In using the expansion joint modeler to insert a tied, flanged, expansion joint in a section of pipe, the CAESAR II modeler creates a rigid element, one end of which is C-Noded to the far end of the exjt. Let's say the pipe segment starts at an anchor at Node 10 and proceeds in the +X direction to another anchor. Inserting an expansion joint in the middle of this pipe segment would result in the modeler creating a separate rigid element from the start to the far end of the exjt assembly. The node at the far end of the rigid would be C-Noded to the end of the exjt and the two nodes would be associated with each other by the following restraints: X, RY, RZ. My question is with regards to the X restraint. Wouldn't the X condition only apply if the tie rod were nutted on either side of the lug attached to the flange? In which case what's the point of the exjt anyway? Or is that restraint definition pending the results of a subsequent CAESAR run, which the analyst would use to check the loads in the rigid to see if it's behaving as expected?

I think the engineer should to adapt the default of CAESAR II to the real situation. If you got nuts just in one side of flange, change the direction for X. If you got nuts at both sides, keep X.

The situation you describe above is a misapplication of a tied EJ. The tie rods are there to restrain the longitudinal pressure thrust and hence prevent any axial movement. Tied joints only absorb lateral movement and therefore must be used perpendicular to the major movement.

For a positive pressure, the nuts are placed on the outboard ends and a second set of nuts are placed inboard but slightly gaped. In this way, the longitudinal pressure thrust is restrained and the inboard nuts are there for any cold dead weight compression.

Dan - Thanks. For sure it was a misapplication, I was just running a simple configuration to see what CAESAR did and to allow me to describe it on this board in as few words as possible. Now that you point it out, I should have chosen a configuration that had only a lateral movement. But I am still wondering about the restraint CAESAR is choosing. The description you provided in your response above makes complete sense - in which case shouldn't the analyst change the CAESAR restraint to say +X, -X with a gap? I can't see how the full X is correct under any circumstance. Also, the RX, RZ restraints really only have meaning if you are applying the tied joint in the right way - as you say a lateral movement. Otherwise what's to prevent a rotation (albeit a slight one allowed by the slop in the bolts, nuts, bolt holes). Thanks for your input.

On the “typical” tied universal joints, the inboard gap is very small, only sufficient to allow the rods to rock in the lug plates before they bind up. So, in the theoretical world, the ±X, Ry & Rz restraints are accurate. In the real world, in order for the joint to absorb axial compression (-X), the axial force in the pipe line would have to exceed the longitudinal pressure thrust which normally doesn’t happen. Also, to absorb Ry & Rz moments, the moment would have to exceed the tensile stress in the tie rod, again the loads are so high that it normally doesn’t happen. We have specially engineered joints that can do this, but it is the except rather than the rule.

Thanks Dan - good to know the manufacturing details from someone at an OEM (I presume).

Goodsalt - Your welcome and yes, I'm principal engineer at Senior-Flexonics, Pathway.

Dan,

You're probably the guy I was looking for.

I just got some data for a universal tied expansion joint (pathway; of course)which I needed to simulate in my Pipe Stress calc.

However, the stiffnesses mentioned are of the total assembly (which includes two bellows).

Is there a way to calculate the individual stiffnesses of each bellow?

Is the springs in series hold good in this case or is there something more to it?

SJ – You should have received a Bellows Design Analysis (BDA) with your drawings. If so, you can send it to me and I’ll run the single bellows BDA which will give you the values for the analysis. If you didn’t get a BDA, give me the quotation number and/or sales order number and I can track it down.

Dan,

I've very vague info obtained from a sales guy in pathway as I needed that for my pipe stress analysis.

Let me know what all details you need to calculate the stiffnesses of each singular bellow used in the universal tied joint?

Thanks.

SJ - Let's take this off-line and communicate via e-mail. My e-mail address is: pipesol@connectnet.com

Ok Dan.

Thanks. I'll email my query.

Dan

I need a small bit of help related to bellows.

I'll email it to your personal email.

Thanks for the favour.