TRANSPORT PHENOMENA BASED MODELING OF COMMON DEFECT FORMATION IN METAL PRINTING
Open Access
- Author:
- Mukherjee, Tuhin
- Graduate Program:
- Materials Science and Engineering
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- October 04, 2019
- Committee Members:
- Tarasankar Debroy, Dissertation Advisor/Co-Advisor
Tarasankar Debroy, Committee Chair/Co-Chair
Todd Palmer, Committee Member
Zi-Kui Liu, Committee Member
Timothy W. Simpson, Outside Member
John C Mauro, Program Head/Chair - Keywords:
- Additive manufacturing
3D printing
Residual stresses
Defects
Lack of fusion
Heat and fluid flow - Abstract:
- Additive manufacturing (AM) or 3D printing of metals allows one step, near net shape fabrication of complex and intricate components that cannot be easily and economically produced by other means. Stainless steels, aluminum, titanium and nickel alloys are commonly printed using mainly directed energy deposition (DED) and powder bed fusion (PBF) techniques. However, printed metallic components often suffer from defects such as residual stresses, distortion, composition change due to selective vaporization of alloying elements and lack of fusion voids. These defects largely degrade the mechanical properties of the parts and in extreme cases lead to part rejection. For example, high residual stresses may result in warping, buckling and delamination of the parts and are detrimental to fatigue properties. Changes in composition can affect microstructure, corrosion resistance and mechanical properties of the components. Lack of fusion defects are known to adversely affect the tensile properties of the printed parts. Formation of these defects is affected by the transport of heat, mass and momentum, which include heat absorption by the feedstock material, formation of molten pool, convective flow of liquid metal inside the pool and cooling down by exchanging heat with the surroundings by convection and radiation. Therefore, fabrication of defect free and reliable AM parts requires a better understanding of the effects of heat, mass and momentum transfer on the formation of defects. The AM process involves rapid heating, melting, solidification and cooling of the part. As a result, different regions of the workpiece experience repeated heating and cooling. The spatially varying thermal cycles result in residual stresses and distortion in the AM parts. Key physical factors responsible for the origin of residual stresses and distortion include spatial temperature gradient, expansion and contraction of the part due to repeated thermal cycles and large coefficient of thermal expansion, solidification shrinkage of molten pool and temperature and strain rate dependent constitutive behavior of plastic material. At very high temperature, alloying elements may vaporize significantly depending on the equilibrium vapor pressure above the molten pool and the total pressure in the depositing chamber. All elements do not vaporize at the same rate because of the difference in vapor pressures of different elements. Such selective vaporization of alloying elements often results in a significant change in the composition of the part from that of the original feedstock. Composition change is affected by vaporization rates of different alloying elements, temperature distribution on the top surface area of the deposit and molten pool volume. Lack of fusion defects may result due to insufficient overlap between neighboring tracks of deposits and are affected by the shape and size of the fusion zone. Shape and size of fusion zone are often controlled by heat absorption by the feedstock, heat transfer through the substrate and surface energy on the top surface of the molten pool. For example, in wire arc based DED processes, fusion zone geometry is governed by arc pressure and vapor pressure on the top surface of the molten pool, droplet impingement and surface tension. Fusion zone geometry is also often influenced by convective flow of liquid metal primarily driven by the surface tension gradient on the top surface of the molten pool. The abovementioned variables related to heat, mass and momentum transfers are required to be estimated in order to predict defects. Therefore, in this research, threedimensional, transient, heat transfer and fluid flow models for both DED and PBF processes were developed and used. These models calculated 3D, transient temperature and velocity fields, fusion zone shape and size, which affect the defect formation. The models solve equations of conservation of mass, momentum and energy in a discretized solution domain consisting of substrate, deposits, feedstock materials and shielding gas. They also consider the effects of liquid metal convection and thus increased the accuracy in temperature field calculations. The computational efficiency for multi-layer, multi-hatch components is enhanced by implementing a novel traveling grid system. The models were rigorously tested using independent experimental data. Based on the calculated transient temperature field, residual stresses and distortion for multi-track components were predicted using a finite element based thermo-mechanical model. This model was used to calculate residual stresses and distortion for stainless steel 316, Inconel 718 and Ti-6Al-4V components fabricated using DED and PBF processes. It was shown that fabrication of a graded joint using laser DED between 2.25Cr-1Mo steel and alloy 800H significantly reduced the sharp change in residual stresses at the interface of the dissimilar joint between these two alloys. Changes in compositions were calculated using a vaporization model based on the top surface temperature and volume of the molten pool, which were calculated using the heat transfer fluid flow models. Lack of fusion defect was estimated based on the deposit and fusion zone geometries calculated using the heat transfer and fluid flow model. Easy to use lack of fusion index and dimensionless strain parameter were proposed for practical use in shop floors to predict lack of fusion defects and thermal distortion quickly. Apart from providing a better understanding of the evolution mechanism of defects based on heat transfer, fluid flow and mechanics of materials, printability or the ability to resist the defects for various alloy-AM process combinations are also evaluated. Quantitative scales are proposed to construct, test and validate the printability of stainless steels, nickel and titanium alloys for DED and PBF processes. It was shown that components printed using PBF are more susceptible to composition change and lack of fusion defects compared to those made by DED. However, fabrication of components with very thin layers made the PBF components less vulnerable to residual stresses and distortion. This thesis research work represents a contribution to the growing quantitative knowledge base in AM of metallic materials. Expansion of this knowledge base is necessary, if not essential, to fabricate defect free, structurally sound and reliable metallic components using AM. For example, printing of new alloys requires creation of printability database based on the knowledgebase for existing alloys employing the method of this research a critical step forward. In addition, future research is needed to find out the hierarchy of the causative variables affecting defect formation where data-driven machine learning approach could be beneficial.