Residual stresses in large size forgings

The presented research project attempts to improve the understanding of the residual stresses arising from both thermal and phase transformations during the heat treatment of large forgings. Particular attention is given to the accurate representation of the cooling stages. The methodology which combines both experimental and simulation techniques was developed and included finding the best practice for CCT/TTT implementation, dilatometric and hardness testing, microstructure observations and full integration of material data into finite element models. In the thesis the overall numerical investigations of stress management during heat treatment of large-size forgings was performed. Modelling of transient temperature field and a new kinetics approach to calculating the phase transformation for large-size forgings was presented. To study the residual stress that arise in the large size forging the author of the project performed complex investigations that included: 1. Determination of the occurrence and duration of influence of different factors and their possible interactions that may affect generation of stresses in the elements from the selected material (SA508 gr.3 steel) during heat treatment after deformation; 2. Analysis of the heating and cooling conditions of the final products during production, causing uneven heating and cooling of different layers of material; 3. Analysis of heterogeneity of microstructure of the chosen material (SA508 gr.3 steel) after heat treatment and phase transformation, which is a source of high internal stresses related to the heterogeneity of deformation in the volume of the processed element; 4. Analysis of residual stress caused by uneven heating and cooling of the different areas of the part leading to dimensional changes; 5. Analysis of residual stresses caused by phase transformations of the structural components of the material, occurring at different temperatures, due to the change of the specific volume that is accompanying the phase change. The analyses were supported with dilatometric and hardness tests, followed by microstructural observations on the optical microscope. The model was validated using the industrial case provided by the company Sheffield Forgemasters Ltd. (Sheffield, United Kingdom). In this case, the cooling of the forged 95-tonne disc shaped ingot was performed. In the applied model, the relative geometry, reflecting the rectilinear 2D profile machined out of the forging before the heat treatment, was used. The model was used to calculate the cooling rates at the quarter thickness location and in the mid-wall of the geometry. The calculated cooling rates were about two times higher inside the part than those presented in literature. The calculated transient temperature field allowed for estimation of the cooling rate at every node in the model at every time step during the process. These data were used in the last part of the work to assign thermal expansion/contraction coefficients obtained from the dilatometric test to the model. It allowed for prediction of the stress distribution within the forging. The results were presented as the changes in the dimensions of the angles between cylindrical and flat surfaces in the rectilinear model geometry and were compared to the values registered during the industry process. The stress and strain intensity values within the model during the cooling process were also presented. Although minor concerns are present, such as the possibility of insufficient accuracy in defining the growth sign of the angles, the obtained results were satisfying. The developed model could be applied as a sub-function in wider investigations. To demonstrate this, in the final part of the work the sub-function was implemented in the set of simplified welding models, which represent its approach in the exemplary practical calculations. The set consists of the six models: three, where the temperature field due to the thermal exchange under different conditions were calculated, and further three models, which used such calculated fields to determine the stress/strain respond of the material. The stresses, which arose due to the thermal expansion and phase transformation and which caused the local volume change, were predicted in the characteristic locations of the geometry of the weld and within the steel table, upon which they were placed on. The models, generally geometry and condition dependent, mostly localised high tensile stresses within the lower parts of the weld at the centre and at the weld contact surfaces, while the compressive stresses were identified between the weld beads. Although the tensile stresses were identified in steel table at the bottom and in the centre of the bulk, the compressive stresses, highest at the contact surfaces, were mostly present. The obtained results were in good correlation with the result from the literature. Besides, the author believes that the implementation of more dilatometric data could significantly improve the predictive abilities of the developed numerical tool. The aim of the work is to broaden the knowledge of the thermomechanical processes and improve the stress management during manufacturing the large-size ingots using the commercial FE modelling programs.