JPH04200822A - Method for calculation corrected dimension, method for adjusting dimension and manufacture of vtr head - Google Patents

Abstract

PURPOSE:To execute high-precision control by obtaining distributions of an amount of a plastic deformation and an integrated amount of deformation through an experiment, etc., transforming the distribution into a membership function. CONSTITUTION:A head base 7 is fixed by a fixing member at a position 6, a relative distance X in the X-axial direction and a relative distance Z in the Z-axial direction of both reference points 11, 12 of a left adjusting part 8 and a right adjusting part 9 of two head chips of a DA head are aligned by deforming a head base 7 plastically and the respective relative distance X, Z in the X-axial direction, in the Z-axial direction are regulated within a certain range of tolerance. In order to make X, Z within the range of tolerance, the pictures of both reference points 11, 12 on the left part 8 and the right part 9 are picked up by a TV camera and compared with X, Z of a good product and controlling forces actuated in the X-axial direction and in the Z-axial direction so that dimensional difference between X, Z measured and X, Z of the good product are within the range of tolerance are calculated and adjusted by an integrated amount of deformation obtained.

Description

[Detailed description of the invention] [Field of application of the product This invention relates to a method for determining the amount of action to be applied directly to a part when the desired adjustment amount does not match the amount directly applied to the part. In particular, the present invention relates to a VTR head adjustment method for aligning the relative pressure of the tip by plastically deforming the base of the VTR.

[Conventional technology]

In the conventional method, as described in Japanese Unexamined Patent Publication No. 1 243218, adjustment is performed by deforming the member, and the desired amount of strain is obtained from a predetermined relational expression between the force applied during deformation and the amount of strain. It is controlled by determining the amount of action necessary for

Further, this member is limited to a magnetic head, and there are two types of positioning parameters for this magnetic head, and each parameter is controlled separately.

[Problem to be solved by the invention]

In the above conventional technology, the amount of adjustment (plastic deformation amount) of the part and the amount of action directly applied to the part during adjustment (total deformation amount) are controlled using a predetermined relational expression. Therefore, for the part you want to adjust (7), how significant is it, that is, apply a certain amount of force to the part,
Even if a part is plastically deformed, the amount of plastic deformation is not necessarily constant, so there is a problem that it is unlikely to perform highly accurate adjustment. In addition, in the above method, the rose to be adjusted
6. There are two or more meters, and each one influences the other, but since the influence of the phase θ is not taken into consideration,
If you adjust one side, the force will go crazy.)
There was a problem.

The purpose of the present invention is to find the most probable part 711 from the distribution of the amount of plastic deformation of the part obtained from the experimental values and the total amount of deformation in a state Q when force is applied to the part, and adjust the part using the total zero deformation sensitivity. It is an object of the present invention to provide a control method in which the parameter to be adjusted is adjusted in consideration of the mutual influence of other parameters. .

[Means to solve the problem]

In order to achieve [Objective 1] above, the amount of plastic deformation (finally the amount of deformation remaining in part 1) determined by experiment and the amount of total deformation (
The distribution of the relationship between the amount of deformation when a force is applied to the part (71, - is converted into a membership function, and the desired amount of plastic deformation is determined from the membership function number - of 22). In order to obtain

l.Also, even if we know that there are two or more parameters that influence each other, we can distribute the correlation between each parameter using experimental data, replace it with a membership function, create a rule, and perform fuzzy inference. Controls were designed to take into account the effects of each.

Furthermore, in order to perform more reliable control, the above-mentioned fuzzy inference was repeated several times to adjust the control considering the correlation of parameters as well as the significance of the control.

[Effect]

By determining the distribution of the amount of plastic deformation and the amount of total deformation through experiments, etc., and converting that distribution into a membership function that is the basis of Bouasii inference, for example, the frequency distribution of experimental values can be transformed into a membership value that represents significance. When determining the value to be controlled, its significance is taken into consideration, allowing highly accurate control.

Also, when two or more parameters influence each other, it is possible to determine the correlation between the parameters by experimentally determining the correlation between each parameter and providing the rules that form the basis of fuzzy inference. , the control that takes into account the interrelationship of parameters is
Becomes Jf Noh.

