CFD Online Discussion Forums

CFD Online Discussion Forums (
-   CFX (
-   -   combined air flow and heat ransfer during welding (

Benfa August 25, 2012 04:25

combined air flow and heat ransfer during welding

I would like to get some suggestions on the following simulation:
There will be a room that has an inlet and outlet for air. This will be done so that dust and heat can leave trough the outlet (mass flow known). There will be welding of bigger part. The welding jet has an total init. pressure of 12 bar and will be ignited so that the temperature of the jet will be around 3000C. Depending on the welding time the jet will introduce heat into the domain (mainly convective heat as I think). But it will be very computational expensive to resolve the jet (shock structures + combustion) itself. So it will not be possible to calculate the jet from first principles. The jet will be importent because of the convective heat source and the jet flow that will hit on a wall where it will be deflected. What would be the easiest steps to do simulation on that problem? Define a convective heat flux source with equivalent mass flux oder velocity of the jet?

thanks in advance

ghorrocks August 25, 2012 06:43

I hate to say it, but you appear to be approaching this simulation completely the wrong way. Rather than saying "I have all this complex physics happening, that is going to be really tricky to model", how about saying "My purpose for doing this simulation is .... (whatever the purpose is)". Are you trying to optimise the welding process? Or check if the welding poses a health and safety risk? Or check if the ventilation of the room is adequate?

When you know why you are doing a simulation you can then work out what is significant and what is not. And at the moment I have no idea what you are trying to do.

Benfa August 25, 2012 07:11


sorry that I did not clearify the purpose of the simulation. The purpose of the simulation is to see the influence of the welding jet and its generated heat on the venting of the room. It is assumed that dust that is created during the welding needs to be transported to the specified venting outlet. Maybe there will be recitculation zones that keep some of the dust. The transport will be influenced by the room venting and the high temperature and velocity welding jet. So you wil have a force convetion by the venting and the jet and a free convection cause by the generated heat in the room. The second question will be the surface and air temperature at some points in the room to ensure that there will be no hot spots. The welding will take about 4 min. So the simulation will be transient. The simulation can be verified on an existing 1:1 Experiment. If we get a model that can predict the general flow patern and the transport of the dust and an realistic temperature distribution we will be able to do simulations including other objects into the room that could be necessary for special welding. So we will be able to simulate what if scenarios and see if we expect problems with the dust transport or to high temperatures. So we could initiate an optimization (mass flow increase for venting, introduce different outlets for the heat and the dust, vaary geometrical effects). I hope this helps to clarify the purpose of the simulation. if not feel free to ask!

ghorrocks August 25, 2012 08:13

Now we know what you are looking for we can focus the simulation on that.

At a simple level the welding jet is a heat source. The next level of complexity is a heat source with hot air entering at a representative velocity to form a jet. And more complex still is a model with combustion to get the species, velocities and heat generated even closer.

I suspect the middle level complexity (a hot air jet) will be close enough, at least to start with. It should capture most of what you need for an idea of what is going on.

Don't forget the welded metal will probably be a signficiant heat source. And with these temperatures radiation will be significant so take that into account.

The welding dust could be modelled either as eularian particles or lagrangian particle tracking. In this application either is probably OK, so I would probably go with Eularian as it is easier to get accurate compared to Langrangian.

Just my comments based on your quick description - let me know if I am on the wrong track.

Benfa August 25, 2012 12:38

Thank you!

I also thought about a hot air jet. But I am not sure about how to approximate the hot jet because for the welding the air has an inital total pressure of about 12 bar. the ambient pressure will be 1 atm. Therefore the jet itself will be shocked and will be highly compressible including Mach >1 and shock structures. From the experiment there is known that the flame temperature is about 3000C. Following ideas:

1) introduce an inlet in the "room" domain where at least the correct mass flux for a shocked nozzle will be given. the inlet temperature would be the 3000C

2) introduce an inlet in the room domain where an average velocity of the jet will be set. But there will be the problem that mainly the jet is supersonic. If I want to introduce a supersonic velocity I will get the complex structure. So hopefully the "average velocity" over a specific volume will be subsonic.

I think the idea 1) will give a more correct result concerning the temperature field because the correct heat mass flux is introduced. Idea 2 could yield a more precise description of the flow field but will increase calculation complexity if supersonic jet will be modeled.

Does anyone have an better idea how to model the jet?

ghorrocks August 25, 2012 20:59

CFX can handle supersonic flow so if that is what you get you should be able to model it.

You may not need to model the complex supersonic component of the jet as from what you describe it sounds like it is small and would dissipate into a subsonic but larger jet, and this is the feature which interacts with the main room flow field. But whether this is a valid approach will, as always, depend on the simulation and what you want to get out of it.

All times are GMT -4. The time now is 11:44.