University of Southern California Dissertations and Theses

3D printing of polymeric parts using Selective Separation Shaping (SSS)

(USC Thesis Other)

3D printing of polymeric parts using Selective Separation Shaping (SSS)

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content 3D Printing of Polymeric Parts using Selective Separation Shaping (SSS) by Hadis Nouri A Dissertation Presented to the FACULTY OF THE GRADUATE SCHOOL UNIVERSITY OF SOUTHERN CALIFORNIA In Partial Fulfillment of the Requirements for the Degree DOCTOR OF PHILOSOPHY (INDUSTRIAL AND SYSTEMS ENGINEERING) December 2018 Copyright 2018 HADIS NOURI ii ACKNOWLEDGEMENTS I would like to express my deepest gratitude to my PhD advisor, Dr. Behrokh Khoshnevis, for providing me with the great opportunity to work under his supervision. I appreciate his continuous support and insightful advice during my PhD studies. His passionate drive to find unexplored solutions paired with creativity and his profound and hands-on advising approach was extremely inspiring and I am truly grateful for having experienced these unique advising styles. I would like to thank my dear family, my sister Dr. Elnaz Nouri, my parents Jaleh Ostadhaji and Hamid Reza Nouri for their kind endless support during my PhD studies. I would like to also thank my dear grandparents, Mohammadesmaeil, Ziba, Leila and Nasrollah. Special thanks to Dr. Ifa Kashefi for her kind endless support. A special thank you to my dissertation committee members, Dr. Qiang Huang, Dr. Steven Nutt, Dr. Yong Chen, and Dr. Qiming Wang for their beneficial guidance and feedback. I also want to thank especially Dr. Steven Nutt who generously gave me access to his laboratory and research equipment. Without his precious support it would not be possible to conduct this research. Very special thank you to Dr. Jing Zhang who kindly transferred his knowledge to me. His endless technical support was extremely helpful in every step of this project. Thank you, Jing. I am also grateful to my friends Ali Kazemian, Dr. Xiao Yuan, Dr. Amir Mansouri, Dr. Ehsan Barjasteh, Dr. Valdimir Gorbach and Xiang Gao, for their support and encouragement during the past years. iii ABSTRACT Additive manufacturing (AM), or three-dimensional (3D) printing has enjoyed a recent surge of attention over the past few years. AM is a process in which digital 3D design data (CAD model) is used as an input to build physical objects by stacking layers of material. A variety of materials have been used by AM techniques and new materials and processing technologies are constantly being developed to improve current processes and to create new ones. 3D printing market size worldwide is estimated to be around $12 billion dollars by the end of 2018 and at 18 billion dollars by 2021 [1]. AM technologies have certain limitations hence there is a high demand for research in this area to help AM become a mainstream manufacturing technique. Selective Separation Shaping (SSS) is a new powder-based AM technology which was developed at USC with the goal of fabricating low cost, high resolution 3D parts. The main advantage of SSS is that this process enables building of fully functional pieces without the need of any intermediate binder or high cost laser operation. This process has been primarily applied to metallic, ceramic and composite materials including concrete with numerous test cases successfully built. Among all library of materials that have been processed by commercialized AM techniques so far, polymers have been the most-widely used material due to its large range of applications with 65 % of market share based on Sculpteo statistics report in 2018 [2]. The main goal of this research is to provide a framework for an effective selection of materials and process parameters in SSS for successful fabrication of polymer parts. Nylon 6,6 has been used as a starting base and several test cases are fabricated to identify the key performance factors in this process. Design of iv experiments is performed and interaction between separation material and base material is investigated to improve the surface quality of parts. Additionally, different heating mechanisms are studied to achieve better control over shrinkage and maintain an effective binding between layers. In the following sections, first, various approaches in AM for fabrication of polymer parts are presented in brief review; then, background and previous research efforts on SSS in literature are discussed. Section 4 encompasses a critique of past approaches and existing research gaps. In the sections following, research methodology and an experimental study of 3D printed parts built by SSS are presented. Finally, results and future research directions are discussed. v Table of Contents ACKNOWLEDGEMENTS ............................................................................................................ ii ABSTRACT ................................................................................................................................... iii 1. INTRODUCTION ................................................................................................................... 1 1.1. Existing commercial AM technologies for polymers ...................................................... 3 1.1.1. Vat photopolymerization based ................................................................................ 3 1.1.2. Material extrusion based ........................................................................................... 5 1.1.3. Binder jetting ............................................................................................................ 5 1.1.4. Powder bed fusion..................................................................................................... 6 1.2. Selective Heat Sintering ................................................................................................... 7 1.3. Selective Inhibition Sintering ........................................................................................... 7 1.4. Material jetting ................................................................................................................. 9 1.4.1. PolyJet ....................................................................................................................... 9 1.5. Overview materials processed by AM technologies ...................................................... 10 2. SSS PRINCIPLE ................................................................................................................... 12 3. STATEMENT OF PROBLEM ............................................................................................. 14 4. RESEARCH METHODOLOGY .......................................................................................... 15 5. BACKGROUND OF STUDY............................................................................................... 16 5.1. SSS – Metal .................................................................................................................... 16 5.2. SSS – Ceramic ................................................................................................................ 17 5.3. SSS - Cement based ....................................................................................................... 18 6. CRITIQUE OF OTHER APPROACHES ............................................................................. 21 vi 7. SELECTIVE SEPARATION SHAPING .............................................................................. 22 8. EXPERIMENTAL TESTS; SSS FOR POLYMERS ............................................................ 25 8.1. Material requirements for SSS for polymers.................................................................. 26 8.1.1. Selection of base powder ........................................................................................ 26 8.1.2. Preliminary experiments ......................................................................................... 31 8.1.3. Discussion ............................................................................................................... 31 8.1.4. Selection of separation powder ............................................................................... 31 8.2. Process parameter requirements for SSS ....................................................................... 38 8.2.1. Deposition of B-powder on the substrate................................................................ 38 8.2.2. Requirements for successive layer spreading and deposition ................................. 39 8.2.3. Separation trench shape in consideration of insertion depth .................................. 42 8.2.4. Nozzle insertion depth in relation to layer thickness .............................................. 43 9. SINTERING AND COALESCENCE ................................................................................... 45 9.1. Radiation ........................................................................................................................ 46 9.2. Convection ..................................................................................................................... 47 9.3. Experiments .................................................................................................................... 48 9.3.1. Bulk sintering in SSS .............................................................................................. 48 9.3.2. Layer by layer sintering setup- Radiation ............................................................... 48 9.3.3. Layer by layer sintering setup– convection ............................................................ 50 9.4. Pre-heating ..................................................................................................................... 51 9.5. Process parameters ......................................................................................................... 52 10. PRELIMINARY PART FABRICATION ............................................................................. 52 vii 11. CHALLENGES IN SSS FOR POLYMERS ......................................................................... 54 11.1. Weak bonding between layers .................................................................................... 54 11.2. Cumulative heating and warpage ............................................................................... 54 11.3. Shift between layers .................................................................................................... 55 12. HYPOTHESES ...................................................................................................................... 56 12.1. Hypothesis 1: .............................................................................................................. 56 12.2. Hypothesis 2 ............................................................................................................... 61 12.3. Hypothesis 3 ............................................................................................................... 64 13. EXPERIMENTAL SETUP AND PROCEDURES ............................................................... 65 13.1. Impact of S-powder deposition on part surface quality .............................................. 65 13.2. Impact of S-powder choice on part surface quality .................................................... 66 13.3. Effect of heat capacity of S-powder on surface quality .............................................. 73 13.4. Shape of separation surfaces created .......................................................................... 77 13.5. Assessment of ease of separation ............................................................................... 79 13.6. Discussion ................................................................................................................... 80 13.7. Development of a model for separator powder deposition in the base material ........ 82 13.7.1. Angle of repose of B-powder .................................................................................. 82 13.8. Development of a model for separator powder deposition in the base material ........ 83 13.8.1. Total amount of S-powder required for any geometry ........................................... 84 13.8.2. Maximum printing speed viable under maximum deposition ................................ 86 viii 14. SELECTIVE HEAT SINTERING IN SSS ........................................................................... 87 14.1. Selective sintering by forced convective sintering ..................................................... 89 14.2. DOE for determining significant factors in forced convective sintering .................... 91 14.3. Micro tensile test ........................................................................................................ 97 14.4. Total shrinkage and density ...................................................................................... 102 14.4.1. Discussion ............................................................................................................. 103 14.5. Reduction of waste material in SSS ......................................................................... 104 14.6. Discussion ................................................................................................................. 106 14.7. Demonstration parts .................................................................................................. 106 15. CONCLUSION ................................................................................................................... 107 16. PROPOSED FUTURE RESEARCH .................................................................................. 108 17. REFERENCES .................................................................................................................... 110 ix List of Figures Figure 1. Triangular meshes generated from CAD model .............................................................. 2 Figure 2: Summary of current polymer fabrication technologies by AM ...................................... 2 Figure 3. Demonstration of Clip process [8] .................................................................................. 4 Figure 4. Three-dimensional printing process with curing light [10] ............................................. 6 Figure 5. An illustration of Selective Laser Sintering [12] ............................................................. 7 Figure 6. Polystyrene test cases built by SIS [16] .......................................................................... 8 Figure 7. Material jetting process [18] ............................................................................................ 9 Figure 8. a) SSS deposition nozzle assembly, b) needle, c) needle opening, d) drawing ............. 13 Figure 9. SSS process steps [12] ................................................................................................... 14 Figure 10. Bronze (left) and Steel (right) parts built by SSS [20] ................................................ 17 Figure 11. Interlocking ceramic parts built by SSS [21] .............................................................. 18 Figure 12. Deposition of thin wall of S-powder within a 12-mm thick layer ............................... 19 Figure 13. SSS micro scale (left) and meso scale (right) nozzle dimensions ............................... 19 Figure 14. Portland cement and gypsum parts built by SSS (above parts are single layer and the lower part is double layer) [22] ..................................................................................................... 20 Figure 15. Nozzle movement inside B-powder. a) top view in X direction, b) top view in Y direction, c) isometric view in X direction, d) front view in Y direction after rotation ............... 23 Figure 16. SSS beta machine modified for fabrication of polymeric parts .................................. 24 x Figure 17. Flow diagram of SSS-polymer ................................................................................. 25 Figure 18. Particle size distribution of Copolyamide and Vestamelt 171 .................................... 27 Figure 19. Particle size distribution of Vestamelt-470 and Vestosint .......................................... 27 Figure 20. SEM images of Copolyamide(top left), Vestamelt 171(top right), Vestamelt 470 (bottom left), and Vestosint (bottom right) ................................................................................... 28 Figure 21. DSC curves of polyamide samples .............................................................................. 29 Figure 22. Chromium Oxide powder (left) and SEM image (right) ............................................. 32 Figure 23. Mechanical interlocking of powders for non-spherical and spherical particles [27] .. 33 Figure 24. 20 Micron Chromium Oxide (source: Praxair co.) ...................................................... 34 Figure 25. Packing density overview versus particle size [27] ..................................................... 35 Figure 26. Deposition rate vs. frequency for Cr2O3 ...................................................................... 37 Figure 27. Controlled Cr2O3 deposition over polymer ................................................................ 38 Figure 28. Gaps created by separation lines ................................................................................. 39 Figure 29. Gap created by movement of nozzle inside base material .......................................... 39 Figure 30. S-powder deposition volume relative to slot ............................................................... 41 Figure 31. Gap shape in relation to insertion depth ...................................................................... 43 Figure 32. Printing failure due to large insertion depth (left), correct needle insertion depth (right) ............................................................................................................................................ 44 Figure 33. Undesired sintered bridges in between walls .............................................................. 45 Figure 34. SEM pictures of powder before(A) and after sintering(B) .......................................... 46 xi Figure 35. Glow distribution of Kanthal (a), Ni 32(b), and Ni 26(c) ........................................... 49 Figure 36. Linear radiation heat setup .......................................................................................... 50 Figure 37. Hot air linear heater setup ............................................................................................ 50 Figure 38. Heaters attached to platform (a) and build tank(b) ...................................................... 51 Figure 39. 3D test case fabricated by Copolyamide and Vestamelt 171 ...................................... 53 Figure 40. Enlarged view of 2.5 gear fabricated by Copolyamide ............................................... 53 Figure 41. Partially sintered sample with weak bonding between layers ..................................... 54 Figure 42. Warpage caused by cumulative heating ...................................................................... 55 Figure 43. SEM images of layers after separation(a) and magnified damage (b) ........................ 57 Figure 44. Gaps created by nozzle with (right) and without(left) edge removal .......................... 57 Figure 45. Top view of 3D part after separation ........................................................................... 58 Figure 46. Lines with S-powder deposition decreasing from left to right .................................... 60 Figure 47. Cylinder test case made by SSS .................................................................................. 60 Figure 48. Surface plot of convection sintering at 11mm distance (left) versus 4 mm (right) from surface layer .................................................................................................................................. 62 Figure 49. SEM image of Vestosint at 150 magnitudes sintered at distance 11mm(left) and 4mm (right) ............................................................................................................................................ 62 Figure 50. Surface of Vestosint under in low (left) and high level heat intensity ........................ 63 Figure 51. Surface of Vestosint under IR heating in low and high-level heat intensity (4 mm distance) ........................................................................................................................................ 63 xii Figure 52. Surface plot of Vestosint under Radiation heating in low and high level heat intensity (4 mm distance)............................................................................................................................. 64 Figure 53. Separation of part after nozzle movement with no S-powder deposition ................... 65 Figure 54. Magnified image of separation line cross section and connected polymer bridges .... 66 Figure 55. Interlocking effect between powders [42] ................................................................... 67 Figure 56. Demonstration of sintering contact line of S-powder candidates with B-powder ....... 69 Figure 57. Top view of sintered layers over the cross section ...................................................... 70 Figure 58. Test cases fabricated by tungsten, black alumina, and soda lime from left to right .... 72 Figure 59. Increase in atomic energy vibrations under heat [44] ................................................. 74 Figure 60. Test cases fabricated by soda lime, chromium oxide and black alumina (from left to right) as S-powders ....................................................................................................................... 76 Figure 61. Sample block test case (left), top view with Micro Vu ............................................... 78 Figure 62. Shape of the separation surfaces from the top view (S-powder: black alumina) ........ 78 Figure 63. Shape of walls from the top view (S-powder: soda lime) ........................................... 78 Figure 64. Force measurement along separation line ................................................................... 79 Figure 65. Blocks built by black alumina (left) and soda lime (right) S-powders ........................ 80 Figure 66. Isometric view and front view of test case fabricated by black alumina ..................... 82 Figure 67. Angle of repose of bronze, Copolyamide, Vestamelt, Vestosint (Left to right, respectively) .................................................................................................................................. 83 Figure 68. Created gap by nozzle with outer diameter of D and insertion depth of L ................. 84 xiii Figure 69. Schematic of S-powder deposition inside the gap ....................................................... 85 Figure 70. Impact of speed on shapes of gaps created: low (left) and high (right) speed ............ 86 Figure 71. Schematic of forced convective heater ........................................................................ 89 Figure 72. Demonstration of sintering by forced convective heater ............................................. 90 Figure 73. Sintering profile around the contour of the part .......................................................... 91 Figure 74. Sintered lines under different heating conditions ........................................................ 94 Figure 75. Half normal plot for layer thickness ............................................................................ 96 Figure 76. Engineering drawing of dog bone used for experiments (all dimensions are in mm) . 