EV2 TL14-170 (Light Mullion)
FIGURE D- 7: EV1 – MEASURED ACCELERATION LEVELS (DB) FIGURE D-8: EV2–MEASUREDACCELERATIONLEVELS(DB)
0
20
40
60
80
100
12.5
16
20
25
31.5
40
50
63
80
100
125
160
200
250
315
400
500
630
800
1k
1.25k
1.6k
2k
2.5k
3.15k
4k
Measured Acceleration Level (dB) - EV1 3R -02,-08,-03
Sill (EV1)_AccelLevel Mullion (EV1)_AccelLevel Glass (EV1)_AccelLevel
0
20
40
60
80
100
31.5
40
50
63
80
100
125
160
200
250
315
400
500
630
800
1k
1.25k
1.6k
2k
2.5k
3.15k
4k
5k
6.3k
8k
10k
Measured Acceleration dB - EV2 3R -02,-05, 03
Sill (EV2)_AccelLevel Mullion (EV2)_AccelLevel
Glass (EV2)_AccelLevel
340
EV1 TL14-167 (Heavy Mullion) EV2 TL14-170 (Light Mullion)
FIGURE D- 9: EV1 – CALCULATED ACCELERATION (M/S2) FIGURE D- 10: EV2 – CALCULATED ACCELERATION (M/S2)
FIGURE D- 11: EV1 – CALCULATED VELOCITY (M/S) FIGURE D- 12: EV2 – CALCULATED VELOCITY (M/S)
0.0E+00
2.0E-01
4.0E-01
6.0E-01
8.0E-01
1.0E+00
12.5
16
20
25
31.5
40
50
63
80
100
125
160
200
250
315
400
500
630
800
1k
1.25k
1.6k
2k
2.5k
3.15k
4k
Acceleration (m/s2) - EV1 3R -02,-08,-03
Sill (EV1)_Accel Mullion (EV1)_Accel Glass (EV1)_Accel
0.0E+00
2.0E-01
4.0E-01
6.0E-01
8.0E-01
1.0E+00
31.5
40
50
63
80
100
125
160
200
250
315
400
500
630
800
1k
1.25k
1.6k
2k
2.5k
3.15k
4k
5k
6.3k
8k
10k
Acceleration m/s2 - EV2 3R -02,-05, 03
Sill (EV2)_Accel Mullion (EV2)_Accel Glass (EV2)_Accel
0.0E+00
1.0E-04
2.0E-04
3.0E-04
4.0E-04
5.0E-04
12.5
16
20
25
31.5
40
50
63
80
100
125
160
200
250
315
400
500
630
800
1k
1.25k
1.6k
2k
2.5k
3.15k
4k
Velocity (m/s) - EV1 3R -02,-08,-03
Sill (EV1) Velocity Mullion (EV1) Velocity Glass (EV1) Velocity
0.0E+00
1.0E-04
2.0E-04
3.0E-04
4.0E-04
5.0E-04
31.5
40
50
63
80
100
125
160
200
250
315
400
500
630
800
1k
1.25k
1.6k
2k
2.5k
3.15k
4k
5k
6.3k
8k
10k
Velocity m/s - EV2 3R -02,-05, 03
Sill (EV2) Velocity Mullion (EV2) Velocity Glass (EV2) Velocity
341
EV1 TL14-167 (Heavy Mullion) EV2 TL14-170 (Light Mullion)
FIGURE D- 13: EV1 – CALCULATED SOUND PRESSURE LEVEL (DB) FIGURE D- 14: EV2 – CALCULATED SOUND PRESSURE LEVEL (DB)
FIGURE D- 15: EV1 – CALCULATED SOUND POWER LEVEL (DB) FIGURE D- 16: EV2 – CALCULATED SOUND POWER LEVEL (DB)
0
20
40
60
80
100
12.5
16
20
25
31.5
40
50
63
80
100
125
160
200
250
315
400
500
630
800
1k
1.25k
1.6k
2k
2.5k
3.15k
4k
Lp (dB) - EV1 3R -02,-08,-03
Sill (EV1) _Lp(dB) Mullion (EV1) _Lp(dB) Glass (EV1) _Lp(dB)
0
20
40
60
80
100
31.5
40
50
63
80
100
125
160
200
250
315
400
500
630
800
1k
1.25k
1.6k
2k
2.5k
3.15k
4k
5k
6.3k
8k
10k
Lp (dB) - EV2 3R -02,-05, 03
Sill (EV2) _Lw(dB) Mullion (EV2) _Lw(dB) Glass (EV2) _Lw(dB)
0
20
40
60
80
100
12.5
16
20
25
31.5
40
50
63
80
100
125
160
200
250
315
400
500
630
800
1k
1.25k
1.6k
2k
2.5k
3.15k
4k
Lw (dB) - EV1 3R -02,-08,-03
Sill (EV1) _Lw(dB) Mullion (EV1) _Lw(dB) Glass (EV1) _Lw(dB)
0
20
40
60
80
100
31.5
40
50
63
80
100
125
160
200
250
315
400
500
630
800
1k
1.25k
1.6k
2k
2.5k
3.15k
4k
5k
6.3k
8k
10k
Lw (dB) - EV2 3R -02,-05, 03
Sill (EV2) _Lw(dB) Mullion (EV2) _Lw(dB) Glass (EV2) _Lw(dB)
342
D.4.4 Analysis Summary
The resultant sound power and sound pressure levels are summarized at the receiving chamber (Table D
– 9).
