Thermal Stress Analysis problem with Cosmosworks

I've reduced my thermal stress analysis problem to the most basic scenario and the results are not reasonable.

I've taken a 1" x 1" x 1" cube of 1020 steel and applied 83 degrees f to all sides, put an immovable restraint on one of the verticies.

The results give me 1,700 psi of stress. There's really no appricable heat. Just 83 degress. How can it figure on this much stress?

My real world problem of a class 300 flange at 750 degress f., and one side restrained, give me material failure results. About 200,000 psi of thermal stress.

Is Cosmosworks reliable for thermal stress, or is there something basic I'm missing?

Pat

Reply to
throwaway
Loading thread data ...

The amount of stress induced depends on the amount of thermal expansion you are restricting on the restrained vertice.

The amount of thermal expansion is dependant on the thermal differential, ie at what temperature was the vertice restrained before applying heat. eg if the block was at room temp (approx 69 deg F) then the expansion would be due to a differential of 83-69=

14 degrees.

1700 psi could easily be the resulting stress level at the vertice.

---------------------------------------------- Posted with NewsLeecher v1.0 Final * Binary Usenet Leeching Made Easy *

formatting link

Reply to
Phil Evans

You don't say clearly how you applied the temperature boundary condition. Likewise you don't say how you have constrained the cube.

I have to ask what problem you are proposing CW answer? Is this setup as a steady state or a transient thermal problem? Since you are modeling a solid with solid elements what is the meaning of restraining just one node? You need at lease three non co-linear nodes restrained just to prevent free body motion. If you wish that restraint to be stress free you must do so in such a way as to allow expansion. Then there is the problem of restraining at a corner which can immediately induce a singularity and give high ficticious stresses.

If you simply want to apply a uniform temperature to the cube I think that is done during setup in properties somewhere.

TOP

Reply to
TOP

I ran into a similar problem: if things are how they were for me, basically the temperature was applied to only the faces. Cosmos defaults the interior of the cube is to 0 deg C so it induces a lot more stress. What you want is to apply the 83 deg temp to the body of the cube, not the faces.

Reply to
esobota

I'm not sure what happened to my message when I tried to reply. Maybe it just went to you, Pat?

I ran into a similar problem. Cosmos assigns a default temperature of

0 deg C to anything that doesn't have another temperature applied. When you apply a temperature to the faces, it still assumes the rest of the body is at 0 deg. What you want is to select the entire body from the SW feature manager, then assign the temperature to that.

See if that helps!

Reply to
esobota

PolyTech Forum website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.