Stress Analysis on Hydraulic Press

[pg471939HMBM]

Hello,
I need to perform a simulation on the press of an extrusion machine for outsoles. The hydraulic cylinder in the press has a design issue, which causes the bolts on the sides holding the flange to break after a few cycles (mechanical fatigue). It is known that the flange bends when the mold is closed.

For the simulation, how should I apply the 200 bar pressure on the cylinder? On the other hand, how should the contacts between the bolts and the cylinder sleeve be modeled?
The model goes attached.
Thanks for the help.

[kacper.suchomski]

Hi

Is this your graduation project?

The strength of bolts is checked/calculated in a few minutes with an engineer's handbook on the knee.
It is not without reason that many models are simplified from bolts, because FEM simulations are not optimal for this; they are intended for the analysis of non-standardized elements.

If you absolutely have to do it in Inventor, that's fine, but generally speaking it's not the optimal path.

Paradoxically, for the needs of FEM analyses, neither joints (as you ask) nor complex structures are defined. You should quickly calculate the force (pressure x area) on a calculator, which you divide by the number of screws. And you should set such a load for the analysis of one screw. It will be 10x faster than creating a mesh and analyzing the entire assembly.

[pg471939HMBM]

Hello Kacper,

Yes, this is my graduation project. I'ts not only about this, because I'm reengeneering the machine, making a new eletronic system, aplying safety measures and other stuff (i'ts an old machine). I already made the calculations for the screw using the goodman criteria and the coefficient of safety was good (about 3). But the truth is that all the machines made like this one have this fatigue problem on the bolts, so I don't know how to proceed to verify this.

Thanks for the answer,

Francisco Silva.

[kacper.suchomski]

---

@pg471939HMBM wrote:

But the truth is that all the machines made like this one have this fatigue problem on the bolts, so I don't know how to proceed to verify this.

---

The first step should be to check whether you have selected the appropriate safety factors during calculations. Remember that the material is calculated differently under constant load than under variable load (e.g. kt, ktj)

As for Inventor - the model for analysis should still be simplified to the bare minimum of geometry. Only geometry that actually affects the force/reaction should be retained.

Another issue is that the standard analysis module in Inventor does not allow you to simulate variable loads, so if you want to play around with more dynamic scenarios, you should use Inventor Nastran.

[pg471939HMBM]

Goodman criteria for fatigue in bolts only considers traction, but in this case there's also bending on the cylinder joint. Maybe it's the combination of both that makes the bolt break. I don't know and I'm a bit lost because my teachers don't really help (looks like they don't know too 😒).

I tried to do the simulation using only the cylinder, using two constrains on the holes where the shafts connect. Then I need to make all the contacts. For bolts, I used grounded where they thread and slinding/no separation where the bolt head sits. Then I aplied the pre tension on bolts and pressure on cylinder. for the contacts between the cylinder rod and cylinder flange, and between cylinder flange and cylinder barrel, what sould I use? Do you think theres is a better way to simulate?

[kacper.suchomski]

One of the basic principles of using bolted connections is the fact that they cannot be used to transfer bending loads.

This is a rule from the area of ​​basic machine design.

Here you will find information about the contacts between components:
https://help.autodesk.com/view/INVNTOR/2025/ENU/?guid=GUID-4CAFC652-873B-4709-B107-D39A399FCE8D

Oh, and here's some constraints:
https://help.autodesk.com/view/INVNTOR/2025/ENU/?guid=GUID-2E36EC5A-8C83-43F0-A58E-41744C5325A0

[hollypapp65]

Which bolts broken?  The red one?

Which parts are fixed, which one moves?

but in this case there's also bending on the cylinder joint

---

 Bending?  Where?  Got photo of the broken parts?

I think teacher just want you to figure out what's wrong.

I have a few guesses.  Telling you will defeat the purpose of this assignment.

I built cylinder to lift  rail bridge.  Also tested up to 15000PSI.