Need Help Understanding Thermal Cycling

[GrizzleBone]

I am using Ansys to model a simple 2" beam welded to a 48" x 24" inch plate(Check Image). The beam is skip welded with 2" welds every 6". The inner surface of the beam is heated continuously to 900 degrees F while the air on the plate is cooled by air down to around 250 degrees F. It reaches a steady state condition. The material is 304L stainless steel with around 1/4" welds. The plate is fixed on all 4 edges. The thickness of the stainless steel is 0.06"

I have seen over 100 units in the field over 3 years that have not failed under 100s of cycles. My Ansys model shows stresses in the hundreds of thousands of psi. Could anyone give me some insight into what is happening? The Ansys model shows a failure but none of the real world units have failed.

Thanks

 

[PriamEngineering]

How do you know your model is correct? What is the max displacement and how does that compare to what you can measure/observe?
if you are looking to predict ther life at the welds you need to look at a specific weld fatigue code (BS7608)

regards

Priam

http://www.priamengineering.co.uk

 

[corus]

If the plate is fixed under thermal loading, then you're assuming that it has no thermal expansion at all at those edges. This would be very surprising and would no doubt lead to very high stresses. You don't show the temperature field but I would guess that for thin plate then the temperatures would be almost uniform through the thickness. If you have a high delta T then you'll get very high stresses too. On top of that you should really look at nominal stresses in the plate, away from the welds and not at peak values given by your model. Compare these values to limits in fatigue design codes, as Priam suggested.

Tata