AT503622A2 - METHOD FOR DETERMINING FORCES ARISING BETWEEN A RAIL VEHICLE AND A RAIL - Google Patents

Description

«·» J · 9 ·· ·· 5

The invention relates to a method for determining forces occurring between a moving rail vehicle and a rail using a strain gauge and evaluation means comprehensive measuring system, in which the Deh-10 voltage gauges are attached to an axle shaft and / or Radschei Ben having wheelset of the rail vehicle, the The measuring system is then calibrated to obtain calibration coefficients, strain changes in the rotating coordinate system are detected by the rotary strain gauges during the travel of the rail vehicle, the detected strain changes in the rotating coordinate system are transformed into a stationary coordinate system yielding strain changes in the rail-fixed coordinate system and the forces can be calculated from the 20 strain changes in the track-fixed coordinate system using the calibration coefficients, with an error error in calibration is performed to suppress a ripple. 25 Such a method is from the article published in ZEVrail Glasers Annalen, 126 - 5 I on pages 190 to 199, entitled "The lateral forces of the high-performance locomotives of the 1016/1116 series". already known by Werner Breuer and Peter-Jürgen Gaede. In the method described there, the forces acting on a wheelset 30 of a rail vehicle are examined. Here, a strain gauge comprehensive measuring system is used, wherein the strain gauges are attached to the wheel discs of a wheelset. Especially at high loads high forces are introduced into the wheel sets, which lead to elastic deformation of the wheel disc. The deformations of the wheel discs are detected by the strain gauges. Assuming a linear relationship between the strains or strains measured by the strain gauges and the forces causing the deformation, the forces can be determined by solving a linear system of equations from the measured strains. To solve the equation system, however, calibration coefficients are to be determined in a calibration step. 10 One of the difficulties of the measurement is that the strain gauges rotate while traveling with the axle or the wheel discs, so that the strain measurements obtained must be converted into a rail-stable or, in other words, coordinate system. By multiplying the strains with a sine or cosine function, which are measured over time, and by appropriate addition of such products, a periodic but unfortunately only a first approximation sinusoidal strain signal is obtained. Rather, the strain signal thus obtained has a so-called residual ripple, which is called in English Ripple. To suppress the ripple a targeted calibration of the wheelsets is proposed. More detailed explanations for targeted calibration, however, can not be found in the said publication. 25

Further methods and devices for determining the forces between the wheel set of a rail vehicle and a rail leading to the wheelset are described in the articles "Radsatzwelle and Radscheibe - the right combination for measuring the forces between wheel and rail". by Hermann Berg, Gustav Gößling and Herbert Zück, published in ZEV + DET Glas. Ann. 120 (1996) No. 2, page 40, "The current state of development of the measurement method" Radsatzwellenverfahren ". for the determination between wheel and rail " by Max Ostermeyer, Herman Berg and

3 * • · · ♦ • · · ·

Heinz-Herbert Zuck, published in ZEV glass. Ann. 102 (1978) No. 2 and "Determining the forces between the wheel and rail from the bending expansions of the wheelset shaft". by Michael Zeilhofer, Günter Sühsmuth Günter von Piwenitzky, published in the ZEV-5 glass. Ann. 96 (1972) No. 12, page 373.

The object of the invention is to provide a method of the type mentioned, with which a ripple can be suppressed. 10

The invention solves this problem according to a first variant in that the calibration includes the determination of error forces as a function of a Radaufsatzpunktes and a rotation angle of the wheelset, wherein the error forces from the difference between the 15 measured between the forces measured during calibration and using the calibration coefficients a Fourier transform of the error forces to obtain ripple order constants, from the ripple order constants 20 are calculated by inverse Fourier transform ripple correction constants, and the ripple correction constants are subtracted from the calculated strain changes in the rail fixed coordinate system. The invention solves this object according to a second variant in that the calibration coefficients are detected as a function of a rotation angle of the wheel set to obtain angle-dependent calibration coefficients, wherein the forces from the strain change are calculated using the angle-dependent calibration coefficients.

