JPH0237523B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Technical Field of the Invention] This invention relates to improvements in non-contact radiation thickness gauges that measure the thickness of various materials by detecting the amount or intensity of radiation transmitted through them. An example of such a thickness gauge is described in US Pat. No. 3,955,086.

Such thickness gauges are applied to various surfaces.
With this thickness meter, it is possible to continuously measure the thickness of sheet materials made of various materials without contacting the material. Due to their fast response, radiographic thickness gauges, such as X-ray thickness gauges, are useful for on-line measurements in rolling lines for parts that require rapid automatic adjustment to obtain continuous and uniform sheet thickness. Can be used. In order to obtain a desired thickness of sheet material in the rolling line, a deviation of the actual thickness value of the produced sheet material from this desired thickness value is required. The present invention relates to a thickness gage particularly suitable for setting a desired thickness value (i.e. a nominal thickness value) of a sheet material and for measuring the thickness of an object to be measured in the form of a deviation from the nominal thickness value.

[Technical background of the invention and its problems]

Conventional non-contact thickness gauges typically use a movable standard piece of a predetermined thickness value to calibrate the thickness gauge.
One or more standards of known precise thickness are selected and inserted into the radiation, and an analog meter receiving the detector output is zeroed for standards of a particular set thickness. During measurement, the deviation value from the nominal thickness value is expressed as the meter deviation. Other prior art examples use complementary standards, where the total measurement range is divided into multiple subranges. Calibration of each sub-range is typically performed on one standard with its maximum thickness value. This one standard piece is called a "reference standard piece." When the reference standard is in the radiation path and the thickness gauge is calibrated to zero, the reference standard can be removed to measure the thickness of a material with a smaller thickness within the thickness range of its sub-range. Some of the standards adding supplementary thickness to the desired strip thickness are then inserted into the radiation so that the total thickness is equal to the thickness of the reference standard. If the thickness is a predetermined value, the analog meter will read zero.

Two-point calibration systems have also been used. In this system, two sets of standards are sequentially inserted into the radiation path during a calibration operation. The first standard piece is selected to have an apparent thickness determined from the nominal thickness of the material and the alloy correction factor. The second standard piece is
The value is selected as the desired apparent thickness plus the planned deviation value from the desired value. In this system, calibration is performed based on two apparent thickness points and relying on the assumption that there is a linear relationship between the output of the detector and the thickness of the material being measured. Such a system is
Therefore, it can be considered as a linear interpolation system.

In conventional systems, calibration is typically performed directly on the object to be measured, ie, on the apparent thickness. Alloy correction is done during the calibration process,
It is not done during the measurement process. In traditional systems,
The nominal thickness of the object to be measured in the radiation path has been changed,
It is necessary to recalibrate each time the settings are made.

Furthermore, in the conventional system, a large number of standard pieces had to be combined in order to construct a standard piece having the apparent thickness of the object to be measured. The need to combine multiple standard pieces also resulted from the reliance of conventional systems on linear approximations. For example, a series of binary coded decimal (BCD) standard strips ranging in thickness from very thin 0.001 mm to 8 mm were used.

[Purpose of the invention]

SUMMARY OF THE INVENTION An object of the present invention is to provide an improved non-contact radiation thickness gauge that uses a highly accurate interpolation method.

Another object of the present invention is to provide a thickness gauge in which alloy correction is performed at the measurement stage rather than at the calibration stage of the thickness gauge.

Still another object of the present invention is to provide a thickness meter that allows the nominal thickness of a measured object to be changed even during measurement of the measured object.

Another object of the present invention is to provide a thickness gauge in which the number of required standard pieces is reduced by selecting a fixed number of calibration points arranged in a geometric series to determine the calibration curve. .

[Summary of the invention]

A preferred example of the invention is a non-contact radiation thickness gauge using a plurality of movable standard pieces of predetermined thickness, having a radiation source, receiving radiation from the radiation source, and measuring the radiation thickness in relation to the level of radiation received. detection means for producing an output signal, and memory means for storing parameters specifying a calibration curve for each of the thickness ranges, where the calibration curves have a substantially geometric series over the entire thickness range. There are calibration points selected according to the method, at least three of which calibration points and respective output signals of the detection means for these calibration points specify each said calibration curve, and the nominal thickness value of the material to be measured and It has a setting unit for setting alloy correction values, and calculates the output signal of the detection means and the signal of the setting unit in relation to the calibration curve stored in the memory when the object to be measured is in the radiation path. The present invention resides in a thickness meter having arithmetic means for determining the thickness of a substance to be measured.

[Embodiments of the invention]

Hereinafter, a preferred embodiment of the present invention will be described with reference to the drawings.

In FIG. 1, a transmissive radiation source, here an X-ray source 2, generates X-rays that pass through a material 6 along a path 4. In FIG. Detection means, including a detector 8, an amplifier 10 and an analog-to-digital (A/D) converter 12, receive radiation from the X-ray source 2 and produce an output signal related to the level of radiation received.
The level of this received radiation is a function of the radiation absorption properties and thickness of the material 6 placed in the radiation path between the X-ray source 2 and the detector 8. Radiation detector 8 is arranged to receive X-rays emitted from X-ray source 2 through substance 6 , converts the received X-rays into an electrical signal, and applies this to amplifier 10 . The amplified signal from amplifier 10 is converted to digital format (preferably 16 bit format) by A/D converter 12 and provided to processor 14.