97 Figure 77. 3D printed dog bones................................................................................................... 98 Figure 78. Deben 5KN micro tensile tester .................................................................................. 99 Figure 79. Stress strain curve of specimen 1,2 ........................................................................... 100 Figure 80. Stress strain curve of specimen 3,4 ........................................................................... 101 Figure 81. Stress strain curve of specimen 5,6 ........................................................................... 101 Figure 82. Stress strain curve of specimen ................................................................................. 102 Figure 83. Waste material in SSS by linear sintering ................................................................. 104 Figure 84. Waste material reduction by spot sintering ............................................................... 105 Figure 85. Polyamide samples built by SSS ............................................................................... 107 xiv List of Tables Table 1. Processed materials by AM technologies ....................................................................... 10 Table 2. AM methods for fabrication of polymer part geometry creation methods ..................... 11 Table 3. Average particle size versus polyamide grade ................................................................ 27 Table 4. Thermal properties of polyamide grades ........................................................................ 30 Table 5. Percentage porosity for polyamide grades ...................................................................... 30 Table 6. Specification of separation powder................................................................................. 33 Table 7. Powder flow for 20 and 40 micrometers powders .......................................................... 35 Table 8. S-powder deposition test ................................................................................................. 36 Table 9. Process parameters in SSS for Nylon 6,6 ....................................................................... 52 Table 10. Maximum deposition setting ........................................................................................ 59 Table 11. Properties of S-powder ................................................................................................. 68 Table 12. Average deviation for each S-powder .......................................................................... 71 Table 13. S-powder thermal properties ......................................................................................... 75 Table 14. B-powder thermal properties ........................................................................................ 75 Table 15. S-powder adhesion to B-powder measurement ............................................................ 77 Table 16. S-powder separation tensile stress measurements ........................................................ 80 Table 17. Factors of importance ................................................................................................... 92 Table 18. Design matrix (1 at high level and -1 at low level) ...................................................... 93 xv Table 19. Average values for three replications ........................................................................... 95 Table 20. Sintering trials ............................................................................................................... 99 Table 21. Stress strain curve of specimen ................................................................................... 103 1 1. INTRODUCTION 3D printing is one of the advancements in rapid prototyping. This technology offers the ability to build intricate geometries by eliminating constraints on machining and tooling. The wide range of techniques such as multi-material printing, multi-color printing, manufacture of complicated geometries, and building high resolution physical models have brought 3D printers to public attention and expanded its usage in a variety of applications. Examples include, in the aerospace industry to fabricate complex metal components such as jet engines, in biomedical engineering for prototyping of human organ implants and reconstructive surgery aids, in dentistry for building high resolution dental crowns, in the construction industry for building shelters and quick concrete structures, in fashion and architecture to test and visualize intricate design concepts, etc. [3]. Although AM has developed rapidly and is receiving more attention these days, it is still in its developing phase. Intensive research efforts have been primarily focused on developing new additive manufacturing technologies since late the 1980s when the first 3D printing technique, Stereolithography (SLA), was commercialized. The other most widely used techniques are Fused Deposition Modeling (FDM), Selective Laser Sintering (SLS), Multijet (Polyjet), and Digital Light Processing (DLP), respectively, based on 2017 Forbes report [4]. The common feature among all AM technologies is that a digital CAD model is sliced and converted to triangular meshes which are stored in the so called STL file format. The geometrical model is then sliced to determine the layers to be printed by the machine (Figure 1). 2 Figure 1. Triangular meshes generated from CAD model According to existing research studies [5], polymer AM methods can be categorized into 5 main groups: i. Material extrusion, ii. Powder bed fusion, iii. Vat photopolymerization, iv. Material jetting, and v. Binder jetting. This categorization does not include polymer composites which can be processed by Direct Energy Deposition (DED) and Laminated Object Manufacturing (LOM) processes as well. In the chart below, each category followed by related AM technologies are summarized. Figure 2: Summary of current polymer fabrication technologies by AM AM Polymer facrication Extrusion based FDM, 3D bio printing Power bed fusion based SLS, SHS Photo- polymerization based SLA, DLP, CLIP Material jetting based Inkjet printing Binder jetting based 3DP 3 1.1. Existing commercial AM technologies for polymers This section gives an overview of existing technologies to address current research gaps. In the process of improvement, it is crucial to be aware of the capabilities of each individual technology to compare current techniques or even for future improvement. In the following section, a brief introduction of AM technologies developed for the fabrication of polymer parts are discussed. 1.1.1. Vat photopolymerization based Stereolithography Sterolithography is a layered rapid prototyping process and was first developed in the late 1980s. The first commercialized RP system based on Stereolithography is called Stereolithography Apparatus or SLA. SLA has always remained as one of the most powerful and versatile rapid prototyping processes [6, 7]. In this technique, ultraviolet laser radiation is used to cure a vat of liquid resin to convert the liquid to solid polymer. The photopolymer is selectively cured on a layer-by-layer basis where the cured area corresponds to the desired cross-section shape taken from the three-dimensional CAD model. The solidification then, takes place within the desired layer thickness. The recoating blade moves across the surface to apply each new layer. This procedure repeats until the final shape of the object is reached. As part of the design, support structures are used to enable the fabrication of hanging parts and parts with complex geometries. Digital Light Projection DLP or Digital light projector is another AM technique developed to fabricate parts using photo-curing material. This process is similar to SLA in that the part is formed in vats of photopolymer as the primary method of building objects. DLP; however, uses a digital projector which is exposed once to a 4 single image of each layer across the platform. This exposure selectively solidifies layers of squared voxel. In the SLA process, the light source of the curing photo-reactive polymer is the laser which draws rounded lines as it moves along. Therefore, DLP is desirable for high speed printing where the high quality in small-scale details is not required. For higher resolution parts with fine details SLA is still preferable; however, it accounts for tradeoffs in cost and speed. Clip- bottom up CLIP or Continuous Liquid Interface Production was patented in 2014 by EiPi Systems Inc. In this process, parts are formed by having continuous control over the ratio of light exposure and Oxygen in contact with each layer. The UV light solidifies the photopolymer while oxygen works as an inhibitor. Maintaining a balance between these two key factors results in smooth attachment of layers upon raising up build platform during the process. Figure 3. Demonstration of Clip process [8] This phenomenon also expedites the printing process by continuous fabrication. 5 1.1.2. Material extrusion based Fused Deposition Modeling FDM is another rapid prototyping process that builds parts on a layer-by-layer basis. In this process, heated thermoplastic material such as filament is extruded from the tip of a nozzle. There are several nozzles used in this technology that are separated for product material deposition and support structure material deposition. The thermoplastic material should solidify upon adhering to previous layers. The STL file is mathematically sliced and oriented by the software. In some cases, support structures are generated. Compared to other AM technologies, the FDM process is highly complicated, thus, it is unlikely to create a variety of theoretical models for the prediction purpose [9]. 1.1.3. Binder jetting Three-dimensional printing Three-dimensional printing, also called binder jet printing was initially developed at MIT in early 90s. 3DP is a powder-based AM process. In this process, upon spreading each layer a binder solution is selectively deposited to join powder particles on desired areas. When binder printing is complete, the whole part is transferred to the furnace for sintering to obtain the desired strength. Finally, unbound powder is removed from the part body. As an alternative heating procedure, layer by layer sintering can be done by inserting a curing light on the top of the platform where it can be exposed to the layer in which binder is deposited (Figure 4). 6 Figure 4. Three-dimensional printing process with curing light [10] Currently 3DP is one of the most commonly used AM techniques due to its platform flexibility to fabricate parts using metal, ceramics, polymers and composites. Moreover, the application of this process is expanded to larger scale fabrications and is even being used in building architectural structures [11]. Ongoing research efforts have been dedicated to increase printing speed in binder jet printing which would help this process overpass other AM techniques in terms of speed and cost. The main disadvantage of this process on the other hand is contamination and impurity caused by binder bounding. 1.1.4. Powder bed fusion Selective Laser Sintering Selective laser sintering is a technique in which particles in a powder bed are selectively sintered layer by layer to form the final part based on the three-dimensional CAD model. SLS is generally meant for rapid prototyping and rapid tooling. In fact, the precision of the SLS rapid tool depends on several factors such as the CAD model, slicing algorithms, data transfer, device motion 7 resolution, powder granulometry, beam offset and shrinkage. A variety of materials in powder form can be applied in SLS. Examples include polymers, ceramics, metal, and composites which sinter under laser exposure. Figure 5. An illustration of Selective Laser Sintering [12] 1.2. Selective Heat Sintering Selective Heat Sintering (SHS) was initially patented in 2009 by Blue Printer Company in Denmark to offer a low-cost alternative to SLS [13]. The principle of this process is based on thermal sintering. Upon depositing each layer, a thermal print-head moves back and forth to selectively melt and solidify the powders. During fabrication, the powder bed is always held at an elevated temperature, i.e. 2 to 5 C below the thermoplastic’s melting point so when the print head is scanned, it only needs to slightly raise the temperature to bond the layers. 1.3. Selective Inhibition Sintering Selective Inhibition Sintering (SIS) process is another AM technique which was developed the early 2000s at USC by Dr. Behrokh Khoshnevis [14]. In SIS, parts are fabricated based on powder sintering. In this process, a sintering inhibition agent which is in liquid form is deposited by an inkjet printer on 8 each layer a commercial inkjet printer at the boundary profile. The inhibitor solution prevents-powder particles from fusing at the treated sections. This process is similar to 3DP in terms of using a liquid agent to form the final part. However, in this process, the printed inhibitor works as thin sacrificial mold which defines the part body within its volume. The main advantage of SIS to 3DP is that the liquid is deposited on the boundary profile, therefore, avoiding any contamination inside the part. In addition, printing speed is increased by only printing on a perimeter instead of the whole layer area. There are three main consecutive steps in making parts by SIS: (I) Powder is uniformly spread on the build tank, (II) inhibitor solution is deposited on the boundary profile on each layer based on 2D cross sections of the 3D model, (III) A heater is moved along the x axis and the layer is sintered (except boundary profiles). Back to step1, a new layer of powder is spread on the previous layer and the procedure is repeated (Figure 5). Finally, sintered and un-sintered regions are separated at the boundary. Variety of materials besides polymers have been used in this process including ceramic, metal. [15]. Figure 6. Polystyrene test cases built by SIS [16] 9 1.4. Material jetting 1.4.1. PolyJet In the PolyJet technique, polymer liquid is deposited (jetted) onto the surface of a build plate in the form of droplets. The Ultraviolet (UV) light is then exposed on the surface and jetted polymer is cured. The build tank is lowered and again this process is repeated until the 3D part is fabricated. Material jetting is very similar to inkjet printing. The difference is that instead of ink, photocurable polymer is deposited on the tray. Once printing is finished, most of the time the part can be used right after without the need for post-processing. If any supports are used, they can be removed by hand or water jet. The main advantage of this process is allowing for multi-material printing, using the same process and the same 3D model. Using multi-materials in one physical object provides additional functionalities to a single part eliminating the need of multiple builds. Material jetted 3D parts can be built with high dimensional accuracy and low residual stress [17]. The application of this process is however limited due to weak mechanical properties of final parts. Figure 7. Material jetting process [18] 10 1.5. Overview materials processed by AM technologies Table (1) provides a summary of different material used by main additive manufacturing techniques. Table 1. Processed materials by AM technologies AM technology Polymer (powder/ resin) Ceramic Metal Sand Composite Plastics (filament) Fast-setting (Cement/Gypsum) SLA Yes No No No No No No SLS Yes Yes Yes Yes Yes No No FDM No No No No Yes Yes No 3DP Yes Yes Yes Yes Yes No Yes DLP Yes No No No No No No DMLS No No Yes No No No No SHS Yes No No No No No No SLM No No Yes No No No No DED No Yes Yes No Yes No No LOM NO Yes Yes No Yes No No SIS Yes Yes Yes No No No No SSS Yes Yes Yes Yes Not tested No Yes 11 SSS will provide a new solution to AM of materials which is based on separation principle. Table 2 summarizes the AM 3D printing fabrication methods for polymer fabrication. Table 2. AM methods for fabrication of polymer part geometry creation methods AM technology 3D printing method SLA UV light beam/Laser SLS Laser FDM Extrusion 3DP Binder DLP Digital projector screen DMLS Laser SHS Heat SLM Laser SIS Separator deposition (Liquid) SSS Separator deposition (Solid) In the next section, SSS is introduced and process principles are thoroughly presented. 12 2. SSS PRINCIPLE Selective Separation Shaping (SSS) was invented at the University of Southern California by Dr. Behrokh Khoshnevis [19]. In this approach, a part is selectively separated in the powder tank from its surroundings with a thin wall of barrier material called “separator” (S-powder). Like other powder- based additive manufacturing processes, initially a uniform layer of base powder is delivered into the build tank. This material is called a base material (B-powder). After one layer is paved, a nozzle is inserted into the base powder and the printing process starts. At his step, the nozzle which is composed of a miniature hollow tube moves along the contour of the part layer. (Figure 8). As this movement is initiated inside the B-powder, a vibrational stimulator is applied to deposit S-powder through the opening of the nozzle into the gaps that the nozzle movement creates. Subsequently, another layer of powder is spread on previous layer and process continues until all layers are processed. Different from a majority of other AM techniques, SSS enables fabrication of a wide class of materials. In other words, any material in powder form that can be solidified upon heating such as ceramics, metal, and polymers may be used. The bonding medium is SSS does not always need to be heat. In fact, SSS can process other materials such as Portland cement (lime based), sorrel cement (magnesia based), gypsum and a multitude of other powder materials that would cure and bond when wetted by water. Unlike common additive manufacturing processes, a different tool path method is required for SSS since this process only considers part boundary and disregards the part core. Nozzle rotation is also programmed into the toolpath which traces the contour of the slices. This rotation helps obtain the most effective filling of gaps along the path. 13 Figure 8. a) SSS deposition nozzle assembly, b) needle, c) needle opening, d) drawing a b c d 14 Figure 9. SSS process steps [12] Parts fabricated by SSS are supposed to possess similar or better mechanical properties as other additive manufacturing techniques that use binder instead of heat. The intricacies of this process, however, demand research on both process-ability of each material and their post-processing steps. In previous SSS research efforts metal and ceramic parts fabrication have been studied. The aim of this research is to expand the application of Selective Separation Shaping for the fabrication of polymeric parts. 3. STATEMENT OF PROBLEM Although AM has developed rapidly and is receiving increasing attention, it is still in early development stage and still significant amount of improvements are expected for AM machines to become mainstream manufacturing equipment. Improvements are expected in the following direction: i. Increasing speed of printing by maintaining quality of printed parts, ii. Online monitoring of process 15 parameters and in-situ sensing techniques to identify deviations and correct them during the process, iii. Apply new materials with desired chemical properties and microstructures. iv. Reducing waste material and further research on reusing recycled materials. v. Fabrication of multi-functional parts and control of material properties and their performance. vi. New AM technologies which can process a wider range of materials in a single platform without a need for retooling. To further expand application of AM to functional engineering parts rather than prototypes, several challenges need to be addressed. There is high demand for a technology that reduces time-to-market and can produce lower cost parts while maintaining high quality. Additionally, a machine with the flexibility in applying multiple classes of material can be more affordable for a firm than multiple machines that each process a certain class of materials. Selective Separation Shaping (SSS) was developed with the goal of fabricating low cost, high resolution parts in a single multi-material platform. In this research, influential factors in successful fabrication of polymeric parts using SSS are studied with the aim of improving the quality of parts made with the process. 4. RESEARCH METHODOLOGY The following steps will make up the core structure of this research study: 1. A review of SSS Background Study: This part includes the literature review of the parts. 2. SSS polymer Test Setup: Includes setting up the right type of experiments, parameters, boundary conditions and assumptions to mimic the real SSS condition. 3. Validation of the Test: To validate the developed tests, some preliminary analysis is conducted. The results can be compared with the existing results from the other researchers. 16 4. Preliminary experiments: The initial experiments include the simplified SSS process. The most simplified condition -two layers on top of each other under different sintering conditions is investigated. 5. Primary SSS experiment: This step includes more complex geometries with multiple layers on top of each other. A series of test cases will be designed to investigate the bonding strength, and the different bonding mechanism. 6. Discussion and analysis of the results: The previous results are investigated in this section to determine the best bonding mechanism that leads to the least shrinkage over the layers that can make the parts comparable with other 3D printing processes. In the following section, a background of SSS and materials that have been previously processed by SSS are discussed. 5. BACKGROUND OF STUDY 5.1. SSS – Metal Initially, metallic parts fabrication using SSS was studied [20]. The motivation behind SSS is offering a technology with low equipment investment, high production rate, material flexibility and potential for larger scale fabrications. In the meso-scale sintering-based SSS process, which has been reported in the literature, a thin wall of high melting point separator powder material (S-powder) is deposited within the metallic base powder material (B-powder) to form a barrier at the boundary of each layer. This barrier creates a separation between the part and surrounding material, which allows for the separation of the part from the surrounding powder after sintering is complete. Finally, the part is removed from the platform and inserted into sintering furnace. After sintering, the part is easily 17 separated from the surroundings as the S-powder remains in powder state due to its relatively higher sintering temperature. Figure 7 represents samples printed with bronze and steel powder material. Both powder particles are within 45 micrometers diameter and S-powder material used is tungsten W90. Figure 10. Bronze (left) and Steel (right) parts built by SSS [20] 5.2. SSS – Ceramic In addition to metal, Lunar regolith simulant (JSC-1A)2 material was processed by SSS and multiple samples were built. In Figure 8, interlocking ceramic tiles are shown. As properties of metals and ceramics greatly differ, their sintering properties also change. Sintering properties dictate proper S- powder material. For lunar regolith simulant, aluminum oxide (𝐴𝑙 2𝑂 3) was found to perform as an effective separator. 