JPH02247536A - Measuring robot - Google Patents

Abstract

PURPOSE:To accurately measure a strain distribution, etc., in a short time by providing a 1st robot, a 2nd robot, a difference arithmetic means, an expression determining means, a degree determining means, and a distortion arithmetic means. CONSTITUTION:The 1st robot 40 places a load on a body 50 to be measured, the shapes before and after placing a load are measured by the 2nd robot 30, and the difference arithmetic means 61 finds the difference between the shapes before and after the load is placed to obtain discrete displacement data. The expression determining means 62 finds the approximate expression of a displace ment distribution by a spline function from the discrete displacement data. In this case, the degree determining means 63 determines the degree of the spline function by a displacement expression obtained from a model of the body to be measured which is formed by simple load addition. The obtained approximate expression is differentiated twice to find the distortion distribution by the distortion arithmetic means 64 from found curvature. The displacement distribution can accurately be approximated by the approximate expression, so not so many measurement points are required. The measurement time is therefore shortened and the curvature is accurately calculated.

Description

[Detailed Description of the Invention] [Summary] A measuring robot that can measure strain distribution etc. with high precision and in a short time, regarding a measuring robot that automatically measures the mechanical properties of flexible structures such as plastic structures. The purpose of this invention is to provide a first robot that can change the direction, position, and load of applying a load to an object to be measured, and to measure the shape before and after applying a load to an object to be measured. The second thing that can be done
In a measuring robot having a robot and a difference calculation means for calculating the difference between the shape before and after applying a load to the object to be measured, at least within a limited range from the discrete displacement data obtained from the difference calculation means. formula determination means for determining an approximate formula for a second-order differentiable and continuous displacement distribution using a spline function, and determining the order of the spline function based on a displacement formula determined by adding a simple load to the object to be measured. It is composed of an order determining means and a strain distribution calculation means for calculating a strain distribution using the second-order differential value obtained from the approximate expression.

[Industrial Field of Application] The present invention relates to a measuring robot that automatically measures the mechanical properties of flexible structures such as plastic structures.

In the field of structural design, in response to demands for higher functionality and lighter weight products, extreme designs are being attempted using plastic molding technology to make materials thinner and provide the minimum necessary strength. ing. Many of these structures are flexible structures that are easily deformed, and are accompanied by not only elastic deformation but also plastic deformation and creep deformation.

In order to develop highly functional and lightweight products, it is necessary to measure and evaluate the mechanical properties of flexible structures and feed back the results into design.

For this purpose, it is necessary to measure the curvature, strain distribution, etc. of the object to be measured accurately and in a short time.

[Prior Art] In order to measure the mechanical properties of flexible structures as described above, the applicant has proposed a "measuring robot" as shown in FIGS. See Application No. 63-08214) a In Figures 6 and 7, 1 and 2 are 3 of XYZ.
The robot 2 is an orthogonal robot having an axial movement mechanism, and the robot 2 has a rotation mechanism 3.

4 is a support stand, and a jig 5 is placed on the support stand 4,
An object to be measured 6 is fixed on the jig 5. A force is applied to the object to be measured 6 via a force sensor 7 with a rod 8 attached to the tip of the robot 1. The displacement of the object to be measured 6 at this time is measured by a displacement detection probe 11 attached to the tip of the robot 2 via a force sensor 9 and a displacement detection section 10.

Discrete displacement data is obtained by determining the difference between before and after the displacement of the object 6 to be measured, and the strain distribution and the like are determined from these displacement data.

The strain distribution is determined by the following procedure. That is, a function expressing the displacement distribution is obtained from discrete measurement data through numerical processing, and the curvature is obtained by second-order differentiation of this function. In general, given the material properties (Young's modulus and Poisson's ratio), if the curvature is determined, stress, etc. can be determined.

Here, the difference method is mainly used to obtain differential values from discrete measurement data. For example, as shown in FIG. 8, when discrete measurement data is obtained in the X direction, the ratio of the difference x2-x1 to the difference Z2-Zl (
z2-zl )/(x2-xl) to find the first-order differential dz/dX, and similarly, the second-order differential d2z/
Find dx.

1 92W In this way, the curvature rx"2x□ was obtained.

[Problem to be solved by the invention] However, in such conventional measuring robots, the curvature was determined by the difference method, so only the curvature at the measurement point could be determined, and the curvature of the object to be measured could not be determined. There was a problem that it was not possible to measure accurately.

