JPH0630740B2 - Defective product detection method and defective product detection / elimination device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION (a) Field of Industrial Use The present invention measures the weight and size of an article to be inspected in an inspection process such as a production line, and the measured value is measured outside a predetermined management area. In some cases, the present invention relates to a method and apparatus for judging the article as defective. Further, the present invention relates to a device for determining whether or not the number of packed articles matches a predetermined number.

(B) Conventional technology Parts and finished products (hereinafter collectively referred to as articles) produced in a production line of a factory or the like are generally packaged as an integrated body of one piece or a plurality of pieces, and an assembly factory. And supplied to the general market. Therefore, an inspection process for distinguishing good products from defective products and eliminating defective products is usually installed in the final process. In this inspection process, although the inspection target and quality differ depending on the type of the inspected article, the weight and size of the article are generally inspected for suitability.

Conventionally, in a weight inspection device for inspecting the weight of an article, a fixed reference value has been used, in which the standard weight of the article is used as a reference value, and the allowable range and control area of non-defective products are set based on the measured variation in weight. The method (abbreviated as fixed standard method) has been adopted.

However, since the basic idea of this fixed standard method is to set the control area by a statistical method assuming an infinite population, if it is applied as it is, the range of acceptable products will be extremely wide. , There is a possibility that a large amount of defective products may be mixed in the product. In order to avoid such a problem, an appropriate management area is set in consideration of the degree of rejection of non-defective products and the risk of mixing defective products, but it is sufficient if the weight of the products varies. It has the drawback that it cannot perform its inspection function.

In particular, if the physical properties of the raw materials supplied to the manufacturing process that affect the weight and size of the product are liable to fluctuate, the fixed standard method should be used to widen the control range or to manually set the standard value frequently. Since it has to be changed and set again, the possibility of mixing defective products into the product increases, or if extra manpower is required and the reference value control timing is incorrect, the defective product is regarded as a good product. Mistakes are easy to make.

Therefore, as a method for solving such problems as described above, "the weight of the produced article is measured, and whether or not the weight is within the control range is inspected in an inspection step to distinguish good articles from defective articles. When rejecting defective products, an average value calculated from the weights of 3 to 20 items which were determined as good products by measuring the weight immediately before is used as a numerical value for setting the above-mentioned control area, and the management set by that is used. A method for detecting defective products in a production line, which is characterized by determining whether the product weight is good or bad according to the area. ”

(C) Problem to be Solved by the Invention As described above, in the conventional technique, the management area is an average value of weights of a plurality of articles determined to be non-defective immediately before (hereinafter abbreviated as a moving average value). However, even if the control area is narrowed, the inspection accuracy is guaranteed to some extent.

However, it has been empirically known that the condition and condition of the production line (device) fluctuate due to various factors such as weather, temperature, humidity and familiarity of mechanical parts, and in particular, they greatly change at the start and end of work. Has been.

The graph shown in FIG. 2 shows this variation.

As is clear from this graph, the management area A at the start of work is located at a high place, while the management area B at the end of work is located at a low place.

Therefore, the data at the end of the day before is not useful at the start of the day. That is, when the management area A and the management area B have a small overlapping portion, if the management area B at the end of the previous day is used as a reference, most of the goods that should be good products will be determined to be defective products. .

In addition, if the width of the management area itself is widened in order to solve such a problem, there is a risk that the defective product will be judged to be a good product.

The present invention has been made in view of the problems of the related art, and by setting the management area to a fixed maximum allowable width only from the start of work to the inspection of a specific number of articles, It is intended to prevent the non-defective product from being judged as non-defective by setting the width of the management area narrow based on the moving average value while reducing the number of non-defective non-defective products.

(D) Means for Solving the Problem A first main invention is to individually measure the weight of a plurality of articles, detect whether the weight of each article is within a predetermined management range, and In the defective product detection method for determining an article located outside the management area as relatively defective, from the start of the weight measurement operation until the weight of two or more first specific number of articles is measured, After the control area is set to a fixed maximum width and the weight of the first specific number of articles is measured, the weight is measured immediately before, and two or more second It is characterized in that new management areas are sequentially set according to the average value of the weight of a specific number of articles.

