Return to search

Motor Control System for Near-Resonance High-cycle Fatigue Testing

This research project develops a low-cost high-cycle fatigue (HCF) testing system comprised of an AC motor, variable frequency drive (VFD), eccentric cam, and feedback controller. The system acts as a forced harmonic oscillator leveraging mechanical resonance to vibrate a specimen at a frequency required to induce the testing's strain amplitudes.
This system depends highly on the material being tested. As such, the controller incorporates material characteristics. A frequency sweep measures the strain amplitude to characterize the specimen. Additionally, other measurements such as acceleration can be used as a proxy control variables for strain. A function converts the control variable to frequency. This function tunes a proportional integral derivative (PID) controller to emphasize stable control. This function, coupled with a tuned PID controller, converts the correction update into a voltage signal that commands a motor speed to reach the desired strain amplitude.
Testing showed that a longer feedback loop time of 5 seconds was necessary to adequately control the system since the control variables are oscillatory by nature and need to be averaged over time to estimate accurate updates. Also, specimens with low damping are more subject to transient effects; consequently, rapid updates degrade system performance.
Overall, the system tested over 250,000 cycles and various specimens. The main limitation of the system is a maximum strain amplitude limited by the specific specimen resonant peak. However, adjusting the system's fixed displacement enables transferring more force to the specimen, changing the shape of the resonant peak.

Identiferoai:union.ndltd.org:ucf.edu/oai:stars.library.ucf.edu:honorstheses-1997
Date01 January 2021
CreatorsArmaly, Samer K
PublisherSTARS
Source SetsUniversity of Central Florida
LanguageEnglish
Detected LanguageEnglish
Typetext
Formatapplication/pdf
SourceHonors Undergraduate Theses

Page generated in 0.0022 seconds