Hello everyone,

I'm carrying out for the first time a dynamic analysis with Caesar and I have some doubts you might solve.

From surge analysis I got the following Force vs. Time plot:

http://img4.imageshack.us/img4/8509/fhs.png

this is the evolution over time of the upstream pressure during the closure of one emergency shut-down valve. During closure of ESD valves, the surge relief valves open.

My big question is: given this excitation shape, what is the best method to determine the response?

I would say Time History, but what about Force Spectrum analysis?

At the beginning I discarded the harmonic analysis but since the excitation seems to have a sinusoidal form with kind of damping, should I re-consider my ideas?

Thank you

Why do you not transfer the Pressure /Time graph not in a force/time graph?
Force = Pressure *inside area pipe

I'm sorry, my post could be confusing.

The image in the link is Pressure vs. Time, however what I'm considering as input is Force vs. Time.

With a force time input you have 2 possibilitys.
1. Modal time history
2. Response analysis
In this case Caesar transfer the force/time input in a DLF/frequency curve

With methode one you get more precisely results.

In force Spectrum analys, DLF spectrum is used as input by Caesar to generate the outputs. Basically from DLF spectrum you can understand at which frequencies the ratio of maximum dynamic deflection to maximum static deflection (which is the DLF) reaches critical values. What I suppose Caesar does to get dynamic loads and displacements, is simply multiplying static loads by the DLF, this will yield dynamic loads.

On the other hand, in time history analysis the response of the system is determined by integrating the dynamic equation of motion in a step-by-step manner (Wilson 0 method). So, the smaller the time-step, the more accurate the solution.

Hoping my understanding is correct, my question is: what makes time history method more accurate than force spectrum analysis? Where is the approximation in Force Spectrum analysis method?

With the spectrum method you will get at every eigenvalue a system answer.
This answers will be combined SRSS, ABS, 10 %, Group method...
This method call "Response Analysis"
At Modal Time History Method ( isnt a direct integration method)you get a solution only at one decided frequency.
I think other eigenvalues ( not decided ) will not included , because answers not in the same time.

Response spectrum (RS) analysis is more a statistical evaluation of a dynamic event. The maximum expected response of each mode is combined (e.g. SRSS)to get an estimate of total maximum response. (Modal) time history (TH) calculates the "true" position in time for each mode and sums them algebraically - it's more exact.
I'd say, for piping, that RS is better for initial design, a more general evaluation, and TH is best for forensic engineering - where you have a well-defined event.

But in the dynamic result TH ( shape movie ) we can see only one shape .

The "time history animation" is how the system actually responds to the applied load, the combination of all of the modes.

In a "response spectrum analysis", your animating a single mode of vibration - which is independant of the applied load.