In addition, the correlation between characteristics and parameters is considered at the same time. −2, repeating fuzzy inference multiple times,
- and - are 111' more efficient, so adjustments to the parts can be reduced.

〔Example〕

Embodiments of the present invention will be described using FIGS. 1 to 11.

Figure 1 shows an overview of the present invention, from the amount of deformation (plastic deformation) to which the target part 1 is desired to be adjusted, to the amount of deformation that can actually be applied to the part (total deformation), which will be explained in detail later. FIG. 3 is a diagram illustrating calculation by fuzzy inference.

FIG. 2(A) shows V'I, which is one of the application targets of the present invention.
The object of the present invention is to fix the head head 7 at position 6 with a fixing member (not shown), and fix the left adjustment portion 8 (
1': (hereinafter referred to as the left part 8) and the right adjustment part 9 (hereinafter referred to as the left part 8)
The relative distance ΔX in the X-axis direction and the relative distance ΔZ in the Z-axis direction of both reference points 1.1.1.2 (referred to as right part 9) (see FIG. 2 (B)) are determined by plastically deforming the head base 7. The purpose of this is to adjust the position by adjusting the positioning angle, and to bring the relative distances ΔX and Δ2 in the X-axis direction and the Z-axis direction, respectively, within a certain tolerance range.

The relative distances ΔX and Δ in the X-axis direction and Z-axis direction, respectively.
In order to put Z within a certain tolerance range, left part 8 and right part 9
Images of both reference points 11 and 12 are captured by a TV camera (not shown).
The measured ΔX was taken in and compared with the ΔX and ΔZ of a good product.
, ΔZ and ΔX, ΔZ of non-defective products, as shown in FIGS.
13, 14.15.16. working in the X-axis direction. The adjusting forces 17.18°19 and 20 acting in the Z-axis direction are adjusted by the total deformation amount obtained by the calculation method described later.

Here, the means for measuring ΔX and ΔZ may be a dial gauge, a capacitance sensor, or the like.

Figure 4 (A) shows the total amount of deformation (a, ) when force is applied to the part, and Figure 4 (B) shows the final amount of deformation when the force applied to the part, IO, is removed. It represents the amount of plastic deformation (b) which is the amount of remaining deformation.

Above this -), the total deformation amount (a) and the plastic deformation amount (1))
and are not the same value due to the influence of springback, etc., so it is necessary to obtain the necessary amount of plastic deformation (b) from the total amount of deformation (a), which is a controllable value.

Figure 5 (A) shows the total deformation 1 in the X direction determined by experiment.
tct (corresponds to the total deformation amount (a) in Fig. 4 (A)) and the plastic deformation amount G in the X/-+ direction (corresponds to the plastic deformation i (b) in Fig. 4 (13)). FIG. In this way, it can be seen that even if the total amount of deformation Gt is made constant, the amount of plastic deformation G is not constant, but has some variation.

Figure 5 (B) shows the values obtained from experimental values, with the total deformation Gt in the X-axis direction and the plastic deformation iH on the right side in the Z@force direction as axes (to This figure shows that even if a certain amount of total deformation is obtained by applying force in the X-axis direction to a part, some plastic deformation will occur in the Z-axis direction. to be called.

In summary, in order to simultaneously bring the relative distances ΔX and ΔZ between the left and right Herad chips 8 and 9 within a certain tolerance range, which is the objective of the present invention, it is necessary to solve the following three problems. Become.

(1) The total amount of deformation (a-) (the amount of deformation when force is applied to the part) and the amount of plastic deformation (+) (the amount of deformation that ultimately remains in the part) are due to the influence of spring bags, etc. Since the values are not the same, the necessary amount of plastic deformation (b) should be obtained from the total amount of deformation (a), which is a value that can be controlled.

(2) Even if the total amount of deformation (a) is constant, there will be variations in the amount of plastic deformation (b) that ultimately remains in the part.

(3) The influence of the relative distance ΔZl\ in the Z-axis direction when adjusting the relative distance ΔX in the X-axis direction and the relative distance ΔX in the X41ilIh direction when adjusting the relative distance ΔZ in the Z-axis direction
impact on.

If the above problems (1) to (3) are solved, it is possible to perform relative positioning between two chips of a VTR, and a high-quality VTR head can be produced with a high yield.

Hereinafter, a control method for solving problems (1) to (3) will be explained in detail.

