Fatigue strength assessment of additively manufactured metallic structures based on a macroscale approach

Wolfgang Schneller

Research output: ThesisDoctoral Thesis

Abstract

Fatigue failure of additively manufactured (AM) metallic structures is associated to immanent imperfections which demand proper assessment. This research work identifies and investigates main influencing factors and their adverse effects on fatigue. To enable universal applicability, specimen series are manufactured by laser powder bed fusion of different materials. Conducting various post treatments allows data spreading of essential parameters. Eventually, a novel methodology in order to estimate the fatigue strength by considering individual aspects as well as their interaction, is developed. Macroscopic bulk material imperfections are assessed by test series exhibiting machined surface condition. As-built surface condition reveals a combinatoric effect of intrinsic flaws residing in the vicinity of notch-like surface roughness features. Significant residual stresses at the failure initiation site alter fatigue performance, hence are viewed as a present mean stress condition. Established methodologies are extended by empirical correlation coefficients derived from experimentally determined data, enhancing accuracy and applicability. Thereby, transferability from conventionally to additively manufactured structures is ensured. The long life fatigue strength is estimated by referring introduced reduction factors to ideal, defect and residual stress free base material strength. Killer defect distribution is determined by fracture surface analysis and considered in terms of a bulk material reduction factor. Concerning observed different defect types inherent to the AM process, no distinction is made and flaws are treated equivalently. Surface roughness features, namely macroscopic notches, are found to be well assessable by the areal pit depth and notch root radius. The basis of deriving roughness metrics form three dimensional, digital optical surface scans. Common stress concentration based models are enhanced and a corresponding reduction factor is derived. To cover the interaction of surface and bulk characteristics an empirical correlation exponent is introduced. The fatigue relevant, macroscopic residual stress condition is elaborately evaluated by X-ray diffractometry and assessing the X-ray spectrum. Residual stresses of first order overlay with load stresses, therefore act as present mean stresses and are further considered by an equivalent damaging parameter. Concluding, a comprehensive and unified fatigue assessment methodology is developed in order to estimate fatigue strength at ten million load cycles for machined and non-machined structures. Verification of presented estimation methodology is performed by investigating several independent test series in both surface conditions. Experimental data matches the analytically estimated fatigue strength well. Throughout initial series, which are used for deriving presented model, and independent validation series, application provides a conservative fatigue strength assessment by revealing a mean deviation of -8%. Feasibility in engineering applications is ensured and thereby significantly contributes to the dimensioning process of AM structures with beneficial effects on lightweight and function integrating designing, saving of resources, exhaustive testing and post processing as well as enhanced reliability of safety relevant, structural components.
Translated title of the contributionErmüdungsfestigkeitsbewertung additiv gefertigter metallischer Strukturen auf makroskopischer Ebene
Original languageEnglish
Awarding Institution
  • Montanuniversität
Supervisors/Advisors
  • Stoschka, Michael, Assessor B (internal)
  • Leitner, Martin, Assessor A (external)
  • Stockinger, Martin, Co-Supervisor (internal)
  • Grün, Florian, Supervisor (internal)
Publication statusPublished - 1800

Bibliographical note

embargoed until null

Keywords

  • Fatigue
  • Residual stress
  • Porosity
  • Surface roughness

Cite this