Elevator Ride Quality
Group Project
Project Mentor: Dr. Reza Tafreshi, TAMU
Spring 2014
This was a semester-long project undertaken as part of the Mechanical Measurements course. We measured the ride quality of the elevators at Texas A&M University at Qatar using ISO 18738-1 standards.
Motivation
This sophomore project was very much an exercise in learning to measure vibrations and building the right test setup. One of the elevators at the university was notorious for sometimes jerking violently at the third floor. We found an ISO standard (ISO 18738-1: Measurement of Ride Quality) which outlines the data collection procedure for measuring acceleration, sound levels etc., and also found some reference values online for these quantities. We decided to test the quality of the elevators at the university by collecting data as per the standard and comparing it to the reference values.
The Project
We first used magnetic, high resolution accelerometers (Piezoelectric accelerometers from PCB Piezotronics). But to our dismay, the data was so noisy that no meaningful values could be obtained from them. We later found out that due to their high sensitivities, the accelerometers were picking up excessive noise. While tit might have been possible to filter them out, the ISO standard demanded a 10 Hz low pass, 2-pole Butterworth filter to be applied to the data and this filter wasn't sufficient.
So instead, we used a 3-axis wireless accelerometer from Vernier ( 'Vernier WDSS') that had a resolution of only 0.04 m/s^2, which was about 10 times worse resolution than the magnetic accelerometers used earlier. Data obtained after filtering are shown in Items 8a and 8b for one of the elevators.
Item 8a: Acceleration of the elevator vs time, while the elevator was going up. SI Units.
Item 8a: Acceleration of the elevator vs time, while the elevator was going down. SI Units.
Consequently, we collected data for all elevators and concluded that with 95% confidence, two of the elevators on campus did not live up to the reference standards on ride comfort.
We supplied a copy of our report to the Building Operations department, who then proceeded to ignore it. :D
The biggest lesson I learned during this venture is that the sensor with the best resolution may not be the best sensor for the task at hand. In our case, the highly sensitive (and expensive!) piezoelectric accelerometers were completely useless while a less precise sensor did the job perfectly.