Processor 14 receives output signals from setting unit 16, that is, a thickness setting signal corresponding to the nominal thickness Tn of the object to be measured and an alloy correction coefficient N. Memory 18 includes memory for storing data that is supplied to processor 14 to obtain a calibration curve by the operations described below.

The thickness indicator 20 receives the output from the processor 14 via a digital-to-analog (D/A) converter 22, and determines a desired thickness value, that is, a nominal thickness value Tn.
Displays the thickness deviation from

Processor 14 calculates and stores calibration curve data in memory 18, calculates the data and other signals, provides thickness deviation as output to thickness indicator 20, and further proceeds with calibration and measurement operations. Generates various signals. For example, processor 14
is determined by the signal supplied through the signal line 24.
The voltage setting of the line source 2 is controlled, and the gain setting of the amplifier 10 is performed via the signal line 26.

The thickness gauge of the present invention has a standard magazine 21 containing a plurality of movable standard pieces of predetermined thickness values.
has. The standard pieces are made of a given material and have precise and different thicknesses. The selected standard strip combination is inserted into and removed from the radiation path 4 under control of the processor 14 via the signal line 28.

(Overview of operation) The operation of a non-contact radiation thickness meter is basically as follows: (1)
It consists of pre-calibration, (2) range calibration, and (3) measurement. Furthermore, this thickness meter allows the thickness setting to be changed even when the substance to be measured is present in the radiation path.

According to the invention, the first operation, pre-calibration, begins with dividing the total thickness range of the thickness gauge into multiple thickness ranges.

The pre-calibration operation is to sequentially create a calibration curve for each divided range. This behavior is approximately 70
It takes seconds and must be done once every eight hours or so. The above-mentioned curve is hereinafter referred to as the pre-calibration curve. Generally, the pre-calibration curve or calibration curve is based on the thickness of the substance to be measured and the output of the detection means (here, the A/D
(output of the converter 12).

The second operation, range calibration, includes a setup and range selection step followed by a drift compensation calibration for the range selected for use in the measurement operation. The setting and range selection step consists of (1) setting the nominal thickness Tn and the alloy correction coefficient N of the material to be measured, and (2) automatically selecting the appropriate thickness range based on the set values Tn and N for thickness measurement. including steps. A subsequent drift compensation calibration step for this selected range is required to correct for any drift that occurred after the pre-calibration. A calibration curve with drift compensated for the range selected in this way is created and stored in the memory 18. A drift compensation calibration is preferably performed whenever a new range selection is made and takes about 5 seconds.

The third operation, measurement, requires that the material to be measured be placed in the radiation path, the thickness of which is measured using the output signal of the detection means and a drift-compensated calibration curve.

In the thickness gauge according to the present invention, the setting of the setting unit 16 can be changed even when the substance to be measured is present in the radiation path. Calculating the drift compensation calibration curve requires inserting a standard piece into the radiation path. Therefore, if the substance to be measured is placed in the radiation path, this calculation cannot be performed, so the drift compensation calibration curve obtained in the previous drift compensation calibration operation for the same range as the selected one is selected. The drift compensation calibration curve selected in this way is read out and used to measure the thickness of the material.

(Thickness range and standard piece) Figure 2 shows the appropriate thickness ranges for aluminum plate (A in the diagram), cold rolled steel plate (B in the diagram), and hot rolled steel plate (C in the diagram), respectively. . Thickness values are taken on a logarithmic scale on the horizontal axis, and the calibration points for each of the above materials are shown as points 30, 32 and 34, respectively.

The ranges overlap each other. For example, as shown in Figure 2B, in the measurement of cold-rolled steel sheets, 9
Consists of two ranges. The second range starts from 0.04mm
Covers thickness value of 0.16mm, third range is 0.08
Covers mm to 0.32mm. The range from 0.08mm to 0.16mm overlaps, and the same goes for other ranges.

The entire range is divided into multiple ranges for each material. For example, in the cold-rolled steel plate shown in FIG. 2B, nine ranges from range-1 to range-9 are shown. Other necessary parameters, namely the tube voltage of the X-ray source 2 and the gain of the amplifier 10, are also shown. These are determined and controlled by processor 14, as shown in FIG. 2A, B, and C, according to the substance to be measured and the selected range.

For thickness gauges for hot-rolled steel sheets (see Figure 2C),
For example, when measurements are made in the Range-1 to Range-4 range, the X-ray source voltage is set to 100 KV and the amplifier 10 is set to a low L gain level. For measurements in range-5, the voltage of the line source is 100KV, but the amplifier 10 is
The gain is set to In range-6, the line source voltage is 120KV and the gain of amplifier 10 is middle M, and in range-7, the line source voltage is 140KV.
and the gain is high level H. These parameters are shown for each range, both for aluminum (see FIG. 2A) and for hot-rolled steel sheet (see FIG. 2C).

