Linearization of nonlinear contacts

  • Last Post 20 April 2020
  • Topic Is Solved
Lukas14089 posted this 18 April 2020

I am looking for alternative methods to linearize nonlinear contacts for a bolted connection in Ansys. Background is to investigate the opportunity to work only with linear contacts for complex assemblies to work more efficient.
At the moment I found only the tool "Linear Perturbation". But are there any alternatives? So that I can compare different approaches. I want to perform static and dynamic analysis.
Attached the first simplified model to start testing different approaches.
Many thanks!


Attached Files

Order By: Standard | Newest | Votes
peteroznewman posted this 19 April 2020

Hello Lukas,

If you want a linear model, then you can't have frictional contact between the plates to support the preload.

My first choice to connect two plates at holes where a fastener goes is to use Beam Connection. This is fast and easy to do in Mechanical. No contact is used between the plates, the Beam is the entire connection.

  • Liked by
  • Lukas14089
Lukas14089 posted this 19 April 2020

Hello Peter,
First thanks for your answer!
I tried to perfom the simulation with Beam connection as you suggested but an error occurs.. (I guess because of the non existing contact between the plates..)
Could you explain in detail which faces to select in the beam connection to avoid the contact between the plates?

peteroznewman posted this 20 April 2020

Here is a discussion that goes into great detail about a Beam Connection.

Lukas14089 posted this 20 April 2020

Ok but if i am using beam connection I have the same contact issue between the plates or am I wrong? You are right using the Beam is a simplification for the bolt but you still have to use the contact right?
I am searching for a method to replace the "realistic" frictional contacts by using linear contacts (for example just using Bonded contact for the half of the area as approximation).

peteroznewman posted this 20 April 2020

No contact is used between the plates, the Beam is the entire connection. This is a conservative approach for evaluating characteristics of structures such as strength and stiffness. This model result will provide the lower bound on strength and stiffness. Engineers often have lower limits on acceptable performance, a minimum load carrying capacity or a maximum deformation (minimum stiffness). The design can be modified until the linear model predicts acceptable performance. When a real physical structure is tested, the experimental measurements will be better than the model predicted. That is a conservative result. You don't want the opposite to happen.

By using a nonlinear model, with frictional contact and bolt pretension, you can get a more accurate model whose result will be closer to the experimental measurements. You can decide to trade off longer model build and solve times for a benefit such as a lighter weight structure.

  • Liked by
  • Lukas14089