Interference fit 2D
Answers
-
I have updated the attached .fem file to include plane strain elements on the properties. Thank you.
0 -
Hi, try connecting the web and the lug using CGAPG elements. You can force interference fit by assigning a negative initial gap opening on pgap. Finally run an NLSTAT analysis.
0 -
Hi @mvass
If you want to create interference fit by specifying a value use PCONT property and provide a clearance value. Reference this created PCONT in CONTACT card.
0 -
Altair Forum User said:
Hi, try connecting the web and the lug using CGAPG elements. You can force interference fit by assigning a negative initial gap opening on pgap. Finally run an NLSTAT analysis.
Thank you. That did the trick, but partially... at least I have a solution. I'll remesh and inform if I have any better results.
Altair Forum User said:Hi @mvass
If you want to create interference fit by specifying a value use PCONT property and provide a clearance value. Reference this created PCONT in CONTACT card.
Thanks for your suggestion. That was my initial plan and I have tried PCONT with several options. Unfortunately, none of these worked and that's why I posted my .fem file without the contact components, hoping that someone will have a look. Now, with reference to the answer given by Girish, I'll work again on the problem, comparing the results with the pin attached to the lug with gap elements and the other pin with pcomp. Should I have any successful runs I'll post on the topic created in the optistruct/solver section.
Thank you both for your contributions.
0 -
Hi,
Can you share the complete pre-processed model with PCONT, maybe we can check for a solution?
You can use the dropbox link in my signature to share the model file.
0 -
Altair Forum User said:
Hi,
Can you share the complete pre-processed model with PCONT, maybe we can check for a solution?
You can use the dropbox link in my signature to share the model file.
Please find attached the requested .fem file. I have tried a number of PCONT options and I am not sure if the one included in the file is the working one, but in any case I either received a message for an un-converged solution, or a result that doesn't make sense.
0 -
0
-
Altair Forum User said:
Hi,
Can you try this model?
First, thank you for you effort. Your suggested model works (i.e. converges to a solution) but the results (for example stress field) do not present a meaningful distribution for a press fit. I had similar results during my simulation runs.
I would expect a picture of stress field like the one attached below, unless of course I am doing something wrong with HV...
Thank you.
PS: In a past post of yours you have prepared an excellent example of press fit using 3D elements. Have a look on that.
<?xml version="1.0" encoding="UTF-8"?>
0 -
Below you will find some pics from my last effort to have a working solution to the problem. As you will see I have created contact surfaces between the two components to define the contact. Respective properties as seen on the attached pics. Please note that in my analysis the MORIENT option is set to 'NORM' not the one shown.
(By the way, why 'OVERLAP' option returns an error that the slave nodes are exactly on the master surface when this is exactly what is happening?)
However no solution was obtained, thus I had to run a normal modes analysis with debug option. Attached is a picture of the results. It seems that a displacement on the z (3) axis is introduced although this is a plane strain problem. Why? So, I decided to constraint the out of plane movement of all nodes and re-run the simulation. Result: Converges. But: No results are seen (everything zero!). Any explanations?
<?xml version="1.0" encoding="UTF-8"?>
<?xml version="1.0" encoding="UTF-8"?>
<?xml version="1.0" encoding="UTF-8"?>
<?xml version="1.0" encoding="UTF-8"?>
0 -
Hi,
The error pops up when there is a negative clearance value. I think 3D elements can give good results over 2D.
0 -
Following this conversation and for the benefit of all interested users, I would like to post my concluding remarks below:
1) The only solution that gave reasonable results (may require further investigation) was the one where gap elements were used with the appropriate property. The working .fem file is attached below.
2) Using contact between the two components (contact surface with pcont property) required an extra constraint on the z axis (3) to run and resulted in infinitesimal displacements. In general, lots of runs were made with contact between the two components and none of them resulted in anything meaningful.
Thank you all for your participation.
File with gap elements:
0
Altair Employee