18 Figure 11. Interlocking ceramic parts built by SSS [21] 5.3. SSS - Cement based Application of SSS for large-scale fabrication is under investigation. A new beta machine has been developed to show the potential of this technology. The target category of material for large scale fabrication is fast-setting material including cement (lime based), sorrel cement (magnesia based) and gypsum-based composites. In a scaled-up version of SSS, thickness of each progressive layer is significantly larger relative to meso-scale fabrication. Suitable layer thickness can be achieved by adjusting the height of the opening on the nozzle. During initial experiments, a nozzle with 12 mm slot has been designed and tested. 10mm 19 Figure 12. Deposition of thin wall of S-powder within a 12-mm thick layer Figure 13. SSS micro scale (left) and meso scale (right) nozzle dimensions Changing the base material from metal to cement requires a different post-processing step. Consolidation at this stage is achieved by applying an adequate amount of the bonding liquid, i.e. water for hydraulically activated powders or any binder, on the surface of the top layer. The liquid gradually penetrates inside the entire part body and part consolidation starts through chemical reaction. Depending on the base material choice, the initial weak consolidation may take within one hour after which the part is strong enough to be removed from the build tank. The curing agent is added periodically to the part to achieve good part strength through a more complete cement hydration process. Finally, the part is separated from the surrounding material as shown in Figure 14. A brush or air blow can be used to remove the extra separator material away from the obtained part. 20 Figure 14. Portland cement and gypsum parts built by SSS (above parts are single layer and the lower part is double layer) [22] Among other S-powder materials, glass-powder with the grain size of 45 µm was tested for the cementitious SSS experiments. Application of SSS can further expand to rapid fabrication of urban structures, public art, building facades and pavement engineering. The application of SSS for building planetary landing pads with inter-locking tiles made of high temperature ceramics using in-situ material was demonstrated in a NASA sponsored project. The technology won The Grand Prize in an international competition sponsored by NASA. 10mm 21 6. CRITIQUE OF OTHER APPROACHES Previous SSS research has offered insight into many unknown parameters in the SSS process. Zhang et al focused on high temperature sintering material and successfully fabricated metal and ceramic parts. Zhang used tungsten and aluminum oxide as the main separator agents for this purpose. He also studied the impact of frequency, shape wave, powder particle shape, size and density on deposition rate. His selection of S-powder is however limited to metallic and ceramic parts and further research needs to be done for polymers. Moreover, bulk sintering approach used in SSS for metal and ceramic has limitations and is not applied to polymers as previously discussed. Therefore, layer by layer sintering approach needs to be carefully studied and implemented. Research also should be conducted to investigate powder interactions and the bonding of layers as SSS parts will obtain comparable strength as other approaches. SSS platform enables fabricating of multiple classes of materials which further expand its application to other areas. This process is not only limited to small scale fabrications, but it also has potential for fabricating on demand large scale structures. For polymeric part fabrication, due to the operation principle in SSS (printing on the boundaries instead of the part body), this process can compete with other processes such as FDM and SLS in printing speed. Multiple nozzles can print at once on the platform without interfering with one another to build multiple objects simultaneously. Compared to other powder-based polymer processes, the end-user product cost will be low as this process does not require laser scanning. The separating material can be purchased at low cost. Additionally, a variety of polymers can be applied to the machine and be tested in powder form. 22 7. SELECTIVE SEPARATION SHAPING An SSS machine which was initially fabricated for metallic parts has been slightly modified to process polymeric material. The electronic system of SSS is composed of a main motor controller which is a KFLOP that constantly communicates with the machine to send motion commands. The motions of the machine consist of XYZ axes the movements along which are controlled by adjusting speed, and the position of each axes. Sensors are also used to send feedback signals to the controller to adjust the initial and end positions. There are two powder material tanks with pistons which have Z axis motion to precisely raise up and go down to feed powder for paving successive layers. There is also a nozzle rotational movement (Figure 15) which is used to align the nozzle along the opposite of the nozzle movements in the XY plane. 23 Figure 15. Nozzle movement inside B-powder. a) top view in X direction, b) top view in Y direction, c) isometric view in X direction, d) front view in Y direction after rotation The other controlled system on SSS is the deposition system which is composed of a piezo disk as well as a vibration generator. The wave length, shape and frequency of vibration are directly controlled using a wave generator. Deposition triggering and stopping is controlled by the controller software. A layer by layer sintering system has been implemented on the machine for initial experiments (Figure 16). There are two temperature controllers used to keep track of temperature during fabrication. One sensor is inserted inside the build tank to monitor the powder bed temperature (the powder bed is actively heated using a heating element to minimize deformation caused by heat differential during a b c d 24 sintering). The second sensor is placed inside the machine chamber to sense the air temperature inside the enclosed machine chamber. Figure 16. SSS beta machine modified for fabrication of polymeric parts The flow diagram of SSS for polymeric material is shown in Figure 17. Initially, the STL file that is generated is applied to the software to generate G code. In the case of printing by SSS, the boundary profile of each 2D cross section is extracted from G code and converted to sliced paths necessary for printing. The rotational movements are also synchronized with XY movements of the nozzle and controlled by the software such that they work simultaneously. Moreover, initially in the SSS machine, the roller which is attached to the nozzle to pave material on the platform was raised up on the build tank to avoid disturbing deposited areas. In SSS for polymers, the sequence has been altered such that the roller is also used to smooth the surface of polymer after printing on, once layer is done which will be further discussed in the following sections. 25 Input: STL file Conversion to G code Boundary path extraction SSS printing start Platform heating up S-powder deposition Build platform down Blade passes Layer sintering Final layer? Printing ends Cooling down Seperation Output: Final part B-powder spreading No Yes Figure 17. Flow diagram of SSS-polymer 8. EXPERIMENTAL TESTS; SSS FOR POLYMERS In this section, we intend to study the basic principles of parts built by SSS for polymers. This research work follows the research methodology of the previous studies on metal and ceramic. Accordingly, the selection of base and separation materials, vibration frequency, and separation gap size are studied. The main goal is to develop a framework for selecting S-powder material and process parameters. Several test cases were designed and specimens were built for the designed experiments. The following sections present the experimental procedure and the results of the experiments. 26 8.1. Material requirements for SSS for polymers One of the key factors in the success of SSS made parts are in appropriate selection of suitable base and separation material. In this section candidate materials to be processed in SSS are discussed. 8.1.1. Selection of base powder There are four characteristics studied: 1. Thermoset/thermoplastic, 2. Powder size distribution 3. Particle shape, 4. Melting and crystallization behavior, 5. Porosity. Nylon has been one of the most popular polymers applied in AM techniques due to its properties and relatively high mechanical strength. It also bends and returns to its original form, rather than fracturing suddenly [23-25]. We have focused on nylon material throughout this research. Different types of PA 6,6 material; however, has been studied to be able to better characterize the impact of the base material properties on the quality of SSS parts. Four Polyamide powders with different material properties were selected and tested. All selected grades are commercial polyamide powders and are among most widely used types. Particle size and distribution Figure 18 and 19 show the particle size distribution for PA-Vestosint, 171-P1 and 470-P1, which have all been purchased from Evonik (a major German polymer material company) and Copolyamide provided by Arkema company. 27 Figure 18. Particle size distribution of Copolyamide and Vestamelt 171 Figure 19. Particle size distribution of Vestamelt-470 and Vestosint The average participle size for each powder is shown in Table 3. Table 3. Average particle size versus polyamide grade Polyamide grade Average particle size (Microns) Vestosint 139.81 Vestamelt 171 150.52 Vestamelt 470 180.52 Copolyamide 155.61 28 Particle shape All types of polyamides have non-spherical shape with sharp edges (cryogenic milled particles). This is more significant in Vestamelt-470. Vestmalet 470 and Vestamelt 171 seem to have a larger ratio of small size particles relative to other two types. In Copolyamide the distribution is more significantly skewed to the left as particles are mostly large with some very small particles. In Figure 20, SEM images for investigated polyamide powders are shown in the same magnification of x150: Figure 20. SEM images of Copolyamide(top left), Vestamelt 171(top right), Vestamelt 470 (bottom left), and Vestosint (bottom right) Melting and crystallization behavior The next step was to obtain information regarding thermal properties of all candidate polymers. Based on existing literature in application of polymers in AM [26], as the window between melting point and crystallization for type of polymer increases, less shrinkage is observed and hence found to be most suitable for SSS processing application. Although all types of polymers show similar particle shapes 29 and relatively close size distribution, it is observed that crystallization behavior plays major role in characterizing their performance in the SSS process. Differential Scanning Calorimetry (DSC) was used to determine the impact of melting and crystallization behavior of Vestosint, Vestamelt 470 and Vestamelt 171 on SSS processability. Samples of 6 mg +/- 0.5 were heated with the rate of 10 C/min up to 130 C followed by the same cooling rate after heating is complete (Figure 21). Figure 21. DSC curves of polyamide samples The thermal properties of Copolyamide are provided by Arkema Co. The degree of crystallinity of all polyamide grades are calculated. The summary of the settings is shown in Table 4. 30 Table 4. Thermal properties of polyamide grades Polyamide Glass transition temperature (C) Melting temperature (C) Enthalpy (J/g) Crystallinity (%) Vestosint 46.76 120.29 40.59 15 Vestamelt 470 50.43 102.9 19.93 7 Vestamelt 171 53.73 112.7 19.49 7 Copolyamide 58 182 65.95 25 Porosity measurements Through another set of experiments, porosity measurements were performed. All polyamide grades were exposed to the heat within distance as applied in the SSS machine. ImageJ platform has been used for image processing and the porosity percentage of layers for four polymer grades under the same heating treatment have been calculated (Table 5). Table 5. Percentage porosity for polyamide grades Polyamide grade Percentage porosity (%) Vestosint 12.72 Vestamelt 171 17.03 Vestamelt 470 26.45 Copolyamide 16.22 31 8.1.2. Preliminary experiments Among all grades of polyamide powders that were tested by SSS machine, Vestamelt 171 and Copolyamide showed the best performance. Vestosint and Vestamlet 470, although showing advantageous mechanical properties, were not successfully processed due to high shrinkage during fabrication. 8.1.3. Discussion Base material properties play a major role in the success of fabrication in the SSS process. Vestosint offers the lowest percentage of density compared to two other grades that were successfully processed. However, application of this powder in the SSS process was limited due to excessive shrinkage. Vestamlet 171 and Copolyamide both show an 84% density on single layer. Several test cases were fabricated and it was observed that as the result of sintering Copolyamide parts obtain higher density and stronger attachment between the layers. Vestamlet 171, shows higher level of porosity compared to Copolyamide. 8.1.4. Selection of separation powder To reiterate, S-powder prevents polymer particles on the two sides of the separation barrier to adhere under heat exposure. The inhibited boundary profile acts as a scaffolding that can be easily removed. Several factors are considered for choosing an appropriate separator material. These include particle size and shape, density, melting temperature, decomposition temperature, flowability (cohesion coefficient), and non-reactivity with base material. The suitable range for S-powder in meso-scale fabrication, which requires a separation region thickness of under 1mm to allow for fabrication of 32 small geometric details such as small holes, is within the range of 20 to 90 micrometers diameter. What needs to be identifies are: a) Powder flowability through the nozzle, b) Stability at high temperature, c) Non-reactivity with polymer, d) Availability in small particle size. Initially, aluminum oxide was selected and tested as the candidate S-powder. However, since pure aluminum oxide powder was only found in white color and it is difficult to visually examine the deposition of this S-powder in white polymer for this research purpose. Other categories of colored ceramic which were considered were nickel oxide (green) and silicon carbide (green). However, these commercial powders are either rarely available within the size range that can be applied in SSS machine, or lack the desirable density to meet the requirement for effective powder delivery, as gravitational force is a major player in helping the S-powder flow down the nozzle pathway. Chromium oxide (dark green) was eventually found by conducting deposition tests to be the most suitable powder for the flowability purpose. Figure 16 shows microscopic images of this powder. Figure 22. Chromium Oxide powder (left) and SEM image (right) 33 The main advantage of this powder is that it allows for a closely controlled deposition rate due to an interlocking effect between its particles. The non-spherical geometry of particles results in better control over leakage of the powder from the nozzle as compared to earlier aluminum oxide which was tested in spherical form. Sharp edges form interlocked chains which prevent powder from flowing down the nozzle tube under gravity even before applying vibration to the nozzle. Figure 23. Mechanical interlocking of powders for non-spherical and spherical particles [27] Table 6 summarizes material properties of the candidate separating material. Table 6. Specification of separation powder Properties Description Chemical composition Chromium Oxide (Cr2O3) Sieve Analysis 40 +/- 11 microns Melting Temperature 2266 C Density 2.6g/cc 34 Chromium oxide is produced in average particles sizes of 20 to 40 micron in diameter. A set of experiments were performed to examine applicability of using 20 micrometer powders. This powder consists of relatively spherical particles with narrow size distribution (Figure 24). Figure 24. 20 Micron Chromium Oxide (source: Praxair co.) Both powders were tested with the nozzle tube gauges of 25, 26 to 30 (minimum hypodermic needle sizes available). 40 microns particle size shows better flowability in both 25 and 30 gauges due to smaller cohesion between its particles. As the ratio of average powder size to nozzle tube diameter decreases, the cohesiveness of powder increases which cause less flowability. Also, according to the available research studies on powder characterization in AM [28], it has been established that packing density increases and reaches it maximum when particle distribution combines both coarse and fine particles, which in our experiments has been the case for 40 +/- 11-micron (fine) chromium oxide powder. 35 Figure 25. Packing density overview versus particle size [27] The best flowability was observed with a nozzle tube ID size of 26 and average particle size of 40 microns (Table 7). Table 7. Powder flow for 20 and 40 micrometers powders Needle gauge Particle size (micro meters) 20 40 25 No No 26 Yes(limited) Yes 30 No Yes(limited) 36 In the following set of experiments the deposition rate for chromium oxide of 40 microns was studied. Initially, the range of frequencies with positive deposition flow was determined and shown in Table 8. For this range of frequencies three wave shapes which have been previously studied by Zhang [30] to be more effective have been selected. Table 8. S-powder deposition test Average deposition rate Wave form Frequency (KHz) Square Sine Triangular 1 0.024 0.022 0.02 2 0.019 0.023 0.027 3 0.016 0.025 0.03 4 0.015 0.018 0.022 5 0.017 0.004 0.008 6 0.002 0.003 0.003 7 0.001 0.007 0.011 8 0.006 0.007 0.007 9 0 0 0 10 0 0 0 37 Deposition rate is recorded by measuring weight of powder deposited in 30-second time interval for all settings. For each wave form Figure 18 demonstrates the changes in average deposition rate versus frequency. Figure 26. Deposition rate vs. frequency for Cr 2O 3 For all wave forms there are local extreme points. For triangular wave and sine waves the maximum deposition rate was captured at 3 Khz and 7 Khz respectively. For square waves; however, maximum depositions were captured at 5 Khz and 8 Khz, respectively. In all cases, the flow starts to increase to reach the second peak after 6 Khz. At 10 Khz; however, there is almost no deposition in any of the cases. This result is also compatible with the research done by Zhang in examining the deposition rate of Tungsten (20 and 40 microns) which he used for fabrication of metal parts. In all cases the maximum deposition rate was achieved at 3 Khz and similar behavior were observed. Deposition rate is dependent on the material properties and need to be experimentally examined to find the best setting. 38 Figure 27. Controlled Cr 2O 3 deposition over polymer 8.2. Process parameter requirements for SSS 8.2.1. Deposition of B-powder on the substrate Movement of the nozzle inside the B-powder results in accumulation of some powder along the edges of the trench that the nozzle creates. A typical shape of the trench is shown in Figure 28. As it can be seen, small portions of B-powder have collapsed into the trench while the remaining are piled up along the contour. 39 Figure 28. Gaps created by separation lines 8.2.2. Requirements for successive layer spreading and deposition Figure 29 demonstrates the rectangular cross section of the gap. As previously studied by Zhang [30], the shape and ratio of powder collapse varies depending on flowability (angle of repose) of the base powder. Using base powder materials with a higher angle of repose leads to a higher ratio of base powder collapsing along the path. Figure 29. Gap created by movement of nozzle inside base material 40 As this approach is based on layer by layer sintering, layer sintering is performed after S-powder deposition is completed by the nozzle on that layer. There are two ways in which layer sintering can be done: I) The heater starts sintering the layer right after deposition: If the heater passes over the B- powder trench after deposition of S-powder, the piles of powder accumulated on the top of each opening will result in non-smooth sintered surfaces because of the raised clump of B-powder being closer to heat source and receiving more heat relative to other points. Consecutively, when the next layer of powder is spread on the unevenly sintered layer, raised solid points on the sintered layer surface will be created which would collide with the nozzle which holds a constant distance with all points on the layer surface while depositing the S-powder. Furthermore, spreading the subsequent powder layer on a non-uniform surface can displace the layer on the platform and move the piece by the layer smoothing mechanism (blade, roller, etc.). II) The blade passes along the layer first and then the sintering process starts: In this case, the surface smoothing roller (or blade) moves along the surface, first to sweep powder clumps at the trench edges that were created by the nozzle. Afterwards, the heater passes over the layer and sintering happens. As the final resolution of parts made by SSS highly depends on the S-powder printing resolution and the performance of the sintering process, all factors involved during these processes should be closely investigated. Figure 30 illustrates how the movement of the roller (or blade) after deposition can affect the sintering process on the cross section of a linear deposited trench. If the gap depth (d) of the trench is equal to the deposition height (h), as the roller (or blade) collides with the B-powder clumps at the edge, it will spread it over the trench filled with the S-powder as shown in Figure 30 (a) and (b) and spread the rest on the surface of the powder (c). Thus, depending on the height of S-powder inside the gap, base material can partially to completely cover the surface of the 41 filled gap. When sintering occurs in the following step, all the powder will diffuse with each other and the next layer will be spread on the current layer. If the gap has been previously covered with base material, this will result in sintering undesired regions and separation is difficult. It is worthwhile to note that this example considers the worst-case scenario since the direction of movement of the roller is perpendicular to the wall of the trench. Figure 30. S-powder deposition volume relative to slot Note in that in the case of over-deposition of S-powder on the layer, if d is smaller than h, as demonstrated in (d), (e), and (f), when the roller moves forward, it will drag the excess S-powder material and spread it over the base powder along the way. The spreading S-powder on the surface will result in blockage of regions of B-powder which are under the spread excess S-powder from sintering and results in poor bonding of those regions, in addition to contamination of the part with S-powder. 42 In the case the deposition height is less than the trench depth (h<d) as shown in Figures (g), (h), and (i) the the B powder clumps created at of the edge of the trench collapse and falls inside the gap by the roller upon spreading the subsequent layer. In this case there is less possibility for the collapsed powders to diffuse to each other under heat as they have larger distance from the heater than the powder particles on top of the layer surface. However, if the gap is not filled, as the next layer is spread on it, the base material will further fall into the gap and completely fill the slot and get sintered by the heater, hence rendering the trench ineffective in separating the part and non-part regions of the powder. 