JPS6138441A - Method for imaging stress distribution of specimen - Google Patents

Abstract

PURPOSE:To enhance the accurate degree in the imaging of stress distribution, by adhering a thermal filter to the surface of a specimen and calculating interpolation data under such a condition that the stress value of the thermal filter comes to zero while accurately correcting the measuring error of temp. data on the basis of the interpolation data. CONSTITUTION:For example, a thermal filter 10 comprising an adhesive tape is adhered to the surface of a specimen 1. A computer 7 consiting of memories 7a, 7b, 7e, 7f, averaging circuits 7c, 7g, a DELTAV and DELTAH determination circuit 7h, an interpolation data determination circuit 7i, a correction circuit 7j and a difference operation circuit 7d is connected to a change-over device 5. Then, load is applied to the specimen 1 and infrared scanning is applied to the thermal filter 10 to detect temp. data which is, in turn, stored in the memories 7e, 7f while interpolation is accurately performed so as to bring the stress value of the thermal filter 10 to zero to calculate correction values DELTAH, DELTAV in horizontal and vertical directions and a temp. measuring error is compensated to send an accurate temp. component to a display apparatus. As a result, accurate image display can be performed.

Description

DETAILED DESCRIPTION OF THE INVENTION (Industrial Application Field) This invention relates to a method for imaging the stress distribution of a test object by performing infrared measurement on a test object and performing computer processing on the temperature data input into a computer. .

(Prior Art) When designing a single part of a mechanical device or a structure, it is an important issue from the viewpoint of safety to know how much stress is applied to which part. In recent years, a method has been proposed for measuring the stress distribution of an object in a non-contact manner in a short time (Japanese Patent Application No. 55-56H1).The principle of this method will be briefly explained below.

The inventor of this application discovered that when compressive and tensile loads are repeatedly applied to a test object, the surface temperature of the test object repeatedly rises and falls around the temperature at zero load in synchronization with the cycle of the load application. .

For example, as shown in FIG. 4(A), when a load is applied to the test object in a sinusoidal manner, the compressive load corresponding to the positive half cycle and the tensile load corresponding to the negative half cycle are synchronized.
The surface temperature of the specimen repeatedly rises and falls in a sinusoidal manner as shown in Figure 4 (B), and also when compressive load or tensile load is applied in a rectangular wave (Figure 4 (C)).
and (E)), and the surface temperature rises or falls in synchronization with these (shown in FIGS. 4(D) and (F)).

It is known that there is a proportional relationship between the amount of change in surface temperature and stress change, so if a load is repeatedly applied to the test object and the intensity of temperature change at a specific point is detected,
You can know the magnitude of stress at that point.

A conventional stress distribution imaging method based on such a principle will be briefly explained with reference to FIG.

The example shown in FIG. 5 is a method of performing infrared measurement by scanning each point (-point) of the object to visualize the stress distribution. A load is applied to the subject L by a load fi2, the scanner 3 is stopped at each point, and each temperature data is read by the infrared detector 4.0For example, in the case of a sinusoidal load, the detected analog The temperature data is switched between the positive half cycle and the negative half cycle using the switch 5.
The data is sent to the D converter 6 (6a and eb), where it is converted into digital temperature data, and then stored in the corresponding memories 7a and 7b in the computer 7, respectively. From this stored temperature data, the first and second temperature data at the two points in time when the load amplitude difference is maximum within the same cycle are read individually and for each cycle, and averaged by the averaging circuit 7C. do. In this example, the first temperature data can be the data at the time of the maximum positive amplitude, and the second temperature data can be the data at the time of the maximum negative amplitude. It is also possible to average the detected temperature data within the temperature range, obtain a weighted average of these averaged values obtained for each cycle, and use these as the first and second temperature data. The difference between these first and second temperature data is calculated by a difference calculator 7d, and this difference, that is, temperature information including stress information is displayed on a display device such as CART! 18, so that the stress distribution of the subject can be displayed as an image.

Note that the load a2, the switch 5, and the A/D converters 6a and 6b.

The timing of the memories 7a, 7b, the averaging circuit 7c, and the difference calculator 7d is determined by a timing signal from a timing circuit 9.

This point measurement method requires time to measure temperature data point by point for the entire image and convert it into an image. Therefore, a method of shortening the measurement time by performing measurement by scanning a line or one screen may also be considered.

