SAND2025-12284R LDRD PROJECT NUMBER: 238667 LDRD PROJECT TITLE: Understanding Geometry and Microstructure Interactions in LPBF by Linking Lab Scale Studies to Real Part Case Examples PROJECT TEAM MEMBERS: Alexander Wilson-Heid, Benjamin White, Hannah Sims, Brandon Langdon, Jonathan Thompkins, and Keaton Collins ABSTRACT: The mechanical properties derived from simple tensile tests are uniquely important in the design and qualification of load bearing metallic components. The tensile properties in standards for wrought metallic materials are generally accessible for design and simulation engineers, but this is not the case for additively manufactured metals. This work investigates the influence of tensile specimen geometry on the mechanical properties of laser powder bed fusion (LPBF) additively manufactured Ti-6Al-4V, particularly in relation to properties measured from specimens excised from an exemplar part. A comprehensive analysis was conducted across various specimen geometries, considering factors such as thermal history, cross-sectional area, shape, and surface roughness. Key findings reveal that sample size, particularly cross-sectional area, significantly affects reported tensile properties, with ASTM E8 flat and round specimens exhibiting differences in elastic modulus (13%), yield strength (17%), ultimate tensile strength (UTS) (14%), and elongation to failure (11%). Additionally, surface roughness was found to have a limited impact on ASTM round geometries, but it becomes critical in thin-walled and small high-throughput samples. The ASTM flat samples closely matched the properties of samples excised from the exemplar part, underscoring the necessity of testing witness samples that accurately represent end- use conditions. Furthermore, mechanical properties of samples built within the stitch zone of the exemplar part demonstrated reduced ductility, highlighting the importance of laser alignment and qualification in multi-laser additive manufacturing processes. Overall, this study emphasizes the need for careful consideration of specimen geometry and testing conditions to ensure accurate reporting of mechanical properties in LPBF Ti-6Al-4V. INTRODUCTION AND EXECUTIVE SUMMARY OF RESULTS: Additive manufacturing (AM) enables the design and production of geometrically complex components that can accelerate production timelines and enable superior component performance simultaneously. However, metal AM is a thermally complex manufacturing process where both material is being altered, melting powder to solid material, and the component is being built, layer- by-layer. This intertwined relationship results in microstructural variation in as-built parts, making it difficult and important to evaluate mechanical properties that are representative of the final components being built for applications with mechanical loading. Typically, in the AM workflow, mechanical test “witness” coupons are built at the same time as the relevant parts and then are test independently of the part and are taken to represent how all the material in the part will perform [1]. Choosing the geometry, surface finish, build location, gauge height, etc. are all important considerations when designing “witness” coupons to best represent the part performance. Sandia National Laboratories is a multimission laboratory managed and operated by National Technology and Engineering Solutions of Sandia, LLC, a wholly owned subsidiary of Honeywell International, Inc., for the U.S. Department of Energy’s National Nuclear Security Administration under contract DE-NA-0003525. Additionally, the thermal history of a part can be very different from the “witness” coupon due to thermal mass differences, cross-sectional area, or geometric complexity. Oftentimes, heat treatments are done to normalize these effects [2]. The disconnect between property evaluation and part performance is a significant gap in the engineering workflow of AM. Mechanical properties are used to simulate part performance, via finite element analysis (FEA), to ensure reliability in applications and properties are also used to set factors of safety in component design. Therefore, it is critical to appropriately determine properties that will best represent that component of interest. One of the most used tests for determining mechanical properties of materials used in the AM process are static, tensile tests [3], [4], [5] . Tensile tests are relatively simple to conduct using standard geometries, the results are commonly reported in material standards and are used to measure properties such as elastic modulus, yield strength, ultimate tensile strength, and elongation to failure, as shown in Figure 1. Although tensile tests do provide important information, there are additional tests that provide even more depth about material mechanical performance, such as fatigue [6], compression [7], high strain rate [8], and