8.2.3. Separation trench shape in consideration of insertion depth Other than flowability of powder which has an impact on the shape of the parts, the insertion depth of the nozzle determines the amount of the S-powder that enters the trench and whether it will sufficiently cover the gap. Assuming the size of the nozzle tube is constant and that there is adequate S-powder deposition to fill the gap with larger insertion depth, it can be observed that a larger insertion depth for the same trench width creates more excessive B-powder along edges. Thus, in case of high insertion depth, the trench is more likely to be completely covered with the base powder material as shown Figure 31 (a), (b) and (c). 43 Figure 31. Gap shape in relation to insertion depth In case of low nozzle insertion depth, since the width of the trench is the same, there is less excess clumped material at the trench edges. Accordingly, when there will be less B-powder at the edge to cover the top of the trench by the roller (or blade) at the next layer spreading. Based on the discussion above, successful S-powder deposition and layer spreading is obtained when the gap is precisely filled with the S-powder material (d=h) and the insertion depth is as low as possible. It should be noted that in all cases it is assumed that there is no collapse of B-powder clumps inside the gap; however, in real case scenarios B-powder can partially collapse in the gap depending on the movement speed and insertion depth. 8.2.4. Nozzle insertion depth in relation to layer thickness In the previous section, the relation between nozzle slot height (h) and nozzle insertion depth (d) was discussed. In SSS for the fabrication of polymeric material, the layer-by-layer sintering will not allow the nozzle to move down inside the base powder if the insertion depth is larger than the predefined layer thickness. In this case, the nozzle insertion depth is more that the layer thickness, the nozzle tip 44 will touch the surface of the previously sintered (hardened) layer and will displace the built section of the part in the loose powder vat. (Figure 32). Figure 32. Printing failure due to large insertion depth (left), correct needle insertion depth (right) If the insertion depth is smaller than the layer thickness, there will be a thin layer of base material in between previously sintered layer and the nozzle tip lacking S-powder deposition creating discontinuity in the separation region. This will eventually prevent the ease of separation of part from surrounding material due to unwanted sintered bridges in between Figure 33. 45 Figure 33. Undesired sintered bridges in between walls The insertion depth should therefore be as close as possible to the predefined layer thickness but never more than it. In metal fabrication by SSS, due to bulk volume sintering, the nozzle may be partially inserted into previous layer of deposited S-powder to assure complete separation between the layers. 9. SINTERING AND COALESCENCE Powder sintering refers to a thermal process used to convert a powder volume into a solid object. Polymer sintering is the formation of homogenous mass from the diffusion of smaller particles. Generally, this process is carried out between the glass transition temperature (Tg) and melting temperature (Tm) of a specific polymer. When polymer particles are heated above Tg, they form a neck at which polymer chains can interact and diffusion occurs. When the entanglements are fully established, this mass will be at equilibrium [31]. Figure 34 shows growth of the contact neck between powder particles. 46 Figure 34. SEM pictures of powder before(A) and after sintering(B) Heat can be transferred between objects by means of radiation, convection, conduction, and microwave [32]. Extensive research work has been done in developing a thermal analysis method for powder metallurgy processes such as Laser sintering [33-35]. Uniform temperature distribution in powder sintering processes can severely affect the final quality of the part. Understanding process mechanisms can significantly contribute to the future development of SSS parts. In the following section, two methods of powder sintering that were examined for SSS of polymeric material are discussed. In the following section two methods of powder sintering that were examined for SSS of polymeric material are discussed. 9.1. Radiation Radiation refers to method of heat transfer in form of electromagnetic waves. The amount of energy that is distributed from an object is calculated using Stefan-Boltzmann law shown below: 47 𝑄 = 𝑒𝜎 𝑇 4 𝐴𝑡 [36] Where e corresponds to the emissivity of an object, 𝜎 is the Stefan-Boltzmann constant (5.6 × 10 −8 𝐽 /(𝑠 𝑚 2 𝑘 4 ), T reflects the temperature at which object is emitting radiation, and A is the object’s surface area and t corresponds to time interval at which energy is radiated. 9.2. Convection Convection is commonly used as heat transfer term for fluids and gases. Convection heating combines the processes of conduction and heat transfer by fluid flow. Heat transfer by convection can be calculated by this formula: 𝑄 = ℎ 𝑐 𝐴 ( 𝑇 𝑠 − 𝑇 𝑎 ) 𝑡 [36] Where ℎ 𝑐 corresponds to the convective heat transfer coefficient, A is the heat transfer cross sectional area, and 𝑇 𝑠 − 𝑇 𝑎 corresponds to difference in air and exposed surface’s temperature. There are two types of convective heat transfer: 1. Natural or free convection: The flow is caused by density changes due to variation in the temperature in absence of any other external source such as pump. 2. Forced convection: Liquid motion induced by pressure difference which happens by means of external sources i.e. fans. 48 9.3. Experiments 9.3.1. Bulk sintering in SSS Bulk sintering refers to the sintering of the entire powder volume upon completion of printing. This type of sintering is suitable in which there is a need for prolonged heating time. One of the factors that limit the bulk sintering for polymer processing is the uncontrollable shrinkage of material that occurs during sintering. The small threshold between the sintering point and melting point of base materials selected, does not allow for successful sintering of the entire powder volume. Due to sensitivity of this process in controlling shrinkage, non-uniform degree of sintering along the part body and low mechanical properties of final specimens, this approach was not found to be practical for polymers. 9.3.2. Layer by layer sintering setup- Radiation Layer-wise sintering is the most common sintering method used for polymer processing in AM technologies [37, 38]. In SSS-polymer, the sintering step occurs in between successive layer depositions, thus, does not affect the printing speed. The SSS-polymer research began with first starting with the radiation sintering approach and then choosing the proper heating element setup. In this approach a heating element wire is wound around a 2 mm diameter alumina tube to create a linear radiation source. Various heat element wires were examined for use in the machine. Figure 28 shows heat distribution for Kanthal wire with gauge of 28, pure Nichrome with gauge of 32 and Nichrome with 26 of gauge. Kanthal wire was observed to glow uniformly and to be resistant to oxidation (Figure 35 a). Nichrome 32 was observed to glow non- uniformly and had high voltage requirement which was no less than 50 V (Figure 35 b). In case of 49 Nichrome 26 no glow was achieved when applying the maximum voltage (Figure 35 c). Finally. Nichrome 28 had uniform radiation distribution; however, this material quickly gets oxidized and turns into a green color). Figure 35. Glow distribution of Kanthal (a), Ni 32(b), and Ni 26(c) Among the two best candidates Kanthal 28 and Nichrome 28, Kanthal provided higher electric power, therefore, it was selected as the main heating element for SSS sintering. The final liner radiation heat setup installed on the machine can be seen in Figure 36. 50 Figure 36. Linear radiation heat setup 9.3.3. Layer by layer sintering setup– convection To possibly improve the sintering effectiveness in SSS and hence improve the final quality of parts a forced convection heating setup was also used (Figure 37). This system is composed of a fan which forces air through an air duct which contains the heating element near its outlet is used. The pump speed, distance of the hot air outlet from powder surface, and heat intensity are all adjustable. Figure 37. Hot air linear heater setup Air duct Pump Heating element Kanthal wire 51 9.4. Pre-heating To avoid shrinkage by the sudden cooling of layers, heating element sheets are installed in the build tank and platform as shown in Figure 38. Infrared lamps are used as well to maintain an elevated temperature inside printing chamber. Before printing starts heaters are switched on and the tank temperature is raised to 60 C, which is slightly below the sintering temperature of Nylon 6,6. The temperature inside the printing chamber is kept constant at 30 C. Figure 38. Heaters attached to platform (a) and build tank(b) Additionally, temperature controllers are used for the build tank, platform, and chamber to continuously control the temperature of the machine during printing process. a b 52 9.5. Process parameters Table 9 shows a summary of optimum process parameters for successful fabrication of Nylon 6,6 material. Table 9. Process parameters in SSS for Nylon 6,6 Properties Description Thickness of each layer 0.09 to 0.25 mm Sintering speed 8.3 mm/sec Sintering temperature 65-70 C Printing speed 40 sec/layer (average) Average wait time between layers 12 Sec Build tank temperature 48 to 60 C Platform temperature 54 C Chamber temperature 30 C Type of polymers Nylon 6,6 Powder size range 90 to 250 micrometers S-powder size range 25 to 45 micrometers 10. PRELIMINARY PART FABRICATION In this set of experiments, more complex geometries with multiple overlapping layers are built by the SSS machine. Figure 39 and 40 demonstrate samples made by Copolyamide and Vestamelt 171. 53 Figure 39. 3D test case fabricated by Copolyamide and Vestamelt 171 Figure 40. Enlarged view of 2.5 gear fabricated by Copolyamide 10 mm 10 mm 10 mm 10 mm 54 11. CHALLENGES IN SSS FOR POLYMERS 11.1. Weak bonding between layers One of the key factors in successful fabrication of polymeric parts with SSS is effective sintering of layers. Increasing layer thickness can result in inadequate transfer of heat energy for diffusing successive layers. Figure 41 shows poor bonding between layers for 200 micrometers layer thickness. Depending on the thickness of the layer, sintering adjustments should be made to achieve effective bonding between layers. Figure 41. Partially sintered sample with weak bonding between layers 11.2. Cumulative heating and warpage Depending on the properties of base material and sintering and cooling temperatures and rates, warpage of the layers can occur as shown in Figure 42. Also, as the number of layers increases this phenomenon becomes more significant. Warpage can occur by penetration of heat into previously 55 sintered layers resulting in over-sintering. Moreover, depending on the melting temperature (Tm) and crystallization temperature (Tc) this warpage can be eliminated in polyamide samples with smaller window between Tm and Tc [39-41]. Figure 42. Warpage caused by cumulative heating Non-uniform heating and cooling can be other reasons for layer warpage. Therefore, attempt is made to keep the temperature inside the chamber at a constant level and to pre-heat the platform before printing starts. 11.3. Shift between layers In successive deposition of layers, it is very important to have control over uniformity in the deposition of B-powder over sintered layer to avoid any displacement of previous layer. Shift between layers is one of the main causes for bad quality in the final part. Shifts can be caused by non-uniform sintering of layers, collision of roller with a sintered layer due to inadequate distance from the layer, high 56 friction between roller and surface of powder in build platform, previous layer’s warped edge which is easily hit by the roller, software failure, and poor bonding between successive sintered layers. Moreover, B-powder’s flowability, particle size and shape as well as roller (or blade) properties such as shape and material, are important factors in successful spreading of powder on previously sintered layer and need to be considered. Finally, Excessive amount of powder fed to the build tank was also observed to be a contributing factor to layer shifting. 12. HYPOTHESES In this section, the following hypotheses are proposed in order to achieve improved quality of fabrication by SSS. 12.1. Hypothesis 1: Ease of part separation, as it relates to the separator powder deposition in the base material, is achieved by preventing base powder material from collapsing inside the gaps. A model may be developed to determine the maximum nozzle linear speed for attaining perfect separation. This approach may be used only when raised edge suppression is used. Figure 43(a) shows a microscopic top view image of layers after separation. As it is shown in the image, polymer particles have diffused to the ceramic powder on the outer side of the separation line which results in poor separation of the surrounding part from the rest. The image on the right shows a magnified area on the part where part surface has been ripped off when separating the two sections at the separation line. In other sections where only, ceramic particles are present separation seems to have taken place without damages. 57 Figure 43. SEM images of layers after separation(a) and magnified damage (b) To address this issue, a raised edge suppression system setup is used to prevent the base powder material from falling inside the gaps. Figure 44 demonstrates two gaps created under the same deposition speed and insertion depth of 100 micrometers by the nozzle. The edge suppression system was used to remove excessive powder that is accumulated on the surface of the powder from the left trench. Figure 44. Gaps created by nozzle with (right) and without(left) edge removal B-powder S-powder Damage on the sintered surface Layer after separation S-powder 58 As it can be seen in Figure 45, base powder particles which have already fallen inside the slot, cannot be removed. The most effective result with SSS is obtained when there is no base powder material dropped inside the gap. This can be obtained by ensuring that the S-powder level completely fills up the opening eliminating chance for base powder to fall inside. As both S-powder and B-powder are competing to enter the gap, if there is a sufficient S-powder deposition rate, it will be unlikely for base powder to drop inside the gap created by the nozzle movement. Figure 45. Top view of 3D part after separation The following set of experiments have been carried out by keeping the nozzle insertion depth at 100 micrometers and varying the deposition rate. The nozzle speed has been fixed in all runs to better examine the influence of S-power deposition. Table 10, shows the set of wave shape and frequencies that have been applied. This selection is based on deposition rate experiments performed in previous experiments. The combination is sorted in ascending order from maximum flow rate to lower settings. B-powder 59 Table 10. Maximum deposition setting Wave shape Applied frequency Triangular 3Khz Sine 3Khz Square 1Khz Triangular 2Khz Sine 2Khz Sine 1Khz The effect of the deposition rate on B-powder collapse is shown in Figure 46. The higher deposition rate demonstrated on the leftmost separation line does not allow polymer particles to collapse. Decreasing the S-powder rate leaves a gap on the top of the slot and increases the B-powder amount inside the gaps. If the gap is too thin, the two powders will diffuse together making separation difficult. 60 Figure 46. Lines with S-powder deposition decreasing from left to right Preliminary result in Figure 47 demonstrate a cylinder test case fabricated by controlled rate of deposition. Ease of separation was achieved by preventing base powder material from covering the gaps. Figure 47. Cylinder test case made by SSS 10 mm 61 12.2. Hypothesis 2 Non-deformed inter-layer adhesion can be achieved by convection based heating at a speed which is at least the same as speed of radiation based sintering which can produce the same result. Further investigation needs to be carried out on base materials properties and the choice of sintering mechanism to further improve mechanical strength and achieve acceptable surface quality. Microanalysis of layers have been performed to obtain a better understanding of the process. The impact of infrared and convection sintering on samples have been studied. To have a better comparison of the two systems, the amount of heat applied to the powder by each method is determined by measuring the temperature at the point where the powder is sintered. Vestosint material has been used as the base powder for the first set of experiments. Sintering speed is kept constant during all experiments. To study the impact of heat flow on sintered layers, a set of experiments were conveyed with variant distance of convection heater to the powder. Figure 48 represents a binary surface plot of layers sintered under convection heating from distances of 11mm and 4 mm from surface of powder. These distances are selected such that 11 mm is the maximum distance from which sintering is possible and 4 mm is the shortest distance possible to the surface of the powder. These plots are created based on Gaussian light distribution of microscopic images. The heat intensity has been constant in both cases. Although in case b, there is a greater amount of heat flowing inside the powder relative to higher distance exposure, it is observed that decreasing distance of the heater from powder has increased porosity on the layer due to high velocity of the flow. Is it observed that in both cases heat penetration depth is relatively high which indicates that in cases where high volume sintering is desirable this approach is suitable. 62 Figure 48. Surface plot of convection sintering at 11mm distance (left) versus 4 mm (right) from surface layer Figure 49. SEM image of Vestosint at 150 magnitudes sintered at distance 11mm(left) and 4mm (right) Increasing the heat intensity under the same condition of distance from surface of powder and flow velocity, results in stronger bonds between particles while maintaining average pore size diameters the same, as opposed to the previous condition in which average pore size was increased. In Figure 50 the surface of Vestosint under low (left) and high (right) heat intensity is shown. 63 Figure 50. Surface of Vestosint under in low (left) and high level heat intensity In the case of radiation, it is observed that increasing heat intensity from low to high level results in complete coalescence of particles on surface (less porosity on surface); however, does not guarantee bottom particle diffusion. Figure 51. Surface of Vestosint under IR heating in low and high-level heat intensity (4 mm distance) In the case of convection sintering, even at high level heat intensity, the energy is absorbed by all particles causing sintering from the bottom up through the whole body. While radiation heat is 64 absorbed in the first layer, any remaining energy is transferred by conduction from top to bottom gradually. Figure 52. Surface plot of Vestosint under Radiation heating in low and high level heat intensity (4 mm distance) Bright sections on these images are indicative of light transfer through the layers and therefore, sintering level. Similarly, in both cases, bottom layers have a higher level of density as moving up to upper layers shows a decreasing sintering ratio. 12.3. Hypothesis 3 Final surface quality can be improved by at least 20 % by proper selection of S-powder in relation to B-powder material. It is hypothesized that proper choice of S-powder is a key factor in final part quality by creating smooth and uniform surfaces. 65 13. EXPERIMENTAL SETUP AND PROCEDURES 13.1. Impact of S-powder deposition on part surface quality This section focuses on improving surface quality of parts built by SSS. Adequate deposition of S- powder is a key factor in the quality of separation lines. Figure 53 demonstrates the impact of sintered base powder on quality of walls without the existence of S-powder. Figure 53. Separation of part after nozzle movement with no S-powder deposition Figure 53 shows cross section of a sintered wall at the separation line. As it can be observed, there are thin B-powder bridges formed between the two trench walls which are mainly caused by B-powder collapse into the gaps. The other reason can be inaccurate nozzle insertion depth into the B-powder. Although nozzle insertion depth is determined and is equivalent to the layer thickness, there is a possibility of slight displacement due to unexpected loose nozzle connection, error in the Z direction mechanical port, or human error in setting up the initial nozzle position. If any of the two causes occur, B-powder bridges would be formed and connect the two sides of separation lines on each layer (Figure 54). 66 Figure 54. Magnified image of separation line cross section and connected polymer bridges Although connected bridges do not fail the separation due to weak bonding, they result in roughness of the surface which requires post-processing to remove extra margins. 13.2. Impact of S-powder choice on part surface quality It has become evident from some experiments that different choices of S-powder exhibit different degrees of S-powder particle interlocking with B-powder particles. Thus, higher surface quality can be achieved upon separation by selecting an appropriate S-powder. According to a study in the literature by Zhang [30] on surface quality of metallic parts built by SSS, the spatial distribution of S-powder in contact with B-powder will be reflected on surface smoothness upon separation. This can be better understood in the Figure 55, which represents the interlocking effect between in-contact powders. B-powder bridge B-powder 67 Figure 55. Interlocking effect between powders [42] Accordingly, a number of S-powder candidates in different sizes and shapes were examined in a set of experiments. Candidates were selected based on their relative density (minimum of 2 gr/cm 3 ), particle size (40, 20 and smaller than 20 micrometers samples), particle shape (spherical and non-spherical) and stability up to 200 C, as shown in Table 11. Since it was not possible to achieve the same amount of deposition for all powder types under vibration in the SSS machine, the following experiment has been conducted. 