Hi Randy
This contraption is just meant for people that accidentally define their material coordinate system in the wrong direction.
If you want your material to be oriented along the x-direction, the "1"direction must be the local x-direction as in this picture. If you happen to want the 1-direction to coincide with the local z, feel free to change the selection.
Hope this helps
Erik
Agnes: The problem is not in cylinders, but in cones.
Mika:
You could try adding a manly pink annotation, but you have to re-enter the modal max and value yourself
Erik
Other option is to change the background scheme of the Results window (from Options dialog). Black color of annotations should change to a contrast color.
I'm stupid. :/
I looked all over. Tried changing all the possibly relevant colors in the System colors menu. Even changed the "Letter" setting about a million times. All the other texts changed color, but the tags remained black. I gave it one more go, and realized that if I change the "Letter" color, the existing tags don't automatically change color. All new tags, however, do.
I could swear I've tried this before, but apparently not. Problem solved.
Mika
At least now you can mark your find as the right answer. Somebody giving the right answers can't be stupid!
Ok, just a little slow then...
Bob,
It's a constant and presumably hardwired.
At least, that's what I have assumed.
I get this message at random times even with sensible material properties.
It seems to occur most often when taking an existing material as a starting point, copying it and then editing the copy.
I may only change the E and nu slightly but it insists that this parameter PTC_TENSILE_YIELD_STRESS is not defined.
So I give up and define the material from scratch and the s/w is happy.
I have been meaning to file a call for a while.
It even happens now and then when you define a new material. No idea what to do about it.
Dear colleagues,
Does any one of you have any "suggested technique"-document, or similar, as for how to set up a stress intensity calculation (i.e. "crack") in Creo Simulate?
I thought I had set up an analysis correctly, but apparently it's not as straight-forward as I had thought. Can't find anything useful in support database... How does one find the "suggested technique"-documents these days?
B.R. Mats Lindqvist
Creo Simulate can't model heat transfer like this. As Charles writes, you enter a "heat transfer coefficient" where heat transfer is proportional to temp difference between surface and surrounding environment. I.e. this "heat transfer coefficient" is an input to, not a result from the analysis.
FloEFD is a CAD-embedded heat/flow simulation package, that can handle radiation, convection, conduction. To my understanding it is a quite competent software to model the coupled flow/heat transfer problem.
https://www.mentor.com/products/mechanical/floefd/floefd-creo/
Just general ramblings if any use at all ...
It is difficult enough trying to determine the correct loads for static analyses. Determining heat transfer coefficients, emissivities and thermal resistances as inputs is very difficult. Yes, using CFD helps (another bag of assumptions) and so does a thermal imaging camera and other experimentation to see how good your estimates for inputs are to calibrate the model better for next time.
In the 'old days' one would use the convection condition in Mechanica to iteratively approximate a radiation load (a technique I think should only work properly for spherical problems,,, but we ignored that otherwise we couldn't do anything). In Simulate we cannot have co-existing (on the same geometry) convection and radiation loads and I wonder whether in the background the convection condition is still used like this but now the s/w deals with the iterating on your behalf; which is good.
A lot of studies are to find out what happens/include the effects of something remote being heated by a radiator. For example, an electronic component that dissipates power in an enclosure; how hot will the case get? We can't do this in Simulate unless we estimate the heat load on the case and input it directly into the case. Simulate radiation doesn't travel through space though an emissivity can be a function of temperature.
We end up being pragmatic. Thermal problems often become a case of 'bounding' likely ranges of answers by tinkering with the input coefficients because we can't find the emissivity for a particular paint for example. So we take a D.O.E kind of approach and amongst the variables try e=0.1 then e=1.0 and see what happens. We find that a lot of the time it is sinking the heat by conduction that overwhelms the other heat transfer mechanisms. It is then down to how good the thermal contact is between components ... it is usually pretty poor.
I would like to be able to shine the sun at the tarmac on a zero wind day and see how hot it gets inside something mounted under a wing. But we always want more for our money.
I agree Charles. How I wish our customers could understand the level of uncertainty involved in heat transfer problems,(or structural problems for that matter) that you need testing to verify input parameters etc. I recall from my graduate course in materials science that the heat transfer coefficient (HTC) during quenching of steel in water, for example, can vary by several orders of magnitude (=several powers of 10), depending on if you get steam bubbles close to the steel surface or not, if you stir or not during quencing etc.
Thanks Erik. I have the material orientation correct but I got a message saying orientation is ignored for isotropic material. Would using the same young's modulus for E1, E2, and E3 give me accurate results? I only have a modulus of 10400 ksi for my material.
Randy
Hi Randy
If you do that, you don't have to do that. To be more clear, if it is correct that E1=E2=E3, than the orientation doesn't matter and you can use isotropic material anbd you don't need a material coordinate system. In stead, you can use that coordinate system to show stresses (σxx ,σyy, σzz ) where x and y and z are according to that coordinate system.
Erik
my mistake indeed,
Well, the SPR is still open and with no news on datecodes for a fix. 'Creo 2 fix planned in mxxx' and 'Creo 3 -' (unknown)
In a previous discussion I presented a contact example to compare ANSYS with Creo. In that case, both results were credible, and Creo's results was somewhat more realistic, but that was with default ANSYS mesh, and no particular effort to refine mesh/convergence in either model.
I have here made a similar example, a pre-loaded bolt that clamps two parts together. The pre-load is generated with a coeff. of thermal expansion= 0.001 for the bolt, and a -1 temp load that makes the bolt contract. The aluminium plates have coeff. of thermal expansion = 0. The bolt is made of stainless steel.
The results here differ more. Creo's contact area is smaller but has higher max contact pressure. ANSYS was run with frictionless contacts and 1 and 10 in normal stiffness factor in two different runs. The ANSYS solution seem indifferent to mesh refinement, i.e. the mesh should be OK. As far as I can see, the material properties are identical. I'm not too experienced with ANSYS, so I struggle with the UI. The Creo solution (SPA, about 7000 elements) took 2+ hours, ANSYS took about 2 minutes. I also ran the Creo model as a "quick check" which was much faster, and did not change the results considerably. Creo's mesh is slightly refined compared to default mesh settings.
ANSYS max contact pressure is about 8 MPa, Creo: 14 MPa.
Which solution is correct?
I found the error! The contacts between bolt heads and plates were "bonded" in the Creo model. With frictionless contacts, the results are more or less idenetical with the ANSYS model. A bit surprising that such a seemingly small difference in the model can produce a significant difference in results. Since neither "bonded" nor "frictionless" contacts in this case.represents reality, the correct results should be somewhere between these solutions.
Best regards...
After correcting the error in the Creo model, how long did it run before a solution was realized?