I was doing a pipe modelling with a Design pressure of 150psi and temperature of 540C. The piping is 30" in size and is connected to a reactor through a nozzle. I have modelled the tower from the skirt to the nozzle as a rigid body.

In the first modelling, I have placed a connecting node between the tower nozzle flange and the pipe flange. The Expansion case indicated that the stress at some part of the piping was at 270% of allowable stress.

In the second modelling, I removed the connecting node between the flanges. The expansion case exceeded the allowable by 370%.

All this while, the sustain loading remains relatively constant.

Questions:
1. I had thought the result would have been similar. Can anyone explain what has changed?
2. What is the significance of the connecting between the flanges in this case?
3. Which will be more representative of the real case?

Thanks.

TO : Hong-Joo, Lee
The CNode makes the piping flange movements to be the same as the vessel flange movements. An anchor restraint on the piping flange will give loads on the vessel, while providing thermal displacements of the vessel nozzle. When you remove the CNode, then the piping flange is fixed - no displacement occurs. Expansion stress is higher with a fixed piping flange compared to the CNode that has upward thermal displacements, reducing the expansion stresses. For your model the CNode configuration is probably closer to actual real condition. Try comparing to another model using the thermal displacements inputted to the anchor restraint of the piping flange, instead of the vessel modeled for thermal displacements. Expansion stresses should be similar.

1.
I think the problem is not the cnode itself. You have to couple the cnode with your nozzle/
tower and give the connected directions as "restraints" X,Y,Z,rx,ry,rz).

Here may be the problem. The stiffness of restraints is a value which comes from your
configuration settings and if the defined values are to small, your nozzles are not "completely"
coupled with the possibility of small difference displacements/rotations.

Try to give a higher stiffness value for each of the restraint directions,
which will overwrite the standard value for the special restraint direction in
your "cnode model".

2.
The node is in the restraint output.

3.
The model without cnode.

There should be no difference if you've simply removed the CNODE anchor - as long as you remember to fix the node numbers so that your pipe system is continous. When you have a CNODE, you will have two different node numbers at the same point in space. The CNODE ties them together using whatever restraint you specify (typically an anchor). If you deleted the CNODE, you need to make sure to delete the restraint all together and renumber the segment so that the node numbers are the same to keep the connection.

So, we really need to know what you mean by "I removed the connecting node between the flanges." I suspect you deleted the CNODE node number, but still have an anchor at the location.

Hi Edward,
I have removed the CNODE number as well as the anchor in the location.

Please review the CAESAR II Applications Guide, pages 3-8 through 3-16. These pages show how to model through a nozzle and include the vessel in your model.

It sounds like you have not included the nozzle flexibilities in your modeling. This is why when you remove the CNODE specification your piping is still anchored.

The CNODE provides the ability to "associate degrees of freedom" between different nodes. There is no point in doing this if you are not going to do something with the CNODE, or define stiffnesses between the CNODE and the restrained node.