Additional bending moments caused by the update parameter command

[akkheng]

There is a staged construction steel-concrete composite beam which is modelled by elasticBeamColumn element. The construction stages are:
1. simply supported steel deck
2. fixed steel deck
3. fixed steel-concrete deck
In the first stage, only the steel deck was constructed. The steel deck is simply supported, the beam ends are fixed only in x and y directions. Under the gravity load of the steel deck, there will be 0 bending moment at the beam ends. In the second stage, I added a new condition to constrain the rotational freedom around the z-axis. Therefore, the beam is fully fixed. In the third stage, I want to simulate the concrete deck construction. Therefore, the stiffness of the beam is increased by using the update parameter command. The problem is that, after I update the stiffness of the beam (the gravity load is not changed), it will cause some bending moment at the beam ends, which should not occur. How can I fix this problem?

Picture1.png (34.38 KiB) Viewed 2105 times

Picture2.png (34.05 KiB) Viewed 2105 times

[STKO Team]

First of all, there is an issue here:

In the secon stage you fix the rotation at 0. So you should see opposite bending moments there. I think this is a problem with the plain constraint handler.
If you use the transformation handler the output should be ok.

Then the 3rd stage should change as well... you changed the stiffness and you should see a difference in the bending moments... or am I missing something?

[akkheng]

Thank you for your reply.
When I use the transformation handler, I can see opposite bending moments in the second stage.
However, I do not want to see opposite bending moments at this stage because this is a construction stages analysis. The structure has already deformed in the first stage, and the deformation is complete. Then, when I fixed the deck, the bending moment on the deck should not change, as there is no further deformation in the second stage. Similarly, I do not want to see any changes in the bending moments in the third stage, as there is no additional deformation during that stage.
Do you have any suggestions？

[akkheng]

In simple terms, I hope that the internal forces remain the same as in the first stage.

[STKO Team]

When I use the transformation handler, I can see opposite bending moments in the second stage.
However, I do not want to see opposite bending moments at this stage because this is a construction stages analysis.

For the second stage you can use the Plain handler that fixes the DOF at the current value, while other constrain handler will impose exactly the constrain value at 0.

Regarding the 3rd stage instead, why are you saying it should not change ?
Let's say you keep the load constant and fix the rotation at the value obtain at the end of stage 1.
At this point you don't have any end-moment.
Now you add the deck, so you add stiffness to your beam. Under the same load the beam will try to reduce its deformation due to the increased stiffness, that's why you see that effect.

I know it is not real, because in reality the deck is applied on top of the existing beam at its deformated configuration.
But to do this, you should physically add the deck as a new beam element, and attach this new beam element to the old one with stiff zeroLengthElements.
You should use the modelSubset command to perform this kind of construction stage analysis.
In this way the new displacement will not be influeced by the intial disp

[akkheng]

You should use the modelSubset command to perform this kind of construction stage analysis.
In this way the new displacement will not be influeced by the intial disp

Thanks for your suggestions.
I have tried the modelSubset command to perform this kind of construction stage analysis, and I obtained the good results.
However, it seems the modelSubset command causes some errors when performing the seismic analysis.
Is it my fault, or is the modelSubset command not suitable for seismic analysis?

[STKO Team]

What error?
Can you describe it and share your file?

[akkheng]

I'm sorry I was too busy last month to reply.
I have resolved the issue using your method. Thank you.

[STKO Team]

You're welcome