ABSTRACT Polyether ether ketone (PEEK) is introduced as a material for the additive manufacturing process called fused filament fabrication (FFF), as opposed to selective laser sintering (SLS) manufacturing. FFF manufacturing has several advantages over SLS manufacturing, including lower initial machine purchases costs, ease of use (spool of filament material vs powder material), reduced risk of material contamination and/or degradation, and safety for the users of the equipment. PEEK is an excellent candidate for FFF due to its low moisture absorption as opposed to other common FFF materials, such as Acrylonitrile Butadiene Styrene (ABS). PEEK has been processed into a filament and samples have been manufactured using several build orientations and extrusion paths. The samples were used to conduct tensile, compression, flexural, and impact testing to determine mechanical strength characteristics such as yield strength, modulus of elasticity, ultimate tensile strength and maximum elongation, etc. All tests were conducted at room temperature. A microscope analysis was also conducted to show features on the failures surfaces. The mechanical property results from this study are compared to other published results using traditional thermo-plastic manufacturing techniques, such injection molding.
Tensile testing was conducted at three raster orientations, 0°, 90° and alternating between 0° and 90°. Average ultimate tensile stresses were determined to be 73 MPa for 0° orientation, and 54 MPa for 90° orientation, with alternating 0°/90° orientations of 66.5 MPa. Compression testing was conducted at two raster orientations, 0° and alternating between 0° and 90°. Average ultimate strength for the single orientation direction was 80.9 MPa with the alternating orientations at 72.8 MPa. Flexural testing was conducted at three raster orientations, 0°, 90° and alternating between 0° and 90°. Ultimate flexural stress was determined to be 111.7 MPa for 0°, 79.7 MPa for 90°, and 95.3 MPa for orientations alternating between 0° and 90°. Finally, impact testing was conducted at three raster orientations, 0°, 90° and alternating between 0° and 90°. Average impact energy absorbed was determined to be 17.5 Nm in the 0° orientation, 1.4 Nm in the 90° orientation, and 0.7 Nm for the alternating 0° and 90° orientations. KEY WORDS: Additive Manufacturing (AM), Polyether ether ketone (PEEK), FDM, FFF, Material Properties
Proceedings of the ASME 2015 International Mechanical Engineering Congress and Exposition IMECE2015
INTRODUCTION Additive manufacturing (AM) has been described by many as the beginning of the “next industrial revolution”, and as such, additive manufacturing processes have gained popularity in the field of aerospace, automotive, biomedical, energy and other fields with the advancement of superior technologies. Today, plastics, metals, ceramics and even glass are being used for prototyping as well as fabricating functional parts. Much of the AM field has advanced very rapidly in recent years, such as hardware and software advancements. However, one of the current limitations of AM is inadequate high quality materials. To advance AM, higher quality engineering materials need to be developed. PEEK (Polyether ether ketone), is currently used in selective laser sintering (SLS) additive manufacturing. PEEK is a semi-crystalline thermoplastic with a linear aromatic polymer structure. This fully recyclable material has excellent mechanical characteristics (wear, fatigue, and creep), chemical resistance properties and sustains very high temperatures. With its inherent low specific gravity (1.3 g/cm3), PEEK compounds can be used to manufacture components with high strength and modulus properties similar to aluminum or steel, but with a much lower weight (approximately 70% weight reduction with similar mechanical properties). Lower density materials with high performance characteristics, such as PEEK, are of special interest to the aerospace and automotive industries, which are always looking to reduce weight without sacrificing structural integrity. PEEK (Polyether ether ketone) is a high performance thermoplastic polymeric material with extremely high melting point (343˚C), wear and abrasive properties superior to steel and titanium, chemical inertness and biocompatibility. Its true scientific name is poly (oxy-1, 4-phenylene-oxy-l, 4-phenylenecarbonyl-l, 4-phenylene) (See
Figure 1).
Figure 1: PEEK's repeating unit The first commercial manufacturing of PEEK (known as VICTREX® PEEK) began back in 1978 by Imperial Chemicals Industries Limited. VICTREX® Ultra-High Purity PEEK™ polymers and VICTREX® High-Flow PEEK™ polymers were introduced in 2005 [1]. Since then, these polymers have been featured in a wide range of applications such as aerospace, automotive, energy and biomedical [2-6].