According to the invention, those skilled in the art are provided with options with which they can already take precautions in the form of applied data in order to calibrate the wheelset to be examined, with which it is possible to suppress the unwanted residual ripple , According to the invention thus a more accurate analysis of the forces between the rail and wheel is possible. The hitherto existing risk that rail vehicles will be denied approval due to excessive forces between wheelset and rail, although the exceeding of the predetermined tolerances was caused only by measurement errors, is thus reduced within the scope of the invention. 10

Advantageously, the forces determined in the context of the invention comprise wheel attachment forces which act in the vertical direction between wheel set and rail, guide forces Y acting transversely to the wheel attachment forces, and tangential forces T acting in the direction of travel, that is to say perpendicular to the plane is spanned by the wheel attachment forces Q and the Radführungskräften Y.

Further expedient refinements and advantages of the invention are the subject matter of the following description of exemplary embodiments of the invention with reference to the figures of the drawing, wherein the same reference numbers refer to the same effecting components, and FIG. 1 shows a schematic representation for defining the forces and used coordinate systems shows

FIG. 2 shows a schematic representation of a wheel disk in a plan view for the definition of 30 variables used,

3 shows the wheel disk according to FIG. 2 in a sectional partial view for the definition of variables used, ························································································ · · · · · · · · ♦ ♦ ··· ··· • * »· · · t · · · ·« · · · ♦ 5 10 15 20

figure

figure

Figure figure figure

Figure 4 is a schematically illustrated wheel disc with a clear number of strain gauges with associated variables, 5 a strain measurement of two strain gauges according to Figure 4, the strain gauges rotate, 6 a convenient addition or multiplication of the strain measurements of Figure 5, 7 a Fourier transform the strain change according to FIG 6, 8 the error forces determined during calibration as a function of the wheel attachment point and the rotation angle of the wheel set, FIG. 9 a Fourier transformation of the error forces according to FIGS. 8 and 10 to FIG. 12 show the calibration coefficients determined during calibration as a function of a rotation angle of the wheel set. 25

FIG. 1 shows parts of two wheel disks 1 and 2 of a wheel set and illustrates a rail-fixed coordinate system Xo, Yo, Zo and wheel contact forces Q, guide forces Y and tangential forces in the direction of travel, which are designated here by Tx. Further, a displacement of a Radaufsatzpunktes 3 is defined with respect to a designated eo measurement plane.

FIG. 2 shows one of the wheel disks 1 according to FIG. 1 in a plan view, whereby the coordinates x and z as well as a rotation angle φ can be recognized.

FIG. 3 shows parts of the wheel disc according to FIG. 2 in a cross-sectional view, wherein the coordinate designated by z lies in the cross-sectional plane eo according to FIG. The wheel base is in the embodiment shown in FIG. 3 shifted out of the measuring plane e0 and designated 3. The wheel contact forces Q are therefore introduced in the direction of the arrow in the designated 3 Radaufstandspunkt. The Ver-5 shift of Radaufsatzpunktes with respect to the measuring plane eo is again denoted by Ay.

FIGS. 4 to 7 illustrate a simplified measurement and its results for a better understanding of the preferred method. Thus, Figure 4 shows the top view of a wheel disc 1 according to Figure 2, wherein strain gauges 4 are arranged at a constant radius with respect to a Radscheibenmittelpunktes 5. In this case, the strain gauges xo and xi80 lie opposite each other on a straight line in the x direction and 15 span with respect to the Radscheibenmittelpunktes 5 at an angle of 180 °. The same applies to the strain gauges zo and ziso, which are arranged opposite one another on the z-axis. In the ferry mode, it comes to the rotation of the wheel disc 1 and thus to the rotation of the Deh-20 voltage measurement strip 4. The measured by the strain gauges 4 strain changes of the wheel disc 1 are thus dependent on the rotation angle Φ of the wheel disc 1.

