TWI467129B - Method for detecting flatness of nozzle - Google Patents

Description

Test method for flatness of casting nozzle

The present invention relates to a method for detecting flatness, and more particularly to a method for detecting the flatness of a nozzle.

In some methods of making amorphous alloy ribbons, a casting nozzle is used to spray the alloy solution onto a metallization roll. Through the rapid condensation of the metal crystallization roll, an amorphous alloy ribbon can be formed. Therefore, the flatness of the surface of the nozzle of the nozzle directly affects the quality of the casting condition of the nozzle. If the casting nozzle is flawed, the process of the amorphous alloy ribbon may be interrupted, or the quality of the amorphous ribbon produced may be poor.

At present, a detection technique for the flatness of the surface of the nozzle of the nozzle is to measure the surface using a three-dimensional gauge bed. Although the measurement accuracy of the three-dimensional measuring bed is high, the measuring time is rather lengthy. Moreover, the three-dimensional volume measurement technology is measured one by one, so the shape of the nozzle cannot be fully described. Furthermore, the three-dimensional volume measurement technology measures the probe in contact with the object to be tested. Therefore, using the probe to measure the nozzle in the red hot state of the upper line will not only damage the probe, but also may Destroy the surface of the nozzle.

Therefore, one aspect of the present invention provides a method for detecting the flatness of a nozzle, which is a non-contact automatic detection method. Avoid damage to the nozzle surface and damage the detector.

Another aspect of the present invention provides a method for detecting the flatness of a nozzle, which uses an optical three-dimensional range finder and a uniaxial moving platform to obtain a three-dimensional coordinate point group of the nozzle surface of the nozzle, and then transmits the image. The way of treatment is to calculate the flatness of the nozzle surface. Therefore, the detection method can quickly obtain a large number of sampling point groups, and helps to fully describe the shape of the nozzle, and the detection speed is fast. In addition, the detection method can further introduce a noise filtering step to reduce the influence of the vibration error signal caused by the three-dimensional range finder on the movement of the structure nozzle, thereby more accurately describing the shape of the nozzle.

Another aspect of the present invention is to provide a method for detecting the flatness of a nozzle, which can quickly detect the manufacturing flaw of the nozzle during the production process of the amorphous alloy ribbon, thereby effectively reducing the incidence of process interruption. And can ensure the quality of amorphous alloy ribbon.

In accordance with the above object of the present invention, a method for detecting the flatness of a nozzle is provided, which comprises the following steps. The surface of the nozzle of the nozzle is scanned by an optical three-dimensional range finder to obtain a plurality of three-dimensional coordinate points. The arithmetic operation is performed using an arithmetic device to fit the equation of the reference plane using at least a portion of the three-dimensional coordinate points. The distance between each three-dimensional coordinate point and the reference plane is calculated using an arithmetic device and using the equation of the reference plane. The flatness detection value of the aforementioned surface is calculated using an arithmetic device and using the distance between the three-dimensional coordinate points and the reference plane and the preset index.

According to an embodiment of the invention, the optical three-dimensional range finder includes a laser and a three-dimensional range finder, and the three-dimensional range finder is coupled to the laser.

According to another embodiment of the present invention, the above-described optical three-dimensional measurement is utilized. The step of scanning the surface on which the nozzle of the nozzle is located includes the following steps. The optical three-dimensional range finder is placed on a uniaxial moving platform. The optical three-dimensional range finder is placed in front of the surface of the aforementioned nozzle. The uniaxially moving platform is moved in a direction parallel to the surface.

According to still another embodiment of the present invention, the above computing device comprises a computer device.

In accordance with still another embodiment of the present invention, the step of performing the logical operation process includes utilizing a linear least squares method.

According to still another embodiment of the present invention, the step of performing the logical operation processing comprises using an arithmetic device and fitting the equation of the rough reference plane by using the three-dimensional coordinate point, performing the filtering processing of the noise, and using the remaining three-dimensional coordinate points. Combine the equation of the reference plane. This noise filtering process consists of the following steps. The distance between each three-dimensional coordinate point and the coarse reference plane is calculated using an arithmetic device and using the equation of the coarse reference plane. The three-dimensional coordinate points that are separated from the rough reference plane by more than a predetermined value are removed.