The standard magazine 21 has a plurality of standard pieces of known materials and known exact thicknesses, which can be used alone or in combination to set the values 30, 32, and 32 at each calibration point.
It is configured to take 34. For aluminum thickness gauges, the standard piece is made of pure aluminum, and for hot-rolled and cold-rolled steel plates, pure iron is used as the standard piece. Each thickness value of the standard piece is preferably chosen as a binary sequence, for example 0.01 mm; 0.02 mm; 0.04
mm；0.08mm；0.16mm；0.32mm；0.64mm；1.28mm；
It consists of standard pieces with thickness values of 2.56 mm; 5.12 mm; 10.24 mm; and 20.48 mm. Here, a standard piece of 20.48 mm is necessary for hot-rolled steel sheets, but not for measuring aluminum and cold-rolled steel sheets.

The standards selected during the calibration operation consist of one or more standards from the standards magazine 21. As the number of required standards increases, the errors introduced due to scattered x-rays due to layered standards also increase. Therefore, each calibration point 3
Thicknesses of 0, 32 and 34 can be constructed from at most three standard pieces. During the calibration operation, one standard piece or a combination of two or three standard pieces are sequentially inserted from the magazine 21 into the radiation path. The operation is controlled by instructions from the processor 14.

(Memory Operation) Data processing using the memory 18 will be described with reference to FIGS. 3 to 5. memory 18
includes read-only memory (ROM) and random access memory (RAM), as shown in FIG. The ROM contains the pre-calibration program (see the flow chart in Figure 10), the range calibration program (see the flow chart in Figure 11), the measurement program (see the flow chart in Figure 12), and the alloy correction program. It includes a common database 42 consisting of data common to all ranges and necessary for calibration and measurements, such as equations, basic quadratic equations, and range selection zone data. The alloy correction formula, the basic quadratic formula, and the range selection zone will be described later. The ROM also includes a range data base 4 with data specific to each range.
4-1, 44-2, 44-3,..., 44-n. The RAM includes first and second tables 46 and 50, and the first table 46 has data range areas 48-1 and 48- corresponding to each range.
2, 48-3,..., 48-n. A second table 50 is used as a work area. In each of tables 46 and 50, each data area includes type 1, type 2, and type 3 data. Type 1 data is ROM
range database 44-1, 44-2,
It corresponds to each range data stored in 44-3, . . . , 44-n. Type 2 data corresponds to the results of pre-calibration of the thickness gauge, and type 3 data corresponds to the results of range calibration. The second table 50 of the RAM is a work area with a capacity capable of storing a set of type 1, type 2, and type 3 data of one specific range.

The transfer of data within memory 18 and to and from processor 14 that advances the operation of the thickness gauge will now be described. At the beginning of the pre-calibration operation, the ROM
Range database 44-1, 44-
Range data from 2, 44-3, ..., 44-n is transferred to the type 1 data area of each corresponding range in the first table 46 of the RAM (see the dotted line arrows in FIG. 3). ).

Fourth to explain the pre-calibration operation of Range-1
In the figure, type 1 data stored in the type 1 data area of range table 48-1 is transferred to the work area of second table 50. This is designated as step-1.
The pre-calibration result data is written to the type 2 data area of work area 50, as shown as step-2. The resulting data is transferred to both the type 2 and type 3 database areas of the first table 48-1 in RAM (step-2) and stored as shown in step-3.

Next, range calibration for range-1 will be explained with reference to FIG. The data in range-148-1 of the first table is transferred to the work area of the second table 50 as in step-1. This is indicated by the arrow at step-1. As the operation progresses, the range calibration result data is written to the type 3 data area of the work area 50 and stored in the range table 48, as indicated by the dotted arrow in step-2.
-1 type-3 data area. As shown in step 3, the data is placed in range table 48-1. Range-1
The measurement of the material thickness in the step shown in FIG.
Type 3 of the first table of RAM shown in 3-3
The process proceeds using the drift compensation curve stored in .

After the measurement operation, when range-1 is selected again by the setting and range selection step in the range calibration operation, if no pre-calibration has been performed in the meantime, range-1 in the first table of RAM is selected. The contents of 1 are the same as the RAM contents obtained in the range calibration for range-1 that was performed immediately before. In other words, the type 3 data in the range table 48-1 in the RAM is updated every time the range is required and range calibration is performed.

(Detailed Operation of Thickness Gauge) The operation of the non-contact radiation thickness gauge, including pre-calibration, range calibration and measurement, will be explained in detail with further reference to FIGS. 6 to 9.

First, pre-calibration will be explained with reference to FIG. The calibration points are chosen to be arranged as closely as possible in a geometric series over the entire thickness range of the thickness gauge. At the same time, the calibration point is preferably chosen such that its thickness can consist of one standard piece or a combination of two or at most three standard pieces from the magazine 21.

FIG. 6 shows the functional relationship between the thickness value and the logarithmically transformed value of the corresponding output signal of the A/D converter 12. The calibration points Tc1, Tc2, Tc3, ..., Tcn are plotted on a logarithmic scale on the horizontal axis. Digital output signal V
is converted into a logarithmic signal by the calculation of the processor 14. These logarithmic signals are taken on the vertical axis.

