modal analysis following a thermal load

[davideaka]

Bye to all,
I am trying to assess the variation of resonance frequencies as a result of an applied thermal load.
the solution offered by statsub does not lead to any result.
I was thinking of using the sol 400 for static analysis and then, in a second step, perform modal analysis, manually inserting in the .bdf the appropriate commands.
However I cannot generate a correct code.
I press that I have a version of patran of 2010.
suggestions?
Thank you.

davide

 

[davideaka]

the structure is formed by elements cbeam

 

[Onda]

I don't get anything back.
in a modal analysis you evaluate the relationship between the masses and rigidities. exemplifying a lot, for a single-dimensional system is root of k/m (k spring stiffness and m mass).
Now if you apply loads (which are thermal, or forces or pressures), do not alter the masses, or stiffness. so modal analysis doesn't change. changes to the change of constraints because of course they act on stiffness.
all this is valid for linear systems, for which the application of loads does not change the geometry and/or rigidities of the components in order to vary considerably the matrix of rigidity, (the masses I suppose do not vary).

in case you have to perform a modal analysis of a non-linear structure, you must first solve by performing a non-linear analysis and then do the modal by making a restart with initial conditions those at the end of the non-linear analysis.
find the procedure in the help related to nonlinear analysis - restart (it seems to me). I don't know how to help.

 

[davideaka]

I don't get anything back.
in a modal analysis you evaluate the relationship between the masses and rigidities. exemplifying a lot, for a single-dimensional system is root of k/m (k spring stiffness and m mass).
Now if you apply loads (which are thermal, or forces or pressures), do not alter the masses, or stiffness. so modal analysis doesn't change. changes to the change of constraints because of course they act on stiffness.
all this is valid for linear systems, for which the application of loads does not change the geometry and/or rigidities of the components in order to vary considerably the matrix of rigidity, (the masses I suppose do not vary).

in case you have to perform a modal analysis of a non-linear structure, you must first solve by performing a non-linear analysis and then do the modal by making a restart with initial conditions those at the end of the non-linear analysis.
find the procedure in the help related to nonlinear analysis - restart (it seems to me). I don't know how to help.

I certainly expressed myself badly:
I'm trying to figure out how the resonance frequencies of a structure vary from beams to temperature change.
if a structure undergoes a delta of temperature such as what can occur between summer and winter will surely deform, going to change its characteristics of stiffness and therefore its own frequencies (certain not very much).
I probably fall in non-linear conditions.
Thank you.

 

[Onda]

That's exactly what I don't understand.
if deformations do not change the characteristics of stiffness. then the modal does not change. if instead you have variations of stiffness, fall into the nonlinear field and may vary the modal. being this determined only by rigidity and masses and not by the imposed loads. temperature differences create deformations that in turn are seen as loads from the structure.

first you have to solve a static, evaluate the deformed, then solve a non-linear and compare the deformed with the static.
At that point you can evaluate if you have important variations between the two analyses and determine if you need to do a non-linear.
if you then have influence of temperature on modal, otherwise the influence is nothing.

I'll give you an example. if you have a light material panel, bound to the four sides by robust beams of different material with different thermal expansion coefficient, the panel at temperature range sees compression loads (or traction). if the loads are of compression the panel could go in buckling and before going it would be in conditions of low stiffness due to the effect of softnening stress and therefore its modal would be different. as you see we are talking about very particular situations.
Maybe if you tell us how your structure is made we can better understand.

 

[davideaka]

It's just a framed beam.
the fact is that if I extend the beam of the amount of which it stretches as a result of the thermal load (obtained with a static analysis) then the frequencies change (though slightly).
I also inform you that the two analyses (static and non-linear static) lead to the same deformation values.
I am analyzing the structures and therefore I could not easily repeat this procedure.

 

[Onda]

Okay. So you are evaluating the fact that being the longer beam its stiffness changes, correct?
that is a beam of 1m of steel that undergoes a delta t of 50° stretches 0.6mm (a=12e-6). and the fact that it stretches of 0.6mm entails a different stiffness. Actually you should also consider that it involves a different density, since the material has expanded or contracted but the weight has remained that.
I think you're evaluating values that are beyond the engineering interest in the sense that you're evaluating a length variation of 6 for ten thousand with a delta t of 50 degrees.
However both, through a normal modal this effect is completely neglected.
If you want to see the change, you may try to make a restart from a non-linear analysis. at that point the matrix of rigidity is that of the deformed structure and therefore the beam is of the correct length. But take into account that it does not seem to me that the mass in the fuction of the change in volume. It's a detail I've never taken into consideration, considering these effects negligible.

 

[davideaka]

Nothing but thanks.
Good day