Prestress Modal Analysis

[Shaumack]

I am trying to prestress the PZT stack for my ultrasonic horn assembly. The picture shows a sectional view of the horn cut along the symmetry plane. PZT stacksaregonna be prestressed by a bolt and nut. Bolt is connected to a connecting body which is turn is connected to ultrasonic horn. My interest about adding a prestress is to use it to run the prestress modal analysis to have stiffening effect. I would like to add prestress of 100 MPa on the stack
<?:namespace prefix = o ns = "urn:schemas-microsoft-comfficeffice" /><O></O>
Here are the few things that I have tried but nothing seems to work right. I tried adding thermal preload using Global Temperature Load after changing the material property for the bolt to add CTE and then successively removed the CTE for the other members to mimic the preload in bolt due to shrinkage and hence prestress in stack.

When running the static analysis using Global Temperature load (GTL), <B style="mso-bidi-font-weight: normal">I constrained the entire circumferential surface for the connecting body.[/B] This was the only constraint I used. I could not use the contact elements as prestress modal analysis does not allow using them. So I ran this setup without contact elements. I adjusted GTL till I got a stress of 100 MPa in my stacks. Then I used these results for my prestress modal analysis to run <B style="mso-bidi-font-weight: normal">unconstrained with rigid mode search[/B] modal analysis but to find that the modal frequency did not changed much. Rather it decreased by 2 units. I expect frequency to go up by atleast 50-60 Hz due to stiffening of stack. I am not sure nowif I shouldtrust prestress modal analysis in Pro/MECHANICA or I am doing somethingwrong. Anyone ran with these issues in Pro/M

<O></O>
Can someone please suggest me how would one go about prestressing stack to use the results in prestress modal analysis? Does it matter with Pro/M if you will use a constraint to run a static analysis and then do not use that same constraint for prestress modal analysis?

<O></O>
Thanks in advance.
- Shaumack



Edited by: Shaumack

 

[KurtStark]

Hi Shaumack


It sounds starnge, but according to theory there is no change of the modal frequency by adding prestress, as long as all your components are linear.
Imagine a simple system consisting of a mass and a spring: the resonance frequency depends of the spring rate and the mass. While the spring rate is constant, it doesn't matter whether the weight of the mass iscompressing orextending the spring.
Without contact Pro/E is calculating linear, that's why I'm not surprised, that you get no difference in the result.


Regards, Kurt

 

[Shaumack]

Thanks Kurt. I understand what you are trying to convey and it makes perfect sense.


In the meantime, I tried running the same analysis by adding number of springs to mimic contact region to take in account for the non linearity and rerunning the premodal analysis but i m not satisifiedwith the results.Resonant frequency dropped by 200 HZ. Iam not sure how could one modelmy situation in Pro/M and get the required result
Edited by: Shaumack

 

[KurtStark]

Even if you add springs, the model stays linear.
I assume by adding springs, you reduced the stiffness of the system which results in lower resonance frequency.


But even if you could add contacts, the system would not get stiffer. Now you have full contact between the components, which is the maximum stiffness. If parts touch each other, Pro/M assumes them to be fully connected (as welded on the whole touching surface).
Two parts touching each other in reality have low touching surfac because the surface is not perfect.By applying pretensionyou enlarging the touching surface resulting in higher stiffness.


Simulation never is the same as reality. It is a question of effort, how close you get.