Fig. 6 shows the distribution map of Fig. 5 (A) projected from the direction of the white arrow and displays the "C data distribution as a triangle. For example, in Fig. 5 (Δ), the total deformation The distribution of experimental data for the amount of plastic deformation G when the amount Gt is 150 μn1 is shown in Histodurham.
The triangle with G=1.50 μm in FIG. 6 is displayed after being converted into a square with σ (σ is the standard deviation).

In this way, if the distribution of each total deformation amount in FIG. 5(A) is made into a dizogonal shape and summarized, FIG. 6 is obtained. This triangle was used as the membership function of fuzzy inference.

In this example, the membership function is set with the height of the vertex of the square as 1, but the method of setting the membership function is to set the height of the -7 diagonal to each probability ・13 ・12・There is a way to do that. For example, in Fig. 7(A), when the total deformation amount Gt is 150 μm, the plastic deformation amount G is 5
The probability of being μm means 0.6.

Similarly, when the total deformation amount Gt is 150 μm,
The probability that the amount of plastic deformation G will be 4.5 is 0.55. Also, regarding the base of the triangle, for example, j 1 ・-3(7
.

t, -3σ (here, σ is the standard deviation) may be defined as 2
It may be set as σ, and can be determined as appropriate.

Moreover, it is not necessarily necessary to set t,=j2. In short, if accuracy is required, the distribution should be close to that shown in FIG. 5, and if a reduction in processing time is strongly desired, the square distribution should be set to be as simple as possible. In this way, various methods of setting the membership function can be considered, but here, to simplify the explanation, we will use the method shown in Figure 6, that is, the probability is 1 at the vertex of the commi-gon, the probability is j at the base, === J″″
This will be explained using the membership function ′σ.

The rules when G=5 μm, which is an example of the membership function of the present invention, are as follows.

fF G=-5p, m then Gt, =1
50μm ・-Ruled 14・ Similarly, focusing on the apex of the triangle of Gt14071m and creating rule 2, IF G-3μm t, 11 (40Gt=1
．． 40μr mouth...Rule 2 In this way, we create a square of the membership function from the measured distribution and create rules for the second square, but the number of rules depends on the calculation time and measurement required. From the perspective of time, the fewer the better, and from the perspective of accuracy and reliability, the more the better.The actual number of rules should be determined depending on each individual case. The number was set to 10.

Next, we will discuss an example of an inference method for finding a solution using a membership function converted from an experimentally determined distribution. In the case of only Rule 1 and Rule 2 mentioned above, the total amount of deformation G when the amount of plastic deformation G is 4 μm
The method for calculating t is shown below.

In Fig. 7(B), when finding the intersections of the dotted line for the amount of plastic deformation (-1-4 μl, the triangle when the total amount of deformation Gt-150 μm, and the square shape when the total amount of deformation G t, = 140 μm is found, a is obtained, respectively. Point, b point.A point, 1-1 point 1
The respective membership values of 5. and .5 are 0.7.0.5.

Below -1- is when there are two rules for the membership function, but if you increase the number of rules to four in the same way, you will get the seventh rule.
As shown in Figure (B), the total deformation amount G t = 130
The intersection of the triangles of rules 3 and 4, which are the rules for μm and 1607zm, and the dotted line for the amount of plastic deformation G-4μm is
It increases from 0 point to d point. Here, rule 3. Using the same expression as Rule 1 and Rule 2 for Rule 4, IF G=2 μm t, hen Gt=]30
7zm -Rule 3IF G= 7 pm the
n Gt=160 μm - Rule 4 FIGS. 8(A) and 8(C) εJ, rule 1° and rule 2 are each represented by solid triangles.

Next, when we want to obtain the amount of plastic deformation G-・4 μ■, the membership value of the intersection with the triangle, which is each rule, is taken as the Kasuri axis, and the total amount of deformation G t is taken as the horizontal axis. Figures (B), (D), and (E).

FIG. 8(B) shows that when the plastic deformation amount G-41zm is desired, the total deformation amount G t ” is 140 μm and the misaligned member.
16. With a bulge value of 0.7, it is b. It was a hailstorm,
Similar to FIG. 8(B), FIG. 8(D) shows the amount of plastic deformation G
・When crossing -4,7zm, total deformation amount Gt = 15
This represents that if it is 0 μm, the membership value will be 0.5.

