[iman]

hi every body
i've started learning and using CFX before one month. i've done almost all of the tutorials provided by ANSYS CFX.
just to make sure that i got it right, the general approach will be:
- building the geometry of the model using WORKBENCH
- generating the mesh
-importing the mesh into CFX-Pre and setting the model characteristics, constrains, material properties...etc
-solving the model using CFX-Solver manager
-post processing the results using CFX-post

is that correct?or am i missing any step?
also, i want to ask how can i test and validate my model using one of the online databases? whihc software i should use for that, is it CFX-Post?

thank you very much for your cooperation and help

[ghorrocks]

The general steps of CFD are:
* Generate geometry
* Generate mesh
* Set up simulation
* Solve
* Post-process

In the ANSYS environment you have lots of options. Geometry can be from a solid modelling package or designmodeller or others, meshing can be done in ansys meshing or icem, setup is CFX-Pre for CFX, or Fluent for fluent and CFD-Post for post processing. But you can import and export at just about any of these steps to all sorts of other applications as well.

Good to hear you are looking for benchmark solutions. I recommend for a beginner you look at a few of the standard CFD benchmarks - flow in a lid driven cavity, flow over a forward and backward step, thermally driven flow in a differentially heated cavity (the "double glazing problem"). If you can get accurate answers to these benchmarks over a range of Re/Ra numbers then you can be confident you know the basics of setting up a simulation and getting accurate answers (lots of people know the setup bit but far fewer have taken the time to actually find what makes CFD accurate or not).

[iman]

Originally Posted by ghorrocks

Good to hear you are looking for benchmark solutions. I recommend for a beginner you look at a few of the standard CFD benchmarks - flow in a lid driven cavity, flow over a forward and backward step, thermally driven flow in a differentially heated cavity (the "double glazing problem"). If you can get accurate answers to these benchmarks over a range of Re/Ra numbers then you can be confident you know the basics of setting up a simulation and getting accurate answers (lots of people know the setup bit but far fewer have taken the time to actually find what makes CFD accurate or not).

thank you very much for your kind reply.
so i ccan use for example the results of flow in channel to test and validate a model of circular pipe???
could you please suggest some of the suitable benchmark cases for validating a simple 2D, circular pipe flow of water. i just need to understand and practice the step of validation and verification of model and simulation results

[ghorrocks]

Simply the development of a boundary layer and a fully developed flow in a pipe would be a very good benchmark case which should have some good solutions around. Do a literature search and you will find it.