Hi,

I'm currently working on a research project looking at the "CFD environment of the future" for aerospace designers/ engineers and I'd welcome your views on this topic.

The project is looking at how CFD can be improved by changing certain processes (i.e. ways of improving 3D mesh generation), through the use of technology (i.e. HPC, different algorithms) and changes to the way engineers interact with these tools and processes.

So your views on the real bottlenecks in the current processes, from a technology and human aspect would be most useful. If you can think of a "wishlist" for improvements you could think of, then that would also be very helpful.

The second aspect looks at how designers/ engineers will interact and visualise these processes in the future. So looking at how you can visualise the data produced (i.e. 3D, virtual environments, Haptic technology etc) and what designers/ engineers really want to be able to see and do that they currently can't.

So if in 10 years time you had the unlimited computing power what would you really want to be able to see and do? Would you like to be able add complexity to your simulations? or would you like a single package that could go from CAD to solver with just a few clicks? and if you could do that would you even want to?!

As experts in the field, I'd welcome your views on how you perceive CFD will change and what you'd really like to get from CFD in 10 years time.

Many thanks for your help.

1 - find a way to get rid of those directors/managers who are not 'hands-on' and have poor organizational and problem solving skills. They have nothing to do in CAE and no employee wants to work with them.

2 - couple cfd code with an automatic and validated thermophysical property calculation library.

3 - more use of the MUSCL algorithm

People will think of connecting(Some companies are started doing already) CFD world to thd CSD (Computational Structural Dynamics) world so that, they can utilize the best features of each side. They may turn to common front and back end with two solver codes.

I am seriously worrying about the current turbulent path(why dont we search for a good algorithm and numerical methods esp. for fluid dynamice instead of using empirical relations). Even it is tough to avoid empirical, we can find more robust techniques for turbulent.

The present CFD companies are pulling CFD to their business side, not thinking of spend some money for research but implementing 90's algorithm now and delivering the products. This is dominating in the CFD world(Due to strong Industries backbone support and purchasing license...) and hope we can find some new strategies if our mind of CFD people will work on OPEN CFD SOURECE CODES...

My answer for your posting.. is after 10 years..

we can find the most simplest cfd package.. like "Only 10 clicks for combustion simulation".. and there will be a slight improvement in solver, algo side.

Many thanks for your comments and insights.

If anyone has any other comments then please feel free as this work is trying to pull in as many views as possible to promotes dicussion amongst the CFD end users.

In particular if there's any functionality that you really want to see in the next generation of CFD tools then let me know!

Thanks