Learning Forum

[Olly]

I am conducting harmonic analysis in ANSYS 2024 R2 to study the vibration response of a rotor-bearing system under different operating conditions. My goal is to validate the simulated frequency response against experimental FFT results to identify bearing defect frequencies.

In the experiment, I tested the system at different speeds (900, 1200, 1500 RPM, etc.) with varying loads (0, 10, 20, 30 N). Using vibration sensors, I collected time-domain data and applied FFT to extract frequency-domain characteristics. Now, I want to replicate this in ANSYS to compare the results.

In ANSYS, I modeled a shaft and rotor assembly supported by ball bearings at both ends. The bearing housings were fixed, ensuring proper constraint of the system. A load(force) was applied to the rotor (in experiment,load was applied by using a hanging weight by some arrangement).

Challenges I'm Facing:
🔹 Applying Velocity Input – How should I define the correct velocity input for harmonic analysis? ( can't have rotational velocity in harmonic)
🔹 Extracting Frequency-Domain Data – What is the best way to extract and process frequency-domain results from ANSYS?
🔹 Experimental Validation – How can I accurately correlate simulated harmonic response data with experimental FFT results?
🔹 Speed & Load Variation Effects – How do these affect harmonic response, and how should I simulate their influence in ANSYS?

HOW DO I DO IT?

[dlooman]

When you specify a rotating force in Mechanical the program automatically specifies a rotational velocity equal to the harmonic frequency. 

After a harmonic analysis you can request "Frequency Response" for displacement at some location in the model.

It will be hard to validate the analysis vs experimental data because the true imbalance isn't known and usually damping isn't known either.  Also, it's not obvious to me how to compare FFT results with harmonic results.

 Harmonic results will peak at the natural frequency.  The unbalance force will be proportional to frequency^2.