Ansys 12 and Workbench
 

Following the reply at the end of this thread from Glenn,

http://www.cfd-online.com/Forums/cfx...lyses-one.html,

Let's hear what people think of 12. I started using 11 in September 2008 after migrating from FloWorks. I have always run CFX Pre, solver and Post separate from WB. I find it's easier to keep simulations and results organized in a way that makes sense to me. Nice and simple to run batch files from command line and easy to wipe something out if you don't want it anymore.

Now recently upgrading to 12 and it appears that Ansys is taking a step in the direction of Floworks and its integration with Solidworks. <edit> for the type of analyses that I do, I don't find this useful <edit>. . . but I have only been using it for less than a week.

For now I will continue to work through CFX Launcher and/or Command Line (except for meshing which seems to have to be done in WB).

Hi,

I don't think generalisations like "this appears to be a step backward to me" are very helpful. V12 made massive advances in a wide range of areas. Many forms of simulations (eg FSI, optimisation, parameteric studies just to name a few) became much easier and more powerful in V12 under WB.

However as I said in the previous thread traceability of simulation input, settings and results is more difficult in WB then the old style command line driven things. V12 has progressed further on this path which means anybody using V12 for professional work needs to have a good think about how they are going to store, archive and use their simulation files. This is not a new problem to V12, it is just that V12 means it needs to be done differently to how it was done previously.

Of course as this forum is on CFX, all the CFX tools are still available in V12 as stand-alones, and can all be run from the command prompt so nothing has been lost in V12, you can still do things the old way if you wish.

Of course ANSYS has packages like EKM which is meant to address the traceability issue - I have never used it, so have no idea how effective it is. Also a wide range of 3rd party PDM software is available. There are plenty of ways to address this issue.

Regards,
Glenn Horrocks

I can see the benefit of tool integration, setting up coupled simulations and running everything within workbench however there is reasoning for having and running the stand alone tools outside workbench. I just hope that both options will be available in the future. Dont change something that works!

As many of you think there is room for improvement in regards to functionality and my view regarding workbench is that the current version should be a release candidate but what surprised me the most is that after all the beta testing ansys didnt pick this up.. Opening and using an older project to make a new run is fine as long as the old run is not messed around. Automatic updating and saving is my main problem with workbench and it is not until ansys address and solves that they will have a lot of angry customers..

The EKM package was discussed in the latest ANSYS magazine, it looks interesting.

Frankly, it seems to me that the complaints about file handling in Workbench R12 result from unfamiliarity and the lack of developing a dedicated R12 workflow that is inevitably different. CFX can still be run stand alone and in batch. Workbench projects can be archived. CFX cases can be imported. There is no need to have duplicate copies of geometries and meshes floating around once these have been "frozen", as I saw mentioned in the other thread. Just import the .cfx case into a new Workbench CFX object (NOT a Fluid Flow (CFX) analysis system) in a new project. Or don't do it in Workbench.

Now, the Workbench R12 "Project schematic" has a lot of drawbacks. It doesn't do a lot of things that I think it should, yet. It's touchy about things going stale very easily without an "undo" available. It's a bit flakey about things needing to be refreshed and updated. There are some real problems coupling anlyses in which you want to share geometry and mesh but NOT the model (thermal-structural, for example). The project file structure on the disk really is a byzantine nightmare. There are no thumbnails (my personal suggestion to ANSYS ;) ). There are other issues. But the claim that the Workbench R12 environment is somehow a step backwards seems excessive to me.