multiaxial testing [9], however these tests are outside the scope of the current investigation. Figure 1. A generic engineering stress versus engineering strain plot from a quasi-static tensile test. Important values that are commonly reported from a stress v. strain curve are identified. The elastic modulus is the slope of the curve within the elastic region. Elongation to failure in this instance is defined as the strain at a stress that is 10% lower than the UTS. The material of interest in the current work is Ti–6Al–4V, which is an α+β titanium alloy that exhibits both high strength and relatively good ductility because of its dual phase nature [10]. The alloy is commonly used in commercial aviation, aerospace, and the medical industries because of the combination of mechanical performance and low density. At Sandia National Laboratories, Ti- 6Al-4V is used in multiple laser powder bed fusion (LPBF) AM machines across both sites. The AM process parameters and heat treatment were co-developed with our production agency partners at KCNSC for one of the mutual LPBF machines. Building on the previous knowledge gained from the partnership with KCNSC, in this work we aim to focus on the geometric effects of samples on reported mechanical properties. The current study has evaluated different “witness” tensile test coupon geometries and related those properties to samples that are excised directly from a large, thin-walled exemplar part manufactured via LPBF AM with Ti-6Al-4V grade 23 alloy powder. It has been found that sample size is an important factor when determining the properties that best represent the properties directly measured from an exemplar part. The properties measured from the thin-walled exemplar part were best represented by testing “witness” coupons that were similarly thin-walled as compared to round samples with larger cross-sectional geometries. Additionally, the smaller geometry samples, both flat ASTM E8 and non-standard high-throughput geometries develop at SNL [4], [11], are more impacted by surface roughness effects, which resulted in lower reported yield and ultimate tensile strength compared to the larger cylindrical ASTM E8 samples. Finally, the larger exemplar part was built in an LPBF AM machine that uses two lasers to increase the rate of production, this study found a clear drop in mechanical properties for samples that were built in regions were the working area of the lasers overlapped, commonly referred to as the stitch zone, and further exploration should be completed to better characterize the effects of the stitch zone on part level performance. DETAILED DESCRIPTION OF RESEARCH AND DEVELOPMENT AND METHODOLOGY: The samples evaluated in this study were manufactured on the Nikon SLM Solution SLM280 Production Series V3 LPBF AM machine located in building 979 on the SNL-CA site. Spherical Ti-6Al-4V grade 23 powder was from the supplier AP&C with a particle size distribution of 20- 63 µm. The tensile samples, listed in Table 1 and shown in Figure 2, were built concurrently on a single Ti-6Al-4V build plate, as shown in Figure 3. The parameters for all sample on the build used a layer height of 60 µm, a laser power of 350 W, and a laser speed of 1000 mm/s. After the build was completed, the ASTM flat tensile samples with the tab (Figure 2b) were removed from the build plate, twelve of the samples were kept in the as-built condition. The rest of the build plate with all other samples and the exemplar part attached, in addition to the remaining 12 ASTM flat with tab samples, were heat treated in a vacuum furnace at 800 °C for 2 h. The ramp rate was 6.25 °C/min., the build was furnace cooled and removed at 100 °C. All samples were wire electrical discharge machined (wEDM) from the build plate after the heat treatment. All sample were test with as-built surface roughness, except for the edges of the samples excised from the exemplar part. Table 1. Sample types and quantities manufactured. Specimen Type Quantity Built Quantity test Excised ASTM E8 – Flat Subsize (1.2 mm thick x 6 mm 37 23 wide) ASTM E8 – Flat Subsize (1.2 mm thick x 6 mm wide) 10 5 ASTM E8 – Flat Subsize (1.2 mm thick x 6 mm wide) - tab 12 5 As-built: 12 As-built: 5 ASTM E8M – Round (6 mm diameter) 5 5 High Throughput W = 0.6 mm 24 0 High Throughput W = 1.2 mm 12 7 High Throughput W = 2.5 mm 12 10 High Throughput W = 6.0 mm 12 0 a b c d Figure 2. Tensile test geometries used to evaluate mechanical properties of LPBF Ti-6Al-4V. (a) ASTM E8 – 24 regular tension test subsize specimen with a thickness of 1.2 mm, (b) modified version of (a) to include a 10 mm tall tab at the bottom to constrict thermal path and make for easy part removal in the as-built condition, (c) ASTM E8 – 24 standard round tension specimen 3 with diameter of 6 mm, and (d) high-throughput drop in grip tensile specimen where specimens with W = 0.6, 1.2, 2.5, and 6.0 were manufactured and