JPH04191609A - Apparatus for measuring wavy abrasion of rail - Google Patents

Abstract

PURPOSE:To accurately measure wavy abrasion by a method wherein a truck is moved by the interval between a pair of the displacement sensors attached to the truck in close vicinity to each other and the distances between said sensors and the measuring places of the head top surface of the rail directly under the sensors are measured at each time when the truck is moved and the differences between the measured distances are cumulated and added. CONSTITUTION:Non-contact sensors 11a, 11b are attached to a truck 2 in close vicinity to each other and an encoder is provided to a wheel 3. Even when wavy abrasion is generated in the head top surface 1a of a rail 1, the displacement difference DELTAy between the measured distances XA, XB from the respective sensors 11a, 11b and respective measuring points P, P' is calculated without being affected by the up and down fluctuation of the sensors 11a, 11b. At each time when the truck 2 is moved by the sensor interval (l), the distances between the sensors and the measuring points are measured and the differences between the measured distances are cumulated and added to calculate the displacement quantity of the head top surface of the rail at each measuring place. By this method, the displacement quantity of the head top surface of the rail can be accurately calculated without receiving the effect of the up and down movement of the displacement sensors.

Description

DETAILED DESCRIPTION OF THE INVENTION (Field of Industrial Application) The present invention relates to a device for detecting wave-like wear that occurs on the top surface of a railway rail in the longitudinal direction of the rail.

(Prior Art) It is known that undulating wear occurs on the top surface of railway rails in the longitudinal direction of the rails. and,
This wave-like wear causes noise and wheel load fluctuations when the vehicle is running, so measuring the wave-like wear is an important matter in railway maintenance management.

Specifically, the measurement of such wavy wear is carried out by measuring the amount of displacement of the unevenness of the top surface of the rail in the longitudinal direction with respect to the reference surface to determine the waveform of the wavy wear. For example, the device shown in FIG. 10 is known as a device for performing inspection.

This inspection device consists of a bogie C that moves on rail a via wheels b, and a capacitance displacement meter, eddy current displacement meter, etc. that is attached to the bogie C at a position directly above rail a. Basically, the distance between the displacement meter d and the top surface e of the rail a is measured by the displacement meter d while the cart C is moved on the rail a, and thereby, The waveform of undulating wear is determined by measuring the amount of displacement of the unevenness of the crown surface e with respect to the reference surface.

In this case, when the cart C moves on the rail a, it moves up and down due to the wave-like wear that occurs on the top surface e of the rail a, the joints between the rails a, etc., and therefore the position of the displacement meter d also moves up and down. Since the distance varies, the amount of displacement of the top surface e of the rail a cannot be accurately measured unless the amount of change is added to the distance between the displacement meter d and the top surface e of the rail a.

Therefore, in this inspection device, an accelerometer f is installed on the truck C, and by double integrating the acceleration of the truck C detected by the accelerometer f, that is, the acceleration of the displacement meter d, The amount of vertical variation of the displacement meter d is determined and added to the distance measured by the displacement meter d.

However, in such a measuring device, since the amount of vertical fluctuation of the displacement meter d is determined by double integration of the acceleration detected by the accelerometer f, for each measurement at each measurement point on the top surface e of the rail a, Errors due to electrical or mechanical drift of the accelerometer f are doubly accumulated, and therefore,
It was difficult to accurately measure the displacement cover of the unevenness on the top surface e of the rail a.

(Problems to be Solved) The present invention eliminates such inconveniences, makes it possible to accurately measure the amount of displacement of the unevenness of the top surface of the rail with respect to the reference plane in the longitudinal direction, and therefore enables An object of the present invention is to provide a device that can accurately measure wave-like wear.

(Means for Solving the Problems) In order to achieve the above object, the rail wavy wear measuring device of the present invention measures the displacement amount of the unevenness of the top surface of the rail with respect to the reference surface in the longitudinal direction of the rail. This is a device for measuring wave-like wear that occurs on the top surface of the head of a truck, and includes a truck that moves on the rail via wheels, and a device that is movably attached to the truck at a position directly above the rail, and a device that is movably attached to the truck at a position directly above the rail. A pair of non-contact displacement sensors are arranged in parallel and close to each other with an interval in the longitudinal direction. a measuring means for measuring the distance between the sensor and a measurement point on the top surface of the rail located directly below the sensor;
For each measurement, calculate the difference between the constant distances on both sides measured by both displacement sensors, and cumulatively add the distance differences obtained by the calculation to calculate the amount of displacement of the top surface of the rail with respect to the reference plane at each measurement location. and arithmetic means for calculating.

