Keywords: diagnostics, measurement, wheel sets, axle sets, railway, circular errors.

This paper presents the results of simulation research on application of encoders for identification of the diameters of railway vehicles in motion. These encoders were developed at the Silesian University of Gliwice [1]. The encoder measures an angle. Identification of the diameter consists in finding such theoretical diameter of the wheel for which theoretical time performance of the angle differs from the experimental as little as possible.

Basic idea of proposed methods based on a roller staying in contact with the axle wheel passing measuring unit. Measuring set is equipped in one or more measuring units. The basic problem for the point of view of application proposed methods is sensitivity of the encoder. The question is as follow: what difference of diameters may be detected using such encoder? For that reason we conducted some simulation using worked out mathematical models and the software Excel.

A strategy of diameter identification of a wheel being in motion consist of two stages:

• preliminary identification of the diameter D and speed V,
• exactly calculation of the diameter D basing on the registered signals from encoder.

Preliminary identification of the speed V may be realized by calculation, for example basing on the passes way between two measurement units and measured time. The second method of diameter identification bases on iteration. Using the geometrical model of alpha(t) we are looking for the best value of diameter D for which the summarized error epsilon, it means summarized difference between theoretical alpha(t_i) and obtained from experiment alpha_m(t_i) is minimum. We have done several experiments [2] measuring the time performance of the angle alpha(t) for different value of speed V. We may observe the experimental time performance, theoretical and difference between them.

As results of computer simulations and experimental verification we can say that:

• It is possible to detect diameter difference by comparing theoretical and real signals from proposed measuring stand.
• Proposed method is easy to application.
• Experimental results show that detection of different angles alpha below 15-20° comparing theoretical and measured signals from encoder is very hard, what means that detection of two diameters differs below 0.5mm is questionable.

1

A. Kolka, "Method of diameter identification of wheel set being in motion", PhD thesis, Silesian University of Technology, Gliwice, Poland, 2002. (in Polish)

2

A. Kolka, "Experimental stand for geometrical research of objects being in motion", University Press of the Department of Machine Technology of Silesian University of Technology, Gliwice, Poland, 2000 (in Polish).

purchase the full-text of this paper (price £20)

go to the previous paper
go to the next paper
return to the table of contents
return to the book description
purchase this book (price £110 +P&P)