Hi all,
I have a high-temperature system with occasional loads like seismic and slug. since it is a high-temperature line I had to use Snubbers to qualify my system in occasional cases.

As far as I know, snubbers get activated for sudden shock loads with a particular velocity, and in operating cases, it will allow displacement. However, I observed that system frequency has significantly changed after I incorporated snubbers in the model. As I said above, In normal operating cases snubbers will not act as a restraint then how snubbers can change system frequency. Please share your experience.

Regards
SB

Hi Jay,

My understanding is that snubber will not act in slow moving loads such as thermal loads but will be active in sudden shock dynamic load. The natural frequency of piping is dependent on both mass (m) and stiffness (k) of the system (sqrt of k over m), which the tendency of the piping system to respond to dynamic loads. Since snubber is a dynamic restraint it will be included in the stiffness term k of the natural frequency calculation but will not be included in static stiffness k so that it is free during thermal expansion of piping system.

Please correct me fellow stressers if my understanding is wrong.

Cheers!!!

Thanks for your response. As we check the frequency of the system in operating cases snubber remains inactive. Hence no restraint/stiffness being included in the system. Then how it is influencing the natural frequency of the system.

I went through few old threads and I understood snubbers are active for all dynamic analysis. Still, it is not clear that How snubber can utilize to increase the natural frequency of the system?

I think the modal analysis results are relevant for dynamic analysis, so considering the snubber active for natural frequency calculation (and corresponding modal shape calculation) make sense. Would the opposite approach be better?

Hi Jay,

I agree with Mariog. If I'm not mistaken, the harmonic, response spectrum and Time History Analysis are all dependent on the Modal Analysis (natural frequency). There is Direct Integration Time History analysis which maybe not depend on natural frequency if I'm not mistaken, but most software utilize the Modal Time History analysis which breaks the system into individual modes of vibration and combine them together to get the total response.

Any other opinion is highly appreciated. Just correct any misleading statement I've made.

Cheers!!

Thanks Borzki and Mariog.

As per my understanding,
for the modal analysis system's restraints are considered from static analysis ( Eg Liftoff supports will not be considered in modal analysis).
In operating cases, the snubber remains inactive as it allows thermal movement. However, in modal analysis (operating case) the snubber considered as restraint even if it is inactive in operating cases. can anyone tell me the reason behind this?

Mr. Borzki and Mr.Mariog. do you recommend using Snubber in the system to increase the system frequency?

Hi Jay Stress,

Just a short response to this inquiry/doubt:

<< As we check the frequency of the system in operating cases snubber remains inactive. Hence no restraint/stiffness being included in the system. Then how it is influencing the natural frequency of the system.>>

Natural Frequency is evaluated and corresponds to a CERTAIN STATUS of the restraints/supports. Under Normal Operating conditions (excluding dynamic actions and vibrations/cyclic excitation), Snubbers are indeed inactive/off.

However, under a dynamic load action (such as earthquake in your case), snubbers' restraints become active and they stiffen the piping system.

As pointed out above, when you define Dynamic Input Caesar II dynamic input processor, you must specify the Snubbers' restraints. Caesar II will consider those restraints (including their "finite" stiffness, as specified) in addition to the existing "permanent" support restraints, so that the dynamic response of the piping system will include the snubbers' restraints influence.

Dynamic analysis is quantifying ONLY the dynamic response of the System to the applied dynamic loads.

If you perform Response Spectrum analysis, for instance, and you need to superpose the Response Spectrum calculated support loads over an Operating Case loads, previously evaluated by the Static Analysis, then those Operating Loads are calculated "without" snubbers' action, while the Response Spectrum loads are quantified "with" snubbers' action included.

Therefore, if snubbers are actually installed in a piping system, and a Dynamic Modal Analysis is performed "without" snubbers (e.g. keeping snubbers inactive), then the Natural Frequency calculated is irrelevant for piping system dynamic behavior description, because Snubbers are actually part of the dynamic response.


Best regards,

Jay

From your last post it seems you consider the modal analysis must always start the math over the operating state. Despite your remark about lifted-off supports is correct, such statement is not particularly true for snubbers case because they are devices intended to not participate to operational cases but to dynamic cases ( mathematically within modal analysis, as well). For dynamics, K matrix must consider the snubbers existence. In fact Caesar just follows in calculation the physical reality, this is the accurate statement.

Thanks all for your responses, still...I am not convinced to use a snubber to arrest the vibration. (Ie increase the natural frequency of the system).
let us take cases of a transfer line system that vibrates in the startup/shutdown cases and used a snubber to minimize the vibration.
In the startup case, the snubber will activate due to vibration and it acts as a rigid support. If the vibration is a continuous phenomenon and the snubber remains active. Since it is a startup case further movement of the pipe due to temperature also will restrict by the snubber and it creates expansion stresses in the system. In such cases, application of snubber can be justified?

I guess nobody wants to convince you "to use a snubber to arrest the vibration, i.e. increase the natural frequency of the system".

Snubbers are rather shock absorbers, which in such applications form restraints between piping and the structure. The displacements due to expansion do not encounter resistance by snubbers, in theory.

True, in calculation one effect is modifying the natural frequency. That's because a snubber is in calculation like a switch: for Non- dynamic case=0 (no Restrain), for Dynamic case=1 (Restrain). What I don't understand is which is your proposal to improve the calculation for shock dynamics cases, since "as it is" does not convince you?

A snubber offers impedance proportional to velocity. Thus, it's conceivable that a snubber could seize and act as a hard stop, axial to the snubber, once a critical impulse is met.

This would be analogous to a building door that would slam open and closed due to pressure variations being solved by adding a door closer, which is a combination spring + snubber.

The problem would be quantifying how to design the snubber for the pipe.

It's one thing if we can measure it, or if we have a solid grasp on how the input converts to piping response, but we don't always have the info necessary to solve this.