Bayesian framework for probabilistic low cycle fatigue life prediction and uncertainty modeling of aircraft turbine disk alloys

• Multi-source information fusion using Bayesian inference was performed for probabilistic low cycle fatigue life prediction.
• A Bayesian framework for uncertainty quantification and probabilistic low cycle fatigue life prediction was proposed.
• The proposed framework provides a good correlation on GH4133's fatigue life distribution over the temperature range.

Probabilistic life prediction of aircraft turbine disks requires the modeling of multiple complex random phenomena. Through combining test data with technological knowledge available from theoretical analyses and/or previous experimental data, the Bayesian approach gives a more complete estimate and provides a formal updating approach that leads to better results, save time and cost. The present paper aims to develop a Bayesian framework for probabilistic low cycle fatigue (LCF) life prediction and quantify the uncertainty of material properties, total inputs and model uncertainty resulting from choices of different deterministic models in a LCF regime. Further, based on experimental data of turbine disk material (Ni-base superalloy GH4133) tested at various temperatures, the capabilities of the proposed Bayesian framework were verified using four fatigue models (the viscosity-based model, generalized damage parameter, Smith–Watson–Topper (SWT) and plastic strain energy density (PSED)). By updating the input parameters with new data, this Bayesian framework provides more valuable performance information and uncertainty bounds. The results showed that the predicted distributions of fatigue life agree well with the experimental data. Further it was shown that the viscosity-based model and the SWT model yield more satisfactory probabilistic life prediction results for GH4133 under different temperatures than the generalized damage parameter and PSED ones based on the same available knowledge.