FIG. 5 shows the strain change in the rotating coordinate system 8rxo of the measuring strip and the change in strain in the rotating coordinate system srxi8o of the measuring strip Χχβο as a function of the angle of rotation Φ. Upon rotation in the direction shown in FIG. 2 by approximately 90 °, the strain gage x0 undergoes its greatest change in strain. At this angle of rotation, the strain gauge Xo adopts the position of the strain gauge Ζχ8ο. At a rotation angle Φ of about 270 °, the strain change in the rotating coordinate system of the strain gauge Xo again passes through a lo- ······················································································· ···································································································································································································································· of 90 ° is greatly attenuated. With a rotation angle of φ 270 °, the strain gauge designated x0 in FIG. 4 assumes the position of the strain gauge Zo shown in FIG. Corresponding strain changes in the rotating coordinate system result for the strain gage xieo ·

FIG. 5 shows only the measurements obtained from the strain gages xo and χχ80. Analogous measurement curves are obtained for the strain gauges z0 and zi8o, but the strain changes are out of phase. When converting to a rail-fixed coordinate system, the deh-15 nungsänderungen in the rotating coordinate system of the arranged on a disk segment offset by 180 ° measuring points are added in a first processing step. The resulting rotational expansion curves srx, o, i8o and srz, o, i80 are shown in FIG. 6 in the upper part of the illustration. It will be appreciated that the resulting spa forms are balanced with respect to their maxima and minima and are periodic over a full rotation of the wheel disc. The

Strain change in the rail-fixed reference frame is determined by transforming the strain changes in the rotating coordinate system according to g = "0> i * o cos (O) - £ r Λθ 180 sin (T). The corresponding curve is shown in the lower part of FIG.

FIG. 7 shows a Fourier transformation or, in other words, the spectrum of the rail-fixed strain changes g * according to FIG. 6. It can be seen that gr has a residual ripple and is composed of different frequencies.

. As can be expected from the £ circulation, the dominates the

Ripple with four times the wheel speed the £ ° r signal.

However, the two-, six- and eight-fold frequencies of the wheel speeds with significant amplitudes in the Restwel- 5 ligkeit involved. The distribution of the multiples of the frequencies of the wheel speed is determined by the position of the measuring points on the wheel disc and, moreover, by the displacement Ay of the wheel attachment point in the transverse direction. For example, measuring points close to the flange lead to higher amplitudes at six and eight times the wheel speed, since here the influence of the wheel contact forces Q on the pulley expansion force is more pronounced than with radii lying further inside. 15 The residual ripple in the measured strain changes in the rail-fixed coordinate system falsify the results of the running-technical measurements, since these strain changes are converted into the desired wheel forces via the solution of special systems of equations. The conversion is generally performed by means of special measuring computers, which solve the equation system in real time during the technical driving admissions. The following equation 1 gives an example of the solution of a wheel disc equation system with known wheel contact force Q: 25 £ \ ~ K. \, Q 'Q ~ / Cl, Y' ^ JCl, Qe 'ö' Av (1) £ l ~ K.1q Unknowns of the above system of equations are the guiding force Q and the position of the point of contact on the running surface Ay, where Ay = 0 corresponds to the measuring circle plane ε0 according to FIG. 1. . Ει, the strain change in the measuring plane 1, and 82, the strain change in the measuring plane 2. kiq, K2ßi tQe and K2Qe are calibration coefficients. With

5 K \ q · Q} ε2 - ε2 κ2 q · Q follows

10 and

If the measurement error caused by the residual ripple is not corrected, a measurement error of up to 10 kN, for example, can occur for the tracking forces Y. Errors on such a scale are undesirable because these orders of magnitude may be of relevance to the registration of the rail vehicle. FIG. 8 illustrates a first method step for eliminating the residual ripple according to a first variant of the invention. For example, the errors in the determination of the tracking forces are determined. This is done by direct measurement of the tracking forces by means of a Kalib-25 riervorrichtung, for example, in ZEV + DET glass. Ann. 120 (1996) No. 2, page 44 and which is intended to be part of the disclosure by this reference. In a second step, the calibration forces Y calculated using the calibration coefficients obtained. The 30 error forces are by simply subtracting this

