Time history analysis response at resonance

Hello,

I have Time history model of structure supporting machine please find the attached zip file.

As you can see the 2 beams parallel to z axis are subjected to 1000 kg of machine dead load which produces 1000 kg of dynamic load in x axis.

Time History example.zip

1) Is my time history analysis definition is right?

2) What is the "integration time step" (plz refer screenshot "Time history analysis")?

3) From the file you can see the excitation frequency matches the dominant mode frequency i.e., resonance has occurred.

At resonance the response is only dependent on the damping, so for the 2% damping the load amplification shall be 25 times the sinusoidal force defined in time history. 

Note: load amplification = [ (1-r^2)^2 + (2*g*r)^2 ]^-0.5 , where r = excitation freq / natural freq & g = damping factor = 2%

But STAAD results (plz refer screen shot "Response at resonance") show only [1134 + 1134] / 1000 = 2.268 amplification.

How does STAAD calculate the response of time history at resonance?

Thanks in advance.

Vishal

  • There are some issues as noted below with the time history loading/analysis. 

    You have accounted for the masses multiple times. Specifying SELFWEIGHT X 1 LIST ALL is good enough to account for the selfweight of the entire structure.

    You should remove the STEP 3 as 3 seconds is too big. Instead use the default which is 1/12th of the time period calculated based on the frequency of the harmonic loading.

    FUNCTION SINE

    AMPLITUDE 1000 RPM 1400 PHASE 0 CYCLES 1 STEP 3

    Also the loading should be applied for more than 1 cycle. Use 100 cycles.

    Integration time step is the time interval at which the structure is analyzed.

    A modified file is attached for your reference Time History example_r.std 



  • Hello Sye,

    Thanks for the clarification.

    As per your suggestions and corrected file, I have implemented time history in real example,

    In this case, there is no resonance, but still the response is very small not even equal to amplitude of the force (370 kg).

    Please find the STAAD file below,

    18-E199 For weight.zip

    One more question,

    In this case, if I have to include seismic load case using Response spectrum analysis (Eurocode)

    1) Do I have to define the inertia masses in the response spectrum load case again, which are already defined in the time history load case?

    2) Can we combine time history load case with response spectrum earthquake? If yes, any special way?

    Since i tried using Define combn command but there was an error 

    " ** Read/write ERROR in Unit No. 16
    subr errord j= 36"

    Vishal

  • Hello Sye,

    Awaiting your response for above details.

    Vishal

  • Hello Community,

    Let me know if someone is looking after the query or not.

    One more point about the 'STEP', I think it should be the multiple of 'integration time step', otherwise the response will be less than the actual value,

    The default values follow this rule, (1/12)/0.001388 = 60

    Am I right?

    Looking forward to quick response.

    Vishal

  • Here are the answers to your questions

    Dynamic masses would have to be defined again within the response spectrum load case. You may simply copy and paste the mass definition from within the time history load case to within the first response spectrum load case. Alternately masses may be defined as Mass type Reference loads in which case all dynamic loads and even equivalent static loads will use that single mass definition.

    You can only combine results from a time history load case with that form a response spectrum load case using load combination. Please ensure that you are using the latest STAAD.Pro Connect Edition ver 21.03.00.146. Also I would suggest that you create a new folder and copy only the .std file into it and rerun the analysis in case you are getting the Read/Write error. The analysis ran just fine for me.

    STEP is used to digitize the harmonic forcing function. It should be small enough so that the digitized data adequately represents a sine/cosine curve. I would suggest that you leave the STEP and also the integration time step as the defaults.

    I added a CUT OFF MODE 25 command to increase the mass participation and responses are now higher.   

    A modified version of the file is attached for your reference

    18-E199 For weight_r.std