According to still another embodiment of the present invention, in the step of performing the logical operation processing, before the step of fitting the equation of the reference plane, the method of repeating the equation and the noise filtering using the three-dimensional coordinate point to fit the rough reference plane is repeated. Except processing at least once.

According to still another embodiment of the present invention, the preset indicator is a maximum value, a minimum value, an average value, or a standard deviation.

100‧‧‧Test method

102‧‧‧Steps

104‧‧‧Steps

106‧‧‧Steps

108‧‧‧Steps

200‧‧‧Optical three-dimensional range finder

202‧‧‧Laser

204‧‧‧Three-dimensional range finder

206‧‧‧ uniaxial mobile platform

208‧‧‧ computing equipment

210‧‧‧Support frame

212‧‧‧Carrier

214‧‧‧ casting nozzle

Section 216‧‧‧

Section 218‧‧‧

220‧‧‧ spout

222‧‧‧ surface

The above and other objects, features, advantages and embodiments of the present invention are made. Can be more obvious and easy to understand, the description of the drawing is as follows:

1 is a flow chart showing a method for detecting the flatness of a nozzle according to an embodiment of the present invention.

2 is a schematic view of a device for detecting the flatness of a casting nozzle according to an embodiment of the present invention.

Since the quality of the casting nozzle will seriously affect the process and quality of the amorphous alloy ribbon, there is no high precision and no damage to the nozzle. Therefore, the present invention proposes a non-contact automatic detection method for the surface flatness of the casting nozzle, which can quickly detect whether the casting nozzle is suitable for the upper wire to manufacture the amorphous alloy ribbon.

Please refer to FIG. 1 and FIG. 2 simultaneously, which are respectively a flow chart of a method for detecting the flatness of a nozzle according to an embodiment of the present invention, and a schematic diagram of a device for detecting the flatness of the nozzle. The detection method of the present invention can be applied to the detection of the surface flatness of various objects. In the present embodiment, this detection method is applied to detect the flatness of the surface 222 where the nozzle 220 of the nozzle 214 is located.

In the present embodiment, when the flatness detecting method 100 of the casting nozzle 214 is performed, the flatness detecting device of the casting nozzle 214 can be provided first. In some embodiments, as shown in FIG. 2, the detecting device mainly includes an optical three-dimensional range finder 200, a uniaxial moving platform 206, and an arithmetic device 208. The optical three-dimensional range finder 200 can primarily include a laser 202 and a three-dimensional range finder 204. The laser 202 and the three-dimensional range finder 204 can be coupled to each other. In an exemplary example, as shown in FIG. 2, the laser 202 and the three-dimensional range finder 204 can be integrated into one. The structure. The laser 202 can scan the surface 222 of the nozzle 214 by emitting laser light toward the nozzle 214 to be inspected. The laser 202 can pass its scanned signal to the three-dimensional range finder 204, thereby obtaining a plurality of three-dimensional coordinate points for the surface 222. The three-dimensional range finder 204 can digitize these three-dimensional coordinate point data.

The optical three-dimensional range finder 200 is placed on the uniaxial moving platform 206. The uniaxially moving platform 206 can be driven by a linear motor to move in a single direction while simultaneously moving the objects disposed thereon in the same direction. That is to say, the uniaxial moving platform 206 can drive the optical three-dimensional range finder 200 to move in a single direction.

The computing device 208 is electrically coupled to the optical three-dimensional range finder 200 and can receive data from the digitized three-dimensional coordinate points of the optical three-dimensional range finder 200. The computing device 208 can utilize these three-dimensional coordinate point data for logical operation processing to fit the equation of the reference plane. The computing device 208 can also calculate the distance between each three-dimensional coordinate point and the reference plane. The computing device 208 can further calculate the flatness detection value of the surface 222 by using the distances according to a preset index. In some demonstrative examples, computing device 208 can include a computer device.