Traditionally, PEEK components have been manufactured with processes such as injection molding and hot pressing. Powdered PEEK has been used with high temperatures in these processes. Manufacturing complex shapes with PEEK has not been possible until additive manufacturing processes were adopted. All previous research [7-11] on AM of PEEK has focused on Selective Laser Sintering (SLS). The SLS process employs computer controlled laser pulses to sinter thin layers of powdered PEEK to form solid objects. A 3-D CAD model is “sliced” into many 2-D layers and a laser path is developed. The laser selectively fuses powdered PEEK, one layer at a time. In between layers, powdered PEEK is reapplied to the top surface of the previous layer, and a new layer is then sintered. The process repeats until the entire object has been created. In the present work, the Fused Filament Fabrication (FFF) technique has been employed instead of the SLS technique. This approach is simpler for the user as handling filament is far easier than powder. Moreover, there is less contamination and/or degradation and safety concerns in using filaments over powder. PEEK filament can be spooled similar to other common materials (ABS, Polylactic acid (PLA), etc.). Furthermore, the FFF technique allows for continuous, controlled deposition of material compared to fusion of discrete particles. This paper presents the results of materials testing conducted on samples produced through Fused Filament Fabrication of PEEK. Material properties, such as yield strength, ultimate strength, modulus of elasticity, modulus of flexure, and impact energies will be presented. The results of this testing are also compared to material properties of parts manufactured using traditional PEEK manufacturing techniques. FUSED FILAMENT FABRICATION (FFF) Fused Filament Fabrication (FFF) AM technology is also known as Fused Deposition Modeling (FDM). Fused Deposition Modeling and it’s abbreviation FDM are trademarked by Stratasys Inc., one of the major market players in the field of additive manufacturing technology. In FFF (or FDM), a spool of filament of a certain material (usually thermoplastic or wax) is pushed through an extrusion nozzle. The nozzle is heated and the material is melted and deposited as necessary to build each layer. Stepper motors or servo motors are usually used to aid the movement of the extruder. The extruder moves horizontally, in the x-y plane to deposit the material on the build plate. After the entire layer has been created, the build plate moves in the z direction to allow for a new layer to be deposited on top of the previous layer. The part is created using the bottom up approach.
SPECIMEN DESIGN AND FABRICATION Four different mechanical tests (Tension, Compression, Flexural, and Impact) were performed in this study. Tensile testing specimens were designed with specifications conforming to ASTM D638 Standard Test Method for Tensile Properties of Plastics (see Figure 3). Compression testing specimens were designed with specifications conforming to ASTM D695 Standard Test Method for Compressive Properties of Rigid Plastics (see Figure 4). Flexural testing specimens were designed with specifications conforming to ASTM D790 Standard Test Methods for Flexural Properties of Unreinforced and Reinforced Plastics and Electrical Insulating Materials (see Figure 5-specimen drawing and finished picture).). Impact testing specimens were designed according to ASTM D6110 Standard Test Method for Determining the Charpy Impact Resistance of Notched Specimens of Plastics (see Figure 6). The specimens were manufactured at raster orientations of 0˚, 90˚, and alternating 0˚/90˚ to quantify material properties at a range of possible raster orientations (see Figure 7 for raster orientation definitions). The models were then sliced at the raster orientations listed above using the build parameters listed in Table 1. An Arevo Labs 3D printer was used to fabricate the specimens. The PEEK used in this study is a proprietary PEEK formulation from Arevo Labs.
Parameter Value Infill 100% Layer Height 0.25 mm Extruder Temperature 340 °C Platform Temperature 230 °C Extruder Print Speed 50 mm/sec Nozzle diameter (flexural, impact, compression)
1.91 mm
Nozzle diameter (tensile) 1.8 mm
EXPERIMENTAL METHODS Tensile testing was performed in accordance to ASTM D638 [12] using an MTS 370 Landmark Universal testing machine with extensometer. This machine has an MTS 100 kN load cell and an MTS 634.31F-24 extensometer with a 20mm gauge length (see Figure 8). A constant displacement rate of 5 mm/min was used to test the specimen. Data points were collected at 99 Hz. Compression testing was performed in accordance to ASTM D695 [13] using the same MTS machine, however, strain was calculated using the built in LVDT for compression testing (see Figure 9). A constant displacement rate of 1.30 mm/min was used to compress the specimen. Standard engineering stress/strain equations were used to calculate stress and strain for both of these tests types. Yield stress calculations were made using the standard 0.2% offset method.