Further, the present invention further includes a sensor output correction means for zeroing the outputs of both displacement sensors when the outputs of both displacement sensors at each measurement point are substantially the same as the outputs of the measurement point immediately before the measurement point. Features.

Furthermore, each measurement point is calculated by subtracting the average value of the cumulative sum of the predetermined number of measurement points before, after, or before the value obtained by cumulatively adding the distance difference to each measurement point. The present invention is characterized by comprising an arithmetic correction means for determining the amount of displacement of the top surface of the rail with respect to a reference surface.

(Function) According to this means, the measurement by the measuring means is performed each time the cart moves by the distance between the two displacement sensors, and when the cart moves up and down during its movement, both the displacement sensors are integrated. The difference between the two measured distances calculated by the calculation means for each measurement is:
While this is the difference between the displacement amounts at two adjacent measurement points on the top surface of the rail, this value does not include the vertical fluctuations of both displacement sensors. Therefore, for example, if the smooth flat part of the top surface of the rail on which the cart is located is set as the reference plane in the initial state before the cart is moved, the cumulative addition of the difference between the two measured distances obtained by the calculation means The value is the amount of displacement of the top surface of the rail with respect to the reference plane at each measurement location, and from this, the shape of the top surface of the rail in the longitudinal direction and, by extension, the waveform of the undulating wear that occurs on the top surface of the rail can be determined.

In this case, the outputs of the two displacement sensors generally include relatively small errors due to electrical or mechanical drift for each measurement. Therefore, if the above measurements are performed continuously, the output of each displacement sensor at each measurement point will be affected by the error at the previous measurement point, and as a result, the error will gradually become 1 in the output of each displacement sensor. In such a case, the output of each displacement sensor will gradually change even in a place where there is no wave-like wear on the top surface of the rail, that is, a flat place.Therefore, In such a case, when the sensor output correction means is provided in the measurement device, the output of each displacement sensor at each measurement point is substantially the same as the output at the measurement point immediately before the measurement point. At a certain time, that is, when the top surface of the rail is a flat surface, the outputs of both displacement sensors are forced to zero, so at a point where the top surface of the rail is a flat surface, the outputs of both displacement sensors are The output corresponds to the flat surface.

Furthermore, in each of the measurements by the measuring means, if both of the measured distances include an error,
The error is also cumulatively added by the cumulative addition of the distance difference by the calculation means, and the error included in the cumulative value, that is, the value obtained as the displacement lid at each measurement point on the top surface of the rail, gradually becomes larger. There is a risk. Therefore, in such a case, when the arithmetic correction means is provided, the average value of the cumulative sum at a predetermined number of measurement points before, after, or before the cumulative sum at each measurement point is subtracted from the cumulative sum at each measurement point. It becomes possible to reduce errors. In other words, the cumulative addition value at each measurement point is the sum of the actual displacement of the top surface of the rail at the measurement point and the error. Therefore, the above average value is the value before and after the measurement point. The value is the sum of the respective average values of the actual displacement amount and error in a predetermined section of . In this case, the actual displacement is
At the wavy wear part of the rail, the amount changes roughly sinusoidally, so the average value is approximately zero,
Furthermore, it is naturally zero at a flat location that serves as a reference surface. Therefore, the average value of the cumulative addition values is approximately the average value of the errors.

Therefore, as described above, by subtracting the average value of the cumulative sums before and after the cumulative sum at each measurement location, the error is reduced.

(Example) An example of the rail wavy wear measuring device of the present invention will be described with reference to FIGS. 1 to 9. Figure 1 is a side view of the measuring device, Figure 2 is a block diagram of the main parts of the measuring device, and Figure 3 is a side view of the measuring device.
6 through 6 are explanatory diagrams for explaining the operation of the main parts of the measuring device, FIG. 7 is a flowchart for explaining the overall operation of the measuring device, and FIGS. 8 and 9. is an explanation interval for explaining the measurement error of the measuring device.

In Figure 1, ■ indicates a rail, and 2 indicates wheels 3 and 4 on the rail 1.
This is a trolley that is movably supported via the in this case,
The trolley 2 is provided below the main vehicle 5 that moves on the rail 1, and is connected to the main vehicle 5 via a connection mechanism (not shown).
move above.