A second main invention is a transfer means for transferring a plurality of articles continuously or intermittently, a weight measuring means for individually measuring the weight of the articles transferred by the transfer means, and a predetermined weight of each article. Detection means for detecting whether or not it is in the management area, determination means for determining an article whose weight is outside the management area as defective, and exclusion means for excluding an article determined to be defective. In the defective product detection / exclusion device, a detection means for detecting whether or not the article is in a fixed fixed initial management area, and two or more first primary detection areas in the initial management area from the start of the weight measurement operation are provided. Counting means for counting the number of the specified number of articles, and average value calculation means for calculating an average value of the weights of two or more second specified number of articles whose weight is measured immediately before and determined not to be defective, Set a new management area based on the average value It is characterized in that the dynamic management area setting means is provided.

A third main invention is to measure the dimensions of a plurality of articles individually to detect whether or not the dimensions of each article are within a predetermined management area, and the dimensions are outside the management area. In the defective product detection method that determines that a product is relatively defective, the maximum amount of fixing that is allowed in the control area is maintained from the start of the size measurement operation to the measurement of the size of the first specific number of two or more products. After significantly setting and measuring the dimensions of the first specific number of articles, immediately after measuring the dimensions, the average value of the dimensions of two or more second specific number of articles determined not to be defective is determined. The feature is that new management areas are sequentially set.

A fourth main invention is a transfer means for transferring a plurality of articles continuously or intermittently, a dimension measuring means for individually measuring the dimensions of the articles transferred by the transport means, and a dimension of each article is predetermined. Detection means for detecting whether or not the article is in the control area, determination means for determining an article having a dimension outside the control area as defective, and exclusion means for excluding an article determined to be defective. In the defective product detection / exclusion device, a detection means for detecting whether or not the article is in a fixed fixed initial management area, and two or more first first detection areas in the initial management area from the start of the dimension measurement operation. Counting means for counting the number of the specified number of articles, and average value calculation means for calculating the average value of the dimensions of two or more second specified number of articles whose dimensions have been measured immediately before and determined not to be defective, Set a new management area based on the average value It is characterized in that the dynamic management area setting means is provided.

(E) Operation In the first main invention, from the start of the weight measurement operation until the weight of two or more first specific number of articles is measured, that is, as shown in FIG. Until data is collected, set the initial management area to a fixed maximum width C,
After measuring the weight of the first specific number of articles, a new moving average value of the weight of two or more second specific number of articles determined immediately before is determined to be non-defective is used to sequentially generate new values. Since the management area is set, the initial management area C includes the management area A at the start of work and the management area B at the end of work.

Therefore, an article that should originally be a good product at the start of work is not determined as a defective product.

Further, since the management area is set based on the moving average value after the initial data is collected, there is no possibility that a defective product is determined to be a good product.

The operation of the other main invention is not substantially different from that of the first main invention.

(F) Embodiment FIG. 3 to FIG. 6 are defective product detection / exclusion devices as a first embodiment of the present invention. FIG. 3 is a schematic front view of the device, and FIG. FIG. 5 is a schematic block diagram, FIG. 5 is a block diagram showing details of the memory, and FIG. 6 is a flow chart for explaining the operation.

In FIG. 3, reference numeral 1 is a main body of the apparatus, a weight measuring unit 2 is provided in the lower center portion thereof, and a weighing conveyor 3 is provided above the weight measuring unit 2. The weighing conveyor 3
4, a load cell 4 as shown in FIG. 4 is connected, and the packed caramel 5 in a box as the inspected article is placed and transferred on the weighing conveyor 3 one by one while the weight thereof is being transferred. To be measured.

A supply conveyor 6 for supplying untested boxed caramel 5 is provided on the weighing conveyor 3 on the upstream side of the weighing conveyor 3 along the transfer direction of the boxed caramel 5. On the downstream side, a sorting conveyor 7 is provided for sorting the in-box caramel 5 into good products and defective products (insufficient or excessive).

The sorting conveyor 7 is provided with a defective product discharging device 8 for laterally discharging the caramel 5 in the box which is determined to be defective.

Photo sensors 9 and 10 for detecting the presence of the caramel in the box 5 are provided near the downstream sides of the supply conveyor 6 and the weighing conveyor 3, respectively. Further, 11 is a metering control unit, and 12 is a level adjuster.

Next, the control circuit will be described mainly with reference to FIGS. 4 and 5.