In this example, each thickness range has seven calibration points, including points at both ends of each range. For example, range-1 is Tc1 to Tc7, range-1 is Tc1 to Tc7, and range-1 is Tc1 to Tc7.
2 has points from Tc4 to Tc10. As is clear from FIG. 6, the ranges partially overlap each other. However, the range selection zones within each range used for range selection operations do not overlap with adjacent range selection zones. For example, range-2 has a range selection zone extending from point TR1 to TR2. The next range selection zone is for range-3 and extends from point TR2 to TR3. In this way, in principle there is no overlap in the range selection zone.

During pre-calibration, for example 1 with total thickness Tc1
One or more standards are inserted into the radiation path 4 and an output signal V1 is generated. This signal, preferably in 16-bit digital format, is provided from A/D converter 12 to processor 14, where an operation is performed to convert this signal into a log ₂ V signal.

During pre-calibration, the magazine 21 is driven under the control of the processor 14 to sequentially transmit Tc1 and Tc into the radiation path.
A standard piece with a thickness of 2, Tc3,..., Tcn is inserted, and the processor 14 sequentially generates a logarithmic signal.
log ₂ V1, log ₂ V2, log ₃ V3, . . . , log ₂ Vn are calculated, and these signals are stored in the memory 18.
The memory 18 stores data (Tc1, log ₂ V1), (Tc
2, log ₂ V2), (Tc3, log ₂ V3), ..., (Tcn,
log ₂ Vn). Thickness point Tc1, Tc2, Tc
3,...Tcn are each stored in the range data base of the ROM and are logarithmic signals log ₂ V
1, log ₂ V2, log ₂ V3, . . . , log ₂ Vn are stored in the type 2 data area of the first table (FIG. 4). Curve 52 in FIG. 6 was obtained by connecting the data points and represents the calibration curve of these points.

A calibration curve generally refers to a curve representing the functional relationship between the thickness of the substance to be measured and the output of the detection means (here, the output of the A/D converter 12).
(Refer to the explanation about pre-calibration in the operation overview section). However, to be more precise, the calibration curve shown in FIG. It approximately represents the functional relationship between

If the number of calibration points is increased to narrow the pitch between adjacent calibration points, the approximation accuracy of the curve obtained by connecting the data points will be improved. However, an increase in the number of calibration points causes various disadvantages such as an increase in the time required for calibration.

In this example, in order to reduce the number of calibration points,
A quadratic curve is used. More specifically, the calibration curve is created by connecting quadratic curves connecting three adjacent calibration point data. By doing this, compared to the method of creating an approximate curve by connecting two adjacent calibration point data with a straight line,
The same level of approximation accuracy can be obtained with approximately one-fifth the number of calibration points. Therefore, according to this embodiment, since the pitch between the calibration points can be made relatively wide, calibration can be performed without using an extremely thin standard plate, and the calibration points can be adjusted by combining up to three standard pieces. Calibration points can be selected to obtain the thickness.

Approximation using a quadratic curve will be described in detail below.

In the calibration operation, a calibration curve is created for each range. Therefore, as many calibration curves as there are ranges are created. Each range is divided into three subranges, and each subrange is specified by three calibration points including calibration points at both ends of the subrange. A quadratic curve, here a parabola, passing through these three calibration points becomes the pre-calibration curve for the subrange. The precalibration curve for each range consists of three quadratic curves connected together. (The various range-1 curves are shown in FIG. 6 with corresponding equation numbers, which are explained below.) The non-contact thickness gage according to the invention uses a logarithmic signal log _2V
and the thickness T _s of the material to be measured which has the same radiation absorption properties as the standard piece. This relationship can be expressed by the following quadratic equation (basic quadratic equation) with desired accuracy for a specific thickness range.

T _s = a ₁ + a ₂ (log ₂ V) + a ₃ (log ₂ V) ² Equation 1 For example, calibration points Tc1, Tc2 in range -1,
The curve for the subrange consisting of Tc3 can be derived as follows.

Tc1 = a ₁₁₁ + a ₁₁₂ (log ₂ V1) + a ₁₁₃ (log ₂ V1) ² equation 2 Tc2 = a ₁₁₁ + a ₁₁₂ (log ₂ V2) + a ₁₁₃ (log ₂ V2) ² equation 3 Tc3 = a ₁₁₁ + a ₁₁₂ (log ₂ V3) + a ₁₁₃ (log ₂ V3) ² Equation 4 Data (Tc1, log ₂ V1), (Tc2, log ₂ V2),
(Tc3, log ₂ V3) is the memory 18 as mentioned above.
Since these are known numbers, the processor 14 solves the above simultaneous equations for the three unknown numbers a ₁₁₁ , a ₁₁₂ , and a ₁₁₃ to determine the curve for the first subrange. Parameter a ₁₁₁ ,
a ₁₁₂ and a ₁₁₃ specify a pre-calibration curve consisting of calibration points Tc1, Tc2 and Tc3.

It is assumed here that the solution to the above simultaneous equations 2, 3, and 4 is expressed as follows.