The trolley 2 has a main body 6 supported on the rail 1 via wheels 3, and an arm 7 extending in the longitudinal direction from the upper part of the main body 6 at a position directly above the rail 1, The arm portion 7 has a distal end supported on the rail 1 via a wheel 4, and is pivoted to the main body portion 6 via a support shaft 8 so as to be able to swing up and down. Then, the trolley 2 has its main body 6
and the tip of the arm part 7 are connected to piston rods 9a and 10a of air cylinders 9 and 10 suspended from the lower surface of the main vehicle 5, respectively, and these cylinders 9 and 10 constantly move toward the rail 1. It is energized and is pressed against the rail 1 via the wheels 3.4.

Further, in the figure, lla and llb are non-contact displacement sensors such as eddy current displacement meters or capacitance displacement meters, and 12 is a rotary encoder (hereinafter simply referred to as an encoder).

Displacement sensors lla, llb are connected to the rail 1 on a horizontal portion 7a formed on the arm portion 7 of the truck 2 at a position between the wheels 3.4.
The rail 1 is attached to the lower end of the mounting member 7b vertically installed toward the
It is fixed to the top surface 1a of the rail 1 with a predetermined distance therebetween, and is also placed directly above the rail 1 in the longitudinal direction of the rail 1 with a predetermined distance therebetween. In this case, both displacement sensors l
La, Ilb interval! (hereinafter referred to as sensor interval l) is a value sufficiently smaller than the interval L of wheels 3°4 (hereinafter referred to as wheel interval) (for example, ff1-10+
+un, 1°-1ooownwn).

These displacement sensors lla and llb each output a signal according to the distance from the top surface 1a of the rail 1 directly below them, and in this case, when the top surface 1a of the rail 1 is a smooth flat surface, the output signal is is initially set to be zero, for example.

The encoder 12 is attached to the axle 3a of the wheel 3, and outputs a pulse every predetermined rotation angle of the wheel 3, that is, every predetermined movement amount of the truck 2.

In this case, the encoder 12 is set to output a pulse every time the truck 2 moves by the sensor interval β.

On the other hand, in FIG. 2, 13 is a displacement sensor II a, 11
Based on the output signal of b and the output pulse of the encoder 12, each displacement sensor lla, llb and the top surface 1a of the rail 1
A measuring device W (measuring means) 14 measures the distance from the distance measured by the measuring means 13 in order to obtain data regarding the shape of the wave-like wear occurring on the top surface 1a of the rail. These are arithmetic processing devices that perform arithmetic processing on data, and these are mounted on the main vehicle 5.

The measuring device 13 includes a filter 15 and a sample hold 16.
, a multiplexer 17 and an A/D converter 18, and a filter 15 from the output signal of each displacement sensor lla, llb.
After removing out-of-band signals by The output signal is sent to multiplexer 17 and A
The data is input to the arithmetic processing unit 14 as digital data via the /D converter 18.

Therefore, each displacement sensor lla by the measuring device 13.

The distance between the rail 11b and the top surface 1a of the rail 1 is measured at intervals of the sensor interval 1 in the longitudinal direction of the rail 1.

The arithmetic processing device 14 has arithmetic processing means 19, a sensor output correction means 20, a first arithmetic correction means 21, and a second arithmetic correction means 22 as its arithmetic processing functions. Each displacement sensor II input from
While correcting the data of the output signals of a and 11b as necessary by the sensor output correction means 20, the basic calculation according to the present invention is performed on the data of the output signals by the calculation means 19, and further, By appropriately correcting the calculation results by both calculation correction means 21 and 22, shape data in the longitudinal direction of the top surface 1a of the rail 1, in other words, waveform data of the wavy wear occurring on the top surface 1a is obtained. . Further, the data obtained in this manner is sequentially recorded on, for example, a magnetic tape (not shown).

Incidentally, the output pulses of the encoder 12 are processed by the waveform shaper 2.
3, the data is inputted into the arithmetic processing bag Wjt 14 as data related to the moving distance of the truck 2, and the data is also sequentially recorded on a magnetic tape or the like.

Next, using this measuring device, the top surface 1 of the rail l is
The measurement principle and the calculations performed by the calculation means 19 when measuring the wave-like wear occurring on the surface a will be explained with reference to FIGS. 3 and 4.

3 and 4 show the truck 2 and the displacement sensor 11a,
llb, etc. are schematically shown using virtual lines and solid lines.

In FIG. 3, the bogie 2 and the like shown in phantom lines are located at locations where the bogie 2 has no wave-like wear or the like on the top surface 1a of the rail 1;
In other words, it shows a state (hereinafter referred to as initial state) in which the top surface 1a is supported on the rail 1 at a location where it is a smooth flat surface lx (hereinafter referred to as reference plane IX), and the bogie 2 and the like shown in solid lines are The figure shows a state in which the trolley 2 has moved on the rail l from its initial state and is supported on the rail l at a location where wavy wear 1y has occurred on the top surface 1a of the rail 1 (hereinafter referred to as a moving state).