Figure 8. Tensile Testing Setup
Figure 9. Compression Testing Setup Flexural testing was performed in accordance to ASTM D790 [14] using an MTS Insight 5 universal testing machine. This machine has a MTS 5kN load cell and the displacement (later converted to strain) was measured using the built in LVDT (see Figure 10). For flexural testing, stress was calculated as shown in Equation 1 and strain was calculated as shown in Equation 2.
σ = Flexural Stress (MPa) P = Load (N) L =Support span (50 mm) b = width of specimen (mm) t = thickness of beam (mm)
𝜀 = 6𝑑𝑡𝐿2 (Eq. 2)
where: ε = flexural strain (mm/mm) d = deflection (mm)
Figure 10. Flexural Testing Setup Impact tests were performed in accordance to ASTM D6110 [15] using the Instron SI 300 with built in sensors to measure pendulum swing angle (see Figure 11).
Figure 11. Impact Testing Setup All samples were exposed to room temperature conditions for several days before testing and all tests were conducted at room temperature of approximately 20° C. RESULTS Tensile Testing Nine tensile samples were tested at three different raster orientation styles. The full results of the tensile testing are summarized in Table 2. Each raster orientation style showed different results and different fracture styles and characteristics. The 0° samples showed the highest ultimate tensile strength, but also showed the lowest 0.2% offset yield strength. Table 2. Summarized Tensile Testing Results
Raster Orientation
Ultimate Stress (MPa)
Yield Stress (MPa)
Fracture Stress (MPa)
Modulus of
Elasticity (MPa)
Max Elongation
(%)
0° 71.36 34.68 70.29 2871.47 5.01
0° 74.49 45.77 73.80 2825.66 3.99
0° 73.19 46.30 72.49 2643.44 4.45
90° 53.91 45.93 53.50 2846.72 2.29
90° 57.69 45.29 57.69 2694.62 2.86
90° 50.63 43.13 50.24 2673.16 2.29
0°/90° 67.75 40.40 67.75 2732.56 3.93
0°/90° 74.46 50.79 72.21 2483.82 6.77
0°/90° 57.35 41.20 57.15 2791.29 2.84
The failure methods for each type of sample was also different. The samples with raster orientation angles of 0° showed a partial failure due to delamination between layers about halfway
through the specimen. One half of the specimen failed with a clean break, while the other half of the sample stayed intact, but seemed to delaminate not only between layers, but also between tool path lines. The remaining sample types failed with a sudden clean fracture. Figure 12 shows the failure patterns for each of the sample types.
Figure 12. Fractured Tensile Specimen (Top 3 samples are 0°, Middle 3 samples are 90°, Bottom 3 samples are 0°/90° alternating) Figures 13 and 14 show the fracture surface of a representative tensile sample after testing. Since the 0° samples didn’t completely fracture, microscope images were not possible.
Figure 15. 0°/90° Tensile fracture surface, 100x zoom Compression Testing A total of six compression samples were tested at two different orientations (the corresponding third orientation to the tensile testing would have been redundant in a cylindrical sample). The samples using a single raster orientation angle had a higher yield stress and ultimate tensile strength. In addition, the single raster orientation samples also had more consistent test results. The improved compressive strength of the single raster is likely from the increased surface area contact between each layer compared to the alternating orientation angle. Table 3 shows a summary of the test results from these samples. Table 3. Summarized Compression Testing Results
Raster Orientation
Ultimate Stress (MPa)
Yield Stress (MPa)
Modulus of
Rupture (MPa)
Modulus of
Elasticity (MPa)
Max Elongation
(%)
0˚
83.61 71.15 83.10 2016.42 6.87
79.67 63.46 79.62 2033.34 6.09
79.33 63.58 78.18 2056.19 6.99
0˚/90˚
69.70 52.35 68.19 1936.97 7.90
64.15 47.03 63.06 1931.96 7.45
84.49 61.63 83.75 2324.40 5.84
Figures 16 and 17 shows the fractured samples. For the single raster orientation angle samples, the fracture pattern was a consistent clean fracture due to buckling. For the samples with
alternating raster orientations, the fractures were more inconsistent and much less of a clean fracture.
