JPS60253830A - Load measuring method - Google Patents

Abstract

PURPOSE:To reduce the error in the measurement of load using a washer type load cell by a strain gauge system, by calculating load by using a weighted addition formula preliminarily calculated from the longitudinal strain in the inner surface side of an annular block and that in the outer surface side thereof. CONSTITUTION:A load cell is constituted so as to detect the longitudinal strain in the inner surface side of a web having an I-shaped cross-sectional area and that in the outer surface side thereof by adhering strain gauges to a plurality places of the inner and outer surfaces of said web in the circumferential direction. Further, the weighted addition formula of longitudinal strains in the inner and outer surface sides is defined by formula I and the output Fr of the load cell is calculated according to formula II by using the weighted addition value epsilon. The load measuring error e3 in this case is imparted by formula III. By performing heavy regression analysis of inner and outer surface side strains epsilon1, epsilon0 when the loads of two kinds or more of load patterns are applied, coefficients a, b of the formula I are calculated. Herein, F0 is load actually applied to the load cell.

Description

DETAILED DESCRIPTION OF THE INVENTION (Industrial Field of Application) The present invention relates to a load measuring method using a load cell, and particularly to a load measuring method using a washer type load cell using a strain gauge method.

(Prior technology) Load cells for measuring loads are widely used in various fields. For example, load cells for rolling mills in the steel manufacturing process are used to measure rolling loads, and are essential sensors for various rolling controls including plate thickness control. be.

Load cells for rolling mills include strain gauge methods, electromagnetic methods, capacitance methods, etc. as load detection methods, and the washer type.

Although there are disk-shaped, rectangular, etc. load cells, the load cell targeted by the present invention is a washer-type load cell using a strain gauge method.

The strain gauge method [Koyoru washer type load cell and its usage are described in, for example, "Nippon Steel Tube Technical Report, A.
48 J (published by Nippon Kokan Co., Ltd., 1972), pp. 49 to 58 f. As described in the article "Considerations on rolling pressure gauges", a load cell with a 7-ring block [cost strain gauge attached] was used.

It is installed between the rolling mill's NAND and the housing.

However, under such conditions of use, what kind of strain occurs depending on the load distribution that the load cell receives, and how measurement errors occur depending on the situation in which the strain occurs has not been sufficiently researched. child.

The present inventors conducted various analyzes regarding the above-mentioned points using the finite element method (FEM)4 regarding washer-type load cells, and clarified the problems in load measurement using conventional washer-type load cells.

(Problems to be Solved by the Invention) Fig. 1 is a perspective view showing a washer-type load cell used by the present inventors for analysis, and the cross section (indicated by the body line in the figure) is an "I" shape. It is. Strain gauges are installed at several locations in the circumferential direction on the inner and outer surfaces of the web in the cross section (○ in the figure).
Show it on the mark? f) The structure is such that longitudinal strain on the inner and outer surfaces of the web can be detected. Table 1 is a table showing the dimensions of each part of the three load cells with different web thicknesses used in this analysis, and the symbols of the valleys in the table correspond to the symbols shown in FIG.

The load applied to the flange surface of the load cell can be divided into uniformly distributed load and unbalanced load. Uniformly distributed load is a load that has a uniform distribution on the flange surface of the o-cell, and unbalanced load is a load that is divided from the total load to the This is the remainder after subtracting the uniformly distributed load.

The shape of the load cell and the rolling mill) is not necessarily uniform in the radial direction of the load cell.

Radial stiffness is not uniform. The stress applied in the longitudinal direction of the load cell generally consists of uniformly distributed stress and non-uniformly distributed stress. The inventors of the present invention installed the load cell shown in Notification 2 in Table 1 between the nut and housing of a rolling mill, applied a load, analyzed the nut and the housing, and conducted a second analysis.
The load distribution on the upper and lower seven run surfaces of the load cell as shown in the figure was obtained. The vertical arrows in the figure represent the magnitude of the longitudinal (Z) direction stress (σZ) at each point in the radial (r) direction of the flange surface. FIG. 3 shows the displacement of each part of the cross section of the load cell corresponding to the load distribution at this time. The solid line in the figure shows the cross-sectional shape when no load is applied.
The dotted line indicates the shape (displacement) under load.