(Tc1, log ₂ V1), (Tc2, log ₂ V2), (Tc3, log ₂ V3)
→a ₁₁₁ , a ₁₁₂ , a ₁₁₃ Equation 5 Then, parameters specifying the curves of other subranges are given as follows.

(Tc3, log ₂ V3), (Tc4, log ₂ V4), (Tc5, log ₂ V5)
→a ₁₂₁ , a ₁₂₂ , a ₁₂₃ Equation 6 (Tc5, log ₂ V5), (Tc6, log ₂ V6), (Tc7, log ₂ V7)
→a ₁₃₁ , a ₁₃₂ , a ₁₃₃ Equation 7 As a result of these calculations, the pre-calibration curve for range-1 has three sets of curve specific parameters a ₁₁₁ , a ₁₁₂ , a ₁₁₃ ,
Identified by a ₁₂₁ , a ₁₂₂ , a ₁₂₃ and a ₁₃₁ , a ₁₃₂ , a ₁₃₃ . These parameters are for type 2 work area 50, as shown in FIG.
is stored in the data area of

Curve specific parameters for the remaining ranges can be obtained in a similar manner. (The boxes attached to the curves in FIG. 6 indicate other ranges and subranges.) It is desirable that the non-contact radiation thickness gauge operate stably. However, non-contact radiation thickness gauges are very sensitive devices and suffer from drift over time. If this drift can be ignored, the thickness of the material to be measured can be determined from the above-mentioned pre-calibration curve. First, the apparent thickness (corrected thickness value) is calculated from the nominal thickness value of the material to be measured and the alloy correction coefficient set in the setting unit 16. A range that has a range selection zone to which this apparent thickness falls (i.e., the apparent thickness falls between the ends of the range's lens selection zone).
A pre-calibration curve of is selected. A material to be measured is placed in the path of the radiation beam 4 in order to obtain a logarithmic signal, and the thickness is determined using a selected pre-calibration curve.

However, since drift in non-contact thickness gauges has been experienced in practical applications, a range calibration operation is performed immediately before measurement to compensate for such drift. By performing the range calibration operation immediately before measurement, the accuracy of the thickness gauge can be maintained on the order of 0.1% over the entire range.

The range calibration operation will be explained with reference to FIGS. 7 to 9. The range calibration operation consists of setting the nominal thickness value and alloy correction value for the material to be measured, selecting the appropriate range, and drift compensation for the selected range.

The nominal thickness of the substance to be measured and the alloy correction coefficient are set in the setting unit 16. Nominal thickness value Tn and alloy correction coefficient N from this setting unit 16
The apparent thickness value Tap is calculated by the processor 14 from the range selection zone to which this Tap belongs, and a range including that zone is selected. As is well known, Tap is calculated using the following formula. Note that N is expressed as a percentage value.

Tap=(1+N/100)・Tn The range calibration operation ends when the selected range is compensated for drift. This will be explained with reference to FIGS. 7 to 9. For range calibration,
What is needed is a relationship between logarithmic signals and thickness values that represents an entire range in one quadratic curve. Therefore, of the seven calibration points used to specify the range in the pre-calibration operation, three calibration points are used: the points at both ends and the center point. For example, range calibration for range-1 is performed depending on the points Tc1 and Tc7 at both ends and the central point Tc4.

Without inserting anything in the radiation path, the data (Tc1, log ₂ V
1), (Tc4, log ₂ V4), (Tc7, log ₂ V7),
It is used to derive a ₁₀₁ , a ₁₀₂ , and a ₁₀₃ from the following equation.

Tc1=a ₁₀₁ +a ₁₀₂ log ₂ V1+a ₁₀₃ (log ₂ V1) ² Tc4=a ₁₀₁ +a ₁₀₂ log ₂ V4+a ₁₀₃ (log ₂ V4) ² Tc7=a ₁₀₁ +a ₁₀₂ log ₂ V7+a ₁₀₃ (log ₂ V7) ² Equation 8 A pre-calibration coarse curve 36 is defined by the solution of Equation 8, and the coarse curve 36 becomes as shown in FIG. Note that 35 indicates a pre-calibration curve. Parameters a ₁₀₁ , a ₁₀₂ , a ₁₀₃
specifies the pre-calibration coarse curve for range-1.

At the same time, the processor 14 processes the thicknesses Tc1 and Tc.
4. The standard pieces with Tc7 are sequentially driven from the magazine 21 and inserted into the radiation, and the current or current data (Tc1, log ₂ V1p), (Tc4,
log ₂ V4p), (Tc7, log ₂ V7p). This current data is based on the current calibration rough curve 3
8 using the same calculation as above, the parameters a ₁₀₁ p, a ₁₀₂ p,
a ₁₀₃ p.

In Fig. 8, the comparison values △logV2,
Δlog ₂ V3, Δlog ₂ V5 and Δlog ₂ V6 are calculated by processor 14 using curves 36 and 38. These comparative values correspond to curves 36 and 3.
This is an approximate value obtained by interpolation using 8, and represents the drift from the previous calibration point to the current point for each calibration point. In fact, the comparison value is
It is calculated by the processor 14 using the specific parameters a ₁₀₁ , a ₁₀₂ , a ₁₀₃ of the curve 36 and the specific parameters a ₁₀₁ p, a ₁₀₂ p, a ₁₀₃ p of the curve 38 .