Figure 16. Fractured Compression Specimen 0°
Figure 17. Fractured Compression Specimen 0°/90° Microscope images (Figures 18 to 20) of each sample type show similar patterns. All fracture surfaces from the single raster orientation samples show nearly identical patterns. The alternating raster orientation samples show interesting fracture patterns. At first glance in Figure 19, it seems as though the filament fibers are coming out of the picture. However, as Figure 20 shows at 100x magnification, those “fibers” are actually the marks left from the layer of material ripped off the previous layer during the fracturing process.
Figure 18. Fracture surface of 0° compression sample at 20x zoom
Figure 19. Fracture surface of 0°/90° compression sample at 20x zoom
Figure 20. Fracture surface of 0°/90° compression sample at 100x zoom Flexural Testing A 3-point bending test was performed to obtain the flexural properties of the specimen. Eight samples were tested at three orientations. As with the tensile testing, the samples with raster orientation of 0° had the highest ultimate stress. However, unlike the tensile testing, these samples also had the highest yield stress, modulus of rupture, and max elongation. The other two raster orientation types, had similar yield stress values, but the alternating raster orientations had a higher ultimate stress. Table 4. Summarized Flexural Testing Results
Raster Orientation
Ultimate Stress (MPa)
Yield Stress (MPa)
Modulus of Rupture (MPa)
Flexure Modulus
(MPa)
Max Elongation
(%)
0˚ 114.16 86.26 114.16 1972.25 10.60
109.18 104.61 109.18 1865.87 8.12
90˚
83.59 65.90 83.59 1954.54 5.81
76.85 65.78 76.85 2039.55 4.33
78.63 65.88 78.63 1979.81 4.70
0˚/90˚
88.70 66.50 88.70 2146.28 6.58
95.22 75.78 95.22 2496.74 5.67
102.10 76.96 102.10 2585.70 8.84
As Figures 21 to 23 show, the 0° samples did not fracture like the other two sample types. The tests were stopped when the samples had deflected so much that they were touching the supports (only data from before contact was used in the analysis).
Figure 21. Fractured Flexural Specimen 0°
Figure 22. Fracture Flexural Specimen 90°
Figure 23. Fracture Flexural Specimen 0°/90° Impact Testing Charpy impact testing was performed with an Instron SI 300 capable of measuring a maximum energy absorption of 300 ft-lb. The notched specimens were exposed to flexural shock by a single swing of the pendulum type hammer and the resistance strength to breakage or crack propagation is measured. A total of 12 specimens were tested in three raster orientation types. Table 5 shows a summary of testing results below. Table 5. Summary of Impact Testing Results
The results of this testing show a clear difference between raster orientations. The 0° raster orientation samples absorbed significantly more energy than the other two orientations. Of the other two orientations, the 90° samples absorbed more than the alternating raster orientations. Figure 24 shows a representative fractured sample from the impact testing at 0° raster orientation. In all of these samples, the material delaminated in at least one location and did not fracture into two pieces. Figures 25 and 26 show the fracture patterns for each of the remaining two raster orientations. In each of these orientations, the samples fractured into two pieces, or nearly did (only hanging on by the outer shell of the sample).
Figure 24. Typical fracture pattern for 0° impact specimen
Figure 25. Fractured specimen for 90° impact specimen
Figure 26. Fractured specimen for 0°/90° impact specimen Figures 27 and 28 show the fracture surface of two representative samples from the 90° raster orientation type and the 0°/90° alternating raster orientation type.