68 Table 11. Properties of S-powder Candidate powder Chemical composition Color Shape Size (Micrometers) Relative density (gr/cm3) Stable up to 200 C Tungsten W Dark gray Spherical 20 10.8 Yes Tungsten W Dark gray Spherical 40 10.8 Yes Chromium oxide Cr 2O 3 Dark green Angular 20 2.6 Yes Chromium oxide Cr 2O 3 Dark green Angular 40 2.6 Yes Aluminum oxide Al 2O 3 White Spherical 3 2.5 Yes Aluminum oxide Al 2O 3 White Angular 3 2.5 Yes Soda lime CaHNaO 2 Light gray Spherical 40 2.53 Yes Alumina Al 2O 3 and TiO 2 balance Black Angular 40 2.6 Yes A thin layer of B-powder with thickness of 100 micrometers was uniformly spread on a flat surface which was in contact with uniformly spread candidate S-powders. The layer thickness was based on current layer thickness used by the SSS machine. Each sample was placed under the SSS heater and sintered at 70 Cover the line where the two powders meet. 69 Figure 56. Demonstration of sintering contact line of S-powder candidates with B-powder Upon separating the sintered B-powder layer, the layer was placed under a microscope to observe form of the separated wall as well as amount of adhesion along the contact wall (Figure 56). Pink color substrates were used as backgrounds for better visibility of powders on sample layers under microscope. Chromium oxide (40 microns) Chromium oxide (20 microns) S-powder B-powder Contact wall 70 Spherical tungsten (40 microns) Spherical Tungsten (20 microns) Spherical aluminum oxide (3 microns) Non-Spherical aluminum oxide (3 microns) Black alumina (40 microns) Soda lime (40 microns) Figure 57. Top view of sintered layers over the cross section It is observed that for smaller particle size, there is less adhesion to B-powder; this is observed for both chromium oxide and tungsten. It is also observed that for spherical particles there is less adhesion to B- 71 powder after sintering which is the case for Tungsten, white alumina and soda lime. For each sample 5 points along the separation line with the longest distance to ideal separation line are selected and deviations are measured using a Micro Vu measuring system. Average deviation for each test case can be seen in Table 12. Table 12. Average deviation for each S-powder Candidate powder Shape Size (Micrometers) Average deviation from separation line (Micrometers) Tungsten Spherical 20 23 Tungsten Spherical 40 55 Chromium oxide Angular 20 154 Chromium oxide Angular 40 175 Aluminum oxide Spherical 3 <5 Aluminum oxide Angular 3 <10 Soda lime Spherical 40 <20 Alumina Angular 40 65 It should be noted that the desirable properties of the S-powder are maximum flowability and highest resulting surface quality. However, there are limitations to controlling S-powder deposition in SSS machining. First, as particle size decreases, cohesion between particles increase which limits flowability for 20 microns and smaller size powders in the SSS nozzle with opening diameter of 260 72 microns. Second, for spherical shapes, due to weak interlocking effect between particles, over- deposition and leakage is unavoidable. Third, powder compaction inside the nozzle barrel which results in faster clogging has been observed to be more significant in spherical cases (tungsten, white alumina and soda lime) as compared to angular S-powders (chromium oxide and black alumina). This result necessitated further experiments with candidate S-powders in the SSS machine. Among the list of candidates, (except chromium oxide which has been previously used), tungsten (40 microns), black alumina (40 microns) and soda lime (40 microns) were selected to be tested. The first step in filtering was to accept only the candidate whose deposition rates were at least 0.5 gr/sec. Thus, all candidates with particle sizes smaller than 40 micrometers were dismissed. In the case of tungsten (20 micrometers), the flow was sufficient. The reason for elimination was due to the free flow from the needle under gravity without vibration. Figure 58. Test cases fabricated by tungsten, black alumina, and soda lime from left to right Test cases were built successfully and evidence of successful separation in all three cases were shown. However, there is a significant difference in the final surface quality achieved (Figure 58). The final 20 mm 73 part quality is highly dependent on the amount of S-powder deposited onto the gaps and for all three cases an attempt was made to ensure equal amount of S-powder deposition while printing. Based on previous experiments, it was expected that higher surface quality would be achieved by spherical tungsten. While the density and shape of particles would not be sufficient for successful fabrication. From the results it can be concluded that uncontrollable leakage of tungsten along the boundary has disturbed the B-powder and hence the rough surface. Another reason for surface roughness by tungsten could be high relative weight ratio of tungsten to B-powder. This could have resulted in undesired movement and penetration of tungsten inside B-powder triggered by the powder spreading roller on each layer. In the cases of using black alumina or soda lime better surface quality and ease of separation was achieved while satisfying the required deposition rate. 13.3. Effect of heat capacity of S-powder on surface quality Another important factor in surface quality of parts is determined by the amount of S-powder residual on the surface of the part. Although for some application cases S-powder can be used as a coating on the surface of parts to provide abrasion protection as well as coloring, it is generally desirable to find a S-powder which can be thoroughly removed from the surface without the need of post-processing. Ease of removal, is dependent on bonding between polymer and S-powder particles which is form under heat caused by thus, density changes. Heat capacity is the ability of molecules to absorb energy. In quantitative terms, the amount of heat required to raise the temperature of one mole of material for one degree. For solids, there are three degrees of translational freedom in a lattice [43]. Lattice vibrations in solids are increased to absorb 74 thermal energy. Figure 59 by Wiley shows changes in atomic vibrations caused by changes in level of energy. Figure 59. Increase in atomic energy vibrations under heat [44] A list of thermal properties of candidate S-powders is given in Table 13. Thermal properties of B- powder are shown in Table 14. 75 Table 13. S-powder thermal properties Candidate powder C p (J/kgk) Coefficient of thermal expansion at room temperature (10^-6 /C) Thermal conductivity k(W/m-k) Energy transfer mechanism Tungsten 132 4.5 180 Atomic vibration Chromium oxide 780 8.5 35 Atomic vibration Aluminum oxide 880 7.6 40 Atomic vibration Soda lime 680 5.5 2 Atomic vibration Alumina (TiO2, Al2O3) 775 7 45 Atomic vibration Table 14. B-powder thermal properties Candidate powder Cp(J/kgk) at room temperature Coefficient of thermal expansion at room temperature (10^-6 /C) Thermal conductivity k(W/m-k) Energy transfer mechanism Copolyamide 1750~1900 135~185 0.12~0.25 Vibration of chain molecules Based on previous experiments, black alumina and soda lime were found to result in better surface quality. It is hypothesized that the amount of residual S-powder on the surface is linked to the S- powder’s thermal properties and can be reduced to less than 5%. The specific heat capacity (cp) and coefficient of thermal expansion (alpha) of suitable S-powder hold a larger gap with cp and alpha of B- 76 powder. Thus, another set of experiments is conducted to examine the impact of thermal properties on final surface quality of the samples. Test specimens in form of stars were built by applying soda lime, chromium oxide and black alumina as S-powder. Stars are selected since they provide 10 surfaces in different directions which can provide sufficient replications for each S-powder. Figure 60. Test cases fabricated by soda lime, chromium oxide and black alumina (from left to right) as S- powders To quantify the amount of S-powder remaining on the surface of the part, images were taken from each of the lateral surfaces of the specimens The Image J software platform has been used for image processing to measure the area ratio of S-powder over lateral surface of each blade, the average value of which is shown in Table 15. 77 Table 15. S-powder adhesion to B-powder measurement Sample S-powder Percentage adhesion (average value for 10 surfaces) 1 soda lime <2% 2 chromium oxide 35% 3 black alumina 5% Initially, weight measurements before and after removing S-powder from the surfaces were used as the criteria to determine the percentage of adhesion; however, due to insignificant changes reflected on overall weight, this method was no longer used. 13.4. Shape of separation surfaces created For better visibility of surface roughness and shape of separation surfaces created, a separate microscopic image of sample blocks built using black alumina and soda lime as S-powder were taken using Micro-Vu precision machine (Figure 61). The placement of blocks under the microspore is to capture a top view which is along the same direction as the nozzle deposits in B-powder. Magnified wall shape for each sample can be seen in Figures 57 and 58. 78 Figure 61. Sample block test case (left), top view with Micro Vu Figure 62. Shape of the separation surfaces from the top view (S-powder: black alumina) Figure 63. Shape of walls from the top view (S-powder: soda lime) 79 It is observed that, with 40-micron particle size and high stiffness of black alumina, stronger separation walls are created post sintering. Angular dents are observed on the surface. Soda lime, on the other hand, has led to smoother surfaces with more round dents along the wall. 13.5. Assessment of ease of separation To evaluate ease of separation for each S-powder choice, a digital hanging scale was used to measure the amount of force required for separation along each side of the two rectangular blocks. Each block was held firmly by clippers from the two sides of a separation line (Figure 64). An increasing pulling force was then applied to the bottom clipper in perpendicular direction to the separation line for complete separation. Part orientation was kept the same during experiments for both blocks (Figure 64 and 65) and reference points on each side helped ensure equal distance between clippers and separation lines. Figure 64. Force measurement along separation line Side2 Separation line 80 Figure 65. Blocks built by black alumina (left) and soda lime (right) S-powders The final measured tensile stress measurements are reflected in Table 16. Soda lime is observed to require smaller force for separation. Table 16. S-powder separation tensile stress measurements σ (N/mm 2 ) S-powder side 1 side 2 side 3 side 4 Block 1 black alumina 0.32 0.41 0.29 0.43 Block 2 soda lime 0.24 0.39 0.21 0.35 13.6. Discussion Proper choice of S-powder can highly impact part quality by creating smooth and uniform surfaces. Based on experiments conducted in this study, black alumina and soda lime are selected as suitable separation powders due to the following favorable properties: Side1 Side3 Side4 Side1 81 1. With sufficient deposition rate, both powders allow for “controlled deposition”. Leakage and over-deposition can badly affect part surface quality since base powder segments outside of the layer can be easily disturbed and fuse to other regions, which is of course not desirable. Comparing the two powders in terms of controlled deposition rate, compaction occurrence is higher in soda lime due to its spherical particle shape. 2. Due to thermal properties of the two powders, the diffusion of particles into B-powder segments under heat is minimized and they can partially (in the case of black alumina) to completely (in the case of soda lime) be removed. 3. Fine separation powder size prevents the diffusion of B-powder through the part boundary in both cases. 4. In using black alumina, stronger separation surface structures are achieved. Stiffness of alumina allow for a more resistant surface thereby preventing any cracks or damages to the part while separating. Depending on the application, any of the two selected S-powders can be applied. In Figure 66, a polymer part surface without any post-processing built by black alumina S-powder is shown. It is observed that the part surface is very smooth and, as measurements have shown in table 16, geometric dimensions along the boundary are within a few microns of target dimensions. 82 Figure 66. Isometric view and front view of test case fabricated by black alumina 13.7. Development of a model for separator powder deposition in the base material In SSS, when the nozzle moves forward inside B-powder and mechanically interacts with B-powder, there are two possibilities that can happen and form the gap wall shapes: 1. The two sidewalls of the gap slide down under gravity force into the gap. 2. The accumulated material on the two sidewalls remain along the edge on the surface and there is no collapse. In the first case scenario which was studied by Zhang [30], the S-powder and B-powder simultaneously compete to fill the gap that the nozzle leaves behind. Assumptions were made for the slope of collapsed B-powder and an analytical model was proposed to determine the separation gap width for a given flow rate. 13.7.1. Angle of repose of B-powder Angle of repose is the internal angel of the slope created by the surface of a pile of powder, which is related to surface area, powder particle shape, and coefficient of cohesiveness of the material. In SSS, it is important to understand the flowability properties of both B-powder and S-powder to achieve a perfect separation. 83 It was learned by the previous SSS-metal research [30] that the angle of repose of the B-powder would determine the shape of the gap walls post nozzle movement inside B-powder. The smaller the angle of repose of a B-powder, the lower the likelihood of powder collapse. A comparative representation of angle of repose of polymers and bronze is shown in Figure 67. Figure 67. Angle of repose of bronze, Copolyamide, Vestamelt, Vestosint (Left to right, respectively) It can be observed that bronze has larger angle of repose, thus, higher flowability compared to polymers. The current polymer used in the SSS machine is Copolyamide, which is the second material from left in Figure 67. The excessive piled up S-powder along the edge is either swept by the SSS roller (which was thoroughly discussed in Section 2.5) or removed by an edge removal suppression system. 13.8. Development of a model for separator powder deposition in the base material A model can be used to determine the exact amount of required S-powder and the maximum nozzle linear speed for attaining near-perfect separation in the SSS-polymer process. This approach can be used under the assumption that raised edge suppression is used. 84 Figure 68. Created gap by nozzle with outer diameter of D and insertion depth of L As previously discussed in Section 4, the required insertion depth of the nozzle inside B-powder is determined by layer thickness. For any given geometry, the exact amount of S-powder required to fill the gaps can be calculated. Figure 68 is a representation of a gap created by SSS nozzle with nozzle diameter of D and insertion depth of L. 13.8.1. Total amount of S-powder required for any geometry Given the diameter of the nozzle, layer thickness and perimeter of each cross section, the total required amount of any S-powder can be calculated by Equations 16.3.1 and 16.3.2 which represent the volume sum of filled S-powder cylinders with diameter of the nozzle (D) and height of nozzle insertion depth (L) along the perimeter of the part contour on each layer (Figure 69). The actual amount deposited can be continuously read on the nozzle syringe gauge. Active monitoring of actual versus expected amount of S-powder provides very useful information in detecting inconsistent S-powder deliveries in printing. Deviations between expected and actual amounts of S-powder used can be caused by partial clogging of the nozzle, insufficient air flow, formation of arc, contamination, or leakage. 85 Equation 16.3.1 𝑆 𝑛 = 𝜌 × ∫ 𝑓 𝑦 𝑑 𝑦 𝑦 =𝑛 𝑦 =1 Equation 16.3.2 𝑆 𝑛 = 𝜌𝐷𝐿 × ∑ 𝑃 𝑦 𝑦 =𝑛 𝑦 =1 Where: D : Diameter of nozzle (𝑚𝑚 ) L : Nozzle penetration depth (Layer thickness) (𝑚𝑚 ) 𝑃 𝑦 : Perimeter of contour of layer y 𝜌 : Density of S-powder ( 𝑔𝑟 𝑚𝑚 3 ⁄ ) Figure 69. Schematic of S-powder deposition inside the gap Following calculations are based on the assumption that there is no collapse of powder clumps into the gaps. In real practice, there may be partial collapses of walls due to vibration of the nozzle as well as increase in nozzle movement speed. Figure 70 highlights the effect of nozzle speed on the shapes of the gaps. As it can be seen, increasing speed has led to higher degree of powder collapse. 86 Figure 70. Impact of speed on shapes of gaps created: low (left) and high (right) speed 13.8.2. Maximum printing speed viable under maximum deposition The maximum printing speed in SSS can be achieved under maximum rate of deposition of S-powder (Equations 16.1.3.1and 16.1.3.2). Deposition rate is a function of wave frequency, wave form, particle size, particle geometry, and density, and nozzle hole wall friction, which is not considered here due to lack information about the nozzle hole surface friction coefficient. Through a set of experiments, as discussed in Section 16.2.1 the maximum deposition rate for any powder can be determined. Equation 16.1.3.1 𝑉 max= 𝑓 ( 𝑅 𝑚𝑎𝑥 , 𝐷 , 𝐿 , 𝜌 ) Equation 16.1.3.2 𝑉 𝑚𝑎𝑥 = 𝑅 𝑚 𝑎 𝑥 𝜌 ×𝐿 ×𝐷 Where: 𝑅 𝑚𝑎𝑥 : Maximum deposition rate ( 𝑔𝑟 𝑠𝑒𝑐 ⁄ ) 𝑉 𝑚 : Maximum printing speed ( 𝑚𝑚 𝑠𝑒𝑐 ⁄ ) 87 Consequently, the minimum printing speed in SSS is obtained by the following equation: Equation 16.1.3.3 𝑇 min(𝑠𝑒𝑐 ) = 𝜌 ×∫ 𝑃 𝑦 × 𝐷 ×𝐿 𝑦 =𝑛 𝑦 =1 𝑅 𝑚𝑎𝑥 In the case of base powder collapse, the separation gap width reduces from D to a smaller value depending on several factors including height of the slot, height of side opening of the nozzle as well as angle of repose of the B-powder which was studied and described by static model in the literature for SSS-metal by Zhang [30]. Thus, the maximum speed in case of powder collapse can be obtained by following equations: Equation 16.1.3.4 𝑉 𝑚𝑎𝑥 = 𝑅 𝑚𝑎𝑥 𝜌 ×𝐿 ×𝑊 𝐷 Equation 16.1.3.5 𝑊 𝐷 = 𝑓 (𝐻 , ℎ, 𝛽 ) Where 𝛽 , H, h and 𝑊 𝐷 represent angle of repose, gap height, nozzle opening height and width of the separation wall (𝑚𝑚 ) after B-powder collapse respectively. 14. SELECTIVE HEAT SINTERING IN SSS Material shrinkage is one of the main sources of part inaccuracy in AM, primarily caused by thermal phase change of the material during layer formation [45-48]. Proper adjustment of process parameters can severely affect the output quality of the products. In terms of process planning, many research efforts have been dedicated to look for the relationship between the output part quality and the input 88 building process parameters [49-52]. Parametric process optimization is used to improve functional requirement of an additive manufacturing system such as accuracy by finding the optimal settings of factors such as light intensity of the machine, laser beam width, light exposure duration, scan speed, layer thickness, and so forth. It is crucial to find the appropriate process parameter settings which lead to the best trade-off among part accuracy and manufacturing build time. Several studies have been dedicated to improving part quality by modifying process variables. Sager et al. [53] studied a stereolithography process and investigated the relationship between laser scan speed and surface finish of the final part. He concluded that appropriate laser scan speed leads to a better surface finish of SLA geometries. Raghunath and Pandey [54] investigated the relationship between process parameters and material shrinkage. They examined laser power, beam speed, hatch spacing, part bed temperature and scan length and adopted Taguchi method for analysis of shrinkage. It was concluded that laser power and scan length are found to be most significant process variables influencing shrinkage in the X direction. Senthilkumaran et al. [55] also focused on shrinkage behavior by analyzing the relationship between shrinkage calibration specimen and various building conditions in SLS parts. In order to reach an accurate estimation, layer thickness, laser power, hatch spacing, scanning speed, part bed temperature and scanning mode were studied as the process parameters set in the experiment and beam offset and scanning position were observed to have the major influence on the shrinkage of the part. The focus of this section is to address shrinkage in SSS using process planning and machine parameter setup adjustment. The following section presents the study performed on the strength of the bond between layers using forced convection and the corresponding results. 89 14.1. Selective sintering by forced convective sintering With an understanding of sintering behavior of base material, it was necessary to determine whether or not an effective layer bonding could be achieved through heat transfer mechanisms other than direct radiation. Based on preliminary experiments by linear convective heater that were discussed in Section 14.3, it was hypothesized that non-deformed inter-layer adhesion of at least the same quality can be achieved by convection-based heating at a speed which is at least the same as the speed of radiation based sintering. In radiation sintering interlayer adhesion of the top two layers is achieved primarily by conduction of heat from the top layer to the lower layer. Conduction takes times but faster conduction can be achieved with higher radiation intensity, but at the expense of layer deformation. More stable layer shape can be achieved at lower heat if conduction is used over an extended time. However, slow sintering prolongs part fabrication time. Convection-based heating can force hot air to penetrate into top two layers through the pores of the top layer and sinter the two layers simultaneously without the delay that conduction heating would require. The first step was to build sample parts using this heat transfer approach. Figure 71. Schematic of forced convective heater 90 A pointwise ceramic heater tube with outer diameter of 6mm and inner diameter of 4 mm was used for experiments. The main components of heater include a ceramic tube, a kantal coil, as well as an air pump which blows air through the coil inside the ceramic tube (Figure 71). It was noticed that switching heater between on and off state modes on each layer is associated with delays in reaching a steady temperature. Thus, week degree of bonding would be achieved at start points as well as undesired sintered areas following end points. To overcome this issue, a heat masking shutter was devised and used to block and unblock hot air exposure when desired (Figure 72). Figure 72. Demonstration of sintering by forced convective heater Upon completion of deposition on a layer, the base material tank piston moves down by an increment equal to half of the layer thickness. The heater is then moved over the surface as shown in Figure 73. Sintering starts as the heater moves along the boundary and inner body areas (Figure 73). 91 Figure 73. Sintering profile around the contour of the part When sintering is completed the heat masking shutter is closed and the base plate moves down by an increment equal to half of layer thickness and the next layer is deposited. The reason for splitting the downward movement of platform to two separate half layer thickness increments is the following: 1. Prevent any collision between roller and the sintered layer, 2. Keep platform and heater as close as possible to each other for more effective heat transfer. Based on experiments performed to determine this distance, the platform should at least move down for half of the layer thickness (which is approximately 60 microns) to avoid any contact with the roller. 14.2. DOE for determining significant factors in forced convective sintering Experiments were designed to allow better understanding of relationships between the process inputs and the response. The DOE technique is used to test the experimental factors at different levels. The motivation to use this approach in this investigation is due to the high efficiency of DOE in analyzing process parameters and their interactions effect in convective sintering. This experiment is especially important as it is indicative of the bond formation between successive layers. Sintering profile Desired body Sintered contour 92 Table 17. Factors of importance A: Distance from powder B: Heat intensity C: Air flow D: Speed of movement Table 17 shows factors involved in the experiment and the design matrix can be seen in Table 18. This experiment was designed for the following objectives: 1. The interaction between distance of the heater from the surface, air pressure of the pump, heat intensity and finally speed of sintering are studied. Optimal setting can be found accordingly for more effective sintering. 2. Effective layer bonding can be examined through evaluation of layer thickness and strength of specimens. 