Additive Manufacturing Fatigue and Fracture: Developing Predictive Capabilities: Fatigue Modeling and Prediction
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 8:30 AM
March 2, 2022
Room: 258B
Location: Anaheim Convention Center

Session Chair: Nima Shamsaei, Auburn University

8:30 AM  Invited
Predicting Structural Integrity of Additively Manufactured Parts Using Probabilistic Damage Tolerance: Robert Mcclung1; 1Southwest Research Institute
    Because fatigue failure in Additively Manufactured (AM) parts is often defect-controlled, conventional predictive methods for structural integrity may not always be appropriate, especially for components of higher criticality. This presentation will discuss the ongoing adaptation of established zone-based probabilistic damage tolerance (PDT) methods and software for safety-critical AM applications. These PDT methods explicitly address the potential size and frequency of material defects remaining in the part when it enters service, including the influences of nondestructive inspection, and they explicitly address the possibility that a fatal defect could occur nearly anywhere in the part. Topics to be addressed will include alternative analysis strategies, challenges for software tools and supporting materials data, case studies, and experimental validation.

9:00 AM  
Predictive Modeling of Fracture in Anisotropic and Porous Materials: Amine Benzerga1; Vigneshwaran Radhakrishnan1; 1Texas A&M University
    Additively manufactured metals often exhibit directional mechanical properties as well as residual processing-induced porosity. Unless production cost is not a factor, anisotropy and porosity are both unavoidable attributes of AM products. Here, recent progress in first-principles modeling of failure in materials with initial porosity is used to lay out a methodology for predictive modeling of failure in AM. The aim is to assess the effects of porosity and plastic anisotropy, taken separately or combined, on a stress-state dependent measure of strain to failure. Published experimental data on materials with varying ductility levels is used to demonstrate the predictive capability of the proposed modeling framework.

9:20 AM  
Information-rich Fatigue Fracture Surface to Evaluate Additive Manufacturing Parameters: David Scannapieco1; Austin Ngo1; Collin Sharpe1; Hunter Taylor2; Joseph Pauza3; Enrique Garibay2; Evan Diewald3; Christian Gobert3; Ryan Wicker2; Anthony Rollett3; Jack Beuth3; John Lewandowski1; 1Case Western Reserve University; 2University of Texas at El Paso; 3Carnegie Mellon University
    Evaluation and optimization of parameters for additive manufacturing (AM) has grown more sophisticated as the industry has expanded. Parameter maps, metallurgical analysis, and predictive formulas are just some of the tools employed in this process. This study introduces a new tool: mechanical and fractographical analysis via 4-point bend fatigue (4PB). The fracture surface of a fatigued AM component provides numerous sources of insight into the details of the AM process for said component. From initiation points, KQ estimates, quantitative and qualitative defect analysis this method creates a roadmap to optimize parameters. Fatigue analysis of AM components is shown to be just as effective on irregularly sized specimens as on ASTM standard specimens. A series of case studies (e.g. Ti-6Al-4V, GRCop-42) are presented demonstrating the process, the information extracted, and the application of identified solutions into a new build to show the method’s success.

9:40 AM  
Effects of Process Parameters on Fatigue Behavior and Defect Characteristics in LPBF Ti-6Al-4V: Austin Ngo1; David Scannapieco1; Hunter Taylor2; Ryan Wicker2; Joseph Pauza3; Anthony Rollett3; Jack Beuth3; John Lewandowski1; 1Case Western Reserve University; 2University of Texas at El Paso; 3Carnegie Mellon University
    Four-point bending fatigue testing was conducted on machined and polished LPBF Ti-6Al-4V mechanical testing specimens. Specimens were built with parameters inside and beyond the optimal process window, resulting in low (baseline) and high (i.e., lack of fusion, keyhole) defect-content builds, respectively. S-N fatigue data was generated for each process parameter set, and specimen fracture surfaces were imaged using OM and SEM. Fractographic analyses consisted of quantifying all defects on the fracture surface, identifying ‘killer’ defects responsible for fatigue failure, and estimating fracture toughness from the crack length at catastrophe. Fracture surface height profiles were generated via laser scanning and compared to visual analyses of crack progression. Different types of process defects were more prevalent depending on a build’s relationship to the process window, which further influenced S-N curves. The effects of process parameters on defects and resulting S-N fatigue will be discussed in the context of a Kitagawa-Murakami-type approach.

