Tailoring Thermal Expansion in Additively Manufactured Titanium Alloys to Enable Functional Grading
Open Access
- Author:
- Hilburn, Skyler
- Graduate Program:
- Materials Science and Engineering
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- July 07, 2020
- Committee Members:
- Todd Palmer, Dissertation Advisor/Co-Advisor
Todd Palmer, Committee Chair/Co-Chair
Allison Michelle Beese, Committee Member
Edward William Reutzel, Outside Member
Timothy W. Simpson, Committee Chair/Co-Chair
Timothy W. Simpson, Dissertation Advisor/Co-Advisor
John C Mauro, Program Head/Chair - Keywords:
- Additive Manufacturing
Thermal Expansion
Coefficient of Thermal Expansion
Functional Graded Materials
Functionally Graded Materials
Titanium - Abstract:
- Dissimilar material combinations are common in complex engineered systems. While these material combinations can contribute to improved performance, they can also lead to a range of interfacial failures, such as those caused by differences in their thermal expansion. Non-uniform stresses generated with changes in temperature can be large enough to lead to distortion and even failure. Given the ability of additive manufacturing (AM) to create novel materials, these interfaces can be replaced by a functionally graded material (FGM) designed to minimize the coefficient of thermal expansion (CTE) mismatch between dissimilar materials and produce more robust and higher performing engineered structure. The design of FGMs using AM processes presents a unique set of challenges, primarily driven by the mixing and solidification of dissimilar materials which have different levels of compatibility. Typical failure mechanisms which must be addressed through design and AM processing include solidification cracking, chemical incompatibilities, such as immiscibility of alloying elements, and CTE mismatch. While most of the failure mechanisms have some mitigation strategies, the current techniques for mitigating the CTE mismatch are largely undeveloped. One potential route, however, can involve the controlled manipulation of CTE through changes in alloying element composition. Since thermal expansion, particularly in hexagonally close packed systems, is anisotropic, wrought commercially pure (CP) titanium and AM CP titanium processed through laser based directed energy deposition were as investigated. While the processing route can alter grain orientation or provide a grain texture, the significant difference in observed CTE was correlated with sample orientation, with the vertically oriented samples displaying lower CTEs for both processing routes. The observed sample orientation CTE differences were determined to be a result of crystallographic texture. Copper and silver were chosen as candidate alloying elements to increase the CTE of titanium beyond that of the wrought material. At 15 at.% copper alloy content, the CTE of commercially pure titanium was increased from 9.45 ppm/oC to 10.5 ppm/oC. The measured CTE values exceeded those predicted by an elemental rule of mixtures (ROME) procedure, which predicted a CTE of 11.9 ppm/oC, which is 12% higher than the 10.6 ppm/oC experimentally measured value. This discrepancy can be tied, in part, to the traditional rule of mixtures (ROM) approach applying to the phases present and not simply alloying element fraction. Additionally, the rapid solidification experienced in AM generated microstructures that were different than what was predicted by the equilibrium phase diagram, further complicating the prediction. Hot isostatic pressing (HIP) was used to reduce build porosity and to homogenize the microstructure, driving the phase fractions closer to equilibrium. HIP reduced the CTE of the 15 at.% copper alloy content from 10.5 ppm/oC to 10.0 ppm/oC. The ability to tailor the CTE value depends on obtaining accurate values for each of the intermetallics or phases in the microstructure. Density Functional Theory (DFT) was utilized to estimate the unknown intermetallic phase CTEs. For the CuTi2 intermetallic, DFT overpredicted the CTE by 7.5%. While this is a reasonable result, to design a final gradient using 0.5 ppm/oC steps, experiments are required. Knowing the CTE of CuTi2, a traditional phased-based ROM was utilized and accurately predicted the CTE within ~1% of the HIPed AM material. Using the phase-based ROM and the equilibrium phase fractions, a route to design functional materials that grade the CTE was demonstrated using a designed gradient from the experimentally measured value of 9.45 ppm/oC of pure titanium to the DFT predicted CuTi intermetallic with a CTE of 11.6 ppm/oC in 5 layers stepping in 0.5 ppm/oC steps. A linear gradient in CTE requires a nonlinear change in composition based upon the intermetallic CTEs and the phase fractions.