Additive Manufacturing Fatigue and Fracture: Developing Predictive Capabilities: Joint Session with Fatigue in Materials Symposium - Microstructure-based Fatigue Studies on Additive-Manufactured Materials
Sponsored by: TMS Structural Materials Division, TMS: Additive Manufacturing Committee, TMS: Mechanical Behavior of Materials Committee
Program Organizers: Nik Hrabe, National Institute of Standards and Technology; John Lewandowski, Case Western Reserve University; Nima Shamsaei, Auburn University; Mohsen Seifi, ASTM International/Case Western Reserve University; Steve Daniewicz, University of Alabama

Wednesday 2:00 PM
March 2, 2022
Room: 258B
Location: Anaheim Convention Center

Session Chair: Orion Kafka, National Institute of Standards and Technology (NIST); Nik Hrabe, National Institute of Standards and Technology (NIST)

2:00 PM  Invited
Designing Printable Alloys for Fatigue Strength: Gregory Olson1; Jiadong Gong; Gary Whelan; 1Massachusetts Institute of Technology
    Sixty years of academic collaboration and thirty years of commercialization by a network of small businesses have delivered a mature technology of computational materials design and accelerated qualification grounded in a system of fundamental databases now known as the Materials Genome. A major focus of current application is the rapid development of the new alloys enabling the much-desired technology of additive manufacturing, with adaptation of the AIM methodology to accelerate qualification of printed components. Examples of printable gear steels and precipitation-strengthened shape memory alloys exploit micromechanical damage parameter simulations to design specifically for improved fatigue strength distributions.

2:30 PM  
Micromechanical Modeling of Porosity Defects in Additively Manufactured Alloys: Krzysztof Stopka1; Michael Sangid1; 1Purdue University
    Additively manufactured (AM) components can offer substantial manufacturing cost and time savings. However, much work remains to develop industry standards for the rapid qualification of AM components for use in fatigue-critical applications. Internal defects are of high concern because they cannot be remedied as easily as other imperfections (e.g., surface roughness by grinding or polishing, residual stresses by heat treatment, etc.). Structural defects produced by laser powder bed fusion (LPBF) may include lack of fusion, gas porosity, and linear pore alignment. In this work, the stress and strain fields around these key defects are investigated using micromechanical crystal plasticity simulations of Ni-base superalloy IN718 in the low cycle fatigue (LCF) regime. The results will aid in the development of a standard for the rapid qualification of AM components alongside future experimental studies and reduced-order modeling.

2:50 PM  
Fatigue Crack Initiation Behavior of Laser Beam Powder Bed Fused IN718: Mohammad Dodaran1; Muztahid Muhammad1; Shuai Shao1; Nima Shamsaei1; 1Auburn University
    The fatigue performance of certain additively manufactured alloys, such as IN718, is known to be less sensitive to the presence of defects. Fatigue failure in IN718 specimens is seen to be caused by cracks initiated from both persistent slip bands (PSB) and volumetric defects. This work is interested in laser beam powder bed fused (LB-PBF) IN718 and sheds light on the competition between the PSB driven and defect driven fatigue crack initiation using crystal plasticity (CP) simulations validated by experiments. It was found that shear localization, a precursor for fatigue crack initiation, is governed by local cyclic strain softening and is sensitive to both micro- and defect-structures. Longer crystallographic slip distances and larger defects can both promote the localization. The LB-PBF induced, large columnar grains can have diameters beyond hundreds of microns and are favorable for PSB governed fatigue crack initiation.

3:10 PM  
Experiments to Enable Expert-informed Machine Learning of Fatigue Performance of DMLM Ti-6Al-4V: Samuel Present1; Laura Dial2; Thomas Straub3; Chris Eberl3; Kevin Hemker1; 1Johns Hopkins University; 2General Electric Research Laboratory; 3Fraunhofer Institute for Mechanics of Materials IWM
    Understanding and predicting fatigue performance is paramount for aerospace applications, and rapid qualification and certification of metal alloys for use in cyclic loading environments is necessary for widespread adoption of additively manufactured components. Fatigue studies of additively manufactured metals and alloys have elucidated the fact that surface roughness and microstructural features can profoundly affect fatigue life. In the current study, resonate high-cycle micro-bending fatigue experiments were employed to identify the number of cycles to, and specific location for, crack nucleation in direct metal laser melted (DMLM) Ti-6Al-4V samples. Cross-correlation with nanoCT scans and EBSD maps, of surface roughness and the underlying microstructure, facilitated identification of critical nucleation sites. These experimental results are being used to underpin finite element simulations and to create training sets for expert-informed machine learning protocols, to enable rapid simulation of thin-wall fatigue performance.

