Thermal Analysis to Simulate Welding Heat Generation

[jdesouza]

I want to weld two parts together on a rotary tig-welder, and one of the components will have a bonded strain gage on it, which can see up to 500 degrees F for a short time before it starts to degrade.

I wish to conduct a thermal analysis to see what the temperature of the part would be after the weld is complete (1 revolution). I should be able to get all the welding parameters from our welder's weld schedule, but before I do that, does anyone know if this type of simulation can be conducted? I am very new to Thermal analysis, but we bought the tool so I thought this would be a good first experiment.

~Jeff~

SolidWorks 2006 SP3.1 on WinXP SP2
Pentium(R)4 CPU 3.60GHz 1.00GB of RAM

 

[Mech151]

This is a transient thermal analysis. This class of problem can be handled with Cosmos Professional or above (which it sounds like you have). This is a pretty straight forward problem if you just want the equilibrium temperature after welding, and can neglect heat loss to convection (i.e., the process is relatively quick). In fact, it would probably be quicker to just do a hand calc and divide the thermal mass by the energy input to come up with a temperature change.

If you are interested in performing a transient thermal analysis, it will be a little more complicated. Please confirm your goals, as well as the size of your part (how many elements, etc).

 

[jdesouza]

Mech151,

Thanks for the response. The way I see it, if I can do a simple problem first I will be good to go with more complicated geometry.

All I wish to know how to do is determine the equilibrium temperature as you stated using a transient thermal analysis on two cylinders of 1" diameter x .25" height stacked on top of each other. Would you be able to explain how such a model could be set up in Cosmos Professional, keeping in mind that the power source for a tig welding operation would rotate around the cylinders for one revolution.

~Jeff~

SolidWorks 2006 SP3.1 on WinXP SP2
Pentium(R)4 CPU 3.60GHz 1.00GB of RAM

 

[Mech151]

I have not used Cosmos professional yet, but what you need to do is apply a thermal load.

I am guessing that you do not have any data as far as temperature or energy input. One idea would be to model the weld and the adjacent parts. Place the weld at a set temperature (you should be able to find this information, or just use the melting temperature plus some nominal amount). Place a 'prefectly insulated' boundary on all external faces and let the parts come to thermal equilibrium.

If you wanted to check your results with measured values, you may want to consider changing the perfect insulation to allow convective heat transfer...

Sorry that I cannot be more specific, as my thermal analysis have been performed using another software.