Figure 27. Fractured surface of 90° impact specimen at 20x zoom
Figure 28. Fractured surface of 0°/90° impact specimen at 20x zoom DISCUSSION AND COMPARISONS The tensile properties of PEEK produced by conventional thermoplastic processing techniques vary at different temperatures as well as different strain rates due to the viscoelastic nature of this material. Limited information is available for material properties of SLS or SLA manufactured PEEK. However, for injection molded components, tensile strength at room temperature is in between 75-220 MPa [16] depending on the blend of PEEK. This compares fairly well to the results presented in this paper of 73 MPa with 0° raster orientation and 54 MPa with 90° raster orientation. The Modulus of Elasticity values presented in this paper vary between 2.6-2.8 GPa, which compares to 3.6 GPa [17] for injection molded PEEK components. Flexural Modulus properties vary in between 1-25 GPa depending on the temperature for injection molded components, while at room temperature flexural modulus results vary in between 1.9-2.3 GPa in this paper. The influence of different raster orientations on the properties is more or less similar to other additively manufactured thermoplastics (such as ABS, PLA) tested at various raster orientations. For the tensile specimens, 0° samples demonstrate stronger tensile properties as the mechanical loading direction and the printing orientation are same. The 90° samples have the weakest tensile strengths as the mechanical loading direction and the printing orientations are normal to each other. CONCLUSION The objective of present work was to investigate the mechanical behaviors of additively manufactured PEEK components using the Fused Filament Fabrication (FFF) technique. Specially produced PEEK filament was used to manufacture the test specimens, unlike most other AM technologies that have been using powdered PEEK. Three different raster orientations (0°, 90°, and 0°/90°) have been incorporated while making the
specimens to study the influence of these different orientations on the properties of the components. Several different mechanical tests (Tension, Compression, Impact, and Flexural) have been performed. Trends in strength results related to printing orientation were shown to be similar to the trends of other Fused Filament Fabrication materials, such as ABS and PLA, yet PEEK filament showed better material properties. ACKNOWLEDGEMENTS The authors would like to thank the Materials Evaluation and Testing Lab (METLAB) and the Mechanical Engineering Department at South Dakota State University for the use of testing equipment. The authors would also like to thank Arevo Labs for supplying the PEEK material and manufacturing all of the samples. REFERENCES [1] Thomas, S., Visakh, P.M., 2011, Handbook of Engineering
[2] Shekar, R. I., Kotresh, T. M., Rao, P. M. D., and Kumar, K.. (2009). Properties of high modulus PEEK yarns for aerospace applications. Journal of Applied Polymer Science, 112(4), 2497-2510. doi: 10.1002/app.29765.
[4] Victrex plc. Victrex Energy Brochure 2013. http://www.victrex.com/en/energy
[5] Kurtz, S.M., Devine, J.N., 2007. PEEK biomaterials in trauma, orthopedic, and spinal implants. Biomaterials 28, 4845–4869.
[6] Williams, D.F., et al., 1987. Potential of polyether ether ketone (PEEK) and carbon–fibre-reinforced PEEK in medical applications. Journal of Materials Science Letters 6, 188–190.
[7] Tan, K. H., Chua, C. K., Leong, K. F., Cheah, C. M., Cheang, P., Abu Bakar, M. S., and Cha, S. W. (2003). Scaffold development using selective laser sintering of polyetheretherketone–hydroxyapatite Biocomposite blends. Biomaterials, 24(18), 3115-3123. doi: 10.1016/S0142-9612(03)00131-5.
[8] Beard, M.A., Ghita,O.R., Bradbury, J., Flint,S. and Evans,K.E. (2012). “Mechanical Characterization of Additive Manufacturing of componets made from a polyetherketone (PEK) high temperature thermoplastic polymer,” Proceedings of the 5th international conference on advanced research and rapid protyping, p. 329-332.
[9] Pohle, D., Ponader, S., Rechtenwald, T., Schmidt,M., Schlegel K.A., Münstedt, H., Neukam, F.W., Nkenke, E. & von Wilmowsky, C. 2007. Processing of threedimensional laser sintered polyetheretherketone composites and testing of osteoblast proliferation in vitro. Macromolecular Symposia 253(1):65–70.
[10] Rechtenwald, T., Eßer, G., Schmidt, M. & Pohle, D. 2005. Comparison between laser sintering of PEEK and PA using design of experiment method. Proceedings of the 3rd International WLT-Conference on Lasers in Manufacturing (LIM 2005) 3:263–267.
[11] Schmidt,M., Pohle,D., Rechtenwald,T., 2007, “Annals of the CIRP” Vol. 56/1/2007. doi:10.1016/j.cirp.2007.05.097
[12] ASTM D638-14, Standard Test Method for Tensile Properties of Plastics, ASTM International, West Conshohocken, PA, 2014, www.astm.org
[13] ASTM D695-10, Standard Test Method for Compressive Properties of Rigid Plastics, ASTM International, West Conshohocken, PA, 2010, www.astm.org
[14] ASTM D790-10, Standard Test Methods for Flexural Properties of Unreinforced and Reinforced Plastics and Electrical Insulating Materials, ASTM International, West Conshohocken, PA, 2010, www.astm.org
[15] ASTM D6110-10, Standard Test Method for Determining the Charpy Impact Resistance of Notched Specimens of Plastics, ASTM International, West Conshohocken, PA, 2010, www.astm.org