In this case, analog temperature data detected by scanning a part or all of the object while the load is being applied is stored in each of the memories 7a and 7b of the computer 7 in the same manner as described above.
to be memorized. This scanning may be a single line scan, several line scans, or a field scan, and then, for example, 256 to 512 points are taken in the line, and the line is scanned at high speed many times, After storing each of the first and second temperature data in the above-mentioned memories 7a and 7b, they are read from the data, an averaging circuit 7c outputs an average value of the first and second temperature data, and a difference calculator 7d calculates an average value of the first and second temperature data. A difference calculation between average values is performed and the result is stored in a one-screen memory (not shown). Then, after the temperature information per screen is recorded, this temperature information is sent to the display device a in the same manner as described above, and an image of the stress distribution is displayed.

Similarly, in single-screen scanning, except for the fact that the subject l needs to be scanned at a scanning speed comparable to that of television scanning, other points are processed in the same way as in the case of line scanning described above. , it is possible to image the stress distribution.

(Problems to be Solved by the Invention) However, the method of imaging the stress distribution of a subject based on such a principle is effective when the temperature gradient of the subject is small.

However, when a load is applied to this test object and compression or tension is performed, in addition to the temperature change caused by the stress, temperature data at another point that deviates from the original measurement point due to positional deviation of the test object is generated. Therefore, if there is a large temperature gradient in the subject for some reason, a large measurement error will occur.

The length of the subject l shown in FIG. 6 is, for example, vertical = 100
It is assumed that the elongation of the specimen 1 is 0.2 + oa + when a load of about ±22 kg/mm 2 is applied thereto.

On the other hand, for a width of ~3Gmm, the temperature gradient of the object 1 is ~
Assuming the temperature is about 30°C, there is a slope of 1"0 at 1+w+a.

By the way, in the method using digitization as described above, the line interval when scanning the subject 1 having a length of about 100 cm is about 1 mm to 0.5 tracks, so at this time If a positional shift error of This temperature error is a very large error compared to the signal. In reality, the temperature difference caused by this thermoelastic effect is 0.4° C. under the above-mentioned load, so the above-mentioned temperature error component becomes an extremely large error.

Therefore, it is necessary to correct the temperature error caused by such positional deviation.

When performing this correction, another problem is how much scanning line interval should be scanned. Assuming that one pixel is 11, moving in units of one pixel means that the center point of one pixel has moved by 1. Similarly to the above, if we consider that there is a temperature gradient of 30°C/30°C towards the subject, then [
In ■, the temperature change due to positional deviation is 1℃, and if you move something with such a temperature change at lam intervals,
Considering the magnitude of the signal, correction cannot be made.

An object of the present invention is to provide a novel method for imaging the stress distribution of a subject, which solves the above-mentioned problems.

(Means for Solving the Problems) In order to achieve this objective, according to the method of the present invention, a load is repeatedly applied to an object, the object is scanned using an infrared detector, and the load is measured. Each temperature of this object during two time periods giving the maximum load amplitude difference for each cycle is detected as first temperature data and second temperature data, and the difference between these first and second temperature data is calculated. When imaging the stress distribution of the subject.

A thermal filter is attached to the surface of the test object, and the thermal filter is scanned with an infrared detector to measure each temperature of the thermal filter during two time periods giving the maximum load amplitude difference for each cycle of the load. and detecting each as temperature data and fourth temperature data.

A step of obtaining interpolated data from the third and fourth temperature data under the condition that the stress value of the thermal filter is zero.

The method is characterized in that it includes a step of correcting an error caused by a one-position shift between the first and second temperature data using the interpolated data before performing the above-described difference calculation.

(Function) With this configuration, even when the temperature of the object becomes high and the temperature gradient becomes large, it is possible to accurately correct large temperature data measurement errors caused by minute positional displacements of the object using interpolated data. Therefore, it is possible to more accurately visualize the stress distribution of the object.

(Description of Embodiments) Hereinafter, embodiments of the present invention will be described with reference to FIGS. 1 to 3. Incidentally, even in the case of this invention, the load waveform shown in FIG. 4 and the circuit shown in FIG. 5 described above can be applied.
The detailed explanation will be omitted since it will be redundant.

Fig. 1 mainly shows the functional blocks of the computer 7 (shown in Fig. 5) used in this invention, Fig. 2 shows a subject to which a thermal filter is attached, and Fig. 3 shows correction of temperature data according to this invention. FIG.