only W = 1.2 and 2.5 were test. Figure 3. (a, b) CAD final build layout of the samples and large, thin-walled exemplar part used for excising Figure 2a sample from the walls. The red region is the “stitch zone” where both lasers are used to build the parts that overlap both boundaries of the red region, only the exemplar part meets that condition (c) As-built samples after manufacturing, the Figure 2b samples with the tab are removed in this image. The exemplar part was modeled to represent a large, thin-walled housing with complex wall angles. This type of part would require a large blank of wrought material, creating a large amount of waste material with subtractive machining. The part could also be traditionally made in multiple components, but that would compromise performance and add to the part count number in a large assembly while adding to assembly time. This type of housing is an ideal part to use AM for except for the fact that thin walls are susceptible to distortion during the AM process. Mitigation techniques, such adding corrugation to stiffen the walls is an option for real-part use cases but was not explored here. For this study, flat walls, accepting that they would be distorted, were built with the end goal of cutting samples out of the walls. Tensile specimens with the same dimensions as the ASTM flat samples (Figure 2a), were cut in the vertical orientation from each of the walls, as shown in Figure 4, using wEDM. A total of 37 samples were cut from the exemplar part. Figure 4. Process of excising tensile specimens from the exemplar part. The digital cut plan in this figure shows horizontal specimens that were not used in the study, instead all samples were cut in the vertical orientation. Mechanical testing of the ASTM flat samples, with and without the tab, were completed on an MTS 810 load frame with a 100 kN load cell. Strain measurements were made with both a non- contact laser-based strain gauge and a clip-on strain gauge, both were 25.4 mm long. Samples were loaded at a strain rate of 0.001 s-1. Data was collected at a rate of 100 Hz. The ASTM round samples were test on an MTS 370 load frame, with everything else being the same as the flat samples. The high-throughput samples were subjected to tensile testing at a strain rate of 0.001 mm/s until failure. This testing was performed using a 3A Stephens load frame (model number: 204m, serial number: 783), equipped with both a 67 kN load cell (model number: SWP20-25k-B000, serial number: 126430A) and a 22 kN load cell (model number: 1210AF-5K-B, serial number: 1174144A). High-throughput grips were employed to securely hold the samples during testing. To facilitate digital image correlation (DIC), for strain measurements, the specimens were speckled along the gauge length surface. The front face of each sample was imaged using a Point Gray camera with a resolution of 4.1 million pixels, capturing images at a rate of one per second. For the W=1.2 mm samples the gauge length was approximately 4.5 mm and for the W=2.5mm samples the gauge length was 10.2 mm. RESULTS AND DISCUSSION: Effect of Sample Size using the High Throughput Geometry Two of the four high-throughput sample geometries that were manufactured were evaluated to determine the effect of size on mechanical properties. The W=1.2 mm sample and the W=2.5 mm samples exhibited different tensile behavior, as shown in Figure 5. The yield strength and ultimate tensile strength were 9% and 8% different, respectively. The largest difference was the value of the elastic modulus, which was 21% different between the two sample geometries. The elongation to failure was only 2% different between the samples, 13.8% and 14.1% average strain to failure for the 1.2 mm and 2.5 mm samples, respectively. To better evaluate the role of the geometry on the properties, the stress data was calculated using the designed cross-sectional area for both sizes, referred to as the effective properties. Measuring linear dimensions using mechanical tools such as calipers and micrometers always overestimates the dimension because these tools measure from the peak of the surface roughness on one side to the peak of the surface roughness on the other. The nominal CAD dimensions do not suffer this problem, but also do not account for potential over/under print, leading to potential confounding with the printer dimensional tolerance. The CAD dimensions however better connect the design intent with the reality of the print, making it a useful metric for design purposes. When comparing effective properties, the yield and ultimate tensile strengths begin to converge, now only 3% and 2% differences, respectively. However, the average elastic modulus difference remains relatively large at 15% difference, 103.2 GPa for the 1.2 mm sample and 120.1 GPa for the 2.5 mm samples, using the effective properties. The large difference in the modulus indicates the