A rule is created for each total deformation amount of 10 μm for sea urchin, J-uni, and the membership function and plastic deformation amount (i
All intersections of H=4μI11, Figure 8(B).

Figure 8 (E) shows a straight line as shown in (D), and a diagonal line drawn in the polygon that connects each membership value.
It is.

Next, the value on the scale of the horizontal axis of the straight line that bisects this polygon, 143 μm shown by the broken line in Figure 8 (g; 4μn1
This means that the total deformation amount Gt is the most suitable for achieving this.

Below-1-1, as an inference method other than that explained, see Figure 7 (B
), several inference methods can be considered. For example, it may be possible to calculate the weighted average as shown below.

・17・ 130Xm, 1(1+140Xm, 4g' 150X
m, ,,6'160Xm, 6゜m+:+o-'m
+4o”1L5o)111+i.

Here, for example, m11° is the intersection point (membership value of 1) of the straight line with the amount of plastic deformation G=4 μm and the triangle when Gt=130 μm.

There is also a method of simply setting the value of the horizontal axis of the vertex of the membership function having the vertex closest to the plastic deformation amount G as the total deformation amount Gt that is closest to the highest point.

So far, the method of controlling ΔX by applying an adjustment force in the X-axis direction has been described. This method can solve the problems (1) and (2) of the present invention described in -, but (3)
) cannot solve the problem of crosstalk. Therefore, in order to take the cross distortion into consideration, the total deformation amount H, , , , H, of the left and right parts is calculated below.

Figures 9 (A) and (C) are basically the same as Figure 6, which shows the membership functions obtained from the distribution in Figure 5 (A). This is a modified version of .

Figures 9 (B) and 1) project the distribution in Figure 5 (g) from the open direction of the white arrow, 7, in the same way as in Figure 6, and show the location where the frequency is greatest. (・-The apex is located, the height is 1, the base is 2σ (σ is the mark (deviation), and there is C of the card represented by a square square. As shown in Figure 9 (Δ) 1, Plastic deformation amount G-3 μm and total deformation amount G t −
140p, 21 which is the intersection with the membership function of m
The membership value of the point is 0.7. Further, when the total deformation amount G t-140 μIll in the X-axis direction, the total deformation amount HII + of the right part is as shown by the solid line cone in FIG. 9(B). That is, the total deformation amount G in the X-axis direction is 1
When the diameter is 40μm, the total deformation of the right part it Hl, 1 is 7μ
The distribution is expressed as a square with m as the vertex.

Using the same number, calculate the total deformation ratio of the left part. When the total deformation in the X-axis direction () is 1407 zm, the total deformation H1 of the left part is calculated as a spherical shape with the vertex at 6 μm. The distribution is shown in Table 1 (see Figure 1 (B)).
19. What this rule 1' means (J, if you want the amount of plastic deformation in the X direction to be 3 μm, , the most appropriate total i′′i deformation Gt in the X Ii direction at that time is 1·10μ
This means that m, (total deformation amount of right part i) (,,,) is 7 μm.Also, if we create a rule for the plastic deformation amount G-5 μm using the same number as rule 1', we get becomes.

The membership functions expressed by rules 1' and 2' have now been found.

Although it is possible to improve the reliability by increasing the number of rules in the same way, it is better to have fewer rules from the viewpoint of calculation time, time required for measurement, etc.

In this embodiment, it is thought that the optimal number of rules is around 10, but to make the explanation easier to understand, the number of rules of the membership function is first assumed to be two, rule 1 and rule 2.

For example, if you want the amount of plastic deformation to be G-4 μm, rule 1'
Based on this, the intersection with the membership function of G t = 140 μm is a·20 point as shown in FIG. 9 (Δ), and its membership value is m−·0.7. In this case, the membership value is 0.7, which means G・-4μm
If Gt = 140 μm, there is a probability of 0.7, which means that G = 4 μm is tυ. next,
2, the total deformation of the right part ff1H, , is, but from FIG. 11(B), the total deformation amount in the X-axis direction G 1: =1
When the thickness is 40 μm, the total deformation amount of the right part is H, , -=
The distribution with the apex at 7 μm (, -. This distribution is the 9th
Since the membership value obtained from i and m in Figure (A) is 0.7, the distribution of 1413, -71t m in Figure 9 (E', ) seems to have a maximum probability of 0.7, so the vertex is .7
Correct it to the rectangular shape (hatched area).

