Hi, I'm working on Abaque.cae, using the step "steady state dynamics" in order to optain a FRF from my FEM model which is a machine tool, and also to simulate the damping effect on the frf, by inserting concentrate damping on grease lubrificated roller linear guides, bearings and nut, modeled as viscous dashpot dampers.
I've discovered that the numerical FRF is similar to the experimental one, only by inserting enourmous values of viscous damping in the model (8000 Ns/mm for roller linear guides which are more influent than bearings and nut as a damping source of the machine tools).
after that I try to insert structural damping in the model and I found that a value of 0.9 inserted on the roller linear guides, matches my numerical FRF to the experimental FRF.
from the abaqus documentation a formula that convert structural damping in equivalent viscous damping is riported. (see the paragraph "Structural damping in modal dynamic analysis" in the link attached)
by applying this equation I've discovered that 0.9 correspond to 40 Ns/mm, which is different than 8000 Ns/mm...
my questions are:
is this convertion equation reliable?
which type of damping is best suited in frequency domain analysis in abaqus? viscous or structural?
what are the typical values of damping in lubrificated roller linear guides in machine tools?
thanks a lot
http://www.egr.msu.edu/software/abaqus/Documentation/docs/v6.7/books/stm/default.htm?startat=ch02s05ath28.html
Hi
I have not used Abaqus since the mid 90's, so I am rusty to put things mildly.
For what it is worth, damping has always been poorly treated in Abaqus as it primarily relies on the Rayleigh damping model. The problem with this model is that it is defined purley for mathematical convenience and is lacking a physical model, i.e. it does not even try to mimic a real life damping phenomenon.
Abaqus allows also the use of modal damping values that are input as % of crtical damping. I would stick to this damping model whenever possible.
It is not possible to follow exactly what you are doing from your description. Ns/mm would be the value for a viscous dashpot damper. But, for the system you are describing, would your damping phenomenon truly be viscous? I imagine viscous damping may appear if you are squeezing liquids through narrow gaps.
If not, Coulomb damping or microplasticity sound more plausible mechanisms to me. These would require non-linear time domain analysis which is right up Abaqus alley.
I do not know if an expanded discussion on damping isof interest, but it can be found here should you wish to read it. http://qringtech.com/2014/06/22/designed-damping-types-mechanisms-application-limitation/
Just my 2 cents
Claes
Many thanks for the answer,
the equation i mentioned in my previous answer works only for Sdof systems and for system whose modal matrices are diagonal...
which is the relation between structural and viscous damping for MDOF system, in the case of coupled(non diagonal) modal matrices?
Hi Matteo
Ccrit = 2sqrt(M*K). 2*C/Ccrit = eta, where eta is the structural damping loss factor.
This relation holds at resonance.
More infor here: http://en.wikipedia.org/wiki/Damping_ratio
Sincerely
Claes
Hello Matteo Ubertini:
I don't like accessing urls on RG hence I have not checked the formula provided by Abaqus. Structural damping is accounted for using complex structural stiffness defined by K* = K(1+ iβ) where β is the loss factor (In most literature it is written as 2β). Viscuous damping force is defined by c*dy/dt where c is coefficient of damping and y is displacement. The relation between c and β is
c = βMωn2/ω
where ω = frequency of forcing function in rad/sec, ωn is the natural frequency of a mode of vibration of the system and M is the mass. This formula allows you to assign appropriate damping for each mode of vibration of the system by using the correct ωn. If M is in unit of Kg, c will have unit of Ns/m. I hope you will find this answer useful.
hi dears
I want to simulate acoustic anechoic layer with loss factor in Abaqus software.
but I dont know how enter loss factor in material property. do I enter loss factor in structural damping value or must use a formula for transfer it to another parameter?
please help me.
best regard.
sohrabi
- Badges
- Science topic