Dear sir:
When we do modal or spectrum analysis,the dynamic eigensolver is circulating in one frequency as following:(just for example)
Frequnecy 86=6.0167 Elapsed time=.....
Frequnecy 87=6.1268 Elapsed time=.....
Frequnecy 88=6.0167 Elapsed time=.....
And you can find there are some low frequency will be repeated at high frequency zone,when we change some snubber or testraint's location,maybe it's changed.But we can not understand why and how to avoid it?Who can explain it to us?
Thanks in advance!

It sounds like you have a very flexible system with a lot of low frequencies. There is no guarantee that the eigen values/vectors will be determined in order (from lowest to highest). Changing the system stiffness (by adding, removing, or moving) a restraint will change the natural frequencies of the system.

You can try and see if these adjustments help things out (note, I'm guessing here ...):

On the Advanced tab of the Dynamic Input ...

Change "estimated numboer of significant figures" to 10
Change "decomposition singularity tolerance" to 1e16.
Change "subspace size" to 8

Thanks for your reply.
Our system is flexible as same as your say,its first frequency is about 0.4566Hz.The pipe's design temperature is 610 deg.C,so we can not add too much restraint to stiffen it because we must consider exp. Stress and save pipe.So,refer to the low netrual frequency system,would you please give us more suggestion?
Thanks for your attention!

I wasn't suggesting you change your system. Have you adjusted the dynamic solution setup as per my previous post?

No,I will try your solution tomorrow because I'm out of office.And thanks for your reply again.By the way,I sent a email to support@coade.com as same question.
Another question is:Refer to water hammer spetrum analysis,the cutoff frequency is about 300Hz for many times,this is suggestion of ceasar's dynamic manual,but for steam hammer,which cutoff frequency should be selected?We use pipenet to calculate hammer force,and input to Ceasar to generate the spetrum,and we find the biggest DLF's value is 30Hz(just example),if we select 30Hz or maybe little more than 30Hz,so is it enough to consider the system's safty when steam hammer happen?If not,and we must select 300Hz as cutoff frequency,but our system is too flexible,what can we do?
Thanks for your relpy!
Best regards!

Please don't do that. One incident report is sufficient.


The high cutoff suggested in the manual is necessary (sometimes) to include any axial modes in long runs of piping. For many jobs you don't have to go anywhere near this high. So you are correct, something past 30hz (say 50hz) will be sufficient for the job you describe.

Your suggestion is sufficient,thanks a lot!