Algorithms for Improved Numerically Controlled Manufacture of Stress Optimal Free Form Shapes

Australia's Defence Science and Technology Organisation (DSTO) has developed technologies to design and manufacture airframe repairs that extend the life of Royal Australian Air Force (RAAF) aircraft. For example a shape optimisation technology has been developed based on iterative finite element analyses (FEA) to determine optimal repair profiles which blend out fatigue cracks and minimise peak local stress under operating conditions. Here previous work has however also shown the need to improve the transformation of optimal shapes as defined by FEA co-ordinates to achieve smooth shapes suitable for effective numerically controlled (NC) manufacture. A capability for efficiently creating intermediate near optimal shapes from bounding optimal shapes is also important. Hence this paper presents algorithms to smooth undesirable local variations in the radius of curvature of typical optimal shapes along with algorithms to create intermediate (or morphed) optimal shapes. The success of these approaches is demonstrated on both generic problems and through application of the procedures to the design and manufacture of repairs to airframe components. The results also show that the abovementioned methods can be used without degrading the optimality of the initial optimal shapes.