I will have a project in which I need to model a free surface flow. The project is on laser cladding, in which laser melts the substrate and a melt pool forms. Meanwhile metal powder particles (maybe solid or liquid) impinge the melt pool. Surface tension is very significant and is caused by high temperature gradients. I need to model both free surface flow and droplets (or solid particles). You can see that phase change (melting and solidification) is involved in the process and the fluid flow is coupled with heat transfer.

I would be grateful for advice on suitable commercial software packages. It is better if the software company provides good training and user support because I am new to this field.

I would be grateful for any input.

Eric Jiang

I think you need to build your own code for this for example in fortran

But using commercial software is the requirement of the company that gave the project to my group. can FLOW3D or CFD-ACE handle this task?

http://www.flow3d.com/

At worst it'll be a very good start. Call 'em and ask.

http://www.flow3d.com/

otd, thank you for your advice!

i never saw in my life any serious code which is capable to deal with 3d free-surface flows.

developing free-surface codes itself is a state-of-the- art research.

so you don't need bother yourself asking this company that company.

kenn: Really? But I need to find a better one which can handle 3d free surface flow as well as other aspects.

We use Flow3D and have validated it successfully on a number of simulations involving FS flows and have obtained excellent agreement.

...Highly recommended

Hi, TITAN: What process did you simulate with FLOW3D? Could you tell me some of your papers on FLOW3D? According to your experience, can FLOW3D handle all the aspects I mentioned? Thank you in advance!

Eric

Dear Eric,

Please take "Mr kenn" comments with a grain of salt. It seem the comment is not well informed about the progress of modeling free surface flows over the last 20 years.

Please feel free to go to

www.ansys.com and check their ANSYS CFX, ANSYS Fluent packages. Both can do free surface flow, and check which one meet your needs.

www.flow3d.com and you will see other package focused on free surface flows with a different approach.

www.cd-adapco.com is also another package that could do what you need.

It is up to you to contact the vendors, and get a demonstration and make an informed decision.

Good luck,

Opaque

Eric,

ur problem is a multi-physics and covering all aspects is itself an state-of-the-art work (i do not know your work is a goal oriented application or a research, what is it?).

Of cource some commersial packages could model multi-physics phenomena, but "this option" of them is almost "young" and not mature. So i guess that upon selection of a commersial psackages you face some limitations. And need consultant of their developer. So take care on your selection and consider their service too.

The best selection could be based on work with full-work demo as opaque suggested.

About fluent, to my kbnwledge (ver 6.0) free surface flow solver could not link to solidification solver.

In my opinion best solution is to work with an source code (open source or commersial), but in this case you need CFD knowledge or a CFD expert collaborator. In this regard "Telluride" code (due to: D. B. Kothe et. al.) is the best one i know, see their web for more details.

I have also some other advices, in the case of interest contact me off-line : rohtav@gmail.com

Are you new to just CFD or to fluid dynamics as well? Even if you are very familiar with the physics associated with your problem and know nothing about CFD, it will still take you a bit of time to become proficient enough with a commercial code (or any code written by someone other than yourself) so that you can have confidence in your results for a problem that contains a lot of features. If you are not very comfortable with the physics, then your task becomes much more difficult because you may not be able to recognize any suspicious behavior in the solution.

Be wary of commercial vendors. They will all tell you that their code can solve your problem. I'm not saying they are lying. But these guys tend to significantly understate the amount of time and level of expertise required to get the result you are interested in for these types of models.

opaque, rt and Jake: Thank you very very much for your advice! rt: My work is a goal oriented application. I will contact you off-line.