Next, when you want the amount of plastic deformation to be G-411m, rule 2
Based on ', H, , , is determined. From Fig. 9(C), when G-4μIT, the total deformation amount G t =150μ■
The membership value of 1 is 0.5. Also, according to Fig. 9 (Ii)), the total deformation amount Gt = -・150 μm
When , the right part is a membership function J that transforms by 8〃■1. This membership function can also be expressed as any 11- and total deformation in the X-axis direction Pi G tl
) 1 = 15071m has a membership value of 0.5, so the membership value at the vertex of the triangle is 0.5. J: Sea urchin (
Shaded area) Correct the triangle.

Below, we created the membership function of H1l+ based on the two rules 1-Rule 1' and Rule 2'. Create membership functions for the total deformation amounts H, , L on the right side when G = 2 μm and G - 7 μm, and overlap them to form the shaded area shown in Fig. 9(E).This superimposed diagonal line To divide the part into two equal parts is H + t,, =
-7.5 μm. In other words, the amount of plastic deformation G=
To obtain 4 μm, we obtain a defeasible solution that the right-hand deformation is 7.5 μm.

From the membership function of the total deformation amount H, , on the right side,
The inference method for finding the fuzzy solution is to stack the triangles that are the membership functions of the comprehensive transformation ff1H, . This was synthesized by synthesizing. Another possible method is to divide the area of the triangle into three equal parts.

・22・ Next, in the same manner as for the right portion, calculate the total deformation amount H,,, for the left portion when the amount of plastic deformation is increased by 4 μm.

Figure 1.0 (A) shows the plastic deformation ftG as shown in (78).
= 1 line of 411 m and total deformation amount C,'r t, =
The membership value of the intersection point a with the membership function of 1.40 p m is 0.73, and the total deformation amount of the left part 1 when the total deformation amount in the X-axis Ji direction G f, -140 μm −T , , , becomes a membership function whose apex is at 87zm and is represented by a triangle whose base is 2σ (σ is the standard deviation). This membership function has a membership value of 0 at its starting point, G t −140μn1.
．． 7, the maximum probability is 0.7, so I-
1, , is a membership function τ whose vertex is 6 μm, and the membership value is 0.7 at the vertex of a certain − diagonal.
How should I complement iE I-? This corrected membership function is shown in Figure 10 (B) as the shaded area.
It was expressed in Below, when the total deformation amount Gt = 15011 m, the corrected membership function is obtained in the same way as shown in Figure 10 (I). It is inferred to divide the shaded area in Fig. 10 (E) into two equal parts, and the inferred value H, . . . = 6.2 μm for .

4B (j) As mentioned in the inference method on the right, there is a method of stacking the shaded parts in Figure 10 (B) and (D), and there is also a method of finding weights (=j times the average). , any suitable method can be considered.

In the following procedure, the total deformation amount Gt, H,,,H,,, to make the plastic deformation amount G・-4μm is Gt, respectively.
=143μm, H++l=7.5μm, T
1. It was inferred that h=6.27zm.

Up to this point, we have described a method for determining the total amount of deformation G t , Hu-, H1, when the amount of plastic deformation G is given. Although this is a basic method, in practice it becomes complicated as these methods are combined.

Hereinafter, the positioning of the relative distance of the VTR head will be explained using the conventional inference method with reference to FIG.

First, the reference position II of the VTR head, which is the target part,
+2 It is difficult to judge whether the error between the measured Z' values of the phase gradient distances, 24, and distances ΔX and ΔZ in the X-axis direction Z4IIllJJ direction and the standard for non-defective products and the J' method is 1.5 pm or less, respectively. Step 10
1) If it is 15 μm or more, it is determined to be good. (Step 102). For the error X' or 15 μm or more, the amount to be corrected, that is, the amount of plastic deformation G, IT, , H, is set based on the difference from the standard J' method for non-defective products (Step 103 ) This, -, - times [As Act 9 of 1, set Karan l- = 1 (step 1.04),
Set the target value as G'-G (Sude knob 105)
. Next i, -: '-7'; From the correlation between the determined plastic deformation amount and the total deformation amount, the total deformation amount Gt is determined when the plastic deformation is desired to be G (step 106). However, as mentioned before, when the total deformation amount Gt is applied to the part, it affects the X-axis direction of the right and left parts, so
Find the influence H,′,H,′ (step 1.07
), the amount of plastic deformation actually desired for both the right and left parts is H
R, HL, but due to the comprehensive deformation tic G 1-, already i·, -H,,'.