In the figure, Ho is the height of the displacement sensors lla and llb relative to the reference plane 1x in the initial state (hereinafter referred to as
H is the height of the displacement sensors lla and Ilb in the moving state relative to the reference 1TiIIIX (hereinafter referred to as the moving state sensor height).

Trolley 2 trolley tiii3. Since it is supported on the rail l via the rail 4, when it moves on the rail 1, it moves up and down due to the wavy wear 1y on the top surface 1a, etc.
Therefore, the value of the moving state sensor height l] of both displacement sensors lla and llb changes with respect to the initial state sensor height H. In this case, since the displacement sensors lla and llb are provided close to each other as described above, the amount of vertical fluctuation can be considered to be approximately the same, and this can be expressed as ΔH (upward direction is the positive direction). ), then H=H, +ΔH (1).

In addition, as shown by the solid line in the same figure, in the moving state, the measurement points of the parietal surface la located directly below each displacement sensor 11a, 11b are set as P, P', and the reference of the parietal surface 1a at the measurement points P and P', respectively. The amount of displacement (upward is the positive direction) with respect to the surface 1x is y, y', respectively, each measuring point P, P'' measured by the measuring device 13 via each displacement sensor lla, lla, and each displacement sensor II a , 11 Let the distance to b (hereinafter referred to as measurement distance) be XA and XB, then
According to the above formula (1), X^=Hy=Ho +Δ)I−y
−・−(2)XB=H−y′=HO+ΔH−
y' ・−・・・・(3) That is, y=Ho+ΔH−XA ・・・・・・(4
)y”-H60ΔH-XB ・・・・・・(
5).

Here, if the displacement difference between the displacement y at the measurement point P and the displacement y' at the measurement point P is Δy (upward is the positive direction), then Δy
-y'-)' -X8-XB
......(6) Or y' = y + Δy = y + (XA-XB)
・・・・・・(You can gain strength.

Therefore, the displacement difference Δy is not affected by the vertical fluctuation of the displacement sensors lla and llb, and the difference between the measurement distances XA and XB at the measurement points P and P' is XA-XB (hereinafter referred to as the measurement distance difference X^
-XB), the displacement difference Δy between the measurement points P and P' is obtained. Furthermore, if the displacement y at point P is known, the measured distance difference XA-
By adding XB, the sensor interval in the longitudinal direction of the rail 1 from point P! The displacement 1y' at the measuring point P'' that has moved by the distance is determined.

Based on the above, next, in Figure 4, the above (
A method for determining the shape of the top surface 1a of the rail 1 in the longitudinal direction using equations 6) and (7) will be described.

In FIG. 4, measurement points Pr (i = 0.1.2.
) is the top surface 1 of the rail 1 in the longitudinal direction.
The measurement points are arranged in order at equal intervals on a, and each measurement point P
The spacing on 88 is the same as the sensor spacing l.

In this case, measurement point P. 1 P, is % Bx of parietal surface la
On the plane lx (standing upright, measurement points Pz, Px.

It is assumed that wavy wear 1y occurs at the positions of ....

When moving the platform II+, 2 or rail 1+, both displacement sensors lla and llb are set at the measurement point PO+, respectively.
The cart 2 starts moving from a position directly above P, and the measuring device 13 measures each displacement sensor Ila.
, llb and the top surface 1a of the rail 1. As described above, the measurement is performed every time the truck 2 moves by the sensor interval l. That is, when the truck 2 starts moving, the truck 2 is moved at the sensor interval l.
Displacement sensor lla and measurement point PO5PI each time it moves
The distances XA, , XA, , . . . , . ...distance to XB6. XBI...
... are measured sequentially.

At this time, similarly to the above, the measurement points P,,P,. 1 (i
=0.1,2. ...), respectively, is Δy, and the amount of displacement of the parietal surface 1a with respect to the reference plane 1x at each measurement point P is y, it is clear from equation (7) above. Also, from the above equation (6), Δ y ,=XA, −XBi
......(9).

Then, by substituting equation (9) into equation (8) and further considering that the measurement point PO is on the reference plane lx of the parietal surface 1a, that is, yo-0, we get ( i=1.2,
3. ) is obtained.