A DC amplifier 13 is connected to the load cell 4, and an analog / digital converter (A / D converter) 15 is connected to the DC amplifier 13 via a filter 14. Further, the A / D converter 15 has an input / output interface (I / D
A microprocessor (MPU) 18 is connected via an O interface) 16 and a bus line 17.

Further, the I / O interface 16 is connected with a ten-key 19 for inputting and setting the number of inspections, reference values, etc., and also with a defective product discharging device 8, a display device 20, a printer 23, etc.

Further, the bus line 17 has a random access memory (RAM) 21 for temporarily storing various data.
A read-only memory (ROM) for storing control programs and fixed data is connected to 22 units.

Next, the operation of this apparatus will be described mainly based on the flowchart of FIG.

First, the operation is started in step S1, and the weight of each caramel 5 in a box is measured by the weighing conveyor 3 in step S2 to collect data. At this time, it is judged whether or not each caramel 5 in the box is in the initial management area C in FIG. 1 (step S3), and the caramel 5 in the box is concerned.
Is within the initial management area C, the data is stored in the data memory DMn (RAM 21) (step S
4) The number of non-defective products is counted (step S5).

Then, this process is repeated until the number n of products reaches the first specific number n0 (specifically, 10) (step S7).

When it is determined in step S3 that the caramel with box 5 is outside the initial management area C, that is, the weight of caramel with box 5 is larger than the initial upper limit value or smaller than the initial lower limit value. In the defective product discharge device 8
And the defective caramel 5 in the box is removed to the outside (step S6). At this time, defective products are not counted.

Then, the average value m of the weight of the caramel 5 in the box which is determined to be non-defective in the above step is obtained (step S8),
The upper limit of the movement management area is m + b obtained by adding the movement management width b to the weight average value m, and the lower limit of the movement management area is mb (step S9). That is, the absolute value of the movement management area is 2b. It should be noted that the movement management width b is fixedly set by manual setting or is obtained from variations in initial data.

Then, in step S10, caramel 5 in each box is further added.
Are weighed and data is collected.

At this time, it is judged whether or not the caramel 5 in each box is within the movement management area (step S11), that is, m-b <xn <.
It is determined whether or not the relationship of m + b is established, and when the caramel with box 5 is in the movement management area, the data is stored in the data memory DMs (RAM 21) (step S12), and the number of non-defective products is counted. (Step S1
3). s is the remainder of n ÷ n0.

Then, until the number n of non-defective products reaches the second specific number n0,
This process is repeated, and the moving average value of the data collected by this process is obtained (step S14).

In the present embodiment, the second specific number n0 is ten, which is the same as the first specific number n0, but the second specific number may be smaller than the first specific number. .

When it is determined in step S11 that the caramel with box 5 is outside the movement management area, that is, the weight of caramel with box 5 is greater than the upper limit value of the control area or less than the lower limit value of the control area. In this case, the defective product discharging device 8 operates in the same manner as described above, and the defective box-flow caramel 5 is excluded to the outside (step S15). Also at this time, defective products are not counted.

In this way, the operation is stopped after the inspection of the predetermined number of caramel boxes 5 is completed (steps S16 and S1).
7).

When the inspection is restarted, the process returns to step S1 (step S18).

Although not shown in FIG. 6, a normal area is set further outside the initial management area C as shown by a chain double-dashed line in FIG. 1, and the weight of the caramel 5 in the box is further than this normal area. When it takes a value that is outside, it may be configured to determine that the device itself is in an abnormal state, notify that the device is in an abnormal state, and stop the entire device.

In addition, in the present embodiment, the weight of the caramel 5 in the box is measured to detect and eliminate defective products. It goes without saying that the present invention can be applied to the inspection of missing parts such as articles and industrial parts such as bolts and nuts.

Next, a second embodiment will be described.

In the second embodiment, instead of measuring the weight of the article to detect a defective article, the poor sealing of the sealed container (specifically, the canned container) is measured by measuring the deformation amount of the sealed container. Since this is a detection process and there is no difference in the data processing from the first embodiment, only the structure of the device will be described.

FIG. 7, FIG. 8, FIG. 9 and FIG. 10 show the device of the second embodiment, FIG. 7 and FIG. 8 are sectional views of the essential parts of different modes, and FIG. 1 is a front view showing
FIG. 0 is a plan view showing the overall structure of the apparatus.