10:00 AM Break

10:20 AM  Invited
Candidate Methods to Assess Structural Integrity of Higher-criticality AM Components: James Sobotka1; Robert McClung1; 1Southwest Research Institute
    This presentation surveys methods to assess the safety of principal structural elements, life-limited parts, and other higher criticality parts manufactured by AM processes. This presentation focuses on parts under cyclic loading that drives fatigue crack growth and fracture in metallic alloys. Here, we consider key features of the structural assessment, provide a consistent terminology, and discuss core concepts that underlie the assessment strategies. Following this background, we discuss possible part requirements on the acceptable limits of flaws produced by AM processes. Based on these considerations, this presentation outline four methods that could be used to demonstrate safe usage: traditional fatigue methods, deterministic damage tolerance, probabilistic Kitagawa diagrams, and probabilistic damage tolerance. We conclude this work by providing an overview of the key enablers that support the usage of the various methodologies.

10:50 AM  
Size Effect on the Ultrasonic Fatigue Behavior of Laser-powder Bed Fusion 316L: Megan Trombley1; Qianying Shi1; John Allison1; 1University of Michigan
    The idea of size effect on fatigue behavior in metals has long been understood as the decrease in fatigue strength with increase in part size. With lab testing often evaluating the material rather than the exact part, higher fatigue strengths are often reported than what is observed in application. The reason for this discrepancy is largely attributed to differences in stress distribution and statistical scatter in strength and microstructure. This statistical scatter in strength is particularly evident in metals with a higher percentage of defects, as seen in additive manufacturing. Evidence of the size effect is demonstrated by cylindrical dogbone specimens of gauge 1.5, 2.5, and 5.0 mm manufactured via laser-powder bed fusion on both a Concept Laser M2 and a 3D Systems ProX DMP 200. An evaluation of the microstructure, defect size, and defect distribution have been conducted to describe the mechanism of size effect.

11:10 AM  
Rapid Characterization and Comparison of the Cyclic Response of Laser Powder Bed Fusion Additive Manufactured Inconel 718 Samples Using Spherical Microindentation: Camilla Johnson1; Aaron Stebner1; Surya R. Kalidindi1; 1Georgia Institute of Technology
    Historically, fatigue is a costly property to assess because of the volume of material required for test coupons and also the time to run each test. Therefore, a critical need exists for the development of novel experimental approaches that can rapidly evaluate the relative changes in cyclic response as a function of alloy chemistry and thermo-mechanical processing history. This becomes especially critical in the materials development efforts as it relates to additive manufacturing, which require systematic exploration of a large materials space. In this work, we present a novel approach using spherical microindentation to deduce stress-strain loops and compare the cyclic response of different heat treatments of laser powder bed fusion additive manufactured Inconel 718. The results show promise for obtaining reliable, high-throughput, quantitative assessments of the cyclic response.

11:30 AM  
Quantifying the Influence of Scan Strategy on the Microstructure and Fatigue Properties of SLM Inconel 718 Thin Walls: Connor Varney1; Paul Rottmann1; Md Imran Noor; 1University of Kentucky
    The as-printed microstructure of additively manufactured parts is a function of many variables that span from scan strategy to part geometry. This is particularly relevant in precipitation strengthened alloys (e.g. Inconel 718), as thermal history—which itself varies across a build—dictates the distribution of precipitates in the microstructure. Elucidation of the complex relationship between geometry, scan strategy, and resultant microstructure is necessary to optimize scan strategies. In this study, a series of thin wall Inconel 718 tensile specimens with thicknesses ranging from 0.1-1mm were printed via SLM. The thin walls were printed using either only contours or a contour+hatching strategy. To obtain a better understanding of the process-structure-property-performance relationship of thin walls the influence of scan strategy and thickness on the microstructure and porosity (EBSD, microCT) as well as mechanical properties (tensile, fatigue) were characterized and compared to bulk specimens from the same print run.