I performed modal analysis using Ansys and experimental modal analysis on a wind turbine blade.
The deviation between FE and experimental measurement for higher modes (third mode shape) is so big
274 Hz from ansys and 306 Hz from experiment.
If you define error as the difference between an exact analytical solution and your finite element solution, then error is larger for higher modes because the element shape functions provide a better basis (approximation) of the shape of low modes and a relatively poorer basis for higher modes.
However, you are comparing results from a finite element model and an experiment. In this case, there are several sources of discrepancy. The usual suspects are (1) boundary conditions that you are using in the model do not accurately reflect the real conditions of the experiment and (2) the model is missing non-structural mass. In your case, these sources of error would probably lead to a higher (rather than lower) frequency. Note, a comparison of mode shapes can be useful to help identify the spatial location in the model that needs to be improved.
Finally, you should at least consider the numerical precision and potential for error in the data analysis performed on experimental data to identify the frequencies.
Thanks Sir,
The first and second modes are not deviated so much
1ST Mode :
FE: 32.59 Hz
Experimental: 32.5 Hz.
2nd mode :
FE: 118.6 Hz
Experimental: 116.3 Hz
So I think if there was problem in the boundary then it would be appeared for all mode shapes so as the missing mass element!
If you have non proportional discrepancies between frequency errors in modal correlation, you have several things to check.
First, how did you pair your modes ? You may have experimentally missed some modes due to poor excitation / observation setups for specific modes. Did you pair using the MAC,...
Second, if you have confidence on the third mode pairing, how clean is the measure, what is the quality of the experimental identification ? (noisy peaks, response level...).
Third what area in your numerical model is bearing the discrepancy (are there coupling approximations, or boundary conditions that are more sollicited by this thrid mode). Do you have oversimplified your model's geometry ....
Remember also that, since the FE are using an implicit shape function, that works as an increase in stiffness: that is why FE values tend to be higher that the real (experimental) ones, even is everything else was perfect. Besides that, I agree with the comments from Darby and Guillaume.
I tend to agree with your III frequency having about 11% error - But I doubt the small deviation of less than 1% for 01st and 02nd frequencies between FEM and experiment - this small variation is valid for a perfect cantilever beam of regular geometry, with beam elements and with perfect B.Cs - again you cannot expect this close accuracy with solid element tetra mesh- did you perform mesh independancy test for FE analysis
Next how did you know what is the third mode shape experimentally to compare with FEM - check for MAC - what is the kind of experiment done and what is the response sensor, howmany and at which locations - this requires more elaborate discussions
It is true, MAC, in any of its variants, requires FRF curves enough to distinguish among the several mode shapes. Any practical FEM simulation will easily provide enough of these, but experimental FRF are a different problem. Anyway, a hint that usually works is to use the FEM-simulated mode shapes to select the number and location of the sensors in such a way that all mode shapes within the frequency range of interest are clearly distinguishable.
Just as Darby said, a comparison of mode shapes can be useful to identify whether the two third-order modes are the same thing. A experimental modal analysis is likely to lose mode information and maybe the third-order is the forth-order. Also, the coincidence of the first two orders cannot prove the correctness of FEM, maybe the first two orders are not sensitive to the wrong FEM, however , the third order mode is probably sensitive.
One of the reasons that may affect the mode shapes pairing is the lack of measured DOFs with respect to the ones of the FE model. This difference can weigh considerably when dealing with higher modes since they mostly reflect local responses of the system, thus the number of measurement points play an important role for the accuracy of the estimation. However, to overcome this drawback you can reduce the FE mass matrix down to the set of test DOFs, by means of a pseudo orthogonality check (POC) for instance.
- Badges
- Science topic