Integrated Processing and Probabilistic Lifing Models for Superalloy Turbine Disks

Currently there is no commercial software modeling capability which correlates the details of a manufacturing process to the probabilistic lifing analysis of a component. Since fatigue life of a nickel base superalloy disk component is greatly influenced by bulk residual stresses, which themselves are functions of prior thermo-mechanical processing (TMP), service conditions, and microstructure features (such as grain size) and material anomalies (such as inclusions and pores), modeling of the evolution of these features during forming and in service will greatly improve fatigue life prediction. Using the integrated process modeling system DEFORM, it is possible to predict the evolution of critical life limiting factors during TMP (e.g. cogging, forging, heat treatment, and machining). Thus, integration of process modeling to probabilistic lifing methods will greatly help the jet engine industry, by improving fatigue life predictions and risk assessments of jet engine components. During Phase I of this project, Scientific Forming Technologies Corporation teamed with Southwest Research Institute to develop a framework to link the process modeling system DEFORM with the probabilistic lifing modeling system DARWIN. At the end of Phase I, we demonstrated a proof of concept model for linking DEFORM and DARWIN, specifically studying the effects of residual stress predictions from DEFORM generated during thermo-mechanical processing and service conditions on probabilistic lifing predictions of DARWIN for a generic jet engine disk. We investigated a modeling framework in DEFORM to effectively link process modeling results with probabilistic lifing method predictions. During Phase II of this project, our team will develop and implement modeling tools that will integrate location-specific grain size predictions, material anomaly orientation and residual stress profiles from processing models with probabilistic lifing methods. Proposed efforts are targeted towards developing techniques to link processing modeling results from DEFORM with the probabilistic component life prediction code, DARWIN. At the end of this program, it is envisioned that an infrastructure in DEFORM will be available to conduct sensitivity analysis specifically to address variabilities in processing conditions, material data and boundary conditions. It is proposed that DARWIN will be enhanced to take into account location specific grain size and material anomaly orientation in its life predictions and risk assessments. Our team is working closely with all the major jet engine OEMs to develop an implementation plan so as to maximize the benefits of linking processing models with probabilistic lifing methods. BENEFIT: It is anticipated that the link implemented in Phase II between process modeling results from DEFORM to DARWIN will enhance the accuracy of fatigue life and risk assessment of jet engine components, thus greatly benefiting the jet engine industry. Integrating process modeling with probabilistic lifing in Phase I has demonstrated the sensitivity of lifing results to residual stresses evolved during thermo-mechanical processing and service. Adding location-specific descriptions of grain size, material anomaly orientation and residual stress variability due to changes in processing conditions, material properties and boundary conditions is expected to provide more accurate predictions of component fatigue life and its variability. Building this link between manufacturing processing models and probabilistic lifing analysis will facilitate a genuine Integrated Computational Materials Engineering (ICME) methodology in analyzing the design and manufacture of a jet engine component. This will make it possible to optimize the design process and improve component performance by directly incorporating material and manufacturing variables into the assessment of component lifing and reliability. This link will also provide a tool to enhance understanding of the interaction between microstructural features and residual stresses on mechanical property response under service conditions. It is expected that this link will help in understanding and optimizing the processing window during the manufacture of jet engine components to push the existing limits of jet engine performance in service. The work proposed in Phase II will also serve as a launching pad for future full scale manufacturing process optimization analysis based on fatigue life of jet engine components under service conditions. It is anticipated that this would help in accelerating the insertion of new materials to service through a better understanding of processing, evolution of microstructural features and mechanical property response.