Induction heat treatment simulation

[sherpa74]

Hello everybody.


I would like to know if someone has experience doing analisys in treated shafts. I am trying to simulate a induction treated area, with different material structure, different hardness...


I will appreciate any idea or comment that would help me on this.

 

[Luis Aguirre]

Hello Daniel,


I Pro/MECHANICA you can create a volume regionwithin a volume. This option is under the "feature" option in MECHANICA. Once you create this volume now you can assign different material properties to each of the volumes.





Luis

 

[lbotez]

Luis,


I don't think it is that simple.Indeed, you can create volume regions with different mat. properties. But can Pro/Mechanica handle anything but elastic linear materials? The issue here will be plastic deformations in the core that will produce high stress in the surface layer. Therefore I don't think Pro/Mechanica is the right tool in this case. Daniel, I think you should look for a more envolved FEA package (ANSYS, Nastran ...) that can handlenonlinear materials.

 

[Luis Aguirre]

Lucian,


You are right! if he is simulating the process by which the material is treated you is a nonlinear analysis. However, if he wants to simulated the area (volume) that was treated which has different material properties then this should be still a linear analysis. Unless the load yields the material then it becomes a nonlinear analysis. My only concern about this analysis is thefact thatonly part of the shaft was treated. Therefore, some residual stress in the boundaries of the treatment were most likely generated. Consequently, depending on the loading condition this residual stress may increase or decrease the yielding propertiesof the material.





Luis



Edited by: Luis Aguirre

 

[lbotez]

Luis,


Residual stresses will usually increase the material strength, as they are in most cases compresive. The only problem with residual stresses is that there's only one way to find them: measurements, and I heard they're not easyto doeither.


On the other hand if you are not concerned that the soft core will yield, than way bother to simulate heat treat since it doesn't have a huge influence on the Ex and NUXY? The only thing that you will have to do after the analisys is solved, is to make sure you compare the stresses to the correct limit according totheir locations (core/surface layer).

 

[sherpa74]

Thanks for the info.


What I would like to know is the max shear stress I have in the end of the spline, and also in the end of the heat treatment area. The main problem could be, as you have wrote before, is the hardness value ramp that the induction creates in the last 5 mm, and how proMechanica can manage that, or if proMechanica can do that.....


I tried with volumes, but it creates a break line between different materials, and the solution I obtained was not so clear and believable.


Talking about residual stress... we have the experience that induction always create compresive residual stresses. If you grind the area before the heat treatment then you could have some problems...and they are not so easy to solve.


Thanks for the replies.

 

[Luis Aguirre]

Hello Guys,


I agree that residual stress are compressive and they usually increase the material strength but only when the external load in the body creates tensile stresses. An example that comes to mindis the material yielding of a pressure vessel that will create residual stress. I realize that MECHANICAmay not bethe right tool to use, for the problem you have Daniel. But ifMECHANICA is the only FEAprogram you have available then assumption and safety factors, must be included in your model.


The following comments are purely academical and only reflect my understanding of the problemgiven the information so far provided. The external load apply to the shaft will create shear Stress ( I assume this is a torsional shaft). If we combinethe residualcompressive stresses with the shear stress of the external loadthe result will be an increaseon themaximum shear stress in the shaft or no significant change will be observed. if the residual stress magnitude is grater in magnitude that the torsion shaft shear stress the Maximum principal shear stress will increase. In the other hand, if the torsional shear stress is much greater than the residual stress the maximum shear stress does not change by much.Therefore, forthe state of stress provided by Daniel's problemthe residual stress does not help but may actually have a decremental effect. Hereare someexamples:


Sxx= Residual Stress, Sxy= Shear torsional stress


|Sxx|>|Sxy|


let's assume that the residual stress is Sx=-1000 psi and the shear do to the torsion of the shaft is Sxy=500 psi the maximum shear stress for this stress combination will be 707 psi. In this case the maximum principal shear stress is increase from 500 psi to 707 psi. There was an increase of 41%.


View attachment 484


let's reverse the numbers |Sxx|<|Sxy| Sxx= 500 psi and Sxy=-1000 psi.


View attachment 485


In this case the Maximum Principal shear increased from Sxy=1000 psi to Sxymax=1030 psi. The increase was less than 1% in the Maximum Shear Stress. Therefore, what these two example show is how residual stress may increase or may not have much of an effect in the shaft Maximum shear stress.


Daniel you mention that the results you got from the model using the two volumes were not very good. Could you elaborate more about what you meant with that?


Luis


By the way if you want to try the Morh's circle program in the web here is the URL:


http://web.umr.edu/~medialab/fipse/preview/Philpot/mohr_stre ss/a_draw.htm






Edited by: Luis Aguirre