TEST
SETUP
MEASUREMENT
NO.
PREDICTED LEVELS MEASURED
Lw (dB)
SILL
Lw (dB)
MULLION
Lw (dB)
GLASS
Lw
(total dB)
Lp rev
(total dB)
1
Lp rev (dB)
2
EV1
EV1 3R-02,-08,-
03
71.7 71 76 78.3 78 65.6
EV2
EV2 3R-02,-05,
03
67 68 75 76.4 75.9 79.5
TABLE D-9: SELECTED VIBRATION MEASUREMENTS FROM TEST EV1 AND EV2
1
(L
p predicted
) with calculation procedure
2
(L
p measured
) in the receiving chamber during the ASTM E90 measurements.
The predicted sound pressure levels (L
p predicted
) are compared to the measured sound pressure level (L
p
measured
) (Figure D-17 and Figure D-18). The results are inconclusive and subject to further study.
FIGURE D-17: EV1, PREDICTED AND MEASURED SOUND PRESSURE LEVELS
0
20
40
60
80
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
Lp, reverberant (dB)
EV2 RCV-02,-05, 03 and TL14-170
Lp,rev - RCV EV1 (predicted) Lp,rev - RCV TL14-167 (measured)
343
FIGURE D-18: EV2, PREDICTED AND MEASURED SOUND PRESSURE LEVELS
D.5 VIBRATION MEASUREMENT SUMMARY
The converted sound power levels (dB re 10
-12
W) from vibration acceleration levels (dB re 10
-6
m/s
2
) at
the receiving chamber for the glass, vertical and horizontal mullion are shown (Figure D-19).
FIGURE D-19: SUMMARY OF SOUND POWER AT EACH UVM MEASURED DURING PHASE 3
0
20
40
60
80
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
Lp, reverberant (dB)
EV2 RCV-02,-05, 03 and TL14-170
Lp,rev - RCV EV2 (predicted) Lp,rev - RCV TL14-170 (measured)
71 70.5
76
68 66
75
0
10
20
30
40
50
60
70
80
90
100
1 2 3
Sound Power Level dB re 10
-12
W
Curtain Wall System Elements
Summary of Sound Power Levels (dB) for EV1 and EV2 Configurations
(Mullion, Sill and Glass at Receiving Chamber)
Vertical Mullion Horizontal Mullion Glass
EV1 EV1
EV1
EV2 EV2
EV2
344
The vibration acceleration levels (dB re 10
-6
m/s
2
) measured at the vertical and horizontal mullion indicate
higher levels of acoustic energy at the receiving chamber than the glass.
Converting these vibration levels to sound power (dB re 10
-12
W) and correcting for the surface area of
each element, the highest level is at the glass. This indicates that the largest excited surface at the
receiving chamber (i.e. glazing) may be the dominating sound flanking path.
The analysis from the vibration measurements provides preliminary insight to what element contributes
most to the energy transmission of the curtain wall system. This work is subject to future refinement and
development in future studies.