In addition, the detecting device can further include the support frame 210 and the carrier 212 according to actual application requirements. As shown in FIG. 2, the carrier 212 can be a frame such that the nozzle 214 to be inspected can be loaded into the frame. A support frame 210 is disposed under the carrier 212 to support the carrier 212 and the nozzle 214 disposed in the carrier 212. In some examples, the support frame 210 has an adjustable height design to adjust the nozzle 214 to be inspected in response to inspection requirements. the height of. In general, the nozzle 214 can include two portions 216 and 218. The cross-sectional shape of the portion 216 may be U-shaped, and the cross-sectional shape of the portion 218 may be U-shaped. The two portions 216 and 218 are butted to form a casting nozzle 214 having a spout 220 therebetween.

In some embodiments, referring to FIG. 1 and FIG. 2 simultaneously, when the detecting method 100 is performed, the casting nozzle 214 to be detected is disposed in the carrier 212, and the optical three-dimensional range finder 200 is set in a single Move axially on the platform 204. Moreover, the uniaxially moving platform 204 and the optical three-dimensional range finder 200 disposed thereon are placed in front of the surface 222 of the nozzle 214 provided with the spout 220, so that the optical three-dimensional range finder 200 can be edged. The surface 222 of the front spout 220 is scanned with laser light in a single direction until scanning of the area to be tested of the surface 222 of the nozzle 214 is completed. In a preferred embodiment, the uniaxially moving platform 204 moves in a direction parallel to the surface 222 of the casting nozzle 214. The height of the nozzle 214 needs to match the height of the optical three-dimensional range finder 200 such that the nozzle 214 is within the scanning range of the laser light of the optical three-dimensional range finder 200. In some examples, the three-dimensional range finder 204 may be instrumentally calibrated prior to optical scanning to improve detection accuracy.

After the installation of the detecting device is completed, the laser light of the optical three-dimensional range finder 200 can be used to emit laser light to the surface 222 of the nozzle 220 of the casting nozzle 214 to scan the surface 222 of the casting nozzle 214 as described in step 102. . The lateral resolution of the scan depends on the rate of movement of the uniaxially moving platform 204, while the longitudinal resolution of the scan depends on the resolution of the laser of the optical three-dimensional range finder 200. The laser 202 transmits the signal obtained by the scan to the light The three-dimensional range finder 204 of the three-dimensional range finder 200. After the three-dimensional range finder 204 processes these signals, a number of three-dimensional coordinate points are available. The three-dimensional range finder 204 can further digitize these three-dimensional coordinate points.

Next, as described in step 104, the equation of the reference plane is fitted through the computing device 208 and using all or a portion of the digitized three-dimensional coordinate points. In some embodiments, since the three-dimensional coordinate points obtained by the optical three-dimensional range finder 200 scanning may not all represent the three-dimensional coordinate points of the surface 220 of the nozzle 214, the surface 220 representing the nozzle 214 may be first captured. The three-dimensional coordinate point is used as the target point for fitting the reference plane. In some exemplary examples, if the movement of the optical three-dimensional range finder 200 is stable without vibration, there may be no or only a few noise points in the three-dimensional coordinate points obtained by the scanning, which may interfere with the fitting result. No need to perform the impurity filtering process. In the general case, in order to improve the accuracy of the fitted reference plane, one or more noise filtering processes are usually performed on these three-dimensional coordinate points.

For example, in the example in which the noise filtering processing is not performed, the program in which the arithmetic device 208 performs the logical operation processing is as follows. Stipulate that the point groups of these three-dimensional coordinate points are And define the fitted plane equation as aX+bY+cZ+d=0. Where { P } is the point group of these three-dimensional coordinate points, and p _i ( x _i , y _i , z _i ) is each three-dimensional coordinate point. The cost function f to be minimized is set to the sum of the distances D _i between the three-dimensional coordinate points p _i and the pseudo-planes:

Using the optimized numerical method, the minimum solution ( a ₁ , b ₁ , c ₁ , d ₁ ) of the cost function f can be obtained, and the parameters describing the plane equation can be obtained. In an exemplary example, this logical operation includes the use of a linear minimum variance method. In this example, the noise filtering process is not performed, so the fitted plane equation is the equation of the reference plane.