■11′ Plastic deformation occurs, so H,,If,,′
.

H. By finding the correlation between
, , H, , is obtained (step tog >. At this time, find the influence G ″ in the X-axis direction (step)
09). Here, it is desirable that G'+G'' becomes (], so Δ-=G (G' + G'') and the error Δ are calculated (, (step 110), the error Δ is within the tolerance range 2ε. (Step 111).

If it is, it is OK, and the total deformation amount Gt, H, .

Set H1 and L and finish (step), 1.2)
,,If it is not included, add the count λ to perform the second inference of I (step 113).In this case, if the count exceeds 5 times, it is determined that it is defective because it cannot converge. (Step 114). If the count is within 5 times, the error △ is multiplied by a coefficient 7 to calculate the correction value E so that the error converges. Here, in addition to the above, E is simply E=f(△ ), or as a function of Δ, such that E=f(Δ.

G, G', G''), but the point is that the settings should be adjusted so that the difference from the standard dimensions of the good product converges through this loop (Step 115).■ Target value G' of the first inference
Adding the cut-out correction value E to [1] of the second L1 inference
Target price ('J') (Sdep 11G).

A similar inference is made by setting the second target value G' in step 105 of the tie J page. Applicable to V T
Needless to say, it is not limited to R heads, but can be applied to anything that satisfies the problems (1) to (3) related to -.

As explained above, according to the present embodiment, each crosstalk can be taken into consideration, so even if the number of parameters to be adjusted increases, it can still be applied.

〔Effect of the invention〕

If the amount of plastic deformation that is ultimately desired to be adjusted cannot be directly controlled, i.e., the amount of plastic deformation and the amount of total deformation do not match, 1) the amount of plastic deformation that is required is Since the total amount of deformation can be estimated by fuzzy inference, highly accurate control is possible.

・27 In addition, even if the total amount of deformation is constant, there will be variations in the plastic deformation amount number depending on the shape, thickness, and type of the object. Since it is converted into a function, this variation can be taken into account and highly reliable adjustment can be made.

Furthermore, even if two or more adjustment parameters influence each other, the rules in fuzzy inference can be determined by experimental values or alternative distributions, so the mutual influence can be eliminated. This has the effect that Ht can be determined and highly accurate control is possible.

Moreover, this allows the number of adjustments to be minimized, which has the effect of reducing material fatigue of parts and improving product VT reliability.

[Brief explanation of the drawing]

FIG. 1 is a diagram showing an outline of the present invention, FIG. 2 is a diagram showing a VTR head to which the present invention is applied, and FIG.
The figure shows the adjustment force for adjusting the applicable VTR head (Figure 7), Figure 4 is an explanatory diagram of the total deformation, 28. shape and amount of plastic deformation, and Figure 5 (A) is the Figure 5 (B) is a frequency distribution diagram showing the variation in the amount of plastic deformation G in the X-axis direction depending on the total amount of deformation (l, C'j) in the axial direction.
Fig. 6 is a frequency distribution diagram representing the variation in the amount of plastic deformation in the X-axis direction for each total deformation amount G in the X-axis direction. Figure 7 (Δ) represents the frequency distribution diagram of Figure 5 (A) as a membership function using the conversion method outside of part C). , FIG. 8 (A) and (C) are diagrams representing the frequency distribution diagram of FIG. , = 150 μm is shown by a solid line (Fig.-78, Fig. 8 (B), (I)), the horizontal axis represents the total deformation amount, the vertical axis represents the membership function, and the members at each total deformation amount are shown. The diagram showing the ship value, Figure 8 (E), is
Figure 9 is a diagram that summarizes the membership values for each total deformation amount, and Figure 9 is an explanatory diagram of the membership function on the right.
FIG. 0 is an explanatory diagram of the membership function on the left, and FIG. 11 is a flowchart of control in this embodiment. ”( ’01 ear~110