In the load measurement using Asahi II'W+Koinj, 1- and 2 load cells, the output of the load cell was calculated once as in the conventional method. That is, using the longitudinal strain detected by strain gauges attached to the inner and outer surfaces of the web.

T-1 (ε.+εka)/2 ・・・・・・・・・・・・
・・・・・・(1) Fm−πptEg ・・・・・・・・・
・・・・・・(2) Here, Fm: Load cell output [N
] ε. Sum of longitudinal strains on the outer surface of the web detected by the Ni strain gauge εL Sum of longitudinal strains on the inner surface of the web detected by the Ni strain gauge E: Young's modulus [N/TIA] D Average diameter of two load cells [■] t: Web The load cell output Fm was calculated as the thickness [-].

However, the actual load Fo applied to the 0-docel9 is a value calculated from the following formula using the integral value of the longitudinal strain ε(r) at each point in the radial σ direction of the web cross section. Therefore, the 7Ho-dossel output (load 9 calculated from the above (2) 2) has a measurement error eI shown in the following equation with respect to the actual load.

e, =(rrDtB5-Fo)/F
o x 100 C%]=...(4) Here, if it is assumed that the change in the radial direction of the longitudinal strain ε(r) at each point in the radial (r) direction of the web cross section is linear, then the above Equation (3) can be expressed as shown below, and the measurement error e2 in this case corresponds to the second term of equation (5).

As can be seen from these formulas (5) and (6) military force,
Only when the longitudinal strain ε(r) at each point on the web cross section is completely equal and uniform in the radial direction, the load cell output calculated by the above equation (2) matches the actual load.

However, in reality, the distribution of the load applied to the load cell is not constant due to differences in rigidity of the rolling mill valley, differences in the slope of the contact surface of the load cell, and differences in surface finish accuracy, and as mentioned above, the uneven load is There is. Therefore, the present inventors experimentally determined the distribution of longitudinal strain ε(r) at the trough point of the web cross section and the measurement error depending on this strain distribution when loads with different distributions were applied to 9 load cells. .

Figure 4 shows the five types of load patterns used in this experiment. Pattern E is a load that gives equal displacement, turn W is a uniformly distributed load, patterns S and R are triangularly distributed loads, pattern B is an experimental example shown in Figure 2, and the total load is 15 MN for each pattern. Figure 5 shows the distribution of longitudinal strain ε(r) when loads of patterns E, W, S, R, and B are applied to the load cell A2 in Table 1 above. is an enlarged view of the strain distribution (deviation distribution) corresponding to patterns E, W, and Bl in Figure 5. Figures 5 and 6 [As is clear from the experimental results shown here,
The strain ε at the trough of the web cross section is not uniform in the radial direction (C, r) and is nonlinear, regardless of the load pattern.

As shown in Table 2, the measurement error e1 when using the conventional measurement method using equation (2) and assuming that the change in longitudinal strain ε(,r) is linear. There is a fairly large difference between the measurement error e2 and the measurement error e2.

Table 2 This difference is shown in FIGS. 5 and 6.

This is caused by the nonlinearity of the longitudinal strain ε(,r). Even for load patterns E and W, the measurement error e
1 is so thick that it reaches -6 to -7%.

As mentioned above, the conventional measurement method for calculating the load cell output is to simply average the longitudinal strain on the inner surface and the longitudinal strain on the outer surface.

It has become clear from the analysis conducted by the present inventors that the load measurement error is large. The present invention is based on the above knowledge,
This invention provides a load measurement method using a load cell that reduces measurement errors.

(Means for Solving the Problems) The gist of the present invention is to separately calculate the longitudinal strain on the inner surface and the longitudinal strain on the outer surface of an annular block in a load measuring method using a washer-type load cell using a strain gauge method.
This is detected by the load cell [Weighted addition formula of the longitudinal strain on the inner surface side and the longitudinal strain on the outer surface side where the load measurement error is minimized using two or more types of eccentric loads with different load distributions] In this load measurement method, the load is calculated using the weighted addition formula from the longitudinal strains on the inner surface and the outer surface, respectively, which are detected separately in actual load measurement.

The present invention will be explained in detail below.

First, in the present invention, a weighted addition formula for the longitudinal strain on the inner surface side and the longitudinal strain on the outer surface side is prepared in advance. This weighted addition formula is defined by the following formula t.

