Is it possible to simulate surface shear stresses on metal contacting surfaces

[Anonymous]

I am trying to simulate surface shear stresses on a part that is sandwiched between two surfaces. Part is under compressive force and is also being pushed out of the contacting surfaces. I have tried to simulate by applying compressive forces and full body push out force. When I solve it I get shear stress values at edges rather than on the surface. What I am trying to establish is to increase the contact surface to decrease shear stress and establish what contact surface I can work with. This is shown as blue region in attached image

[John_Holtz]

Hi @Anonymous . Welcome to Fusion simulation.

Can you attach an archive of your model? ("File > Export" and attach the .f3d file.) There are many questions that will be answered easily by inspecting the model but difficult to ask/answer by forum post. For example,

1. How is the model constrained?
2. How is the contact defined between the parts?
3. What type of analysis? (Ok, this question is easy to answer by posting to the forum. 🙂)
4. What are the results!? (My guess is the stress is high at the edge and low everywhere else. That result could be accurate, or it could be an artifact of the analysis in which case the high result just needs to be ignored.)
5. There are probably other questions that I have not thought to ask.

[Anonymous]

1. How is the model constrained? I am using two fixed constrained shown in the model
2. How is the contact defined between the parts? It is sliding contact
3. What type of analysis? (Ok, this question is easy to answer by posting to the forum. 

   Static Stress

4. What are the results!? (My guess is the stress is high at the edge and low everywhere else. That result could be accurate, or it could be an artifact of the analysis in which case the high result just needs to be ignored.)
5. There are probably other questions that I have not thought to ask.

[John_Holtz]

Hi @Anonymous. Your setup does not make sense. Due to the problems with the setup, the solver is making corrections to the analysis, so the results do not represent anything that you want.

Here are the problems that I see.

1. Body 4 is fully constrained on the bottom face. Therefore, the force on the top is only compressing body 4. None of the load is being transferred to body 3.
2. Body 1 is fully constrained on the top face. Therefore, the force on the bottom is only compressing body 1. none of the load is being transferred to body 3.
3. Body 3 is using sliding contact with bodies 1 and 4. Sliding contact means the part is free to slide. Body 3 has a force in the Y direction that is trying to make it slide. Since it is free to slide, there is no static solution in real life. Body 3 should move an infinite distance! The facts that the displacement is only 0.3 mm and that the analysis did not fail are indications that the solver is changing the setup and solving a different setup.
4. Sliding contact is created only where the bodies are in contact at time 0. It does not matter how much compression you apply to the parts, the surface area in contact does not change.
5. This is not a problem yet, but eventually you will need a much finer mesh on the top of body 3. With the current coarse mesh, the contact area can only be the original 50% of the top area, 75%, or 100% of the top area. Is this too limiting? Will you be compressing the top so much that 75% or 100% of the face comes into contact? What about the increasing contact area on the bottom face?
6. I am also guessing that the force required to increase the contact on the top face of body 3 will yield the pointy edges on the bottom of body 3. You are not including any of those effects in the static analysis. (If you need to include yielding, then you need to be using a nonlinear static stress analysis and make sure the stress-strain input is accurate for your analysis.)

Based on my limited knowledge of what you need to include in the analysis, this would be my suggestions:

1. Eliminate body 1.
2. Constraint the 2 edges on the bottom of Body 3 in the vertical direction (Z) and the sliding direction (Y). We are going to assume that body 3 does not slide relative to body 1 (and therefore is not needed in the analysis).
3. Use bonded contact between the flat on body 3 and body 4. This implies the friction force is large enough to prevent sliding.
4. Apply the compression force 1 and sliding force.
5. Run the analysis. This will show the shear stress on body 3.
6. Based on the displacements, you can see how much of body 3 should come into contact with 
7. Change the model to include the "new" flat area of body 3 based on the previous result.
8. Repeat steps d through g, increasing the compression force 1 to see how much additional area comes into contact.

These steps probably need to be modified depending on what the first analysis shows. Let us know what you find out.

[Anonymous]

Hi John:

I am trying to simulate a piece of metal strip body 3 which is being gripped between body 1 & 4. During actual loading process axial force (y direction) is applied on body 3 either pulling it out of the grips or pushing it down. The idea was to assess amount of shear stresses generated during loading process and if there will be slippage meaning body 3 will move out of the grips. Hence I kept contacts as sliding as my understanding was that bonded contact are more for where contact surfaces do not move such as in the weld joints.

Thanks for your suggestions I will modify design based on that and see what results I get.

Regards;

Fahd

[John_Holtz · Accepted solution]

From a pure physics perspective, body 3 will not slide of the compressive force multiplied by the coefficient of friction is larger than axial force. The shear stress has nothing to do with this. No simulation is required!

Perhaps in reality, the sharp edges on the bottom of body 3 "dig into" body 1, and the amount of deformation prevents body 3 from moving. That would be a different phenomenon that will be hard to simulation in Fusion. Or maybe there is another physical effect that occurs in real life that I did not see or missed in the model.

If the shear stress is the real determination of whether the body slides or not, then the approach is to assume that is does not slide so that you can get a static solution. (If body 3 is free to slide, you will not get an accurate static solution.) Based on the results, you check if the assumption is valid (it does not slide) or invalid (it should slide). This gives the answer.

The steps would be:

• Include all 3 bodies.
• Constrain body 1 on the bottom in all 3 directions.
• Constrain body 4 in the horizontal directions (X and Y), assuming that something holds body 4 in real life. If body 4 moves with body 3 if and when it slides, then put no constraints on body 4.
• Define "separation no sliding" contact between the top of body 3 and the bottom of body 4. Define "separation no sliding" contact between the bottom of body 3 and the top of body 1. (This simulates the assumption: that body 3 does not slide.)
• Run the analysis.
• Do the results support the assumption that body 3 does not slide?