3. Heat transfer mechanisms are compared, and better understanding is obtained. Levels are decided such that, in case of high distance from the surface, it is assumed the main source of sintering is forced convection while in case of low distance and low air pressure, radiation is expected to be the main source. It is important to note that in all cases, there is heat transfer through conduction. However, it is assumed that the impact of conduction is equally the same in all cases. 93 Table 18. Design matrix (1 at high level and -1 at low level) Trial Distance from powder Heat intensity Pump intensity Speed of movement 1 -1 -1 -1 -1 2 -1 -1 -1 1 3 -1 -1 1 -1 4 -1 -1 1 1 5 -1 1 -1 -1 6 -1 1 -1 1 7 -1 1 1 -1 8 -1 1 1 1 9 1 -1 -1 -1 10 1 -1 -1 1 11 1 -1 1 -1 12 1 -1 1 1 13 1 1 -1 -1 14 1 1 -1 1 15 1 1 1 -1 16 1 1 1 1 94 For each of the trials shown in the matrix in Table 18, a path with a length of 40 mm and diameter of the heater, 5 mm, was sintered (Figure 74). The width of the paths varies depending on the distance of the heater end from the powder surface as well as on air pressure. Figure 74. Sintered lines under different heating conditions Each run was repeated 3 times and to reduce bias, experiments were done in a randomized order. The average width, thickness, and strength of each specimen was recorded and are reflected in Table 19. 95 Table 19. Average values for three replications Trial Layer thickness (mm) Strength (N) Layer width (mm) 1 0.22 0.63 8.77 2 0.12 0.03 6.37 3 0.33 6.76 9.03 4 0.18 0.32 7.60 5 0.28 1.42 7.33 6 0.13 0.00 7.23 7 0.37 7.01 10.07 8 0.23 0.42 8.73 9 0.00 0.00 0.00 10 0.00 0.00 0.00 11 0.25 0.90 8.73 12 0.14 0.07 7.27 13 0.00 0.00 0.00 14 0.00 0.00 0.00 15 0.17 0.26 8.10 16 0.10 0.08 4.93 96 Figure 74 is a visual representation of the level of sintering achievable for different levels of heat intensity, air pressure, speed and distance of the heater end from the powder surface. Comparing values, highest layer thickness and strength was achieved in runs 7 and 3, respectively which relates to high air pressure, low sintering speed. The half-normal plot shown in Figure 75 represents significant factors and interaction in determining layer thickness. As it can be seen in the figure, the interaction between air pressure and speed plays a more important role in determining layer thickness than interaction between heat intensity and speed. Figure 75. Half normal plot for layer thickness Greater layer thickness shows better heat delivery to bottom layers. Higher penetration of heat through pores is achieved by convective heating. Effective heat transfer results in homogeneous densification and affects microstructure of sintered samples. Therefore, in the following section tensile tests are performed as well to examine mechanical properties of samples. 97 14.3. Micro tensile test Functional part fabrication was the primary focus of this research; therefore, it was necessary to study mechanical properties of fabricated parts. Effective bonding between layers is thermally driven and determines final mechanical properties of the resultant parts. The bond quality under each heating condition was assessed through measuring tensile strength of test cases. Small grip dog bone specimens were fabricated according to the design shown in Figure 76. Dog bone specimen with overall length of 31mm, width of 7.1mm, depth of 2.54 mm, gage length of 4 mm and 3mm wide and fillet radius of 1.5 mm. Figure 76. Engineering drawing of dog bone used for experiments (all dimensions are in mm) All specimens were centered in the build volume and printed under the same condition and remained at room temperature for 48 hours (Figure 77). 98 Figure 77. 3D printed dog bones For each run, three specimens were built in order to be able to avoid variations caused by different testing conditions. For tensile test a constant speed of tension of 0.5 mm/min was applied to all specimens (Figure 78). Test specimens were built out of Copolyamide and sintered under three various heating conditions. High-pressure convection and low infrared exposure, low-pressure convection and high infrared exposure, and equally distributed exposure of radiation and convection on the surface of the polymer. This experiment was designed for the following purposes: 1) creating a basis of comparison of different heating transfer methods for improving part strength, 2) comparing density of parts under each heat transfer method, and 3) evaluating total shrinkage after fabrication of each specimen. An infrared thermometer was positioned to point at a target location at the center of the platform and monitor the temperature for all heating treatments on the surface of the powder tank. 99 Figure 78. Deben 5KN micro tensile tester It was observed that the strength of the 3D printed part is highly dependent on the sintering process conditions. Sintering conditions are summarized in Table 20. Stress strain plots provide visual representation of final strengths. Table 20. Sintering trials Intensity Radiation Convection Sample1 Medium Medium Sample2 Medium Medium Sample3 Low High Sample4 Low High Sample5 High Low Sample6 High Low 100 It was observed that combination of having radiation and convection at medium levels, results in manufacturing of samples with best mechanical properties (sample 1, sample 2). Samples 1 and 2 both have convection and radiation at medium levels but the convection degree for sample 1 was higher than for sample 2. This is also the case for samples 3 and 4 which combined low radiation and high convection, but convection degree was higher for sample 3 than for sample 4. Weakest bonds were achieved under high intensity IR and low-pressure convection (samples 5 and 6) which could have been caused by insufficient penetration of hot air into the powder. Sample 6 received higher radiation than sample 5. Figure 79. Stress strain curve of specimen 1,2 0 2 4 6 8 10 12 14 16 0 0.1 0.2 0.3 0.4 Stress (MPa) Strain(mm) Samples 1, 2 Sample 1 Sample 2 101 Figure 80. Stress strain curve of specimen 3,4 Figure 81. Stress strain curve of specimen 5,6 Finally, in case of increasing convection heating in all samples, there was an increase in the overall strength. The stress strain curve for all samples can be seen in Figure 82. 0 1 2 3 4 5 6 0 0.1 0.2 0.3 0.4 Stress (MPa) Strain(mm) Samples 3, 4 Sample 3 Sample 4 0 0.5 1 1.5 2 2.5 3 0 0.1 0.2 0.3 0.4 Stress (MPa) Strain(mm) Samples 5, 6 Sample 5 Sample 6 102 Figure 82. Stress strain curve of specimen 14.4. Total shrinkage and density The total part shrinkage along X, Y and Z axes and density of each specimen has been recorded in Table 21. 0 2 4 6 8 10 12 14 16 0 0.1 0.2 0.3 0.4 Stress (MPa) Strain(mm) All dog bone specimen Sample 1 Sample 2 Sample 3 Sample 4 Sample 5 Sample 6 103 Table 21. Stress strain curve of specimen Density of samples were measured using Archimedes method in water. Copolyamide’s relative density in powder form is reported to be 1.150 gr/cm 3 by manufacturing company. 14.4.1. Discussion While radiation heat has been the main source of sintering in all SIS (successor technology for SSS) polymer research, its application in SSS was limited due to uncontrollable shrinkage. Unlike radiation which only impacts the top exposed powders, hot air can penetrate inside the power and result in deeper sintering, hence better interlayer adhesion as well as less deformation. The preferred heating condition is found to reduce shrinkage. Total shrinkage of % 0.3 along X and % 0.1 along Y were achieved. Additionally, convective heating has improved mechanical properties of the parts. Sample Density (gr/cm 3 ) Shrinkage along X (mm) Percentage shrinkage along X (%) Shrinkage along Y(mm) Percentage shrinkage along Y (%) 1 1.74 0.08 0.3 0.21 0.1 2 1.33 0.98 3.2 0.31 4.3 3 1.23 0.38 1.2 0.01 0.1 4 1.38 0.18 0.6 0.11 1.5 5 1.26 0.28 0.9 0.21 2.9 6 1.53 0.38 1.2 0.31 4.3 104 14.5. Reduction of waste material in SSS Amount of waste created by SSS-polymer has been reduced by at least 70 %. By the use of the spot heating approach. In the initial set of experiments a block of 1.5 inch×1.5 inch× (part height) was sintered and parts were extracted out of the block as shown in Figure 83. This was independent of the geometry or size of the part, thus, the remaining body of the block had to be discarded. Based on the current sintering algorithm, a narrow shell of polymer surrounds the part which is easily removed (Figure 84). Figure 83. Waste material in SSS by linear sintering 105 Figure 84. Waste material reduction by spot sintering Assuming that the distance between the heater and powder surface is constant, then the exact amount of waste by the spot heater can be calculated: Equation 9.5.1: W = 𝜌 𝑏 × ∫ 𝑓 (𝑦 )𝑑𝑦 𝑦 =𝑛 𝑦 =1 Equation 9.5.2: W = 𝜌 𝑏 𝐷 ℎ 2 ⁄ 𝐿 ∫ 𝑃 𝑦 𝑦 =𝑛 𝑦 =1 Where: W: Total amount of waste (𝑔𝑟 ) 𝐷 ℎ : Diameter of heater (𝑚𝑚 ) L : Nozzle penetration depth (Layer thickness) (𝑚𝑚 ) 𝑃 𝑦 : Perimeter of contour of layer y (mm) 𝜌 𝑏 : Density of B-powder ( 𝑔𝑟 𝑚𝑚 3 ⁄ ) 106 A larger outside boundary margin for sintering was desired and considered for this work. This allowed for better visualization over cross section, measurement of force required to separate the part from undesired regions as well as ensuring full sintering along the profile boundary. For future experiments this margin can be further reduced to minimize the amount of powder waste. 14.6. Discussion Suitable layer shape may be achieved at lower heat delivered over extended time to allow heat conduction from the top to lower layers. However, slow sintering prolongs fabrication time of the part. Convection-based heating can penetrate heat into top the two layers through the pores of the top layer and sinter the two layers simultaneously without the delay that conduction heating would require. 14.7. Demonstration parts The purpose of specifying a new S-powder as well as implementing a new heating mechanism was to ultimately enable the SSS fabrication process to produce credible polymeric parts. Previous SSS research on metal and ceramic parts provided a viable proof of concept to build high resolution parts. To demonstrate the capability of the technology in fabrication of polymeric parts near ideal conditions as determined in this research were created by selecting proper candidate base and separation powders and sintering regimen to fabricate some representative polymeric parts as shown Figure 85. 107 Figure 85. Polyamide samples built by SSS To minimize deformation the fabricated parts were left in the build tank for gradual cooling and then removed. The excess powder was then carefully blown away and the parts were extracted. There is not any post-processing required for black alumina and soda lime. For other S-powders a brush can be used to remove excess separation material from the part surface. 15. CONCLUSION This research began with modification of the existing SSS-metal machine to implement layer-by-layer sintering. The effort was then followed by identification of candidate base powder material for effective separation. Chromium Oxide was selected as successful separating agent and parts were produced for proof of concept. A comprehensive study on S-powder interaction with B-powder was conducted to improve part surface quality. The average geometric deviation of parts from expected measurements was reduced to less than 20 microns. Moreover, to achieve an effective bonding 108 between layers and reduce shrinkage, a new sintering method based on forced convective heating was implemented. Tensile tests were performed to study mechanical properties associated with samples and to find optimal setting of process parameters to maximize part strength. A sintering pattern was used which reduced the material waste by at least less than 70 %. The research was successful in establishing a framework for printing polymers with improved surface quality, higher strength and geometrical accuracy through shrinkage control. 16. PROPOSED FUTURE RESEARCH Although the current SSS research has addressed successful fabrication of metallic, ceramic, and polymer parts, there still remains a lot to be learned. Future research will help obtain a better insight on the physical aspects of the SSS process. The following research directions are suggested for SSS- polymer: 1. S-powder delivery: While surface quality was improved by the currently identified S-powders, it was noticed that better result could be achieved using finer and more spherical powders. However, there were limitations in depositing smaller than 40-micron powders using the current setup. Adjustments can be made and customized nozzle tubes with smaller inner diameter can be used to allow for finer powder particle size delivery. Moreover, although spherical powders result in smoother surface of parts, the application of spherical powders was limited due to uncontrollable leakage. Novel nozzle designs may be conceived to prevent leakage of spherical powders. 109 2. Selective heat sintering: Algorithm for spot sintering can be further improved for faster sintering intervals. Moreover, under constant distance from powder, decreasing the diameter of the heater will provide more focused heat transfer to the powder and decrease amount of waste material. 3. Base and separation powder choices: Although this research began with selecting the best base powder among a limited set of possibilities, there are still a large group of other polymers that may perform better in SSS and hence should be tested. A framework for selecting the suitable separating agents for best choices of B-powder may also be developed to eliminate exhaustive or blind search. 110 17. REFERENCES 1. “IDC Forecasts Worldwide Spending on 3D Printing to Be Nearly $12 Billion in 2018.” IDC: The Premier Global Market Intelligence Company, www.idc.com/getdoc.jsp?containerId=prUS43487718. 2. “Uses for 3D Printing 2015-2018 | Statistic.” Statista,www.statista.com/statistics/560271/worldwide- survey-3d-printing-uses/. 3. Stansbury, Jeffrey W., and Mike J. Idacavage. "3D printing with polymers: Challenges among expanding options and opportunities." Dental Materials 32.1 (2016): 54-64. 4. https://www.forbes.com/sites/louiscolumbus/2018/05/30/the-state-of-3d-printing- 2018/#7886cfbd7b0a 5. Dimitrov, D., K. Schreve, and N. De Beer. "Advances in three dimensional printing-state of the art and future perspectives." Journal for New Generation Sciences 4.1 (2006): p-21. Melchels, Ferry PW, et al. "Additive manufacturing of tissues and organs." Progress in Polymer Science 37.8 (2012): 1079-1104. 6. Campbell, Thomas, et al. "Could 3D Printing Change the World?."Technologies, Potential, and Implications of Additive Manufacturing. Washington, DC: Atlantic Council (2011). 7. Hague, R., et al. "Materials analysis of stereolithography resins for use in Rapid Manufacturing." Journal of materials science 39.7 (2004): 2457-2464. 8. Tumbleston, John R., David Shirvanyants, Nikita Ermoshkin, Rima Janusziewicz, Ashley R. Johnson, David Kelly, Kai Chen et al. "Continuous liquid interface production of 3D objects." Science (2015): aaa2397. 111 9. Wong, Ken CL, Ronald M. Summers, Electron Kebebew, and Jianhua Yao. "Pancreatic tumor growth prediction with elastic-growth decomposition, image-derived motion, and FDM-FEM coupling." IEEE transactions on medical imaging 36, no. 1 (2017): 111-123. 10. Roberts, Ibiye Aseibichin, et al. "A three-dimensional finite element analysis of the temperature field during laser melting of metal powders in additive layer manufacturing." International Journal of Machine Tools and Manufacture 49.12 (2009): 916-923. 11. De Beer, N. "Advances in three dimensional printing-state of the art and future perspectives." Journal for New Generation Sciences 4.1 (2006): 21-49. 12. Dimitrov, D., K. Schreve, and N. De Beer. "Advances in three dimensional printing-state of the art and future perspectives." Journal for New Generation Sciences 4.1 (2006): p-21. 13. Baumers, M., C. Tuck, and R. Hague. "Selective heat sintering versus laser sintering: comparison of deposition rate, process energy consumption and cost performance." (2016). 14. Khoshnevis, Behrokh, et al. "SIS–a new SFF method based on powder sintering." Rapid Prototyping Journal 9.1 (2003): 30-36. 15. Khoshnevis, Behrokh, Mahdi Yoozbashizadeh, and Yong Chen. "Metallic part fabrication using selective inhibition sintering (SIS)." Rapid Prototyping Journal 18.2 (2012): 144-153. 16. Asiabanpour, Bahram, et al. "Advancements in the SIS process." 14th International Symposium on Solid Freeform Fabrication, Austin, TX. 2003. 17. Chua, Chee Kai, and Kah Fai Leong. 3D Printing and Additive Manufacturing: Principles and Applications (with Companion Media Pack) of Rapid Prototyping Fourth Edition. World Scientific Publishing Company, 2014. 112 18. Fischer, Fred. "FDM and Polyjet 3D printing." Popular Plastics & Packaging 60, no. 6 (2015). 19. Khoshnevis, Behrokh, and Jing Zhang. "Selective Separation Sintering (SSS)-An Additive Manufacturing Approach for Fabrication of Ceramic and Metallic Parts with Application in Planetary Construction." AIAA SPACE 2015 Conference and Exposition. 2015. 20. Zhang, Jing, and Behrokh Khoshnevis. "Selective Separation Sintering (SSS) A New Layer Based Additive Manufacturing Approach for Metals and Ceramics." Solid Freeform Fabrication Symposium. 2015. 21. Zhang, J., and B. Khoshnevis. "APPLICATION OF SELECTIVE SEPARATIONSINTERING IN CERAMICS 3D PRINTING." Advanced Processing and Manufacturing Technologies for Nanostructured and Multifunctional Materials II: Ceramic Engineering and Science Proceedings, Volume 36 6 (2015): 151. 22. Khoshnevis, Behrokh, Xiang Gao, Brittany Barbara, and Hadis Nouri. "Selective Separation Shaping (SSS)–Large-Scale Fabrication Potentials." 23. Chung, Haseung, and Suman Das. "Processing and properties of glass bead particulate-filled functionally graded Nylon-11 composites produced by selective laser sintering." Materials Science and Engineering: A 437.2 (2006): 226-234. 24. Ajoku, Uzoma, et al. "Investigating mechanical anisotropy and end-of-vector effect in laser- sintered nylon parts." Proceedings of the Institution of Mechanical Engineers, Part B: Journal of Engineering Manufacture 220.7 (2006): 1077-1086. 25. Kim, J., and T. S. Creasy. "Selective laser sintering characteristics of nylon 6/clay-reinforced nanocomposite." Polymer Testing 23.6 (2004): 629-636. 113 26. Strobl, Gert. "Crystallization and melting of bulk polymers: new observations, conclusions and a thermodynamic scheme." Progress in polymer science 31, no. 4 (2006): 398-442. 27. 'Optimizing Metal Powders for Additive Manufacturing: Exploring the Impact of Particle Morphology and Powder Flowability', https://www.additivemanufacturing.media/blog/post/optimizing- metal-powders-for-additive-manufacturing-exploring-the-impact-of-particle-morphology-and-powder- flowability 28. Scott, G. D., and D. M. Kilgour. "The density of random close packing of spheres." Journal of Physics D: Applied Physics 2, no. 6 (1969): 863. 29. Khoshnevis, Behrokh, et al. "SIS–a new SFF method based on powder sintering." Rapid Prototyping Journal 9.1 (2003): 30-36. 30. Khoshnevis, Behrokh, and Jing Zhang. "Selective Separation Sintering for Metallic Part Fabrication." 31. Pokluda, Ondřej, Céline T. Bellehumeur, and John Vlachopoulos. "Modification of Frenkel's model for sintering." AIChE journal 43.12 (1997): 3253-3256. 32. Oghbaei, Morteza, and Omid Mirzaee. "Microwave versus conventional sintering: a review of fundamentals, advantages and applications." Journal of Alloys and Compounds 494.1 (2010): 175-189. 33. Zeng, Kai, Deepankar Pal, and Brent Stucker. "A review of thermal analysis methods in Laser Sintering and Selective Laser Melting." Proceedings of Solid Freeform Fabrication Symposium Austin, TX. 2012. 34. Roberts, Ibiye Aseibichin, et al. "A three-dimensional finite element analysis of the temperature field during laser melting of metal powders in additive layer manufacturing." International Journal of Machine Tools and Manufacture 49.12 (2009): 916-923. 114 35. Incropera, Frank P., and David P. De Witt. "Fundamentals of heat and mass transfer." (1985). 36. Shi, Y., et al. "Effect of the properties of the polymer materials on the quality of selective laser sintering parts." Proceedings of the Institution of Mechanical Engineers, Part L: Journal of Materials Design and Applications 218.3 (2004): 247-252. 37. Ngo, Tuan D., Alireza Kashani, Gabriele Imbalzano, Kate TQ Nguyen, and David Hui. "Additive manufacturing (3D printing): A review of materials, methods, applications and challenges." Composites Part B: Engineering (2018). 38. Adekanye, Adefemi, Rasheedat Modupe Mahamood, Esther Titilayo Akinlabi, and Moses Gbadebo Owolabi. "Additive manufacturing: the future of manufacturing." Materiali in tehnologije 51, no. 5 (2017): 709-715. 39. Goodridge, R. D., C. J. Tuck, and R. J. M. Hague. "Laser sintering of polyamides and other polymers." Progress in Materials Science 57, no. 2 (2012): 229-267. 40. Verbelen, Leander, et al. "Characterization of polyamide powders for determination of laser sintering processability." European Polymer Journal 75 (2016): 163-174. 41. Yusoff, W. A. Y., et al. "Influence of molecular weight average, degree of crystallinity, and viscosity of different polyamide PA12 powder grades on the microstructures of laser sintered part." MATEC Web of Conferences. Vol. 26. EDP Sciences, 2015. 42. Pokluda, Ondřej, Céline T. Bellehumeur, and John Vlachopoulos. "Modification of Frenkel's model for sintering." AIChE journal 43.12 (1997): 3253-3256. 43. Gopal, Erode. Specific heats at low temperatures. Springer Science & Business Media, 2012. 115 44. Kieffer, Susan Werner. "Thermodynamics and lattice vibrations of minerals: 1. Mineral heat capacities and their relationships to simple lattice vibrational models." Reviews of Geophysics17, no. 1 (1979): 1-19. 45. Senthilkumaran, K., Pulak Mohan Pandey, and PV Madhusudan Rao. "Shrinkage compensation along single direction dexel space for improving accuracy in selective laser sintering." Automation Science and Engineering, 2008. CASE 2008. IEEE International Conference on. IEEE, 2008. 46. Wang, Xiangwei. "Calibration of shrinkage and beam offset in SLS process."Rapid Prototyping Journal 5.3 (1999): 129-133. 47. Senthilkumaran, K., Pulak Mohan Pandey, and PV Madhusudan Rao. "Shrinkage compensation along single direction dexel space for improving accuracy in selective laser sintering." Automation Science and Engineering, (2008). CASE 2008. IEEE International Conference on. IEEE, 2008 48. Raghunath, N., and Pulak M. Pandey. "Improving accuracy through shrinkage modelling by using Taguchi method in selective laser sintering." International Journal of Machine Tools and Manufacture 47.6 (2007): 985-995. 49. Dai, K., and L. Shaw. "Distortion minimization of laser-processed components through control of laser scanning patterns." Rapid Prototyping Journal 8.5 (2002): 270-276. 50 Anitha, R., S. Arunachalam, and P. Radhakrishnan. "Critical parameters influencing the quality of prototypes in fused deposition modelling." Journal of Materials Processing Technology 118.1 (2001): 385-388. 51. Sood, Anoop Kumar, R. K. Ohdar, and S. S. Mahapatra. "Improving dimensional accuracy of fused deposition modelling processed part using grey Taguchi method." Materials & Design 30.10 (2009): 4243 116 52. Górski, Filip, Wiesław Kuczko, and Radosław Wichniarek. "Influence of process parameters on dimensional accuracy of parts manufactured using Fused Deposition Modelling technology." Advances in Science and Technology–Research Journal 7.19 (2013). 53. Jacobs, Paul F. "The ExpressTool Process." Rapid Tooling: Technologies and Industrial Applications (2000): 161. 54. Senthilkumaran, K., Pulak M. Pandey, and P. V. M. Rao. "Influence of building strategies on the accuracy of parts in selective laser sintering." Materials & Design 30.8 (2009): 2946-2954.