In this invention, it is necessary to finely correct one position and to determine the amount of correction. The fineness of this position correction is set to such a degree that the temperature data required for the finally obtained temperature distribution image is not insufficient.

In this invention, as shown in FIG. 2, a thermal filter 10 is attached to the surface of the subject l. This heat filter 10
As a result, the temperature distribution of the specimen 1 becomes similar to that of the specimen 1, and the temperature gradient of the specimen appears on the surface, but the temperature change caused by the stress caused by the load application is not transmitted, and the And use one that expands and contracts in response to tension. As this thermal filter 10, for example, adhesive tape, vinyl tape, or any other material that satisfies the above-mentioned conditions can be used regardless of the material, size, and thickness. Further, the location where this thermal filter IO is attached can also be arbitrarily selected.

Next, when a load is applied to the subject 1 to which the thermal filter 10 is attached, the temperature distribution of the subject 1 itself appears in the part of the subject 1 where the thermal filter 10 is attached, but the temperature change component due to stress does not appear. This thermal filter 10 is scanned by infrared rays to detect third and fourth temperature data, and these data are stored in the other memories 7e and 7f of the computer 7, respectively, via the switch 5, as shown in FIG. do. Now, the value of the position correction of the subject 1 of one pixel or less is M
As shown in Figure S3, ΔH in the horizontal direction and ΔV in the vertical direction. In FIG. 3, Pji is a point on two-dimensional coordinates that is smaller than one pixel, and four points around this Pji point, that is, temperature data Dji, Dj
i+1.0j◆11+1, D The rectangular area surrounded by the points represented by j+Ii is one pixel. So, for example, consider the case of vertical tension.

Since ΔH can be 0, the third and fourth temperature data taken into the memories 7e and 7f are averaged by the averaging circuit 7g, and then the ΔV and ΔH determining circuit 7h calculates ΔV very finely, for example by scanning. 1/100 unit of line spacing (
(1/100 or less of one pixel) to measure the temperature distribution component due to the positional deviation of the subject l itself. In the obtained temperature distribution data, the stress value is set to zero since no temperature change component due to stress appears in the portion of the thermal filter IO.

If the interpolation is performed accurately, a correction value ΔV that makes the component zero can be found. Generally, ΔH and ΔV are obtained as correction values through similar processing.

Next, these correction values ΔH and Δ■ are sent to the interpolation data determining circuit 61, where, based on these correction amounts ΔH and ΔV, interpolation data=DjiX(1-ΔH)X(1-ΔV+D
ji+1× (ΔH)X(1-ΔV)+Dj+1i
+1×(ΔH)X (ΔV)+Dj+1iX (1
-Δ)()X(ΔV) Interpolated data at point Pji is calculated from the formula: -Δ)()X(ΔV).

Next, the first and second temperature data obtained by performing normal measurement of the object in advance or after that are stored in the memories 7a and 7b via the switch 5, and each stored data is read out. After being averaged by the averaging circuit 7C, it is sent to the correction circuit 71. On the other hand, if the interpolated data obtained from the interpolated data determination circuit 71 described above is supplied to the correction circuit 7j, and each of the averaged first and second temperature data is corrected using this interpolated data, the movement will be corrected. The positional correction component of the temperature data due to the temperature data is completely compensated for, and an accurate temperature component is obtained.

Thereafter, the corrected first and second temperature data are sent to a difference calculation circuit 7d, where a difference calculation is performed, and the result is sent as temperature information to a display device (indicated by 8 in FIG. 5). Displays an image.

Through the processing in the computer 7 described above, even in the case of an object with a large temperature gradient, the positional deviation of the temperature data is corrected by interpolation, so that the load amplitude difference to be calculated by the subsequent computer processing is maximized. There is almost no measurement error component in the first and second temperature data at the two points in time.