geometry has an effect on the intrinsic properties of the material, while the strength differences better indicate the effect of surface roughness or reduction of load bearing area. Figure 5. Engineering stress vs. engineering strain data for the high-throughput 1.2 mm and 2.5 mm samples exhibiting clear variation in both yield and ultimate tensile strength. Effective properties, begin to converge, but still convey a size-effect indicating that geometry is driving the material properties beyond microstructural influences. Surface roughness can have a significant impact on the evaluate properties of a sample, due to the stress concentrations that the rough features can cause resulting in reduction of strength and lower ductility. The effect of the surface roughness varies depending on the overall sample geometry as shown in Figure 6, where smaller samples will be more effected by the surface roughness, although a large enough roughness can begin to effect larger samples if the quotient of the surface roughness and cross-sectional area surpasses 0.001 µm-1. Sample thickness (mm) Fatigue samples ASTM E8flat samples 6.25 mm size effect UT 2.5 mm size effect UT 1.2 mm size effect UT 0.6 mm size effect UT Samples in this stud y Surface Roughness / Cross -sectional area (µm-1) Figure 6. Evaluation of the surface roughness as a function of sample thickness and cross-sectional area of various sample geometries compared to samples in this study. A rough threshold of 0.001 µm-1 was found to indicate that the surface roughness would have an impact on material properties. Effect of Heat Treatment and Thermal Isolation/Build Height The 800 °C, 2-hour heat treatment, as expected, has a significant impact on the material properties of the LPBF Ti-6Al-4V material, as shown in Figure 7. The yield and ultimate tensile strengths of the as-built samples are over 100 MPa greater than the heat-treated samples. With the increased strength is a trade off in elongation to failure, where the heat treated with tab samples on average didn’t fail until 13.2 ± 0.2 %, while the samples heat-treated without the tab failed at 10.1 ± 2.0 %, and the as-built sample with a tab failed at 5.5 ±0.2 %. The reduction in ductility between as-built and heat-treated samples was an expected result, however the difference in the spread of the two heat-treated geometries was an unexpected result. The tab at the bottom of the samples was partially intended to constrict the thermal pathway for cooling the sample during manufacturing thus retaining more heat in the sample and converting the brittle α phase to the ductile β phase in situ. A much more thorough examination of the samples, that was not done in the current iteration of the study, would be required to verify this effect. Other potential causes for the variation in ductility would be build plate effect, potentially stochastic porosity was more prevalent in the regions where the samples without a tab were built and finally, the 10 mm height difference may be enough to change the thermal history of the two samples by changing the laser on time in the layers where the gauge section was being built. In our previous work with KCNSC in developing Ti-6Al-4V parameters, we observed an effect of an overcrowded build plate that resulted in worse mechanical properties in samples when a significant portion of the build area was being melted. Further investigation is required to rule for or against each of these potential causes, but it is a noteworthy finding. 1100 1043.1 a c As-Built with tab 1000 932.3 Stress(MPa) 900 865.8 874.5 Heat treated with tab 805.0 809.3 800 Heat treated on build plate 700 600 Yield Strength (MPa) UTS(MPa) 15 13.2 b (%) 120 Engineeting Strain 100.2 101.4 10.1 d (GPa) 10 100 86.4 Elastic Modulus 5.5 80 5 60 40 0 Elongation To Failure (%) 20 0 Youngs Modulus (GPa) Figure 7. (a) Engineering stress vs strain for the three ASTM E8 flat specimen variations. (b) Elongation to failure is significantly lower in the as-built samples, while the (c) strength is much higher. (d) Indicates that the elastic modulus is improved with the heat treatment an important consideration in applications where yield strength will not be approached. Effect of Specimen Shape and Surface Roughness Although the ASTM E8 flat and round tension specimens were generated from the same standard their resulting properties were significantly different for the heat-treated LPBF Ti-6Al-4V with as- built surface roughness, as shown in Figure 8. The elastic modulus is 13% different between the sample geometries, while the yield strength is 17% different, the UTS is 14% different, and the elongation to failure is 11% different. This across the board >10% difference in properties is indicative of both intrinsic material change as a function of the geometry and again the role of surface roughness on samples with smaller cross-sectional areas, as shown in Figure 