Abstract (if available)
Abstract
The glass curtain wall is a cohesive design of glazing, aluminum framing, structural silicone, and neoprene gasketing. These components for unitized and non-unitized systems are technologically sophisticated and work together as a complex dynamic system. The intricate design accounts for the architecture of structural deflections, thermal properties, acoustic performance, moisture control, fire and smoke protection, amongst others. Therefore, the design and installation of these of these components should work cohesively together to provide an effective building performance, including provisions for good sound isolation. The sound isolation between occupied adjacencies located at the façade is highly influenced by the architectural composite of the various elements. Additionally certain parts within the composite can transfer sound energy more efficiently than others via sound flanking transmission paths. ❧ Sound flanking transmission that exists at the façade curtain wall and interconnecting partition presents a design challenge that reduces acoustic privacy and sound isolation design targets. This intersection can create an acoustic weakness within the curtain wall assembly and where it fastens to the building. Three architectural elements commonly contributing to this weakness are the curtain wall infill glazing, the aluminum mullion extrusions, and the partition connection joining the mullion to an interconnecting partition. The behavior of sound flanking transmission paths across each of these curtain wall elements is currently not well understood for all systems. These architectural mechanisms can create lateral sound paths and degrade the overall sound isolation integrity of the composite architecture. This is especially an issue when a high sound isolation performance between adjacent spaces is expected from an acoustically rated partition. ❧ Research on sound flanking paths in curtain wall systems has been carried out in theoretical statistical energy analysis (SEA) models, sound isolation prediction simulations, and with physical measurements on laboratory and field installations. However, most of these studies are composite and do not necessarily investigate the specific behavior of the separate system components. Typically the sound isolation performance of mullions measured in a laboratory or field may not clearly identify which curtain wall element most significantly contributes to overall performance. Acoustic products for curtain wall systems are emerging with the intent to improve overall sound isolation performance, which is an indication that this problem impacts the architectural practice. In order to improve the acoustic performance of the curtain wall system, the critical components attributed to the sound flanking transmission paths must be better understood, particularly at the mullion. Three elements associated with the architecture of the curtain wall system were selected and studied through a series of laboratory test measurements and sound isolation prediction calculations to determine potential improvement: the partition connection, the vertical mullion, and the curtain wall glazing. The testing method proposed is in accordance with ASTM E90, an acoustic testing procedure that measures the transmission loss (TL) of a building specimen between two reverberant test chambers. The single figure STC classification per ASTM E336 is obtained from these measurements. The STC is a commonly used amongst architects in practice to identify the sound level resistance of walls and floors. Approximately 80 acoustic laboratory tests were performed on select curtain wall elements and modified to identify the highest practicable acoustic performance that may be achieved. Additionally, an auxiliary set of vibration measurements were conducted at one of the test stages in order to examine the acoustic energy injection at mechanically connected elements of the curtain wall system. ❧ Comparisons between the independent test elements were examined in order to understand construction and performance benefits associated with achievable performance. The sound transmission loss data obtained from the laboratory test procedure is analytically calculated with the performance of an interior partition assembly to understand composite effects. Results from this testing method indicate how the performance of individual components influences a composite system and identifies elements that limit the achievable sound isolation. Although global variations of curtain wall designs exist in practice, the conclusions developed from the proposed experiment method are relevant to the specific curtain wall specimen typology measured and have relevance to similar systems. ❧ The research aims to enhance façade tectonic cohesion specific to acoustic design integration and to inform building engineering design and performance decisions.
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
The acoustic performance of double-skin facades: a design support tool for architects
PDF
Unitized double skin façade assemblies: achieving the needs of future envelope systems through construction and performance advances
PDF
Skin fit and retrofit: challenging the sustainability of curtainwall practice in tall buildings
PDF
Simplified acoustic simulation - Rutabaga Acoustics: a Grasshopper plug-in for Rhino
PDF
Evaluation and development of solar control screens: using daylight simulation to improve the performance of facade solar control screens
PDF
Aerophysics of double-skin facades: simulation-based determination of pressure coefficients for multi-story double-skin facades
PDF
Facade retrofit: enhancing energy performance in existing buildings
PDF
Mitigating thermal bridging in ventilated rainscreen envelope construction: Methods to reduce thermal transfer in net-zero envelope optimization
Asset Metadata
Creator
Valmont, Elizabeth
(author)
Core Title
Lateral sound flanking transmission at curtain wall mullions: an empirical investigation to identify controlling mechanisms
School
School of Architecture
Degree
Doctor of Philosophy
Degree Program
Architecture
Degree Conferral Date
2015-08
Publication Date
07/30/2015
Defense Date
05/11/2015
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
facade sound flanking transmission,glass curtain wall sound transmission,laboratory transmission loss testing,lateral facade sound isolation,OAI-PMH Harvest,unitized mullion sound isolation
Format
application/pdf
(imt)
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
Noble, Douglas (
committee chair
), Christoff, Jerry (
committee member
), Moore, James Elliott, II (
committee member
), Schiler, Marc (
committee member
)
Creator Email
elizabethvalmont@gmail.com,evalmont@usc.edu
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-c3-616695
Unique identifier
UC11303781
Identifier
etd-ValmontEli-3765.pdf (filename),usctheses-c3-616695 (legacy record id)
Legacy Identifier
etd-ValmontEli-3765.pdf
Dmrecord
616695
Document Type
Dissertation
Format
application/pdf (imt)
Rights
Valmont, Elizabeth
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Access Conditions
The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...
Repository Name
University of Southern California Digital Library
Repository Location
USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA
Tags
facade sound flanking transmission
glass curtain wall sound transmission
laboratory transmission loss testing
lateral facade sound isolation
unitized mullion sound isolation