I am working on the FE modeling of a beam-to-column bolted end-plate connection in ABAQUS, and I am facing an issue with the application of bolt pretension.
I tested two methods to introduce the preload:
Apply Force" Method:
- I directly apply the preload force calculated according to the Eurocode (Fp=319 kN).
- To verify the force in the bolt, I take a cross-section at the middle of the shank, compute the average axial stress S33 , and multiply it by the bolt's cross-sectional area A=452 mm²
- This results in a force of ~324 kN, which is close to the theoretical value.
Adjust Length" Method:
- I apply a shortening based on ΔL=FpL/EA.
- However, when I check the force at the middle of the bolt, the stress is much lower (~386 MPa), corresponding to a force of only 174 kN, which is almost half of the expected preload.
Why does the "Adjust Length" method result in a significantly lower preload force compared to the "Apply Force" method?
Any recommendations on how to properly model bolt pretension using "Adjust Length" to achieve values consistent with preload bolt force.
The issue is in their nature of load application mechanism. The "Apply Force" method is a force control approach where the preload force is directly applied, resulting in straightforward stress calculations (I think you are calculating it from nodes as you mentioned for S33). In contrast, the "Adjust Length" method is a displacement control approach, where a length change is specified to achieve the desired preload, introducing geometric nonlinearity and potentially leading to lower effective preload forces (Your calculations comes at the stage where you are getting the resultant displacement based forces which will be different). When you specify a length change to achieve the desired preload, ABAQUS adjusts the bolt length accordingly. However, this adjustment involves geometric nonlinearity, which can lead to stress redistribution and lower effective preload forces compared to directly applying the force. In order to check where you are getting the capacity, you will need to automate your process with Python for each frame to get all curves at each frame in .odb for the specific nodes of your interest and extract values from them, you will find there will be at some point approximate capacity of your desired results.
Dear Sara Bachir,
I hope this message finds you well.
According to your issue, which length of the bolt have you considered in calculating the bolt reduction length? I think if you try the grip length (under bolt head till nut top), it may work.
Best Regards,
Reza Khani
- Badges
- Science method
- Similar topics
- Biological Science
- Ecology