Therefore, in order to improve accuracy, it is necessary to make the measurement points denser to some extent, and in this case, the number of measurement points increases, so
One problem was that it took a long time to measure.

The present invention has been made in view of these problems, and an object of the present invention is to provide a measuring robot that can measure strain distribution etc. with high precision and in a short time.

[Means to solve the problem] FIG. 1 is a basic configuration diagram of the present invention.

In FIG. 1, 40 is a first robot that can change the direction, position, and load applied to the object to be measured 50, and 30 measures the shape and final shape of the object to be measured 50 before applying the load. 61 is a difference calculating means for calculating the difference between the shape before and after applying a load to the object to be measured 50; 62 is a second robot capable of calculating at least one of the discrete displacement data obtained from the difference calculating means 61; Formula determining means for determining an approximate formula for a continuous displacement distribution that is second-order differentiable using a spline function within a limited range; 63 is the object to be measured 5;
An order determining means determines the order of the spline function based on a displacement formula obtained by adding a simple load to 0, and 64 is a strain distribution calculation means for determining a strain distribution using a second-order differential value obtained from the approximate formula. .

[Operation] In the present invention, the first robot applies a load to the object to be measured, the second robot measures the shape before the load is applied and the shape after the load is applied, and the difference calculation means calculates the load. Discrete displacement data is obtained by calculating the difference between the shape before and after the addition.

From the obtained discrete displacement data, an approximate expression for displacement distribution using a spline function is determined by an expression determining means. In this case, create a model by adding a simple load to the measured object,
The order of the spline function is determined by the order determining means using the displacement equation obtained from this model.

Then, the curvature is determined by second-order differentiation of the obtained approximate expression, and the strain distribution is determined from the curvature by the strain distribution calculation means.

Since the displacement distribution can be accurately approximated by the approximation formula, the number of measurement points does not need to be so large. Therefore, measurement time can be shortened. Further, since the curvature can be calculated with high accuracy, strain distribution, stress, etc. can be calculated accurately, and mechanical evaluation of the material can be performed accurately. Furthermore, since the order of the spline function is limited, calculation processing can be made faster.

[Example] Embodiments of the present invention will be described below based on the drawings.

2 to 5 are views showing one embodiment of the present invention.

First, to explain the configuration, in FIG. 2, 40 is the first
The robot 30 is the second robot, and the first robot 4
The robot 0 and the second robot 30 constitute an orthogonal robot capable of linear movement in three axes of XYZ.

At the tip of the second robot 30, there is a rotation mechanism 32 that allows rotation around the Z axis. A displacement detection mechanism 33 is installed beyond that via a six-axis force sensor 31. A tip (probe) 34 of the position detection mechanism 33 is linearly guided with respect to the position detection mechanism 33 so as to be slidable in two directions. The distance that the probe 34 has slid relative to the position detection mechanism 33 can be measured by a linear encoder installed inside the position detection mechanism. 1st
A rod 42 is installed at the tip of the robot 40 via a six-axis force sensor 41 . A sample (object to be measured) 50 is fixed to a fixing jig 51.

Here, a control circuit 60 constituted by a microcomputer is connected to the first robot 40 and the second robot 30.

The control circuit 60 includes a difference calculation section (difference calculation means) 61, a formula determination section (formula determination means) 62, an order determination section (determination means) 63,
and a strain distribution calculation section (strain distribution calculation means) 64.

The difference calculation unit 61 calculates the difference between the shape of the sample 50 before and after the load is applied, and outputs discrete displacement data. The formula determining unit 62 determines an approximate formula for a continuous displacement distribution that is second-order differentiable using a spline function at least within a limited range in the XY force directions from the discrete displacement data.

The order determining unit 63 determines the order of the spline function based on a displacement formula obtained by simply applying a load to the sample 50. The strain distribution calculation unit 64 calculates 2 from the approximation formula.
Obtain the strain distribution using the differential value.

Here, the approximation formula for approximating the displacement distribution must satisfy the following conditions.

■Be able to calculate the 22nd order of magnitude.

■When considered physically, the obtained curvature is continuous.

■The obtained curvature accurately approximates the actual curvature.

■There must be a means to easily calculate it.