6. The ASTM E8 flat and high-throughput W=1.2 mm samples have close agreement in properties, as shown in Figure 8b, showcasing the high-throughput geometries ability to replicate the properties derived via an ASTM standard geometry. Notably, the W=1.2 effective properties, and although not shown the W=2.5 mm sample, are similar to the round ASTM E8 geometry. This highlights that effective properties can map to sample geometries that are minimally affected by surface roughness. Figure 8. (a) Engineering stress vs strain curves for the ASTM E8 flat with tab and ASTM E8 round specimens, where the round specimens were significantly stronger than the flat geometry and (b) comparison of the data in (a) to the 1.2 mm high-throughput geometry as measured and effective properties. The W=1.2 mm sample in the as-measured was similar to the ASTM 38 flat geometry and the effective properties were similarly close to the round geometry. To better understand the effects of surface roughness for the round ASTM E8 geometry, the data in the current study was compared to data from Ti-6Al-4V samples built on the same AM machine with the same heat treatment from the parameter development work with KCNSC. The samples from the previous study were machined from cylinders that were directly built. Figure 9 reveals that LPBF Ti-6Al-4V in the round ASTM E8 specimen geometry is only minimally different in the as-built and machined surface roughness conditions. 1200 1100 1059.7 (MPa) 1007.7 Engineering Stress 1000 980.9 959.2 900 800 700 600 Yield Strength (MPa) UTS(MPa) Current study: ASTM E8 Specimen 3 with as -built surface roughness (n=5) Previous study : Cylinders machined to ASTM E8 Specimen 3 (n=15) Figure 9. Comparison of ASTM round samples in the as-built surface roughness condition compared to the same geometry machined from a cylinder. Minimal effect of the surface roughness for the round geometry samples with a large cross-sectional area. Tensile Results from Excised Specimens The exemplar part was designed and manufactured with 1.2 mm wall thickness, identical to the thickness of the ASTM E8 flat and W=1.2 mm high-throughput geometry. The ASTM E8 geometry was excised from each of the eight unique walls in the part, as shown in Figure 10. Samples from the walls were relatively consistent for walsl that were not built across the stitch zone on the AM machine. The average yield strength and UTS from specimens excised from walls I-VI compared to the ASTM E8 flat samples with a tab was 854 MPa vs 809 MPa and 916 MPa vs. 875 MPa, respectively. The elastic modulus for the two sample sets were 104 GPa (excised) and 101 GPa, while the elongation to failure for the non-excised samples was 13.2% compared to 10.3% for excised samples. The ASTM E8 flat specimen with the tab best mapped to the excised parts although they were not exactly the same. The increased strength in the excised samples could be from a variety of factors including that the heat treatment for a large part needs to be for a longer time to alter and homogenize the microstructure to the more ductile β phase, additionally the exemplar parts walls were geometrically distorted and may have retained the residual stress from the manufacturing process. Figure 10. (a) Top-down schematic of the exemplar part labeling each wall, where the n= refers the number of samples that were cut out of each wall. All samples were test from walls (b) I-III, only three samples were test from walls (c) IV-VI, and four samples were test from walls (d) VII and IX. Properties from walls I-VI were very consistent and had slightly higher strength and lower ductility than the benchmark ASTM E8 flat sample. Scatter in the data appeared for walls VII and IX due, which is primarily attributed the stitch zone. Table 2. Summary of all properties from the current study. Youngs Yield Elongation UTS Sample Geometry and Condition Modulus Strength To Failure (MPa) (GPa) (MPa) (%) W=1.2 mm high Average 89.8±9.3 778.2±10.8 850.4±12.2 13.8±1.0 throughput (n=7) W=1.2 mm effective Average 103.2±9.6 894.0±5.3 976.0±2.4 13.8±1.0 (n=7) 2.5 mm high throughput Average 111.0±5.8 854.5±16.9 920.4±19.1 14.1±0.7 (n=10) W=2.5 mm effective Average 120.1±6.3 924.5±3.2 995.6±1.4 14.1±0.7 (n=10) ASTM E8 flat with tab Average 86.4±0.8 932.3±7.8 1043.1±8.1 5.5±0.2 (as-built) (n=5) ASTM E8 flat Average 100.2±0.3 805.0±4.9 865.8±6.8 10.1±2.0 (n=5) ASTM E8 flat with tab Average 101.4±1.0 809.3±11.7 874.5±7.3 13.2±0.6 (n=5) ASTM E8 flat - excised Average 102.4±5.8 845.0±27.1 918.9±7.6 10.4±0.9 from exemplar * (n=20) ASTM E8 round with as- Average 115.6±1.9 959.2±3.9 1007.7±3.7 14.7±0.3 built surface (n=5) ASTM E8 round with Average 115.9±4.0 980.9±5.3 1059.7±2.0 15.6±1.2 machined surface (n=15) (Previous work) *Ignores stitch zone samples to calculate average and standard deviation. A notable result from the evaluation of excised samples was the effect of the stitch