Hi,

Can anyone help me to clarify the below?

When a trunnion support has all directional restraints (X, Y, Z) with zero gap(hence no rotations are allowed), why Caesar doesn’t consider moments for that support?

For the trunnion length, there should be moments acting at the end of the trunnion because no rotations are allowed.

We generally know that there will be moments for only anchor supports since there won’t be any rotations. Similarly, for a support at trunnion end where it has all directional restraints without allowing any rotations, it is not clear why caesar does not consider moments. Is Caesar converting the moments into equivalent forces by using M/Z?. If so, the same method may be applicable for anchor supports also. But for anchor supports, we are getting moments seperately.

Can anyone please clarify?

Regards,
Gopalsamy

Gopal,

Have you specified rotational restraints ((Rx,Ry & Rz) at trunnion end?

If not, then specify them to get moments because even if you specify directional restraints with no gap, there would still be rotations allowed at the restraint node (can validate this with your displacement summary report where you'll see some rotations in X , Y & Z directions).

Keep Smiling..

It always helps.

SJ

I am presuming your reference is to a vertical trunnion that is welded to the pipe and typically restrained at the base by angle steel guides.

Can you please explain your statement ..."with zero gap(hence no rotations are allowed)" ?

What influence has a zero gap on any rotational restraint ?

As SJ says, if you don't fix rotations at a restraint, you don't get moments developed. But what are you actually trying to model ?

Thanks to SJ and Movertz.

Movertz,

Yes, you are correct. It is for vertical trunnion that is welded to the pipe and the base is restrained in all directions. The gaps for guide,line stop and rest+hold down are zero. These type of supports are commonly used.

My question is, for these type of supports, though there is no chance for rotations Caesar does not consider moments because we usually don't input RX, RY and RZ and hence we do not get the moment values for support design.

I feel since these supports are similar to anchors, caesar may treat as anchor support and provide the moments as conservative. Otherwise our support design may not be correct without moments. I think, usually nobody input RX, RY and RZ for these type of supports.

I hope, you understand my concern.

Regards,
Gopalsamy

Movertz,

I forgot to respond to your query ' What influence has a zero gap on any rotational restraint ?'

Since the base is usually a rectangular or square plate, with zero gap in all directions the plate can not rotate in any direction.

Regards,
Gopalsamy

If the support is welded, then an "XYZ" restraint is the wrong way to model this location - you want an "anchor".

As stated above, you need to consider exactly what you're trying to model. The nodes in CAESAR II have 6 degrees of freedom, and each must be addressed. If you only restrain "XYZ", then the rotations are free.

This strikes me as a classic case where you could model the trunnion in as a piece of pipe with the base as an anchor. This would reflect the actual stiffness of the trunnion.

Quote .... Since the base is usually a rectangular or square plate, with zero gap in all directions the plate can not rotate in any direction....

OK, assuming the trunnion is vertical it will not rotate about its own vertical axis.

But what stops the end plate rotating about both horizontal axes ? Surely, only welding it out or bolting it down would do that ?

Gopalsamy,
richard is right, actually you should specify as ANC in the restraint iput, or X,Y,Z, RX,RY,RZ. If not you will not get the moment.

Caesar doesn't know or care if your are modeling a fully restrained base support or a ball joint. It's your responsibility as the engineer to build the model to simulate the system.

As the others have mentioned, if you believe that the support arrangement you have for your system will provide six degrees of restraint, then you need to tell Caesar that. It's not a mind reader.

GOOD way of judging to run analysis for Trunnion to consider Anchor or X,Y,Z,RX,RY,RZ or +Y etc. or not ........is to view, one time physically at site while hydro testing or steam blowing with respective theoretical calculation from CAESAR II comparing practical operating testing condition results.

If you notice the difference then you know what you’re doing.

Thanks

In addition to query from Gopal , For vertical trunnion that is welded to the pipe and typically restrained at the base by angle steel guides.

There is a welded plate at the bottom of the trunnion which is subjected to guiding and all other restraints.

If You have Bolted down( or Hold down) a trunnion , besides guides in X and Z , all with zero mm. How will Rotations ever happen??

The Six restraints (Rotational restraint) thing is valid for any point connection. And this is not a case when trunnions are involved.

So is it so, that we should model rigid Anchor in the above to cases i mentioned.

Anyone?

Richard , please reply.

I think that is already written above. You have to judge yourself. If it is bolted it is a full anchor, if it is guided with zero gap, maybe is, maybe not, you do not know for sure, so you should be conservative.

Thanks Danb,

So, in any case of support arrangement of REST+GUIDE+HOLD-DOWN (all with zero mm gap) ,the conservative approach will be to model as Anchor. Since, the Loads and Moments would further be utilized in designing the Supports.

Is that So?

Coz, sometimes a surprising value of Moments appear in this case.

For your support, probably it is conservative, but for the piping system maybe not.

Deedee,

You have mentioned your support functions as ' Rest+Guide+HoldDown' with zero gaps to consider as an ' anchor' support as a conservative.

But, there should be a ' LineStop' function also to consider as 'Anchor' as a conservative support.

Regards,
Gopalsamy