Claims (1)

[Claims] 1. In a dimension adjustment method for adjusting the dimensions of an object by plastically deforming the object, the distribution of the amount of plastic deformation for each total amount of deformation is determined in advance, the distribution is stored, and the Convert the distribution into a function, measure the dimensions of the object, compare the measured results with the standard dimensions, and set the amount of plastic deformation so that the object matches the standard dimensions. . A corrected dimension calculation method, characterized in that the most probable total deformation amount for obtaining the set plastic deformation amount is calculated based on the converted function. 2. In a dimension adjustment method in which the dimensions of an object are adjusted by plastically deforming the object, the distribution of the amount of plastic deformation for each total amount of deformation is determined in advance, the distribution is memorized, and the memorized distribution is converted into a function. Then, measure the dimensions of the object, compare the measured results with standard dimensions, set the amount of plastic deformation so that the object matches the standard dimensions, and set the amount of plastic deformation so that the object matches the standard dimensions. A method for adjusting dimensions, comprising: calculating the most probable total amount of deformation for obtaining the set amount of plastic deformation, and adjusting the dimensions of the object using the calculated total amount of deformation. 3. In a dimension adjustment method that adjusts the dimensions of an object by plastically deforming it, there are two parameters to be adjusted.
If there are more than one parameter, and each of them influences each other, the distribution of the amount of plastic deformation for each total amount of deformation and the distribution of the amount of plastic deformation of other parameters for each total amount of deformation of one parameter are determined in advance, and the distribution of the plurality of , convert the stored plurality of distributions into a plurality of functions, measure the dimensions of each parameter of the object, compare the measured results with each reference dimension, and The amount of plastic deformation is set so that each parameter matches each standard dimension, and when the amount of plastic deformation of one parameter is given, the total amount of deformation of one parameter and the amount of other parameters are calculated based on the converted multiple functions. A dimension adjustment method comprising calculating a total amount of deformation and adjusting the dimensions of an object using the calculated total amount of deformation. 4. In a dimension adjustment method in which the dimensions of an object are adjusted by plastically deforming the object, the distribution of the amount of plastic deformation for each total amount of deformation is determined in advance, the distribution is memorized, and the memorized distribution is used as the membership. converting it into a function, measuring the dimensions of the object, comparing the measured results with the standard dimensions of the object, and setting the amount of plastic deformation so that the object matches the standard dimensions, Create a polygon by stacking the transformed membership functions that intersect with the set plastic deformation amount, calculate the total deformation amount of the value that divides the area of the polygon into two, and calculate the calculated total deformation amount. A dimension adjustment method characterized by adjusting the dimensions of an object by quantity. 5. In a dimension adjustment method in which the dimensions of an object are adjusted by plastically deforming the object, the distribution of the amount of plastic deformation of the object at each total amount of deformation is determined in advance, the distribution is memorized, and the memorized distribution is used as a function. , measure the dimensions of the object, compare the measured results with the reference dimensions,
Set the amount of plastic deformation so that the object matches the reference dimensions, calculate the most likely total amount of deformation to obtain the set amount of plastic deformation based on the converted function, and calculate the most likely total amount of deformation to obtain the set amount of plastic deformation. Adjust the dimensions of the object based on the total amount of deformation, measure the dimensions of this series of objects, compare the measured results with the standard dimensions, measure the amount of plastic deformation, calculate the amount of total deformation,
A dimensional adjustment method characterized by repeating adjustment using the calculated total deformation amount at least twice. 6. In a method of manufacturing a VTR head by plastically deforming a head base to adjust the dimensions of the VTR head having two head chips so that the relative distance between the two head chips becomes a reference dimension. , obtain the distribution of the plastic deformation amount of the VTR head for each total deformation amount in advance, memorize the distribution, convert the stored distribution into a function, measure the dimensions of the VTR head, and calculate the Compare the results with reference dimensions, set the amount of plastic deformation so that the object matches the reference dimensions, and determine the most likely amount of plastic deformation to obtain the set amount of plastic deformation based on the converted function. A VTR head manufacturing method comprising: calculating a total amount of deformation, and manufacturing a VTR head by adjusting the dimensions of the VTR head using the calculated total amount of deformation.