Abstract (if available)

Abstract Additive manufacturing (AM), or three-dimensional (3D) printing has enjoyed a recent surge of attention over the past few years. AM is a process in which digital 3D design data (CAD model) is used as an input to build physical objects by stacking layers of material. A variety of materials have been used by AM techniques and new materials and processing technologies are constantly being developed to improve current processes and to create new ones. 3D printing market size worldwide is estimated to be around $12 billion dollars by the end of 2018 and at 18 billion dollars by 2021 [1]. AM technologies have certain limitations hence there is a high demand for research in this area to help AM become a mainstream manufacturing technique. ❧ Selective Separation Shaping (SSS) is a new powder-based AM technology which was developed at USC with the goal of fabricating low cost, high resolution 3D parts. The main advantage of SSS is that this process enables building of fully functional pieces without the need of any intermediate binder or high cost laser operation. This process has been primarily applied to metallic, ceramic and composite materials including concrete with numerous test cases successfully built. ❧ Among all library of materials that have been processed by commercialized AM techniques so far, polymers have been the most-widely used material due to its large range of applications with 65% of market share based on Sculpteo statistics report in 2018 [2]. The main goal of this research is to provide a framework for an effective selection of materials and process parameters in SSS for successful fabrication of polymer parts. Nylon 6,6 has been used as a starting base and several test cases are fabricated to identify the key performance factors in this process. Design of experiments is performed and interaction between separation material and base material is investigated to improve the surface quality of parts. Additionally, different heating mechanisms are studied to achieve better control over shrinkage and maintain an effective binding between layers. In the following sections, first, various approaches in AM for fabrication of polymer parts are presented in brief review

