Modeling of residual stress distribution around fastener holes in thin plates after symmetric cold expansion

Abstract

The fatigue failure around fastener holes in metal structures is particularly dangerous, since it leads to inevitable accidents. The presence of residual compressive hoop stresses around these holes, imparted by means of mandrel cold working methods closes the existing cracks and impedes the formation of new ones and thereby extends the fatigue life of corresponding component. Although this type of pre-stressing of fastener holes has been of great advantage over recent years it has some disadvantages: significant and nonsymmetrical with respect to the plate middle plane axial gradient of the residual hoop stresses due to axial force flow passing through the plate, considerable surface upset, etc. The article presents a patented method and a tool ensuring symmetric cold hole expansion—introducing near-uniform residual hoop stresses around the hole along its axis having minimized and symmetric gradient with respect to the plate middle plane. Finite element (FE) simulations of the new process have been carried out. The residual stresses and out-of-plane deformations (surface upset) have been analyzed and compared with those obtained by mandrel cold working method and thus the advantages of the patented method have been shown. The symmetric cold expansion process has been realized experimentally and modeling of residual stress distribution around cold expanded holes in carbon steel specimens has been made. The residual hoop stresses have been measured by means of X-ray diffraction technique. The obtained generalized empirical mathematical model of the residual hoop stress distribution is polynomial with respect to the distance from the hole edge and has “coefficients” that are functions of the governing factors. On the basis of the information obtained from FE simulations, suitable fractional factorial experimental design has been synthesized and used for obtaining the polynomial “coefficients”. The verifications of the generalized model prove its authenticity.