Manpopp/Multiscale-Concurrent-Topology-Optimization-for-Lattice-Material-by-SIMP-on-Spark

Multiscale concurrent optimization on spark

Multiscale-Concurrent-Topology-Optimization-for-Lattice-Material

Mutiscale-Optimization

• Multiscale optimization of lattice materials is a hot topic in the field of structural topology optimization,the picture below shows the difference of topology optimization between continuum structure and lattice material

Distributed Parallel Optimization Method

• Our distributed parallel optimization is based on spark,the DAG is as follows:

Numerical Example

• Boundary conditions of cantilevel beam and the topology result

* Boundary conditions of MMB and the topology result

How to use?

step1: you should have a spark environment
step2: config the input in the parameter.xml which include mesh mode,boundary condition,cpu numbers to use,path for data writing and more.
step3: Packaging the java code as jar file ,and submit to spark . There are two kinds of running mode of spark program.

• --local mode:in this mode,you can run the program on windows or linux
• --cluster mode:in this mode ,you can run the program only on linux
  

if you have any questions please write them on the issue page