Probabilistic risk assessment method considering machining-induced random residual stress

• A probabilistic risk assessment model is proposed to evaluate the impact of machining-induced residual stresses. • The residual stress dispersion is quantified using a scaling parameter. • Quantify residual stress dispersion by measurements and orthogonal cutting simulations. • Establish crack life database under varying residual stresses for efficient probability calculations. • Failure probability with compressive residual stress is lower than that without. The aero-engine disks inevitably have manufacturing-induced anomalies and machining-induced random residual stress (RS) in localized and critical areas, which cause a severe threat to the safety of the aircraft. Traditional structural design of the disks fails to establish a quantitative correlation between the machining process and the failure risk. Therefore, this paper proposes a probabilistic model considering random RS to quantify the influence of machining RS subjected to low-cycle fatigue. The RS dispersion is quantified using a scaling parameter, obtained by X-ray diffraction measurements and orthogonal cutting simulations. The crack life database under varying RSs is established for efficient probability calculations. Results show that the coefficient of variation (COV) of the RS on the same machined surface with the same processing parameters is 7.62% in the local area and 13.74% in the whole machined surface. The risk results show that the probability of failure (POF) considering the deterministic RS is 2-4% lower than the POF without RS, owing to the extension of fatigue life by compressive RS. Furthermore, the POF considering the random RS is almost the same (difference less than 0.6%) as the POF considering the deterministic RS because the depth of the machining RS is around 0.2 mm. The proposed method predicts the POF more accurately and is thus valuable for the safety assessment of an aero-engine titanium disk.