zone on material properties, namely elongation to failure, as shown in Figure 11. For both walls, the samples that were cut right at the interface between the two lasers was picked to be test – these samples all had significantly lower ductility due to the stitch zone defects. This finding is an area that is ripe for further exploration as there is more desire for large scale LPBF parts that are built on machines with more than two lasers. Figure 11. Highlighted effects of the stitch zone in walls VII and IX, which led to reduced elongation to failure from samples directly cut at the stitch. ANTICIPATED OUTCOMES AND IMPACTS: The study on the effects of tensile specimen geometry on the mechanical properties of LPBF Ti- 6Al-4V provides significant insights into the complexities of AM and the critical importance of accurately measuring mechanical properties. As AM technologies continue to evolve and gain traction in the NNSA, understanding the nuances of how specimen geometry influences material behavior becomes essential for ensuring the reliability and performance of AM components. One of the primary impacts of this study is its contribution to the ongoing discourse regarding mechanical testing size effects in AM. Traditional mechanical testing methodologies often rely on standardized specimen geometries, such as those defined by ASTM E8. However, the unique characteristics of AM processes, including thermal gradients, layer-by-layer construction, and inherent microstructural variations, necessitate ability to test alternative geometries. By systematically investigating a range of specimen geometries, this study highlights the need for tailored testing approaches that account for the specific conditions under which AM parts are produced. The findings of the study reveal that sample size plays a crucial role in determining the reported tensile properties of LPBF Ti-6Al-4V. The observed differences in between ASTM E8 flat and round specimens underscore the sensitivity of mechanical properties to geometric variations. This insight is particularly relevant for engineers and researchers involved in the design and qualification of AM components, as it emphasizes the necessity of selecting appropriate specimen geometries that accurately reflect the intended application and service conditions. The current study sheds light on the influence of surface roughness on mechanical properties. While surface roughness was found to have a limited effect on ASTM round geometries, it became a critical factor in thin-walled and small high-throughput samples. This finding is significant, as it suggests that surface roughness can markedly affect the performance of AM components, particularly in applications where mechanical integrity is paramount. Engineers must therefore consider surface roughness as a design parameter, ensuring that the manufacturing process is optimized to achieve the desired surface characteristics without compromising mechanical performance. The alignment of ASTM flat samples with the properties of flat samples excised from the exemplar part further emphasizes the importance of representative testing. The study indicates that testing witness samples that closely mimic the end-use conditions can lead to more accurate assessments of mechanical properties. By advocating for the use of representative samples, the study encourages a shift towards more realistic testing methodologies that can better predict the behavior of AM parts in real-world applications. Another noteworthy outcome of the study is the observation regarding the mechanical properties of samples built within the stitch zone of the AM part. The reduced ductility observed in these samples highlights the potential challenges associated with laser alignment and qualification in production AM machines. As AM technology matures, ensuring consistent and high-quality production processes becomes increasingly important. The findings of this study serve as a reminder that even minor variations in the manufacturing process can lead to significant differences in material performance. This underscores the need for rigorous quality control measures and process monitoring to mitigate the risks associated with stitch zones and other critical areas in AM components. The evaluation of the stitch zones in multi-laser LPBF systems should be further studied to ensure reliable part performance. In addition to its technical contributions, the study also has broader implications for the AM community. As the demand for AM parts continues to grow, there is an increasing need for standardized testing protocols that can accommodate the unique characteristics of AM materials. The insights gained from this research can inform the development of new testing standards that better reflect the realities of AM processes, ultimately leading to more reliable and consistent material property data. This, in turn, can