Development and field testing of a prototype hybrid uniaxial strain transducer

Hosin Lee; Hae-Bum Yun; Brian Maclean

doi:10.1016/S0963-8695(01)00035-4

A Uni-Axial Strain Transducer (UAST) is a Micro-Electro Mechanical System (MEMS) with characteristics such as high resolution and high sampling rate, absolute encoding, no calibration requirements, no drift over time, and less measurement noise than analog-based strain sensors. The research goal is to develop a prototype Hybrid Uni-Axial Strain Transducer (Hybrid UAST) that includes non-volatile RAM to store strain cycling history (e.g. tracking how many times the UAST crosses each of specified strain thresholds across its dynamic range), and to temporarily store the preprocessed data. The objective of this research is to determine the potential of the Hybrid UAST as a new tool to continuously monitor, analyze, and store the strain history of components such as rail. The resulting strain data can be periodically downloaded and used for such purposes as measuring rail stress induced by axle or thermal loadings. The prototype Hybrid UAST consists of three parts: a UAST sensor, a networking controller box, and a communication cable. A load cycle counting algorithm is integrated into a microcontroller, which is programmable using configuration switches. A prototype Hybrid UAST package suitable for field application has been fabricated. In addition, two computer programs were developed for data acquisition and data analysis. The data acquisition program was used to operate the Hybrid UAST from a laptop computer, and the data analysis program was used to implement the peak searching and cycle counting algorithms. Raw data were collected to verify the cycle counting algorithms implemented in the Hybrid UAST in an outdoor operating environment. A prototype Hybrid UAST design and load cycle counting algorithm has been tested using actual strain data taken from a rail at the field test sites in Salt Lake City and Iowa City. The raw data collected at 290 Hz (14-bit mode of UAST) without a train consistently showed a standard deviation of around 1.25 με. In other words, approximately 99.7% of background noises are less than 3.75 με. This level of error can be considered small relative to a peak strain range of 400 με caused by a typical trainload. The raw data collected at the same frequency with a train were also processed by the Hybrid UAST to accurately determine not only the number of load cycles but also the magnitudes of the peak loads, which are consistent with both laboratory measurements and theoretical calculations. In conclusion, field testing of the prototype Hybrid UAST with respect to its repeatability, accuracy, and viability in hybridization can be considered a success. The positive results from this study warrant a continued research effort to develop a more refined commercial-grade Hybrid UAST that is sufficiently rugged to withstand harsh railroad operating environments.

Development and field testing of a prototype hybrid uniaxial strain transducer

View Online

Abstract

Details

Metrics