Therefore, each time the trolley 2 moves by the sensor interval l, the distances between each displacement sensor 11 a , 11 b and the measurement point p, , p, , located directly below it @XAi , XB
If i is measured respectively, the distance differences XA and -XBi are calculated, and the distance differences XA and -XB are cumulatively added,
The amount of displacement y of the crown surface 1a with respect to the reference plane IX at each measurement point P is determined, and thereby the shape of the wavy wear 1y, etc. is determined.

The calculation means 19 sequentially performs such calculations, that is, the calculation of equation 00) when the cart 2 moves.

Next, the correction performed by the sensor output correction means 20 will be explained with reference to FIG.

FIG. 5 shows the shape of the top surface 1a of the rail 1 in the longitudinal direction, the changes in the output signals of the displacement sensors lla and llb corresponding to the shape of the rail 1 when the trolley 2 moves, and
It also exemplarily shows how the measured distance difference XA-XB changes. In this case, the state of change in the output signal and the state of change in the measured distance difference χA-XB are shown by broken lines for the case where the sensor output correction means 20 is not provided, and for the case where the sensor output correction means 20 is provided ( This example) is shown by a solid line.

Each of the displacement sensors lla and llb outputs an output signal when the top surface 1a of the rail l is displaced with respect to the reference plane IX, and generally, the output signal includes electrical or mechanical input, although it is small. This is accompanied by errors (hereinafter referred to as output errors) due to mechanical drift, etc. Therefore, if an output error occurs at one measurement point on the top surface 1a of the rail 1, that output error will be transferred to the next measurement point. Therefore, as the measurement progresses, the output errors δa and δb of the displacement sensors lla and ilb gradually accumulate and increase almost linearly, as shown by the imaginary lines in the figure. I have something to go.

In such a case, the output signals of the displacement sensors lla and Ilb are connected to the top surface 1 of the rail 1, as shown by the broken line in the figure.
In the wavy wear portion of a, the output error δa and δb gradually increase while changing in a wavy manner, and the output error δa also increases in the flat portion.

It increases approximately linearly with δb.

In addition, in this case, in the measurement distance difference XA - , Generally, the output errors δa and δb are slightly different, so the measured distance difference XA-χB is also the same as the output error δa as shown by the broken line.

It gradually increases according to the difference in δb.

The sensor output correction means 20, especially in the flat part of the top surface 1a of the rail 1, has an output error of δa,
The output error δa is calculated based on the fact that when it gradually increases with δb, the amount of increase is very small and substantially the same at adjacent measurement points.

This is for removing δb.

That is, the sensor output correction means 20 adjusts each displacement sensor lla at one measurement point P as shown in the figure.
, llb output signals as Oa, Ob, and the previous measurement point P°
, the output signal is Oa'.

Assuming Obo, for a comparison amount ε of an appropriate size set in advance, 10a-Oa' l <ε--(II) 10b-Ob
"1<ε...02), the output signals of each displacement sensor lla, llb are forcibly set to zero.

In this way, as shown by the solid line in the figure, each displacement sensor lla, l
The output signal of lb becomes zero, and the output errors δa, δb
is removed. At this time, the calculation means 19
After the sensor output is corrected, the measured distance difference X^-XB is calculated, and therefore, even in the measured distance difference XA-XB, the output errors δa and δb are Errors are removed.

Next, the calculation correction performed by the first calculation correction means 21 will be explained according to FIG. 6 with reference to FIG. 5.

FIG. 6 shows the change in the displacement y of the top surface 1a obtained by cumulatively adding the measured distance difference XA - and roughly, the first calculation correction means 2 for the displacement y.
1, the change in displacement Hy after the correction described below is performed is shown by a solid line.

In the correction by the sensor output correction means 20, as described above, although the output errors δa and δb are removed in the flat portion of the top surface 1a of the rail l, the output errors δa and δb are still removed in the corrugated worn portion. δb remains. Therefore, the amount of displacement y obtained by cumulatively adding the measured distance difference XA-χB including the error in the corrugated worn part includes the cumulative error δ as shown by the broken line in the figure, and the cumulative error δ is the cumulative calculation result of the cumulative calculation. There is a possibility that it will increase rapidly as the disease progresses.

The first arithmetic correction means 21 reduces this cumulative error δ as described below.

That is, the first arithmetic correction means 21 calculates the value obtained at a predetermined number of previous measurement points, for example, from the cumulative addition value (displacement amount y) of the measurement distance difference XA-χB obtained by the above formula 00 at each measurement point. The average value of the cumulative addition values of the measured distance differences XA-χB is subtracted, and the value obtained thereby is newly set as the amount of displacement Y of the parietal surface 1a at each measurement point.