facilitate the adoption of AM technologies across various interest groups within the NNSA and DOE, as stakeholders gain confidence in the performance and reliability of AM components. In conclusion, the study on the effects of tensile specimen geometry on the mechanical properties of LPBF Ti-6Al-4V has far-reaching implications for the field of additive manufacturing. By elucidating the impact of specimen geometry, surface roughness, and manufacturing conditions on mechanical properties, the research provides valuable insights that can inform the design, testing, and qualification of AM components. As the AM landscape continues to evolve, the findings of this study will play a crucial role in shaping future research and development efforts, ultimately contributing to the advancement of reliable and high-performance AM technologies. CONCLUSION: This study explored the effects of tensile specimen geometry on reported mechanical properties for LPBF Ti-6Al-4V and how those properties relate to the properties measured from specimens excised from an exemplar part. By testing a wide range of specimen geometries with varying thermal history, cross-sectional areas, cross-sectional shape, and surface roughness conditions a complete picture of the factors influencing reported properties was revealed. The primary findings for this study include: ▪ The sample size is an important factor when reporting tensile static material properties for LPBF Ti-6Al-4V that was manufactured in this study. ASTM E8 flat and round specimens resulted an elastic modulus that was 13% different between the sample geometries, while the yield strength was 17% different, the UTS was 14% different, and the elongation to failure was 11% different. ▪ Surface roughness has a limited effect in ASTM round geometries, whereas in thin- walled samples and small-high throughput samples surface roughness is very important to consider. ▪ The ASTM flat samples were most aligned with the flat samples excised from the exemplar part, indicating the importance of testing witness samples that are representative of the end use case. ▪ The mechanical properties for sample built within the stitch zone of AM part have noticeably reduced ductility, and signify the importance of laser alignment and qualification for production AM machines. REFERENCES: [1] R. Shrestha, N. Shamsaei, M. Seifi, and N. Phan, “An investigation into specimen property to part performance relationships for laser beam powder bed fusion additive manufacturing,” Addit. Manuf., vol. 29, p. 100807, Oct. 2019, doi: 10.1016/j.addma.2019.100807. [2] Y. Lakroune, D. Connétable, J. Hugues, P. Hermantier, P. Barriobero-Vila, and M. Dehmas, “Microstructural evolution during post heat treatment of the Ti–6Al–4V alloy manufactured by laser powder bed fusion,” J. Mater. Res. Technol., vol. 23, pp. 1980–1994, Mar. 2023, doi: 10.1016/j.jmrt.2023.01.123. [3] E28 Committee, ASTM E8/E8M: Test Methods for Tension Testing of Metallic Materials. doi: 10.1520/E0008_E0008M-24. [4] A. M. Roach, B. C. White, A. Garland, B. H. Jared, J. D. Carroll, and B. L. Boyce, “Size- dependent stochastic tensile properties in additively manufactured 316L stainless steel,” Addit. Manuf., vol. 32, p. 101090, Mar. 2020, doi: 10.1016/j.addma.2020.101090. [5] D. June, J. R. Mayeur, P. Gradl, A. Wessman, and K. Hazeli, “Effects of size, geometry, and testing temperature on additively manufactured Ti-6Al-4V titanium alloy,” Addit. Manuf., vol. 80, p. 103970, Jan. 2024, doi: 10.1016/j.addma.2024.103970. [6] P. Li, D. H. Warner, A. Fatemi, and N. Phan, “Critical assessment of the fatigue performance of additively manufactured Ti–6Al–4V and perspective for future research,” Int. J. Fatigue, vol. 85, pp. 130–143, Apr. 2016, doi: 10.1016/j.ijfatigue.2015.12.003. [7] Z. Li, M. Chuzenji, and M. Mizutani, “Compression and fatigue performance of Ti6Al4V materials with different uniform and gradient porous structures,” Mater. Sci. Eng. A, vol. 873, p. 145030, May 2023, doi: 10.1016/j.msea.2023.145030. [8] S.-D. Salehi, R. Beal, and O. T. Kingstedt, “Dynamic behavior and thermomechanical characterization of laser powder bed fusion and wrought Ti–6Al–4V,” Int. J. Impact Eng., vol. 176, p. 104552, Apr. 2023, doi: 10.1016/j.ijimpeng.2023.104552. [9] A. E. Wilson-Heid, S. Qin, and A. M. Beese, “Anisotropic multiaxial plasticity model for laser powder bed fusion additively manufactured Ti-6Al-4V,” Mater. Sci. Eng. A, vol. 738, pp. 90–97, Dec. 2018, doi: 10.1016/j.msea.2018.09.077. [10] G. Lütjering and J. C. Williams, “Alpha + Beta Alloys,” in Titanium, G. Lütjering and J. C. Williams, Eds., Berlin, Heidelberg: Springer, 2003, pp. 177–232. doi: 10.1007/978-3-540- 71398-2_5. [11] B. C. Salzbrenner et al., “High-throughput stochastic tensile performance of additively manufactured stainless steel,” J. Mater. Process. Technol., vol. 241, pp. 1–12, Mar. 2017, doi: 10.1016/j.jmatprotec.2016.10.023.