Dear All,

Our company is seriously considering purchasing a high end CFD package in order to model the fluid flows in our plant, mostly impingement mixing and foaming.

Since we do not have clearly defined problems (and our problems change all the time anyway), we need something general purpose.

Does anybody have any comments about the above packages, or others we should consider. Salespeople from the above companies have all had their chance to speak to me directly(except CFD-Ace who will soon). I'm looking for user comments.

Thanks

Jeff

PS this is definitely one for the now empty FAQ

It is hard for me to give your some comments Since you do not have clearly defined problems. You can tell me more detail about your problems.

cfd_99

(1). To have CFD codes on the system is not a problem. The problem is you need to define the CFD problem, and learn how to use a CFD code to solve it. (2). If you have three engineers ready to work on it, it would be ideal to have three codes on the system so that you can double check your results. It is more reliable that way. (3). If your resources are limited, then just pick a cheaper one to begin with. After 3 months working on it, your engineer should be able to tell whether he needs to try another code or not. (4). So, in 9 months, you should have your solution. (5). In general, the critical part of the equation is the "experienced CFD engineer", not the code itself. This is because you are looking for a general CFD code. (6). On the other hand, if you are looking for a special feature in CFD, then it is worthwhile to shop around.

Hy,

here are some points for you to consider.

1. I think, cfx is often used in turbomachinery whereas Fluent is used if you have reacting things/combustion.

2. I'm not sure, but I think, cfx only allows structured meshes. If you have complicated geometries, the use of cfx is not possible (If this is wrong please correct me !!)

tom

Recent work by S. Lo at AEA puts CFX's 2 phase capabilities way beyond both Fluent and CFD-Ace.

J.

J. Curtis:

Could you show us the results? or you could not say like this.

cfd-99

Hi Jeff,

Some of the main issuses that you have to consider when both talking to the sales guys and to this group are:

1. Types of physical models needed eg. multiphase, combuston, turbulence models, free surface, particle tracking, etc.

2. Accuracy and robustness of the solver including the discretization schemes, coupling or segregation of variables, linear solver performance (ie. multigrid or other acceleration techniques), etc.

3. Level of training, documentation, and user support. Since engineering time can add-up very quickly for a confused newcomer to CFD.

Once you have defined some of these needs, you will be in a better position to get some relevant advice! Another really good idea is to attend a conference which has presentations done by INDEPENDANT industry people with the various CFD codes. Most compamies will also give you some references of people who have licensed the code!

So, get a bit more focused and good luck!

Regards........................................... .Duane

This is in response to Tom's e-mail about CFX. I would like to point out that Tom is correct in pointing out that CFX-TASCflow is the leading CFD package for turbo-machinery. It has been used successful to model compressors, turbines, pumps,... I also would like to add that CFX-4 is well established in the multiphase chemically reacting flows as mentioned by Curtis. In summary, CFX-4: well established in the CPI industry, handles multiphase, chemically reacting flows and other applications very well. CFX-TASCflow: the turbo-machinery code of choice as well as other applications.

Hope this helps.

Hi all

About the comparison between FLUENT and CFX, could someone give more detailed information about the properties of CFX?

I'm trying to model a fuel cell flow channel. That means that there is air and water vapour entering the channel. Oxygen of the air is consumed in the fuel cell reaction, and the product is water, which in some stage starts to condensate. Geometry of the problem is quite simple, however there's a porous diffusion layer between the flow channel and the reaction site.

I've been studying the use of FLUENT 5, the problem is that it is possible to model:

- mixtures with changing composition - phase changes - multiphase flows

BUT not at the same time.

Could someone tell if it is possible to do that with CFX 4.2?

Janne