More specifically, the cumulative error at each measurement point P1 is δ8, the actual displacement amount of the parietal surface 1a at each measurement point P1 is y8'', the displacement amount obtained by the correction by the first calculation correction means 21 is Y8, Assuming that the number of measurement points for calculating the above average value is m, the cumulative addition value (displacement amount yt) of the measurement distance difference XA - XB obtained by the above formula 00 is Vi = 'ji'' + δi ... ...03)
Also, the displacement position Y at each measurement point P4 is expressed as Y, -y, -(5ぢ1?/m)...04)
It is expressed as Furthermore, by substituting formula 03) into formula 04), Y; −(yt'−(X'y,'/m):
1+[δi −(,<−”,−/m) )・・・・・・
05) is obtained.

Here, the second term in the first curly bracket of equation 05) is the rail 1 at each measurement point P and the m measurement points before it.
is the average value (hereinafter referred to as average value By) of the actual displacement amount y, ゛ of the top surface 1a of ” is considered to change roughly sinusoidally, and therefore its average value B
y is approximately zero.

Furthermore, if the plurality of measurement points P are located on the flat part of the rail I, the average value By-yi'-〇 is clearly obtained,
Therefore, the term in the first curly braces of equation 051 is zero.

On the other hand, the second term in the second curly bracket of 05) 2 is the average value (hereinafter referred to as average value Bd) of the cumulative error δ at each measurement point P1 and the m measurement points before it. Therefore, the term in the second curly bracket is the value obtained by subtracting the average value Bd from the cumulative error δ8 at each measurement point P, and at least at each measurement point P.

The cumulative error δ in is smaller than δ.

From the above, the above equation 04) or the above equation 00 essentially becomes Y, -y,"+(δ, -Bd)...(1ω), and therefore the above equation expressed by equation 04) The cumulative error δ is reduced by the correction calculation of the first calculation modal means 21. Then, when the correction calculation according to equation 04) is applied to the waveform shown by the broken line in FIG. is obtained, and this waveform almost matches the actual shape of the top surface 1a of the rail 1.

In addition, in this correction calculation, from the cumulative addition value of the measurement distance difference XA - XB obtained by the above-mentioned formula 00), for example, the cumulative addition of the measurement distance difference XA - It is also possible to calculate the amount of displacement Y of the parietal surface 1a by subtracting the average value of the values. In this case, among the number of measurement points for which the average value should be calculated, the number of measurement points before each measurement point P8 is p, and the measurement point P is the number of measurement points before each measurement point P8.

If the number of measurement points after is q, the displacement iYi is...
...(+4)', and it goes without saying that the cumulative error δ is reduced in this case as well.

Next, the correction calculation performed by the second calculation correction means 22 will be explained with reference to FIG.

The arithmetic processing unit 14 includes displacement sensors lla and llb.
Add the data of XB to the measured distance obtained by
Since the arithmetic processing is performed digitally according to the above-mentioned equation 00 or even equation 04), a floating error occurs. For this reason, for example, when cumulative calculations are performed according to formulas 00) and 04), this floating error is accumulated, and the amount of displacement Y of the top surface 1a of the rail 1 determined by formula 04) is larger than the actual value. It can become large.

Therefore, in the second calculation correction means 22, the displacement i
When Yi exceeds a preset comparison amount, the displacement amount Y is forcibly set to zero. That is, if the comparison amount is Ymax, then l Y 1> Ymax --OT)
When this is the case, the amount of displacement Y=0.

Next, the overall operation of such a measuring device is shown in FIGS. 1 and 2.
The process will be explained according to the flowchart of FIG. 7 with reference to the drawings.

When measuring the wave-like wear on the rail l, the cart 2 is moved along the rail 1 together with the main vehicle 5, and at the same time, the measuring device 13 measures each displacement sensor 11 a ,
11 Distance XA between b and top surface 1a of rail l,
The measurement of B is performed every time the trolley 2 moves by the sensor interval l. In this case, the movement of the trolley 2 starts from a place where the top surface 1a of the rail 1 is a smooth flat surface, for example, on a rail 1 that is not normally used, and starts from the top surface 1a of the rail 1.
is the reference plane IX.

The measured distances XA and XB measured by the measuring device 13 via the displacement sensors lla and Ilb are sequentially input to the arithmetic processing device 14 as digital data.

Then, in the arithmetic processing unit 14, the arithmetic means 19 calculates the measured distance difference XA-XB from these data as described above, and further performs the cumulative calculation of the measured distance difference x 8-χB according to 00)2. be exposed.

At this time, the output signal Oa of each displacement sensor 11a, 11b.