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Selective separation shaping: an additive manufacturing method for metals and ceramics

PDF

Selective Separation Shaping (SSS): large scale cementitious fabrication potentials

PDF

Metallic part fabrication wiht selective inhibition sintering (SIS) based on microscopic mechanical inhibition

PDF

Enhancing the surface quality and dimensional accuracy of SIS-metal parts with application to high temperature alloys

PDF

Deformable geometry design with controlled mechanical property based on 3D printing

PDF

Fabrication of ultrasound transducer and 3D-prinitng ultrasonic device

PDF

A data driven software platform for process automation, planning and inspection of Contour Crafting large-scale robotic 3D printing system

PDF

Statistical modeling and machine learning for shape accuracy control in additive manufacturing

PDF

Scale-up of vapor-phase deposition of polymers: towards large-scale processing

PDF

Energy control and material deposition methods for fast fabrication with high surface quality in additive manufacturing using photo-polymerization

PDF

Hybrid vat photopolymerization: methods and systems

PDF

2D ultrasonic transducer array’s design and fabrication with 3D printed interposer and applications

PDF

Hybrid vat photopolymerization processes for viscous photocurable and non-photocurable materials

PDF

Object detection and recognition from 3D point clouds

$Metasurfaces in 3D applications: multiscale stereolithography and inverse design of diffractive optical elements for structured light$

PDF

Metasurfaces in 3D applications: multiscale stereolithography and inverse design of diffractive optical elements for structured light

PDF

3D deep learning for perception and modeling

PDF

An analytical and experimental study of evolving 3D deformation fields using vision-based approaches

PDF

Towards building a live 3D digital twin of the world

PDF

3D printing and compression testing of biomimetic structures

PDF

Slurry based stereolithography: a solid freeform fabrication method of ceramics and composites

Asset Metadata

Creator Nouri, Hadis (author)

Core Title 3D printing of polymeric parts using Selective Separation Shaping (SSS)

School Viterbi School of Engineering

Degree Doctor of Philosophy

Degree Program Industrial and Systems Engineering

Publication Date 12/12/2018

Defense Date 07/12/2018

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag 3D part fabrication,3D printing,additive manufacturing,agent,alumina,AM,angle of repose,boundary,Cement,ceramic,chromium oxide,composite,controlled deposition,convection,deposition,design of experiments,differential scanning calorimetry,flowability,forced convective heater,functional,gap depth,gap width,gypsum,heating element,high resolution,hot air,hydraulically activated powders,inhibition,inhibitor,insertion depth,low cost,macro scale,mechanical interlocking,meso scale,Metal,move along contour,movement inside powder,multiple classes of materials,Needle,New,nozzle,nylon,OAI-PMH Harvest,opening,particle size,penetration,piezo disk,polyamide,polymer,polymeric,porosity,powder-based,Radiation,rotation,selective,SEM image,separation,separator,shaping,sintering,soda lime,stress strain curve,surface quality,tensile test,thin wall,vibration generator

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Khoshnevis, Behrokh (committee chair), Chen, Yong (committee member), Huang, Qiang (committee member), Nutt, Steven (committee member), Wang, Qiming (committee member)

Creator Email hadis.noori@gmail.com,hnouri@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c89-113859

Unique identifier UC11675756

Identifier etd-NouriHadis-7015.pdf (filename),usctheses-c89-113859 (legacy record id)

Legacy Identifier etd-NouriHadis-7015.pdf

Dmrecord 113859

Document Type Dissertation

Format application/pdf (imt)

Rights Nouri, Hadis

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