Can anyone recommend the best commercial CFD packages in terms of price and performance?

Thanks.

(1). I am not sure I understand your question. (2). If you are interested in displaying the results, it is better to get a post-processing package instead of a general cfd package. (3). If you are in the turbomachinery business, it is a good idea to consider mesh generation package specifically written for such applications. (4). The price can be defined as the license fee only. Or it can include the technical support. But the price does not guarantee the success of getting your problem solved. (5). The performance is very hard to evaluate, unless you have a set of rules in mind. The performance must be related to the task the code is going to perform, that is to solve your problems. (6). So, I think, it is possible to find out which code is better suited for your problem. But then, your will have to disclose your problem first. (7). I guess, you know that a code can not run by itself, so, who is going to run this high performance code? A good Indy 500 car is necessary to win, but without a good driver it is no good. (8). My suggestion: go through vendor's home page, find the example from the applications, then contact the vendor for further information.

can some one show me books talking a bout differincing schemes in time and space (explicit, implicit ...); (upwind, central, hybrid, quick...) and there application with finite volume method in CFD, particulary in discretised transient convection-diffusion equations. thanks

John,

Thanks for your response. I'm looking for the whole package. For example I've looked at ANSYS's Flotran and have also talked to a company called AEA who markets a package called CFX. These packages are basically do the whole thing from mesh generation, calculation and graphical outputs. The software is basicaly a windows type environment with GUI for ease of use. What I wanted to know was if anyone had evaluated these packages (perhaps there's a review somewhere which compares the packages) and which ones are thought to be the best in terms of price or performance. Peformance could be accurage of the solutions provided or speed in which the program solves a problem.

Nick,

Despite the fact that I work for a company that develops commercial CFD software, I have to agree, at least to some extent, with John Chien. In order to answer your question about which code is "best", you have to really frame it in terms of what kind of problems you want to solve.

I have used Flowtran from Ansys in the past, and it is less expensive than our code. So you could say that in terms of price it is "better", however, Flowtran is really suited for doing 2D analyses. It is extremely difficult (at best) to generate good 3D CFD meshes because the package is really geared toward structural analysis.

Our code, CFD-ACE+, is a very good general purpose code for doing fluids, heat transfer, electro-physics, structural interaction, chemical reactions, etc. I'd like to say it's the best (and not just because they pay me to), but that's hard to say without know what you want to do with it.

If your application is something that is very out of the ordinary, it is possible that there is a code that was specially developed for that application that would clearly out-perform everything else.

You also need to consider your criteria. Generally price is an easy one to define. How much will it cost to license the software? That's a number you can pin down.

Performance is a bit more difficult. You should consider both how long it will take you to set up the problem and how long it will take to run. Often problems take longer to set up than they do to run.

I agree with John that the best approach is to spend a few minutes framing your problem in your mind. Then spend some time browsing web sites looking for similar problems. Then e-mail the candidates looking for specific information. They should be happy to help. If their sales folk aren't helpful, then just think how their customer support people will be.

Regards, Alton

(1). In terms of accuracy and benchmark testing, each vendor should be able to provide you some information about their codes. (2). A few years ago, there was also a paper published in AIAA (?) which provided some results from various vendors using same benchmark test cases. I don't have the direct access to that paper right now. So, you may want to do some checking. (3). Since the publication of that paper, many newer versions of the same code, or brand new codes have been appeared. So, it is hard to make direct comparison without testing information. (4). In genreal, you should be able to get fairly good pressure solution to your problem. The velocity field will be somewhat harder to get accurately. The skin friction , drag and loss will be difficult to obtain accurately, because of the state of the art of the turbulence modelling. Then, the surface heat transfer will be the most difficult wall parameter to "predict" directly. (5). The other reason why it is hard to do direct comparison is that we as the user of commercial codes normally are dealing with very complex problems, therefore, it is not practical to "waste" the time to explore the relative accuracy issue, even though I do have access to several commercial codes right now. (6). So, the relative accuracy issue among various commercial codes is still an open question.

Dear Sir,

Subject: Acquisition of information on Software

Department of Mechanical engineering of NED University of Engineering & Technology, Karachi, Pakistan, intends to develop a Simulation & Modeling Laboratory as such the following equipment has been ordered which includes:

- Sun Ultra 10 Workstation - HP workstation - Window NT Workstation - Window NT Server - HP Printer - HP scanner etc.

We would appreciate if you kindly share your experiences on the utilization of various software interconnected with your equipment or your institution is facilitated with such system. This will enable us to select specific software, which indeed shall give us an idea of their reliability to envisage the capital amount in procurement of the essentially desirable software.

Awaiting to receiving an early response & to entered your cooperation with the NED University.

Imran Majeed Assistant Professor Mechanical engineering Department NED University of Engineering & Technology Karachi, Pakistan.

(1). What kind of software? (2). If you are interested in commercial CFD codes, then, you can visit the vendor's web pages quite easily and directly from Internet. (2). From there, you should be able to get more information about the required system software.

If you need the benchmark test document of commercial cfd codes, I can give you for your convinient.

This paper written by Japanese and published in the society of Japanese automotive engineering.

I made a document file with this paper.

The 11 commercial cfd code, which were sells in Japan now, vendors were attend this pjt.

Regards,

turboguy

Hi Mehdi, its me again.

I've found the state of the art book on CFD :

"COMPUTATIONAL METHODS FOR FLUID DYNAMICS" Second Edition, Authors : J.H. Ferziger & M. Peric Published by Springer. .