ε knee (aε.+bE;,) ・・・・・・・・・ (
7) However, a and b are coefficients.The output Fr of the load cell is calculated by the following formula using the weighted addition value ε of the longitudinal strain determined by the formula (7).

Fr-πDtE order... (8) The load measurement error e3 in this case is given by the following formula: e3-=(πDtEε
-F. ) /FoX 100 [%) −(9) Above (
7) The coefficients a and Ll in equation Strain ε. It is possible to perform multiple regression analysis. In the uninvented embodiment G, loads of various patterns of γ shown in FIG. The regression coefficients a, l) were calculated. The results are shown in Table 3.

Table 3 In calculating the regression coefficients a and l), the coefficients a,
b must be stored for each load cell of different size and shape. 1.1 It is difficult to apply the load distribution as shown in Fig. 4 to the flange surface of the load cell, but in short, pattern 3. R By multiplying the coefficients a and br, the coefficients a and br can be obtained without any practical problem. Table 3 shows the load measurement error e1 when using the conventional method and the load measurement error C3 when using the method of the present invention, using the first measurement data for regression coefficients a and b. . As is clear from this result, according to the method of the present invention, the load measurement error is less than 1%, which is much smaller than that of the conventional method.

FIG. 7 shows one of the configurations of an apparatus for carrying out the method of the present invention.
It is a figure which shows an example. In the figure, 1 is a bridge circuit for detecting strain on the inner cylindrical knitting, which includes strain gauge resistance on the inner side of the web, and 1o is a bridge circuit for detecting longitudinal strain on the outer side, including strain gauge resistance on the outer side of the web. The longitudinal strain ε on the inner side of the web and the longitudinal strain ε on the outer side of the web, respectively. Outputs an electric signal for one priest. 2 is an arithmetic unit which calculates the above-mentioned (force equation) and equation (8) and outputs Fr of the equation (8). A function to set regression coefficients a and b is added.

As described above, the apparatus for implementing the present invention may be configured to separately detect longitudinal strain on the inner surface of the web and on the outer surface of the web, and the load cell itself may be one of the conventionally used ones.゜Can be used.

In the above explanation, we have described a load cell in which strain-in gauges are attached to the inner and outer surfaces of the web of the washer type load cell, but it is also possible to drill holes in the inner and outer surfaces of the web and embed strain gauges. hand,
Even if the load cell is designed to detect longitudinal strain on the inner and outer sides.

Of course, the method of the present invention can be applied.

(Effects of the Invention) As stated above, the method of the present invention significantly reduces load measurement errors as shown in Table 3 even in the presence of unbalanced loads that are unavoidable in actual use of washer-type load cells. This invention is extremely effective in practice.

[Brief explanation of drawings]

Figure 1 is a perspective view showing the appearance of the washer type load cell, Figure 2 is a diagram showing an example of the load distribution between the mouth and the deserge, and Figure 3 is an example of the displacement of the valley of the a-decel cross section. Figure 4 is a diagram showing various load distribution patterns used in the experiment, Figures 5 and 6 are diagrams showing an example of longitudinal strain distribution in the cross section of the 7C load cell based on the experiment, and Figure 7 is a diagram showing various load distribution patterns used in the experiment. 1 is a diagram showing a device configuration for implementing the present invention. IL, 1...Bridge circuit, 2...Arithmetic unit. Patent Applicant Representative Patent Attorney Tomoyuki Yafuki (1 person) Part II! @ Figure 2 L-'+ , ---jσ2 z l 1 Question 3 Figure 4 Figure 7 Figure 1 Figure 5 Figure 6 figure

Claims (1)

[Claims] Strain gauge method [This is a load measurement method using a washer-type load cell that separately detects the longitudinal strain on the inner surface and the longitudinal strain on the outer surface of the annular block. Using two or more types of eccentric loads with different load distributions applied to the load cell, a weighted addition formula for the longitudinal strain on the inner surface and the longitudinal strain on the outer surface that minimizes the load measurement error is determined, and then the actual load is determined. A load measuring method characterized in that the load is calculated using the weighted addition formula from the longitudinal strains of the inner surface and the outer surface, which are detected separately in the measurement.