Thus, the new modified logarithmic signal log ₂ Vnc
is expressed as follows.

log ₂ V1c=log ₂ V1p log ₂ V2c=log ₂ V2+△log ₂ V2 log ₂ V3c=log ₂ V3+△log ₂ V3 log ₂ V4c=log ₂ V4p log ₂ V5c=log ₂ V5+△log ₂ V5 log ₂ V6c = _log2V6 + _{Δlog2V6log2V7c} = _log2V7p These corrected logarithmic signals are stored _in memory 18.

From these compensated logarithmic signals a compensated curve 40 is derived (FIG. 9). That is, the following calculation is performed in the same way as was done when creating the previous calibration curve.

(Tc1, log ₂ V1c), (Tc2, log ₂ V2c), (Tc3, log ₂ V3c
) → a _111C , a _112C , a _113C Formula 9 (Tc3, log ₂ V3c), (Tc4, log ₂ V4c), (Tc5, log ₂ V5c
) → a _121C , a _122C , a _123C Formula 10 (Tc5, log ₂ V5c), (Tc6, log ₂ V6c), (Tc7, log ₂ V7c
)→a _131C , a _132C , a _133C Equation 11 From these calculations, a drift compensation calibration curve 40 consisting of a connection of three quadratic curves can be drawn as shown in FIG. Curve specific parameters a _111C , a _112C , a _113C , a _121C ... of the drift compensation calibration curve are memory 1
8 type 3 data areas. Similar drift compensation calibration curves for other ranges can be similarly determined.

Finally, a measurement operation of the non-contact thickness gauge is performed.
A material to be measured is placed in the radiation path, and the thickness of the material is calculated by processor 14 using the drift compensation calibration curve determined above. The deviation value ΔT from the nominal thickness value is also calculated by the processor 14,
The obtained deviation value ΔT is displayed on the indicator 20.

As an example of the measurement operation, it is assumed that range-1 is selected in the range selection step described above. As mentioned above, range-1 drift compensation calibration curve 4
0 consists of three quadratic curves corresponding to each of the subranges that make up range-1. Therefore,
First, it must be determined which of the three quadratic curves to use to calculate the thickness of the material being measured. The connection points of the three quadratic curves are (Tc3, log ₂ V3c) and (Tc5, log ₂ V5c)
It is. The logarithmic signal when matter is present in the radiation is therefore compared to log ₂ V3c and log ₂ V5c to determine which of the three quadratic curves to select. Once a curve is selected, the processor 14
calculates the material thickness from the curve and the logarithmic signal.

Since the thickness Tcd obtained in this way is obtained from the drift compensation calibration curve, this value is
When the composition of a substance differs from that of the standard piece, it is an apparent value, not a true value. This is an apparent value because the drift compensation calibration curve was created based on measurements of standard pieces. True thickness T
is expressed by the following equation.

T=Tcd/1+N/100 Therefore, the deviation from the nominal thickness value of the material is:

△T=T-Tn=Tcd/1+N/100-Tn This calculation is performed by processor 14.

Figures 10 to 12 show pre-calibration,
A flow chart of each operation of range calibration and measurement is shown. Each of these flow charts is
Figure 2 summarizes the operation of the non-contact radiation thickness meter according to the present invention.

As shown in FIG. 10, the pre-calibration operation is initiated by a pre-calibration command 52, and under the control of the processor 14, an initialization step 54 transfers data from the ROM of the memory 18 to the type 1 of the first table 46 of the RAM. It is transferred to the data area (Figure 3).

The next step is condition setting 56. Data for one selected range is transferred from the first table 46 to the type 1 data area of the work area 50 (FIG. 4). Depending on the range selected and the substance being measured, processor 14 sets the radiation source voltage and amplifier gain to certain predetermined values.

Next is data collection 58. Standard pieces from the standard magazine 21 are sequentially inserted into the radiation path under the control of the processor 14. The digital signal of the A/D converter 12 is read,
It is converted into a logarithmic signal by processor 14, and this logarithmic signal is stored in a type 2 data area of work area 50.

Parameter calculation 60 is the next step in the pre-calibration operation. Parameters [e.g. a ₁₁₁ , a ₁₁₂ , a ₁₁₃ ,
a ₁₂₁ , a ₁₂₂ , a ₁₂₃ , a ₁₃₁ , a ₁₃₂ , a ₁₃₃ ] are calculated by the processor 14 and stored in the type 2 data area of the work area 50.

Finally, in a data transfer step 62, data is transferred from the type 2 area of the work area 50 to the type 2 and type 3 areas of the first table 46 (FIG. 4).

These steps are repeated until all parameters for all ranges are obtained.

A flow chart of the steps in the range calibration operation is shown in FIG. First, when a range calibration command 64 is given, the thickness gauge is initialized in step 66. Next, read the setting value 68
Then, the nominal thickness Th and the alloy correction coefficient N are read out from the setting unit 16.