3:30 PM Break

3:50 PM  Invited
Microstructure-based Fatigue Studies on Additive-manufactured Materials: Jiadong Gong1; Gary Whelan1; Abhinav Saboo1; Greg Olson1; 1Questek Innovations LLC
    Additive manufacturing (AM) offers an exciting new direction enabling the production of otherwise impossible to achieve components. A great deal of progress has been made in the development of AM materials using integrated computational materials engineering (ICME). However, fatigue remains a major hurdle for qualification of new AM materials. In the present work, ICME approaches for microstructure-sensitive fatigue modeling for AM Ti64 are presented. The ICME-based fatigue model was employed to consider structure-property linkages, accounting for such structural attributes as the microstructure of the printed material, including phase fractions, grain size, and most importantly, crystallographic texture, as well as internal defects such as subsurface porosity, and external defects such as surface roughness. The focus of this work is high-cycle fatigue. The fatigue model is applied to predict worst-case fatigue life using extreme value distributions of fatigue indicator parameters as a surrogate measure of fatigue crack driving forces near subsurface defects.

4:20 PM  
Comparison of Statistical Predictors of Additive Manufacturing Process-induced Defects Using Fractography and Metallography: David Scannapieco1; Austin Ngo1; Collin Sharpe1; Mahya Shahabi2; Sneha Narra3; John Lewandowski1; 1Case Western Reserve University; 2Worcester Polytechnic Institute; 3Carnegie Mellon University
    Additive manufacturing (AM) is often challenged by its large variation in material quality across different process parameters and machines. In some cases (e.g. non-fracture-critical applications), the design freedoms and rapid prototyping afforded by AM can provide immediate implementation opportunities. However, to establish confidence in the material’s quality, particularly for fracture-critical applications, statistical predictors are needed which can determine, to a needed level of confidence, the reliability of a material built in a particular way. This study examines several predictive measures from extreme value analysis and extrapolated regression analysis across datasets from both metallography and a novel fractographical procedure conducted on fatigue-tested samples. Reliability, advantages, disadvantages, and comparison to the current ASTM standards are addressed. Challenges that arise regarding the construction of high confidence intervals, necessary for fatigue-critical applications, are presented for Ti-6Al-4V samples processed in different regimes to purposely create different types of process-induced defects.

4:40 PM  
A Method to Predict Critical Pore/Defect Size in Laser Powder Bed Fusion Additively Manufactured Ti-6Al-4V Parts: Mahya Shahabi1; Austin Ngo2; David Scannapieco2; John Lewandowski2; Sneha Prabha Narra3; 1Worcester Polytechnic Institute; 2Case Western Reserve University; 3Carnegie Mellon University
    A major factor in the fatigue life of fracture-critical parts is the effect of process-induced defects and the critical pore/defect size. Prediction of critical pore/defect size in different process regimes of a laser powder bed fusion (L-PBF) processed part could provide invaluable information for the widening application of additive manufacturing. This study uses extreme value analysis to predict critical pore/defect size in Ti-6Al-4V bend bar samples using the 2D cross-sectional porosity data. The results confirm that the pore/defect density and the required model precision determine the data required to characterize part porosity, the maximum pore/defect size prediction from process conditions used for one sample applies to another sample with similar porosity distribution, and the peaks-over-threshold model predicts critical pore size reasonably well. An analysis framework is presented and is used to demonstrate its applicability to predict the critical pore size in fatigue samples, with comparison to fracture surface observations.

5:00 PM  
Fatigue Modeling Approaches for Additively Manufactured Ti-6Al-4V: Sushant Jha1; Matthew Krug2; Luke Sheridan2; Patrick Golden2; Mark Benedict2; Nathan Bryant1; Jessica Orr1; 1University of Dayton Research Institute; 2US Air Force Research Laboratory
    Additive manufacturing (AM) represents a significant opportunity for on-demand printing of complex aerospace parts, thereby adding to the sustainment supply-chain, increasing the operational readiness, and reducing cost. In order to optimize an AM process for fatigue behavior and develop fatigue design curves, a large amount of testing is often required using specimens printed under varying AM and post-processing parameters. Physics-based fatigue modeling approaches can significantly reduce the burden of vast test campaigns. Here, a fracture mechanics-based method was used to determine minimum fatigue life bounds for laser powder bed fusion AM Ti-6Al-4V. The method was applied to model the shifts in minimum life bound with surface condition and post-processing treatments. In addition, an equivalent initial damage size (EIDS) distribution was developed for the same surface and processing conditions for a damage tolerance analysis. The proposed methods can play a key role in enabling the use of AM in fracture-critical applications.