In these figures, a can 101 for packing is placed on a positioning table 107 fixed near the periphery of a circular rotary index table 106, and above it is a flat cylinder having an opening on the lower surface. A suction chamber 108 is provided so as to be vertically movable by an air cylinder 109. A capacitance type distance detection sensor 110 is fixed to the inner center of the suction chamber 108. Further, an air suction pipe 111 connected to a vacuum pump (not shown) is inserted and fixed in the suction chamber 108 for sucking the air in the suction chamber 108 to reduce the pressure. 1
12 is a case.

Then, when the suction chamber 108 hangs down from the state shown in FIG. 8 and is brought into pressure contact with the upper surface of the can 101, and the air in the suction chamber 108 is sucked and depressurized in this state,
The pressure in the space defined by the suction chamber 108 and the upper surface of the can 101 decreases. Then, a stress that raises the upper surface of the can 101 is generated.

At this time, when the can 101 does not have a large depression or a poor sealing, the pressure inside the can 101 is also low, and the upper surface of the can 101 does not rise so much.

However, when the canister 101 is poorly sealed, the internal pressure of the canister 101 becomes high and in some cases becomes equal to the atmospheric pressure, so that the upper surface of the canister 101 is lifted by 1.

Therefore, first, in the state before the air in the suction chamber 108 is sucked, the can 1 is moved from the lower end of the distance detection sensor 110 to the can 1.
The distance from the lower end of the distance detection sensor 110 to the upper surface of the can 101 is measured again while the air in the suction chamber 108 is sufficiently sucked and decompressed. If the difference between the measured value and the stored value is calculated, the difference is
Becomes

If the difference value l is greater than or equal to a predetermined value, it is notified that the can 101 is defective in sealing. That is, if the canister 101 is poorly sealed, the pressure inside the canister 101 becomes high as described above, and the value l of the difference becomes large accordingly, so that the value l becomes a predetermined value or more. This means that the can 101 has a poor sealing.

In addition, even when a large depression is formed in the can 101, the internal volume of the can 101 decreases, and the internal pressure rises in inverse proportion to the internal volume. Similarly, the upper surface of the can 101 is lifted by l. Therefore, in this case as well, it is detected in the same manner as described above.

It should be noted that it is difficult to determine whether the canister 101 has a sealing failure or a large depression is formed. However, since it is a defective product in any case, it is not necessary to strictly determine both.

In the present embodiment, the difference between the distance from the distance detection sensor 110 to the upper surface of the can 101 at the time of decompression and the distance from the distance detection sensor 110 to the upper surface of the can 101 at the time of non-decompression is calculated, and the difference is predetermined. If the difference is different from the predetermined value, it is notified that the can 101 has a poor sealing or the can 101 is deformed. Therefore, as in the pull-top can, it is displayed on the upper surface of the can 101. Even if unevenness is formed, it can be sufficiently detected.

Next, the overall configuration of the apparatus will be outlined with reference to FIGS. 9 and 10.

Reference numeral 113 denotes an apparatus main body, and a drive unit, a vacuum pump, and the like are stored inside the apparatus main body 113. The device body 11
The above-mentioned rotation index table 106 is mounted on the upper surface of the No. 3 table. The rotary index table 10 is also provided above the rotary index table 106.
A cylinder holding table 114 having the same diameter as 6 and rotating synchronously is provided.

The cylinder holding table 114 has a positioning table 107 fixed to the rotation index table 106.
The air cylinders 109 are fixed at positions corresponding to the piston rods 1 of the air cylinders 109, respectively.
The suction chamber 108 is fixedly provided at the tip of 15.

Rotating tables 116 and 117 called star wheels are rotatably attached to the supply portion and the discharge portion of the can 101 of the rotation index table 106, respectively. 11
Reference numerals 8 and 119 denote rotary shafts of the rotary tables 116 and 117, respectively. Recesses 116a in which the cans 101 can be inserted one by one are provided in the peripheral portions of the rotary tables 116, 117.
117a is cut.

In addition, a linear conveyor 102 extends from the supply section to the discharge section of the can 101, and the can 101 conveyed by the conveyor 102 is provided on the upstream side thereof with the rotary tables 116 and 117. Recesses 116a, 117 of
A pitch adjusting screw conveyor 120 is provided to match the pitch of a.

