I have few queries-
1. I understand that two different spectra can be defined in spectrum definition option, say I have HORZ and VERT as one set of spectrum which is applicable for few nodes in the system ( say which are below 5 m level) . Few more modes are above 5 m ,where I want to apply different spectra, 5HORZ & 5VERT . (Basically vertical spectra is derived from horizontal one). I define different spectra and then while applying it in dynamic case identify node numbers against each spectrum ( i.e. start node and end node) . Is this approach OK ?

2. If suppose I want to apply ISM for third set of spectrum IHORZ and IVERT. This ISM value of displacement will be only magnitude ( not + ve or -ve ). Please advice. Again I will be identifying the nodes for which this spectra is to be applied.

3. I am having a system which is on an unloading platform. Do I have to apply the ISM at all support nodes or only at anchor nodes ? There are 4 unloading tie-ins ( anchors) . Is it sufficient to apply it only at anchors ? Or I have to apply the unloading platform movement at all support nodes? When I apply at all support nodes there is very high failure in stress.

4. The unloading line then moves to tresstle portion where trestle displacements are different than unloading platform. How to handle such scenarios in modeling?

Regards,
bag_piper

In fact Contract document says that differential support motion to be applied in response spectrum. Does this mean if unloading platform moves by 15 mm and trestle moves by 25 mm then difference of 10 is to be modeled as ISM for support nodes on trestle only? Response from Dave or Richard is most sought after.
regards,
One worried Stress engineer

Not that easy...
The platform could be moving to the left 15mm while the trestle is moving to the right 25mm.
Response to each individual ISM set - in CAESAR II - is evaluated alone and the calculated stresses are combined either ABS or SRSS with other components of the seismic load.
When defining more than one shock spectrum in the same direction and in the same load case be sure to enter data in the anchor movement column. Empty cells here will cause CAESAR II to calculate a displacement based on the acceleration associated with the highest frequency defined in the shock spectrum. If the entered spectrum data does not anticipate this use, you can get some pretty wild numbers. These displacements are used for all boundary conditions (in the same direction associated with the spectrum defined in the shock) in the node range included in the spectrum definition.

This means if one spectra is for ground sleepers where structural displacement is zero, do we have to mention zero for nodes corresponding to that spectrum in ISM column? Entering Ism displacements, when nonzero with sign will alter the results?
differential displacements only to be modeled as 40 mm at all support nodes of trestle? Or only at anchors?

Whenever more than one spectrum in a shock definition is defined in same direction, all ISMs that are not specified will be calculated as stated above.
No sign is associated with these ISMs. Your dynamic input is linear. Each ISM (group) is evaluated independent of all others and summed with the others either absolutely or SRSS.
All restraints defined in the spectrum node range and in the same shock direction as the spectrum is used will have an ISM assigned. (Ignoring friction, I wouldn't expect any response at a Y support in an analysis of an X shock.)

We have a case where there are severe unloading platform displacements . We find that when these are superimposed in response spectrum analysis as ISM at every support , there is sever failure.
How such cases are handled in actual practice.

If those ISMs are correct, add flexibility.
Your values stated earlier do not seem large (25mm & 15mm). Did you enter the ISMs or were they offered by CAESAR II based on your spectrum. If they were offered by CAESAR II, please reread my initial response.

I entered those ISMs, got from civil contractor.
no way to add flexibility due to layout constraint.
tried with tied universal bellows. Not much success.

When we apply these displacements in dynamic response module , we get the total stresses in the case where sustained and this dynamic case is combined . This case is defined as "OCC". This is a net result of seismic stresses due to earthquake effect on mass + displacement effect combined.

Now 31.3 allows us to compare displacement induced stresses ( be it any type of displacement ) with EXP case . In above handling how to segregate stresses purely due to displacements ? Since the response module where displacements are handled has no provision to segregate displacement induced stresses. Hence it is not possible to compare displacement effect purely in above module.

Please advice.

Regards,
bag_piper

The plant under discussion is no nuclear ,( it is in fact LNG regasification terminal) , can we take help of provisions of Nuclear codes ? NC / ND 3665 & NB 3656 ,under faulted service conditions allow primary stress upto 3 Sm and secondary upto 6 Sm.
Can we compare the induced stresses to these values ?
Please advice.

Regards,
bag_piper