If a displacement distribution passing through n measurement points is expressed by one function over the entire domain, the function becomes complicated and tends to be jerky. In particular, the solution tends to oscillate in the differential value. However, the domain is a finite interval (
For example, if it is divided into (n-1 intervals when considered in one dimension) and a polynomial is applied to each interval, it can be relatively easily expressed using a low-dimensional polynomial. By making each piecewise polynomial continuous at each point of contact (boundary between sections), a low-order continuous function can be easily obtained.

Among the functions expressed by such an m-th polynomial, an m-th order differential value exists in the entire domain and is continuous, which is an m-th spline function. Therefore, in this case, if the spline function is at least cubic or higher, then the above ■,
■Can be met. By the way, when using spline functions, it can be seen that there are many that satisfy the conditions (1) and (2). When approximating with a spline function, the results obtained will differ depending on the order. The higher the degree, the more delicate the curve can be approximated.

However, if the order is high, it will be affected by calculation errors. In particular, since measured values include measurement errors, the influence of measurement errors tends to be large. To determine the order of the spline function, create a simple model and determine the order of the spline function based on the displacement equation obtained from this model. Let us now consider a beam model as such a model. When calculated using a beam model, the displacement distribution can be accurately approximated by a cubic equation. When bending a flexible member, it is possible to treat the object to be measured in the same way as a beam when viewed partially.

Therefore, in this case, it is preferable to select a spline function so that the displacement distribution is expressed by a cubic equation. Furthermore, since the third-order spline function has a low order, the influence of calculation errors can be suppressed to a small degree.

Next, the measurement method will be described.

First, the first robot 40 is moved away from the sample 50. In this state, the shape of the sample 50 is determined using the second robot 30 according to the following procedure. The probe 34 is moved above the sample 50. Next, the Z-axis is lowered by a specified distance, for example, about 1 mm. If the probe 34 and the sample 50 come into contact, the probe 34 will move upward with respect to the position detection mechanism 33, which can be detected from the probe movement signal of the position detection mechanism 33. If the probe 34 is not in contact with the sample 50, the second robot 30 is further lowered by 1 mm, a check is made to see if contact has been made, and the same procedure is repeated until contact is made. If it is determined that the probe 34 has made contact,
The height of the sample 50 in seven directions can be determined from the position of the second robot 30 in the Z direction and the moving distance of the probe 34.

The position of the first robot 40 in two directions is
A built-in position detector may be used to drive the motor in two directions. Next, the Z axis is moved upward, the probe 34 is separated from the sample 50, the first robot 30 is moved horizontally by a specified distance, and the height of the sample 50 in the Z direction is determined again. By repeating the same procedure, the height of the sample 50 at the specified X and Y points can be determined.

Next, the second robot 40 is moved so that the tip of the rod 42 is at a specified position on the sample 50, and the rod 42 is pressed against the sample 50 with a specified force. In order to generate a specified force, it is possible to control the second robot 40 based on the force detected by the force sensor 41. The shape of the sample 50 at this time is measured in the same manner as described above. However, in this case, the height of the sample 50 at the location where the rod 42 is pushing the sample 50 cannot be measured using the first robot 30, so the height at that location is measured by the second robot 40.
Determine from the position in the direction. In this way, the shape of the sample 50 before and after applying the load can be determined.

Next, a difference calculating section 61 calculates the difference between the two to obtain discrete displacement data. An actual measurement example is shown below.

In addition, in order to demonstrate the effectiveness of the present invention, the measurements were carried out in cases where theoretical handling was possible, and consideration was given to making comparisons with theory possible. As shown in Figure 3,
A metal flat plate 70 is used as a sample, one end of which is fixed with a fixing jig 71, and a load is applied to one point P on the flat plate 70. The displacement data at this time is illustrated in FIG. 4. A plane 72 shown only as an outline in the figure is a plane with zero displacement. 73 is a curve representing the displacement of the flat plate 70.

In the following, for the sake of simplicity, a one-dimensional case will be described. As shown in FIG. 3, the x and y directions are taken, the value in the y direction is fixed, and only X is considered as a variable. Note that the following discussion holds true even when x and y are variables. - In general, the displacement formula differs depending on conditions such as whether the load is concentrated or distributed. As shown in Figure 3, when a bending load is applied to a relatively thin sample at one point, the displacement equation is, at least within a limited range, a cubic equation as shown in equation 0. Accurate approximation is possible.

z=ax3+bx" +cx+d ---■ Therefore, the order determining unit 63 determines the order so as to select a spline function whose displacement distribution is expressed by a cubic equation.

