The uncertainty elimination in discrete topology optimization of compliant mechanisms

The uncertainty in topology optimization leads to ambiguity caused by different topology solutions. This uncertainty is caused by either having point connections or grey cells which are eliminated by using hybrid discretization model for discrete topology optimization. Mesh dependence also contributes to topology uncertainty. When the design domain of a topology is discretized differently, its solution depends on the type of discretization which leads to uncertainty caused by mesh dependence. To address the problem of mesh dependence, the genus based topology optimization strategy is introduced in this thesis. This approach deals with controlling of the genus of an optimized compliant mechanism in the process of optimizing the topology. There is no uncertainty caused by mesh dependence when this strategy is implemented with hybrid discretization for topology optimization. An example is considered to demonstrate the introduced approach.