Tracking personnel received. In FIG. 8, the error forces are ······························································································ , Depending on the position of the Radaufsatzpunktes, which is removed at the axis 6, and depending on the rotation angle of the wheelset, on the Axis 7 is shown removed. FIG. 9 shows the standardized Fourier transformation of the error forces shown in FIG. 8, wherein the order of the residual ripple of the tracking forces Y is plotted on an axis 8. The axis 6 again corresponds to the position of the Radaufsatzpunktes. The Fourier coefficients 10 obtained in this way are stored as parameters in a memory unit of a computing unit so that the residual ripple on the basis of the stored Fourier transformation data can be suppressed in a later real-time measurement. FIG. 10 shows an exemplary embodiment of the error correction according to the second variant of the present invention. In this case, the calibration coefficients are determined not only for a specific rotational angle .phi..sub.phi of the wheel set but for different rotational angles .phi. Of the wheel set. FIG. 10 shows the cablization coefficients Κ 1, γ obtained for the calculation of the tracking forces as a function of the angle of rotation Φ.

11 shows the calibration coefficients KI (Q for the wheel contact forces Q as a function of the rotational angle Φ and FIG. 12 the calibration coefficients κΙ> 0β for calculating the torque which results from the transverse displacement Ay of the vertical force Q on the tread. By taking into account the residual ripple already in the determination of the calibration coefficients, the ripple in the real-time measurement is almost completely suppressed, so that a corresponding error is minimized.

Claims (3)

· · ··· 1. A method for determining forces occurring between a moving rail vehicle and a rail with the use of a strain gauge (4) and evaluation means comprehensive measuring system, wherein the strain gauges a wheel axle of the rail vehicle having an axle shaft and / or wheel discs (1, 2) is subsequently calibrated, the measuring system is subsequently calibrated to obtain calibration coefficients, strain changes in the rotating coordinate system are detected by the rotating strain gauges during the travel of the rail vehicle which detected Strain changes in a stationary coordinate system can be transformed to yield strain changes in the rail-fixed Ko-15 ordinate system and the forces from the strain changes in the rail-fixed coordinate system using the calibration coefficients, wherein an error correction to suppress a residual ripple is performed during calibration, characterized in that the calibration comprises determining error forces in dependence on a wheel attachment point and a rotation angle of the wheelset, the error forces being from the difference a Fourier transform of the error forces to give ripple order constants, from the ripple order constants are calculated by an inverse Fourier transform ripple correction correction constant and the ripple correction constants are calculated from the calculated strain changes in the rail-fixed coordinate system are deducted. ····· ·········· • ♦ · • ·· ·· 2. A method for determining forces occurring between a rail vehicle and a rail using a strain gauge (4) and evaluation means comprehensive measuring system, in which the strain gauges (4) on a 5 axle and / or wheel discs (1, 2) having wheelset of Rail vehicle, the measuring system is subsequently calibrated to obtain calibration coefficients, changes in the rotating coordinate system are detected by the rotating strain gauges during the travel of the rail vehicle, the detected strain changes in the rotating coordinate system are transformed into a stationary coordinate system yielding strain changes in the rail-fixed coordinate system and the forces from the strain changes in the rail-15 fixed coordinate system are calculated using the calibration coefficients, wherein an error correction to the undershoot is performed during calibration CKEN a ripple is carried out, - characterized in that 20, the calibration coefficients are detected as a function of a rotational angle of the wheel set yielding angle-dependent calibration coefficients, wherein the forces from the change in elongation in the rail-fixed coordinate system is calculated using the angle-dependent calibration coefficients are 25th 3. The method according to any one of the preceding claims, characterized in that acting in ver tical direction between wheel and rail wheel 30 tower forces acting in the transverse direction to executives Y and acting in the direction of travel of the wheel set tangential forces T are calculated.