121 is a control box.

The can 101 conveyed by the linear conveyor 102 has its pitch adjusted by the screw conveyor 120, and is supplied to the rotary table 116.
The can 101 rotated about 180 ° on the rotary table 116 is the rotary index table 106.
It is transferred onto the positioning table 107 of.

And, the can 101 is a rotary index table 1
During the rotation of about 320 ° along the rotational movement of 06, the presence or absence of the sealing failure is detected as described above.

The can 101 after detection is supplied to the second rotary table 117, finally placed on the downstream side of the linear conveyor 102, and further transported in the downstream direction. At this time, the can 101, which is judged to be defective due to the presence of sealing failure or the like, has an air cylinder 122 for discharging defective products.
02 is discharged to the outside (side).

Next, based on the block diagram shown in FIG. 11, a brief description will be given of the operation of detecting a sealing failure or the like. In this figure, 1
Reference numeral 23 denotes an MPU, and the ROM 125 and the RAM 126 are connected to the MPU 123 via a bus line 127, respectively. The bus run-in 127 has an I /
Distance detection sensor 11 via the O interface 124
Controlled devices such as 0 and the cylinder 122 are respectively connected.

The ROM 125 stores a predetermined value corresponding to the difference value 1 of the non-defective can 101, and the RAM 126 stores the distance detection sensor 110 which is transferred from the distance detection sensor 110 when the pressure is reduced. The data on the distance to the upper surface of 101 and the data on the distance from the distance detection sensor 110 to the upper surface of the can 101 when the pressure is not reduced are temporarily stored. Then, the difference between the two data is calculated by the MPU 123, and the difference is stored in the ROM.
Compared with a predetermined value stored in advance in 125, when the difference is larger than the predetermined value, it means that there is a sealing failure or the like in the can 101 as described above, so the can 101 is excluded. A signal for doing so is supplied to the air cylinder 122 for discharging the defective product.

As described above, the processing of the inspection result data is performed by the first
In the same manner as in the above embodiment, those located outside the initial management area and the movement management area are excluded as defective products.

(G) Effect of the Invention In the present invention, since the management area is sequentially set using the movement reference value, it is possible to narrow the management area and prevent defective products from being judged as non-defective. Therefore, there is no risk of defective products being mixed in the product.

In addition, the control area is set to a fixed maximum allowable width only during the inspection of a specific number of items from the start of work, so that even if the condition of the device fluctuates, the number of defective products judged to be non-defective at the start of work can be reduced. It is possible to improve the inspection efficiency.

Further, when the normal range is set to be wider than the maximum width of the management range and it is determined that the weight or size of the article is outside the normal range, the device is in an abnormal state is notified. In addition, if the apparatus is configured to be stopped, it is possible to recognize a state other than good or defective of the article such as a failure of the apparatus.

Furthermore, if the inspection result is fed back to the manufacturing apparatus (for example, a filling machine), more accurate control becomes possible.

[Brief description of drawings]

1 is a graph showing a change in weight of an article in the first embodiment of the present invention, FIG. 2 is a graph showing a change in weight of an article in a conventional example, and FIG. 3 is a first embodiment of the present invention. FIG. 4 is a schematic front view of the apparatus, FIG. 4 is a schematic block diagram thereof, and FIG.
FIG. 6 is a block diagram showing the details of the memory, FIG. 6 is a flow chart, FIG. 7 is a sectional view of the essential parts of a second embodiment of the present invention, and FIG. 8 is a sectional view of the essential parts of a mode different from FIG. Figure, No. 9
The figure is a front view of the overall structure, FIG. 10 is its plan view, and 11th.
The figure is a block diagram. 1 ... Device main body, 2 ... Weight measuring unit, 3 ... Weighing conveyor, 4 ... Load cell, 5 ... Box caramel (article), 6 ... Supply conveyor, 7 ... Sorting conveyor, 8 ...
… Defective product ejector, 9, 10… Photo sensor, 11…
... Measurement control unit, 12 ... Level adjuster, 13 ... DC amplifier, 14 ... Filter, 15 ... A / D converter,
16 ... I / O interface, 17 ... bus line,
18 ... MPU, 19 ... Numeric keypad, 20 ... Display device, 21 ... RAM, 22 ... ROM, 23 ... Printer, 101 ... Can, 106 ... Rotating table, 107 ...
... Positioning table, 108 ... Suction chamber, 109 ... Air cylinder, 110 ... Distance detection sensor, 111 ... Air suction pipe, 112 ... Case, 113 ... Device body,
114 ... Cylinder holding table, 115 ... Piston rod, 116, 117 ... Rotating table, 118, 1
19 ... Rotary axis, 120 ... Screw conveyor, 12
1 ... Control box, 122 ... Air cylinder, 123
...... MPU, 124 ...... I / O interface, 125
... ROM, 126 ... RAM.