The range selection step 70 determines the apparent thickness of the material.
It includes calculations by Tap's processor 14.
A range is selected by determining the range selection zone to which the apparent thickness Tap belongs, that is, by determining the range of the range selection zone to which Tap corresponds.

Then, in a condition setting step 72, type 1, type 2,
Type 3 data is transferred to the corresponding type in work area 50 (FIG. 5). The radiation source voltage and amplifier gain are set to predetermined values depending on the selected range and material type.

The next step in the range calibration operation is a substance check 74, which determines whether the substance to be measured is present in the radiation path. If the substance to be measured is not present, the process proceeds to a data collection step 76 where standard pieces of selected thicknesses (eg, Tc1, Tc4, Tc7) are sequentially inserted into the radiation path and the current reading of the logarithmic signal is obtained. (For example, log ₂ V1p, log ₂ V4p, log ₂ V7p)
is obtained.

The final step is calculation step 78, in which the previous calibration rough curve parameters (e.g., a ₁₀₁ , a ₁₀₂ , a ₁₀₃ ) and the current calibration rough curve parameters (e.g., a _101p , a 102p , a _102p ,
a _103p ). Deviation value (for example,
△log ₂ V2, △log ₂ V3, △log ₂ V5, △log ₂ V
6) is calculated, and a corrected logarithmic signal is calculated and stored in the work area 50. Drift compensation calibration curve parameters (e.g. a _111C , a _112C ,
a _113C , a _121C , a _122C , a _123C , a _131C , a _132C , a _133C )
is calculated by the processor 14 and stored in the work area 50 of the memory 18. Finally, the corrected logarithmic signal and drift compensation calibration curve parameter data are transferred to the type 3 area of the first table in the RAM (FIG. 5). If the substance to be measured is present in the radiation path after substance check step 74, steps 76 and 78 are performed.
is bypassed.

During operation of the thickness gauge according to the present invention, the measuring operation is performed in the first
This is shown in the flow chart in Figure 2.

In standby mode 80, measurement operation is performed under condition setting 8.
Starting at 2, a standard piece is first extracted from the radiation path under the control of processor 14. The timing pulse 83 occurs, for example, every 10 milliseconds,
Applied to AND gate 84. This timing pulse 83, together with the condition setting 82,
Start the measurement sequence.

A logarithmic signal reading step 86 reads the logarithmic signal for the substance currently present in the radiation path for each operation of the processor 14.

A thickness calculation 88 is performed. A quadratic curve is selected that forms part of the drift compensation calibration curve for the selected range. The thickness Tcd is calculated using the quadratic curve thus selected. Then, the true thickness T is determined by the following formula.

T=Tcd/1+N/100 Finally, the deviation ΔT is obtained by calculating ΔT=T−Tn.

〔Effect of the invention〕

The thickness gauge according to the present invention uses a curve to represent the calibration curve, so it can achieve high precision measurement, and it also reduces the number of standard pieces inserted into the radiation path during calibration. Accuracy can be improved, and thus measurement accuracy can be improved. Furthermore, the settings of the thickness and alloy correction values can be changed even during measurement of the substance to be measured. Also,
The built-in standard piece has the advantage of not having to use an extremely thin standard piece that is difficult to create with high precision and is easily damaged.

[Brief explanation of drawings]

FIG. 1 is a diagram showing a block diagram of a non-contact radiation thickness meter based on a preferred embodiment of the present invention, and FIG. 2 is a diagram showing an aluminum plate,
Figures 3 to 5 are diagrams showing preferred ranges of thickness values for measuring cold-rolled steel sheets and hot-rolled steel sheets, and Figure 6 is a diagram showing data transfer and handling in the memory of the same embodiment. , FIGS. 7 to 9 are diagrams for explaining range calibration, and FIGS. 10 to 12 are diagrams showing pre-calibration curves for several ranges and range selection zones, respectively.
FIG. 3 is a diagram showing a flow chart of pre-calibration, range calibration, and measurement operations. 2...X-ray source, 6...Measurement substance, 8...Detector, 10...Amplifier, 12...A/D converter, 1
4...Processor, 16...Setting unit, 18
...Memory, 20...Indicator, 22...D/A converter.

Claims (1)

