About stiffness of snubber

Dear all:
Who can tell me how to define the stiffness of snubber in dynamic?The stiffness of snubber in our stylebook is defined as follow:
K=2*Fn/Sb
K is the stiffness of snubber;
Fn is the specified force when the snubber locked;
Sb is the max displacement when the snubber act,the value is about 6~10mm;
So we can not confirm that if the K above is the stiffness of snubber we defined in CAESAR.
Thanks for your attention in advance!

First of all, I am wondering why you need the stiffeness of snubber. For what.....to calculate natural frequency or thermal ?

I would like to assume the axial stiffeness of snubber is rigid when snubber is locked during dynamic situation.
If you select proper type of snubber(hydraulic cylinder type) by considering enough stroke, installation within the allowed movement(normally 4 degrees angulation or rotation) with spherical joints at rear bracket and clamp joints, you can ignore the flexibility of snubber for thermal analysis, but you can assume snubber is axially rigid for modal analysis. All you need to consider is to add a half of the weight of snubber assembly in your system when snubber is installed horizontally.

Probably reality may be different to my assumption, i.e. it will have a little stiffeness. However, I never heard/experienced any field problem from the sites designed by above approach.
Let's see other member's valuable experiences.

Regards,

Dear Sun Wee,thanks for your reply firstly.
I assumed the stiffeness of snubber as rigid previously,but it's not reflect the reality.As far as known,when snubber is locked,the cylinder will move slowly,not stop absolutly,so it's not rigid.According to our snubber stylebook,we can find that different cylinder refer to defferent stiffeness.If the locked snubber is rigid,maybe CEASAR need not to define this value.
In our dynamic analysis,more big stiffness of snubber,more force will act on snubber,so if we use rigid as snubber,maybe it's not right way.

tWING,

Why don't you consider the stiffeness of structure and support elements every where ? And...hot modulus, gaps beteen pipe and supports, how about real friction factor after 10 year operation(i.e, corroded condition of frictional surfacee ? Caesar II is software...not a stress engineer.

There are always some gaps between real world and theory. In engineering design and analysis of piping systems, more practical and simplified methods are used with experienced engineer's sound judgement. These typical design and analysis ways have been used last over 50 years and proven by most of existing plants in the world.
I would like to point out that we rely on computer program and numbers too much. Im many cases, I have seen the systems desiged by experienced engineers without detail calculation is much more appopriate than new FEA experts.
Thanks.

Dear Sun Wee,Tnanks your reply again!
I got your means,but before I use CEASAR to stress analysis,I should know how to define the value in this software.There are many stress analysis programs just like CEASAR,pipestress,autopipe etc.Many of these programs can define the stiffness of rigid hanger,the default value in CEASAR is 0.175127E+13 N./cm.,but the max stiffness of snubber in our stylebook is about 0.2E+8 N./cm,so if we assume two above value is correct,we do dynamic stress analysis,we presume the dynamic load acting on snubber is 10N and the dynamic movement is 10mm,but if we use 0.175127E+13 N./cm as snubber's stiffness,the dynamic movement may be 0mm and the dynamic load maybe 1000N or more,and the natural frequency of pipe system will change more bigger.So would you please tell me which result is correct?
When we build the model of pipe,we'll input the elastic modulus of pipe material,stiffness of spring hanger(maybe software will sugguest it),stiffness of rigid(valve),stiffness of rigid hanger etc.All of these will influence the movement and the load in pipe system,the stiffness of snubber will act as same as above in occasional case.So I think this value is important.

Okay compare the stiffness of the snubber or strut to the structural member its attached to, my guess is the snubbers k rate is a hell of a lot biggerthan what you are attaching to, if so the most important k rate is the structure.

Dear John C. Luf ,thanks your kind reply.
My means is that the snubber in dynamic load will act as same as spring not rigid,so more stiffness will lead to more load acting on the snubber.The correct input is not my final purpose,the deferent input(0.2E+8 N./cm or 0.175127E+13 N./cm) will lead to deferent dynamic load.when we use 0.2E+8 N./cm as stiffness,the result of pipe stress analysis is conservative,but we can not confirm that the dynamic load is enough to structure because it's smaller than result of 0.175127E+13 N./cm;When we use 0.175127E+13 N./cm,we can use this result to design conservative structure,but we can not confirm the pipe is enough safety because the snubber will not behavior as same as rigid in dynamic load.
Dear Richard Ay,Would you please tell me your choice?I can not find more detail explain in CAESAR's user guide.
Thanks for your kind reply in advance.

You're right, a real snubber won't be rigid. But, if you model it as rigid, it will see more load than the real snubber. What is this load, 500, 1000 lbs? Then pick a snubber to handle this load, acquire its stiffness and then put it in your CAESAR II model. Make a run to verify your selection.

Thanks for your kind reply,Dear Richard Ay.I got your means except a little trouble as following:when I do dynamic analysis,one method is I change the stiffness of snubber to match the load just like select spring?the other is we put the max stiffness snubber into model directly,if the pipe system is safety and load is smaller than the allowable load of snubber,it's ok.which is the correct?
Thanks again.

I wish it was that easy. Keep in mind that the stiffness you use may affect the dynamic response of the system - changing the natural frequencies, the magnitude of the resulting dynamic load and the response to that load.
From a purely analytical view, you may want to run it both ways to seee how sensitive the reponse is to this variable. If it doesn't change the numbers, no worries...

Dear Dave Diehl,thanks your reply.
I'll try yours and Richard's suggestion.
I think we can input the max stiffness snubber,and we select this snubber according the stiffness,so the snubber can support more load than the result of analysis.For example,if we input the stiffness as 0.2E+8 N./cm , the result of dynamic analysis is 10000N,but the max allowable load of snubber(corresponding 0.2E+8 N./cm) is 1000000N.When we design the support,we select the snubber according to its stiffness 0.2E+8 N./cm,and we design the structure according the result load 10000N.Maybe there are some mistake,but I can not find out,would you please tell me where is wrong?