Dynamic Frequency Analysis

[stevetto]

Hello

I would like to analyse how a working chain lift's extitation influences the
dynamical properties of a frame structure.
The excitation is caused by the
chain lift's unequal climbing movement.
I determined it's unequal speed and
acceleration function.

This is a periodic excitation, so I would need to use the dynamic
frequency analysis. I have no other loads, so in my opinion I have to
investigate it as a Base excitation.

I also calculated the Fourier series
of the acceleration. This is my excitation function.

How can I use this function in Mechanica
to define a Dynamic frequency analysis?

Here is the function:

 

[ghgarzon]

The first step is to run a modal analysis including at least the first 6 natural modes of vibration. Then you run a Dynamic Frequency analysis using a table to enter the excitation force profile. It's important to define a measure or measures defining points where you'd like to see the response as displacement vs. time.


The function is entered as rows on a table that has two columns, Time and magnitud; therefore you have to decide how many rows you'll need to define the function; I'd suggest 6 rows/cycle would give you a good representation. By the way, you'll need to define a force on your model, (at at point/surface); Mechanica will then apply the defined profile on that force.

 

[stevetto]

The first step is to run a modal analysis including at least the first 6 natural modes of vibration. Then you run a Dynamic Frequency analysis using a table to enter the excitation force profile. It's important to define a measure or measures defining points where you'd like to see the response as displacement vs. time.
The function is entered as rows on a table that has two columns, Time and magnitud; therefore you have to decide how many rows you'll need to define the function; I'd suggest 6 rows/cycle would give you a good representation. By the way, you'll need to define a force on your model, (at at point/surface); Mechanica will then apply the defined profile on that force.

Hello ghgarzon

Thank you for your answer, but unfortunately I can't agree with you.
You were writing about the Dynamic Time Analysis, instead of the Dynamic Frequency.
In a D. Frequency analysis the firs column is the amplitude and the second is the frequency.
And this is the way how Mechanica use these data (from PTC’s Knowledge base):
Suppose the values a1 for frequency f1, a2 for f2 etc., are entered in the table defining the input acceleration, in the Dynamic Frequency analysis definition form. This value represents a1*cos(2*Pi*f1*t)+a2*cos(2*Pi*f2*t)+a3*cos(2*Pi*f3*t)+... in the time domain.
In other words, the values that are input in the frequency domain are actually amplitudes of cosine functions of component frequencies. Sine components are NOT represented in Pro/MECHANICA's Dynamic frequency analysis input data form.

This is my problem. If you see my function it contains only sin components. So I think I have to use symbolic definition instead of table. But I haven’t been able to define my function in this way yet.

I also think that I don’t have to define loads. There are no loads in my model. I think this is a Base Excitation problem. In this situation loads don’t need to be defined.
Six modes can be a good starting, but we can’t be sure that they are enough. It depends on the m.p.f (mass participation factor), that can be calculated during the D. Freq. Analysis. It has to be at least 80%. If it doesn’t reach it, we have to increase the number of the calculated natural modes.

 

[ghgarzon]

You're right,I was thinking it was a Dynamic Time Analysis. For Dynamic Frequency Analysis my only experience is with pseudo random inputs that can easily be defined by straight lines in the Amplitude vs, Frequency domain.


You don't have to define loads, it can be defined as a base excitation.


Since you have the Fourier series, can you input the amplitude vs.f centered at a box of width dF? (dF a bandwith that you defiene arbitrarily). Then you get a profile of small step functions of width= dF. Yes, it's an approximation so it's 99% of what we do as engineers.


Good luck