What is the Theoritical Background for Topology Optimization and Lattice Optimization in HyperMesh - OptiStruct ?

Hello, 

I am wondering the exact theorical methodology used for Lattice optimization and topology optimization. Let me explain in detail. When I check topology optimization and OptiStruct Guides, they always say The SIMP method is used to conduct topology and lattice optimization problems.

I am looking for references The SIMP method used in OptiStruct. When I checked different articles, I see that there could be some modifications on SIMP method to prevent checkerboard effect. For instance, when MINDIM, MAXDIM are defined in topology optimization, what kind of filter is implemented ( density or sensitivity etc.) ? 

In articles, modified SIMP approach is mentioned to get rid of singularity as well. Is modified SIMP approach used in Optistruct algorithm or does it use different methodolgy to ban singularities ? 

Thank you for your answers 

Best Regards