As mentioned above, if a load is applied to the test object and the test object generates heat for some reason and the temperature gradient becomes large, even a small positional change of the test object can cause a large temperature data measurement error. However, according to the present invention, this measurement error can be corrected. Therefore, to summarize the method of imaging the temperature distribution of a subject according to the present invention, (a) the temperature change due to the thermoelastic effect is not transmitted to the surface of the subject, but the temperature distribution is almost equal to that of the subject itself, and the temperature distribution of the subject is Attach a thermal filter made of a material that can accommodate the expansion and contraction of the object, import the temperature data of the thermal filter attached to the test object into a computer, and use computer processing to calculate the amount of positional deviation correction for the temperature data that will make the stress value zero. (b) a step of importing temperature data into a computer for a subject to which a thermal filter is not attached; and (c)
Using the positional deviation correction amount obtained in step (a), (b)
) The positional correction of the temperature data of the test object without a thermal filter obtained in step 1 is performed by computer processing by interpolation of one pixel or less, for example, to about 1/100 of one pixel; and a step of performing imaging by performing a difference calculation using the temperature data of the subject to which no thermal filter is attached.

In the above-described embodiments, metal was used as the object to be examined, but the present invention can be applied to other materials having a large temperature gradient.

Furthermore, in this invention, even if the scanning is point scanning,
Line scanning or single screen scanning may be used.

(Effects of the Invention) As is clear from the above explanation, according to the method of the present invention, even when a load is applied to the test object, which generates heat and the temperature gradient becomes large, the minute temperature gradient of the test object can be maintained. Since large temperature data measurement errors caused by temperature fluctuations can be accurately corrected, there is an advantage that the temperature distribution of the subject can be imaged more accurately.

[Brief explanation of the drawing]

FIG. 1 is a functional block diagram of a computer for explaining one embodiment of the method of the present invention, FIGS. 2 and 3 are diagrams for explaining the embodiment of the present invention, and FIG. 4 (A ) to (F) are line diagrams and signal waveform diagrams showing the subject part for detailed explanation of the conventional method and the present invention. FIG. 5 is a diagram showing a conventional system and an apparatus system for explaining the present invention, and FIG. 6 is a diagram for explaining the drawbacks of the conventional method. l...Object to be inspected, 2...Loading device 3...Scanner, 4...Infrared detector 5...Switcher, 6.6ia, 8b-A/rJ converter 7, -
Computer, 7a, 7b, 7e, 7f - memory 7
c, 7g...Averaging circuit, 7d...Difference calculator C+a (deaf Yoro) 7h...ΔV and ΔH determining circuit 71...Interpolation data determining circuit 7j...Correction circuit, 8...Display device 9...
・Timing circuit, lO...thermal filter.

Claims (1)

[Claims] 1. A load is repeatedly applied to a test object, and the test object is scanned using an infrared detector to determine the test object during two time periods that give a maximum load amplitude difference for each cycle of the test object. Each temperature of the specimen is detected as first temperature data and second temperature data, and the difference between these first and second temperature data is calculated to image the stress distribution of the specimen. A thermal filter is attached to the surface, and the thermal filter is scanned with an infrared detector to obtain third temperature data and third temperature data for each temperature of the thermal filter in two time periods that give the maximum load amplitude difference for each cycle of the load. a step of detecting each as four temperature data; a step of obtaining interpolated data from the third and fourth temperature data under conditions such that the stress value of the thermal filter is zero; and before performing the difference calculation, A method for imaging stress distribution in a subject, comprising the step of correcting errors due to positional deviation in the first and second temperature data using the interpolated data. 2. In the method for imaging stress distribution in a subject as set forth in claim 1, the material of the thermal filter is heated to a temperature comparable to the temperature distribution of the subject so that the temperature gradient of the subject is An image of stress distribution in a test object, which is characterized by a material that does not transmit temperature changes caused by stress caused by load application and expands and contracts in response to compression and tension of the test object. method. 3. In the method for imaging stress distribution in a subject as described in claim 1, when Pji is a point on two-dimensional coordinates smaller than one pixel, the interpolated data at point Pji is interpolated data = Dji×(1-ΔH)×(1-ΔV+Dj
i+1×(ΔH)×(1-ΔV)+Dj+1i+1×(
ΔH)×(ΔV)+Dj+1i×(1-ΔH)×(ΔV
), in this case, Dji, Dji+1, Dj+1i
+1, Dj+1i are four points around the point Pji such that the rectangular area surrounded by these points forms one pixel, and ΔH and ΔV are from the point Pji to one side of the rectangle and this side. A method for imaging stress distribution in a subject, characterized in that ΔH and ΔV are calculated from conditions under which the stress value of the thermal filter is zero. 4. In the method for imaging stress distribution in a subject as set forth in claim 1, the scanning by the infrared detector is performed by any one of point scanning, line scanning, or single screen scanning. Imaging method for stress distribution.