Learning Forum

[thornyak]

🛈 This post originally contained file attachments which have been removed in compliance with the updated Ansys Learning Forum Terms & Conditions

Hello,

I modeled a pretensioned bolted connection in AIM 2020 R1 and have some comments and questions regarding on the workflow/results.

Mostly I have shell elements with share topology in the model.

First, I idealised the bolt's shanks with solid elements and applied bonded contact between the shank and hole edges and frictional contact between the plates.

Second, I idealised the bolt's shanks with beam elements and applied share topology between shells and beams. I deleted the bolt holes because the connection cannot be created between a beam’s vertex and a hole’s edge, see my previous post (still waiting for answers): https://discoveryforum.ansys.com/t/y4hf53g/contact-between-vertex-and-edge-not-working-in-aim with the beam's vertex-to-edge

On the lug holes, I applied remote forces inducing bending moment (see 1st figure). I applied inertia load for taking into account self-weight. Then I run a multi-step analysis, bolt pretension was only applied in the 1st loadstep (see 2nd figure). Large Deflection was set to ON.

A. Comments/questions to model with bolts as solid elements:
A1) I applied the pretension forces on every solid bolts easily with the help of „Model as: One bolt per location” (see 3rd figure). The pretension forces displayed well on the screen on each bolt position.

A2) I generated a Working Load Contour Plot. I checked the correctness of the application of the pretension forces: the force values in the Summary panel on the left are close to the pretension value (135 kN), see 4th figure. I think the display of the bolt forces are not the best, there are only small coloured dots at every bolt position. How can I measure the bolt foce in one position?

A3) If I generate a Vector result for one bolt position to check the bolt force, I don not understand the result (see 5th figure). Can you explain how to use this Vector result? Or which result setting calculates the bolt forces?

A4) After solving the model and displaying the results for the first time, the contour plot results appearing very-very slow. This is a little model and I have a workstation PC fast enough, but the Equivalent stress and Displacement Contour Plots appear only in a couple of seconds on my screen. After these are generated once and the second time I retrieve these, there is no problem with the displaying speed. How to solve this issue? Can I somehow display the shells without thicknesses?  There is thick shell representation in the results right now (see 6th figure) and I think this decreasing the displaying speed.

B. Comments/questions to model with bolts as beam elements:
B1) A little strange behaviour during pretension defining: Pretension force need to be multiplied by the number of bolts (26*135kN) for correct results, see 7th figure. The pretension forces do not displayed appropriately on the screen on each bolt position.

B2) Working Load Contour Plot displays nothing (see 8th figure). Summary panel shows only circa the correct overall pretension values: 3161 kN instead of 26*135 kN = 3510 kN. Can you explain this please?

B3) Generated Axial Force results look good and I can measure the forces for all positions  (See 9th figure). Beam representation looks good as well. Maybe it would be better to display the beam and it’s mesh already under “Geometry” or “Mesh” task instead of displaying this only under “Results” task.

B4) Displaying contour plot results is still very slow.

Thank you in advance.

[Subashni Ravichandran]

Hello Tamas

I'm looking into this issue. I'll get back to you on this today

[Subashni Ravichandran]

Hello Tamas

Here are my responses to your queries/observations.

A2) You can check the minimum and max values from on the panel on the left side, Although as the contour or the vector display is on the mesh elements, unfortunately, there is no better display available there.

A3) The Force variable shows all the forces generated in the selected region. The Vector display shows the Force vector(arrow). You can alter the number of vectors displayed from the appearance options.

A4) I’m not sure what could be causing the slowdown. Initially, the solver may take some time to solve and display the results. I was able to run the files shared at a fairly good speed. I don’t think there is any way to turn off thick shell representation. I will need to check this with my team.

B1) With one bolt for all locations option the value entered for the axial load will be divided for all locations. So if 135kN is input it will be 135/x kN for x number of locations. So correct value for your simulation would be 26*135kN.

B2) I think that is a 0.1% inaccuracy due to meshing and other factors of FEA and it can be ignored. Although I can check back with my team on this.

B3) You can turn on/off mesh at any point during the workflow using the “Show mesh” option on the top right.
 

[thornyak]

Hello Subashni

 

Subashni said:
A3) The Force variable shows all the forces generated in the selected region. The Vector display shows the Force vector(arrow). You can alter the number of vectors displayed from the appearance options.

 Okay, but this Vector result displays a lot of force vector and doesn't show the exact bolt force in the selected solid. Maybe it shows the summarized nodal forces per each node, that's why more then one vector appears. Can I somehow calculate (generate result) for one solid-bolt position only?

 

Subashni said:
A4) I’m not sure what could be causing the slowdown. Initially, the solver may take some time to solve and display the results. I was able to run the files shared at a fairly good speed. I don’t think there is any way to turn off thick shell representation. I will need to check this with my team.

 I will upload a video/gif how slow is the result generation on my PC.

Subashni said:
B1) With one bolt for all locations option the value entered for the axial load will be divided for all locations. So if 135kN is input it will be 135/x kN for x number of locations. So correct value for your simulation would be 26*135kN.

 Yes, I set 26*135kN as pretension with "One bolt for all locations" in case of BEAM-bolts. If I set 135kN and "One bolt per loaction" then the analysis fails to run. 

In the model containing SOLID-bolts, it works vica versa. I need to set 135 kN and "One bolt per loaction" to get correct results.

It is strange, that in the BEAM-bolts model the pretension forces do not display on the screen in each position as in the SOLID-bolts model. See previously attached pictures.

Subashni said:
B2) I think that is a 0.1% inaccuracy due to meshing and other factors of FEA and it can be ignored. Although I can check back with my team on this.

 This is not 0.1% deviation but 10%. 3161kN divided by 3510kN equals 0.9.

 

B3) I mean that one beam only appears with its line representation if I'm inside the Geometry/Mesh task, and I only see beams with cross section inside the Results task.

[thornyak]

Hello Subashni

I attached two gif files about the slowdown: one for the SOLID-bolt model and one for the BEAM-bolt model. As you can see if I select one result for the very first time, the displaying is very-very slow. For the second time the displaying is faster.

My configuration:

OS: Windows 10 Pro, Verision 20H2, Build 19042.964

CPU: Dual Intel(R) Xeon(R) CPU E5-2680 0 @ 2.70GHz

GPU: NVIDIA Quadro K5000 with up-to-date driver version (Version 466.11, Date 04.24.2021)

RAM: 64GB

SSD: Crucial MX300 2x525GB (RAID1)

You said  that you did not experience slowdown. Can you share your PC configuration and your AIM settings, please?

Thank you.

🛈 This post originally contained file attachments which have been removed in compliance with the updated Ansys Learning Forum Terms & Conditions

[thornyak]

Hello Subashni

Did you watch the GIFs and experience the same slowdown?