What is the difference between the stress results and why is it important?

[WillemBouwer]

I have to run a FEA analysis on a nozzle check valve body DN700... Its been to a professional engineer and he has completed the study so I am trying to do it myself to get to know the software better... I'm using PTC CREO 2 mechanica as the software...

Attached find a picture of the valve body.

I have constaint the valve at the hole locations on the flanges, as pin constaints. Where the angular constraint is free and the axial constraint is fixed.

The valve is rated at 40 bar, therefor should be tested at 1.5 time working pressure so 60 bar... So the load I used was a pressure load on the surfaces of the entire inside of the body for full open condition, and on the surfaces up to the point where the disc should close the flow off for the fully closed condition.

I would like to analize the valve in both fully open and fully closed conditions. With the disc of the valve pressing against the diffuser and the body when fully open, and pressing against the centre piece and the body at fully closed position...

I have set the material to be SG42 Cast iron and meshed the body...

At fully open condition I get the Maximum Von Mises stress to be 224 MPa at the bolted area
At fully closed condition I et the Maximum Von Mises stress to be 221 MPa on the fins of the body...

However a picture of the professional engineers findings have been attached. He finds the Max Von Mises stress to be 312.5 MPa which differs from my analysis.

Now the question is: Why is his stress more, what am I not thinking about? Can anyone help please, thanks... It is complex to explain what I did but ask eny questions, haha...

 

[Vadar2012]

Since your value is lower, it's most likey just a case of your mesh size being too small at the areas of maximum stress. Try increasing the mesh size at the obvious areas.

 

[AlephZero]

It's impossible to guess what the difference is. The most likely two things to check are the lioads and the boundary conditions.

FWIW, in the left hand picture it looks likke the vanes supporting the central disk are bending axially. There is a stress gradient from one end to the other similar to a cantilever beam. In the right hand picture, they are doing something different.

Plotting the deformed shape of the model might show you if something crazy is happening.

I don't have any experience of this sort of pipework analysis, but it could make a lot of difference if you restrain only one flange rather than both flanges. Think about what each end the valve is attached to - for example a stiff pressure vessel or a relatively flexible pipe.