[Claims] 1. A thickness meter for non-contact measurement of the thickness of a material in a predetermined thickness range using radiation; comprising a radiation source; detection means for producing an output signal V related to the level of incident radiation; wherein the level of the incident radiation is a function of the radiation absorption properties and thickness values of a material placed in the radiation path between the radiation source and the detection means; standard piece setting means for selectively inserting a standard piece having a thickness value of a point into the radiation path, the plurality of calibration points are distributed and determined within the predetermined thickness range, and the predetermined thickness range is It consists of a plurality of ranges, each range including three or more calibration points; it has memory means for storing parameters for specifying a calibration curve for each range, and said calibration curve is defined by said output signal V and said calibration curve. It approximately represents the correspondence relationship with the thickness of the material placed in the radiation path, and the thickness value ranges covered by each calibration curve in adjacent ranges partially overlap; the nominal value of the material to be measured. a setting means for setting a thickness value and an alloy correction coefficient; each output signal V when a standard piece having a thickness value of each calibration point is inserted into the radiation path by the standard piece setting means; and the thickness values of the calibration points corresponding to these, the parameters for specifying the calibration curve for each range are calculated and stored in the memory means, and the parameters for specifying the calibration curve are calculated from among the stored parameters for specifying the calibration curve. A parameter in a predetermined range is selected in response to a signal from the setting means, and the thickness of the substance of unknown thickness introduced into the radiation path is adjusted to match the selected parameter and the substance of unknown thickness is introduced into the radiation path. A non-contact radiation thickness meter comprising processing means for calculating from the output signal V of the detection means when placed. 2. The non-contact radiation thickness gauge according to claim 1, wherein the plurality of calibration points are distributed in a geometric series within the predetermined thickness range. 3. The standard piece setting means includes a standard piece magazine for selectively inserting standard pieces of different known thicknesses into the radiation path, and the calibration points are such that each thickness thereof is at least as large as the number of standard pieces in the standard piece magazine. The non-contact radiation thickness meter according to claim 1 or 2, characterized in that the thickness meter is selected so that it can be configured with three sheets. 4. Claim 1, wherein the calibration curve is represented by a quadratic curve whose thickness value is determined by a quadratic function of the logarithm of the output signal V.
3. The non-contact radiation thickness meter according to item 1 or 2. 5. The calibration curve is represented by a parabola whose thickness value is determined by a parabolic function of the logarithm of the output signal V.
3. The non-contact radiation thickness meter according to item 1 or 2. 6. The non-contact radiation thickness meter according to claim 5, wherein the calibration curve for each range comprises three connected parabolas. 7. The non-contact radiation thickness gauge according to claim 6, wherein each range has seven calibration points. 8. The calibration curve represents the correspondence relationship between the output signal V and the thickness of a substance placed in the radiation path having the same radiation absorption characteristics as the radiation absorption characteristics of the standard piece, and From the obtained apparent thickness value Tcd, use the relational expression T=Tcd/1+N/100 (where N is the percentage value of the alloy correction coefficient set in the setting means) to calculate the thickness T of the substance of unknown thickness. The non-contact radiation thickness meter according to claim 1, characterized in that the radiation thickness meter calculates the following: 9. The non-contact radiation thickness gauge according to claim 1, wherein the thickness of the substance of unknown thickness is expressed as a percentage deviation from a nominal thickness value set in the setting means. 10 having a radiation source and detecting means for receiving radiation from the radiation source and producing a detection output signal related to the level of radiation received, wherein said level of radiation received is determined by said radiation source and said detection means; A method for calibrating a non-contact radiation thickness meter having a plurality of movable standard pieces having a predetermined thickness that is a function of the radiation absorption properties and thickness of a material placed in the radiation path between the means and the radiation path, the method comprising: ) sequentially inserting sets of the movable standard pieces corresponding to a plurality of scheduled calibration thickness points into the radiation path;
With this sequential insertion, the detection output signals corresponding to each of the calibration thickness points are sequentially obtained, and calibration point data consisting of a combination of each calibration thickness point and the detection output signal corresponding to each of these points is obtained. , the calibration point thickness points are distributed within a predetermined thickness range, the thickness range being composed of connected subranges, each subrange being characterized by at least three calibration thickness points; (b) (c) determining parameters for specifying a pre-calibration curve formed by concatenation of curves determined from the calibration point data belonging to each sub-range; (d) The set of movable standard pieces corresponding to the calibration thickness points determined by the calibration point data selected in (c) above are sequentially (e) The current calibration rough curve determined from the new calibration point data in (d) above. The current calibration rough curve is obtained by drift-compensating the previous calibration rough curve, and (f) the previous calibration rough curve is drift-compensated by the difference between the previous calibration rough curve and the current calibration rough curve. A method for calibrating a non-contact radiation thickness gauge, comprising the steps of determining specific parameters of a drift-compensated calibration curve compensated for. 11. The method of calibrating a non-contact radiation thickness gauge according to claim 10, wherein the subrange is characterized by three calibration thickness points. 12. The method of calibrating a non-contact radiation thickness meter according to claim 10, wherein the pre-calibration coarse curve is specified by three calibration point data. 13. The method of calibrating a non-contact radiation thickness gauge according to claim 11, wherein the pre-calibration rough curve is specified by three calibration point data. 14. The method of calibrating a non-contact radiation thickness gauge according to claim 11, wherein the thickness range is characterized by seven calibration thickness points. 15. The method of calibrating a non-contact radiation thickness gauge according to claim 13, wherein the thickness range is characterized by seven calibration thickness points. 16 The step f calculates a comparison value of the detection output signal at each calibration thickness point from the previous calibration rough curve and the current calibration rough curve, and compares these comparison values at each calibration thickness point with the calibration determined in (a) above. Compensated calibration point data is obtained from the point data, and parameters for specifying a drift-compensated calibration curve are determined from these compensated calibration point data.
Item, Item 11, Item 12, Item 13, Item 14,
Or the method for calibrating a non-contact radiation thickness meter according to item 15.