Next, the measurement points in the directions are set to XO, Xl, X2. ..., Xn
−1, xn. At this time, the interval [xo, xn] is [
XO, Xl], [X1. X2 ], ---,
When divided into [xn-1°xn], the spline function w(x, y) representing the displacement distribution is divided into division polynomials wl, w2 . ..., represented by wn.

wl(x)=a1 x3+bl x2+c1 +dl(
XO≦X≦XI) w2(x)-a2 x3+b2 x2+c2 +d2(
xl≦X≦x2 ) wn(x)=an x3+bn x2+cn 10dn(
X n-1≦X≦xn)...■ Note that each coefficient a1. a2. --- an, ---, d
l, d2. ---, dn can be obtained from the condition that the displacement distribution w(x, y) is continuous up to the second derivative in the interval [xo, xn].

Therefore, the curvature 1/rx is the displacement distribution w1. w2,・
...It is obtained by partially differentiating wn.

wl(x)”=6al x+2bl (xo≦X≦x
1) w2(x)”=6a2 x+2b2 (x1
≦X≦x2 )wn(x)”=6an x+2b3
(xn-1≦X≦xn) The curvature in the
/rX is shown at 74 in FIG. Plane 75 represents a surface with zero curvature. Curved surface 76 represented only by contour lines
is the theoretically determined curvature.

As is clear from FIG. 5, it can be seen that the curvature can be determined with sufficient accuracy using discrete displacement data. Further, in FIG. 4, the displacement data is a value of an 11×11 point, whereas the curvature in FIG. 5 is a value of a 50×50 point. In this way, values at points other than the measurement points can also be easily determined. Similarly, for displacement, interpolated values for values other than the measurement points can be easily obtained. Therefore, since the displacement distribution can be approximated with high accuracy, accurate strain distribution,
Stress distribution can be obtained.

In this case, since the number of measurement points may be small, the measurement time can be shortened. Furthermore, since the order of the spline function is limited, calculation time can be reduced.

[Effects of the Invention] As explained above, according to the present invention, the displacement distribution can be accurately approximated from the measured discrete displacement data, so the curvature can be calculated accurately, and the strain distribution can be accurately approximated. , mechanical evaluation such as stress distribution can be performed accurately. In this case, since there is no need to increase the number of measurement points, the measurement time can be shortened. Furthermore, since the order of the spline function is limited, the calculation speed can be increased.

[Brief explanation of drawings]

Figure 1 is a diagram showing the basic configuration of the present invention, Figure 2 is a diagram showing an embodiment of the present invention, Figure 3 is a diagram showing a measurement example, Figure 4 is a graph showing displacement data, and Figure 5 is curvature. FIG. 6 is a plan view of the conventional example, FIG. 7 is a front view of the conventional example, and FIG. 8 is an explanatory diagram of the differential method. In the figure, 30...Second robot, 31.41...Force sensor, 32...Rotation mechanism, 33...Displacement detection mechanism, 34...Probe, 40...First robot, 50 ...sample (object to be measured), 51...fixing jig, 60...control circuit, heather! ! 61...Difference calculation unit (difference calculation means), 62...Formula determination unit (formula determination means), 63...Order determination unit (order determination means), 64 ... Strain distribution calculation unit (strain distribution calculation means), 70 ... Plane plate, 71 ... Fixing jig, 72 ... Plane with zero displacement, 73 ... Diagram representing displacement data, 74 ...・Curvature, 75...Plane, 76...Theoretically determined curvature.

Claims (1)

[Claims] a first robot (40) capable of changing the direction, position, and load applied to the object to be measured (50);
A second robot (30) capable of measuring the shape before and after applying a load to the object to be measured (50), and a second robot (30) that can measure the shape before and after applying the load to the object to be measured (50) In a measuring robot having a difference calculating means (61) for calculating a difference, a second-order differentiable and continuous displacement is calculated by a spline function at least within a limited range from the discrete displacement data obtained from the difference calculating means (61). formula determining means (62) for determining an approximate formula for the distribution; and order determining means (62) for determining the order of the spline function based on a displacement formula determined by simply adding a load to the object to be measured (50).
63); and strain distribution calculation means (64) for determining strain distribution using the second-order differential value determined from the approximate expression.