Measurement and finite element prediction of residual stresses in aluminium alloy 7010 forgings

The finite element code ABAQUS is used to simulate the quenching of aluminium alloy 7010 in an attempt to predict the residual stress distribution that develops in simple shapes. All of the thermal and mechanical property data is input into the ABAQUS code as a function of temperature. The problem is sub-divided into a heat transfer problem and a stress/displacement problem. The rate of heat transfer from the material is determined by using the finite element method to predict the heat transfer coefficient from surface cooling curves achieved experimentally from quenching into water at different temperatures. Predicted internal cooling rates are then verified for varying geometries using data measured with deeply buried thermocouples for quenching in water at room temperature (<40°C). The remaining thermal and mechanical data for the computer model is taken from literature. The predicted residual stress distribution is compared to values measured using the hole drilling strain gauge method, the X-ray diffraction technique and a mechanical dissection technique. These three verification techniques were used to ensure accurate verification of the computer predictions and to ensure that the techniques are comparable. A thorough attempt is made at measuring the X-ray elastic constants for the X-ray diffraction technique to ensure accurate residual stress measurement. It is concluded that these constants calculated using theoretical techniques are more accurate than those measured. Given that the hole drilling strain gauge method can only measure residual stress magnitudes that are less than 50% of the materials yield strength, the material was aged to a T6 type temper. This resulted in no measurable decrease in residual stress magnitude but with a large increase in yield strength. The predicted values compare well with those measured for larger geometry blocks. However, stress magnitudes are underestimated for smaller models as shown by the Xii ray diffraction technique and the mechanical dissection technique used. The error in poor prediction for smaller blocks can be attributed to the plasticity data used. The effect of ageing treatments on residual stress magnitudes is studied. The largest stress reductions measured occur at higher temperatures where the material has long soak times. It is therefore concluded that stress reduction through ageing treatments occurs as a result of thermally activated plastic flow. The stress reductions achieved are generally minimal (20%), offering no substantial advantage in stress reduction between one ageing treatment and another.