When Ob satisfies the equation (11) and the zero equation, it is set to zero by the sensor output correction means 20, and the cumulative calculation is performed based on this.

Next, the cumulative calculation value at each measurement point obtained in this way is processed by the first correction calculation means 21 as described above.
is corrected according to the formula 04), and thereby,
The amount of displacement Y with respect to the reference plane IX at each measurement point on the top surface 1a of the rail l is determined. At this time, the displacement Y is
When the 0η formula is satisfied, the second arithmetic correction means 22 sets it to zero.

Then, the top surface 1 of the rail l obtained in this way
The amount of displacement Y of a is recorded by the arithmetic processing unit 14 on a magnetic tape (not shown) or the like.

After this, the magnetic tape etc. are attached to the main vehicle 5, for example.
Further, data on the amount of displacement Y at each measurement point of the rail 1 is read by an analysis device (not shown), and outputted as the shape of the top surface 1a of the rail 1 by a CRT, block, etc. (not shown). As a result, the wave-like wear of the rail 1 is measured. In this case, in this embodiment, after reading the data of the displacement Y, the average value of the displacement 1i1Y at a predetermined number of measurement points before and after each measurement point is calculated according to the formula 04)' for each measurement point. The amount of displacement at the measurement point is subtracted from the first
Perform the same correction as the correction calculation by the calculation correction means 21,
This further improves the accuracy of the displacement Y data.

Next, measurement errors other than the cumulative error δ and the like described above in the measuring device of this embodiment will be explained with reference to FIGS. 8 and 9.

First, Figure 8 schematically shows the truck 2, etc.
As mentioned above, the trolley 2 is connected to the rails via the wheels 3 and 4.
For example, as shown in the figure, the trolley 2
While moving on rail 1, wheel 3 moves on rail 1.
the displacement sensor lla falling into the joint IZ between them;
llb may tilt together with the truck 2.

In such a case, for example, on the flat reference plane IX of the top surface 1a of the rail 1, the displacement sensor I
A difference occurs in the measured distances XA, XB between la, llb and the top surface 1a (reference surface lx) of the rail 1, and an error δX (hereinafter referred to as inclination error δX) occurs in the measured distance difference XA-XB.

Here, as shown in the figure, displacement sensors 11a, 1
When the inclination angle of 1b is θ, the amount of fall of the wheel 3 from the reference plane lx is ΔL, the radius of the wheel 3 is R, and the maximum clearance of the seam 1z is C, the following equation holds true.

Δt = C/8 R...08) δx-1! ,t
a nθ −j! jan (s in-' (Δt/L))...
...09) In this case, assuming that the radius R and the maximum clearance C are respectively practical values, R = 50 mm and C = 20 (company).
From the equation θ8), Δt=1mm, and furthermore, the sensor interval! ! = 10mm, wheel spacing L = 1000m,
δχ−0,0111Iwl is obtained from the equation θ9).

Generally, the maximum displacement due to wave-like wear on the rail 1 is about 1 mm, and the value of this inclination error δX (0,
Olmm) is sufficiently small compared to the maximum displacement (1 mm) of wave-like wear. Therefore, this tilt error δX can be ignored in this embodiment.

Note that the bogie 2 may also be tilted at the wavy wear location of the rail l, but in this case, as mentioned above, the maximum displacement due to the wavy wear is about 1 mm, so both axles of the bogie 2 3.4, the amount of depression on one side is about 1 mm at most, as described above, and therefore, the tilt error in this case can be ignored.

Next, FIG. 9 shows the wave-like wear 1y of the top surface 1a of the rail 1 in a sinusoidal manner.
The displacement amount of a is measured. For this reason, for example, as shown in the figure, if measurement points P and P' adjacent to each other are located on both sides of point Q, where the amount of displacement of the wave-like wear 1y is maximum, the amount of displacement of the crown surface 1a obtained by the measurement The maximum value may become smaller than the actual maximum value, and the difference may appear as an error (hereinafter referred to as sampling error). The sampling error is calculated at the measurement point P as shown in the figure.
, P' is maximum when it is at a position equidistant from point Q, and if this value is large, it becomes impossible to reproduce the shape of wavy wear from the amount of displacement of the crown surface 1a at each measurement point.

In this case, as shown in the figure, the wave-like wear 1y is assumed to be a sine wave, its period is T, the amount of displacement at point Q is Ym, and the measurement point P
, the displacement amount at P' is Ym', and the sampling error is δ
If y=Ym-Ym", then Ym"-Ym-sin
[9090・(l/2)/(l/4)]-Ym-cos
(90.2.kuru) ......Q■.