Claims (10)

[Claims] 1. The weights of a plurality of articles are individually measured to detect whether or not the weight of each article is within a predetermined management area, and the weight of the article outside the management area is determined relative to each other. In the defective product detection method that determines that the product is defective, the control area is set to a maximum allowable fixed width from the start of the weight measurement operation until the weight of two or more first specific number of articles is measured. However, after measuring the weight of the first specific number of articles, the weight is measured immediately before, and the average value of the weights of two or more second specific number of articles determined to be not defective is sequentially updated. Defective product detection method characterized by setting various control areas. 2. The defective product detection method according to claim 1, wherein the second specific number is 3 or more and 20 or less. 3. The defective product detection method according to claim 1, wherein the first specific number is the same as or larger than the second specific number. 4. A transfer means for transferring a plurality of articles continuously or intermittently, a weight measuring means for individually measuring the weight of the articles transferred by the transfer means, and a weight control area for controlling the weight of each article. The detection means for detecting whether or not the article is present, the determination means for determining an article whose weight is outside the management area as defective, and the elimination means for eliminating the article determined to be defective. In the non-defective item detection / exclusion device, a detection means for detecting whether or not an article is in a fixed fixed initial management area, and a first specific number of two or more which are in the initial management area from the start of the weight measurement operation. Counting means for counting the number of articles, average value calculating means for calculating an average value of the weights of two or more second specific numbers of articles, the weight of which has been measured immediately before and determined not to be defective, and the average value Management to set up a new management area based on A defective product detection / exclusion device, comprising: a region setting means. 5. The defective product detection and elimination apparatus according to claim 4, wherein the weight of each article is measured on a weighing conveyor to which a load cell is connected. 6. The defective product detection / exclusion device according to claim 5, wherein a supply conveyor for supplying the articles onto the weighing conveyor is provided upstream of the weighing conveyor along the article transfer direction. 7. A sorting conveyor for receiving the weighed articles from the weighing conveyor is provided on the downstream side of the weighing conveyor along the article transfer direction, and defective sorting means for rejecting defective articles is provided on the sorting conveyor. The defective product detection / exclusion device according to claim 5 or 6, which is provided. 8. The dimensions of a plurality of articles are individually measured to detect whether or not the dimensions of each article are within a predetermined control area, and the articles whose dimensions are outside the control area are relatively measured. In the defective product detection method that determines that the product is defective, the control area is set to a maximum allowable fixed width from the start of the dimension measurement operation until the dimension of two or more first specified number of articles is measured. Then, after measuring the dimensions of the first specific number of articles, a new value is sequentially determined by an average value of the dimensions of two or more second specific numbers of articles which are determined to be defective by measuring the dimensions immediately before. Defective product detection method characterized by setting various control areas. 9. A transfer means for transferring a plurality of articles continuously or intermittently, a dimension measuring means for individually measuring the dimensions of the articles transferred by the transport means, and a control area in which the dimensions of each article are predetermined. A detection means for detecting whether or not the product is present, a determination means for determining an article having a dimension outside the control area as defective, and an elimination means for eliminating an article determined to be defective. In the non-defective item detection / exclusion device, a detection means for detecting whether or not an article is in a fixed fixed initial management area, and a first specific number of two or more existing in the initial management area from the start of the dimension measurement operation. Counting means for counting the number of articles, average value calculating means for calculating the average value of the dimensions of a second specific number of two or more articles whose dimensions were measured immediately before and were determined not to be defective, and said average value Management to set up a new management area based on A defective product detection / exclusion device, comprising: a region setting means. 10. The defective product detection method according to claim 8, wherein the article is a hermetic container, and the defect is detected by measuring a deformation dimension that is deformed when the hermetic container is pressurized or depressurized from the outside. The defective product detection and elimination device according to claim 9.