[kmgraju]

Dear all
CFX is commercially available software.How to provide benchmark cases for validation of software for a given problem ( experimental work not conducted )

Thank you

Govind

[ghorrocks]

First of all well done for recognising the importance of benchmarking your simulation.

The normal approach for validation and verification is to find a test case similar to your model, or at least containing important elements of your model, for which you can find high quality benchmark results on. Model the benchmark simulation and check you get an accurate result within a tolerance you are happy with.

Examples would be:
* For bluff body flows the Ahmed body has lots of published experimental and simulation results.
* For shock waves check that you can model the 1D shock wave in a tube simulation. This has an analytical solution.
* For heat transfer check that you can model heat flow in a simple body with an obvious answer (eg constant heat difference across a rectangular body)