Here, as a practical value, inter-sensor defect 1. =10mm
, circumference 3'J] T = 100 mm, (from formula 201, Ym'' = 0.95Ym, and furthermore, δy = Y m
-Y m ', so δy'; 0.05Ym.

Therefore, the amount of displacement Ym measured at the measurement points P, P''
” and the sampling error δy correspond to approximately 95% and 5% of the actual displacement Ym at point Q, respectively, and at these values, the shape of the wave-like wear can be sufficiently reproduced.

(Effects) As is clear from the above description, according to the rail wave wear inspection device of the present invention, when the carts are moved on the rails via the wheels, the carts approach each other. Each time a pair of displacement sensors mounted on the rail move by an interval, the measuring means measures the distance between each displacement sensor and the measurement point on the top surface of the rail located directly below it, and the calculating means measures the distance. By cumulatively adding the difference in measurement distance,
By calculating the amount of displacement of the top surface of the rail with respect to the reference surface at each measurement location, even if the displacement sensor moves up and down along with the cart when the cart moves, the cumulative calculation allows the top of the rail to be measured. The amount of displacement of the surface can be determined with high accuracy.

At this time, even if the output signal of each displacement sensor contains an error due to drift, etc., if a sensor output correction means is provided, the output of both displacement sensors at each measurement point will be the same as that of the previous measurement. By setting the outputs of both displacement sensors to zero when the output is approximately the same as the output at that point, it is possible to eliminate the error at a flat point on the top surface of the rail, thereby ensuring that the top surface of the rail The correct amount of displacement can be found.

Furthermore, even if the measurement distance includes an error and there is a risk that the error may be accumulated by the above cumulative calculation, when a calculation correction means is provided, it is possible to By subtracting the average value of cumulative sums at a predetermined number of measurement points before, after, or before the cumulative sum from the cumulative sum, it is possible to reduce the error and suppress the accumulation. The amount of displacement of the top surface of the rail can be determined with high accuracy.

By being able to accurately measure the amount of displacement of the top surface of the rail in this way, it is possible to measure in detail the wavy wear that occurs on the top surface of the rail.

[Brief explanation of the drawing]

FIG. 1 is a side view of an example of the rail wave wear inspection device of the present invention, and FIG. 2 is a block diagram of the main parts of the inspection device.
3 to 6 are explanatory diagrams for explaining the operation of the main parts of the inspection device, FIG. 7 is a flowchart for explaining the overall operation of the inspection device, and FIGS. Figure 9 is an explanatory diagram for explaining the measurement error of the measuring device, the first
FIG. 0 is a schematic side view of a conventional measuring device. 1...Rail 1a...Top of head 1x...
・Reference surface 1y... Wave-like wear 2... Trolley
3, 4...Wheels 11a, 11b...Non-contact displacement sensor 13...Measuring means 19...
・Calculation means 20...Sensor output correction means 21...Calculation correction means XA, XB...Measurement distance y...Cumulative addition value Y...Displacement amount Patent applicant Zen Kami Ne Kazu Tokyo Sen Co., Ltd. Instrument Research Institute

Claims (1)

[Scope of Claims] 1. A device for detecting wavy wear occurring on the top surface of a rail by measuring the amount of displacement of irregularities on the top surface of the rail with respect to a reference surface in the longitudinal direction, the device comprising: a pair of non-contacting carts that are movably attached to the cart directly above the rails and are arranged adjacent to each other with an interval in the longitudinal direction of the rails. Measurement in which the distance between each displacement sensor and a measurement point on the top surface of the rail located directly below is measured by each displacement sensor, each time the trolley moves by the distance between both displacement sensors when the trolley moves. By calculating the difference between the measured distances measured by both displacement sensors for each measurement, and cumulatively adding the distance differences obtained by the calculation, the reference plane of the top surface of the rail at each measurement point is determined. 1. A measuring device for measuring wave-like wear on a rail, comprising: arithmetic means for determining the amount of displacement for a rail. 2. At each of the measurement points, the sensor output correction means is provided to set the outputs of both displacement sensors to zero when the outputs of the two displacement sensors are substantially the same as the outputs of the measurement point immediately before the measurement point. The rail corrugation wear measuring device according to claim 1. 3. By subtracting the average value of the cumulative sum of the predetermined number of measurement points before, after, or before the distance difference from the value obtained by cumulatively adding the distance difference to each measurement point, 2. The rail corrugation wear measuring device according to claim 1, further comprising arithmetic correction means for determining the amount of displacement of the top surface of the rail with respect to a reference surface.