On the other hand, in the example in which the three-dimensional coordinate points are subjected to the noise filtering processing, the program for performing the logical operation processing by the arithmetic device 208 is as follows. Fit a rough reference plane first. When fitting a rough datum plane, as in the above example, let the point groups of these three-dimensional coordinate points be And define the fitting equation is a rough reference plane aX + bY + cZ + d = 0. Where { P } is the point group of these three-dimensional coordinate points, and p _i ( x _i , y _i , z _i ) is each three-dimensional coordinate point. The cost function f to be minimized is set as the sum of the distances D _i between the three-dimensional coordinate points p _i and the pseudo-rough reference planes:

Using the optimized numerical method, the minimum solution ( a ₁ , b ₁ , c ₁ , d ₁ ) of the cost function f can be obtained, and the parameters describing the equation of this coarse reference plane can be obtained. In an exemplary example, this logical operation includes the use of a linear minimum variance method.

Since the three-dimensional coordinate points used to fit the calculated rough reference plane described above may include some points such as those occurring in a periodic manner, in addition to the points representing the surface 222 of the nozzle 214, the calculated rough reference is thus calculated. Although the plane passes substantially through the surface 222 of the nozzle 214, the accuracy is Still affected by these noises. These noises may be due to the optical three-dimensional range finder 200 being subjected to a one-way moving platform vibration during scanning. Therefore, in order to obtain a more accurate reference plane, next, the noise filtering process is performed.

In the noise filtering process, the arithmetic device 208 can be used, and each three-dimensional coordinate point is calculated using the equation a ₁ X + b ₁ Y + c ₁ Z + d =0 of the rough reference plane fitted above. The distance between the rough reference planes. The point group {Q} of the noise is defined as a point beyond the predetermined reference value from the rough reference plane: , | D _j |> default value.

The preset value of the distance can be defined according to the magnitude of the vibration and the distribution range of the noise. In an exemplary example, the preset value is set to 0.2 mm. Next, the noise {Q} is filtered out, that is, the three-dimensional coordinate points whose distance from the rough reference plane exceeds the preset value are removed. After the noise {Q} is filtered out, the point group {R} of the remaining three-dimensional coordinate points is obtained.

Then, the arithmetic device 208 is used, and the remaining three-dimensional coordinate points are utilized, and the equation a ₂ X + b ₂ Y + c ₂ Z + d =0 of the reference plane is fitted as described above by fitting the rough reference plane. In some exemplary examples, the noise filtering process can be performed only once. In other exemplary examples, one or more procedures for fitting the equations of the rough reference plane and the noise filtering process may be repeated according to the actual detection requirements.

Referring again to FIGS. 1 and 2, after fitting the equation of the reference plane, each of the three-dimensional coordinate points can be calculated using the computing device 208 and using the above-described fitted reference plane equation as described in step 106. The distance from this datum plane. Taking the reference plane equation a ₂ X + b ₂ Y + c ₂ Z + d =0 obtained after filtering through the noise point as an example, and the distance between each three-dimensional coordinate point and the reference plane is set to D _k , then:

Next, as described in step 108, the flatness detection value of the surface 222 of the casting nozzle 214 is calculated using the computing device 208 and utilizing the distances between the calculated three-dimensional coordinate points and the reference plane and the preset index. In some exemplary examples, the preset indicator may be a maximum value, a minimum value, an average value, or a standard deviation, that is, a maximum value, a minimum value, an average value, or a standard deviation of a distance between the three-dimensional coordinate point and the reference plane. For example, if the standard deviation is used as the preset index, the smaller the value of the flatness detection value calculated by the standard deviation, the higher the flatness of the surface 222 where the nozzle 220 of the nozzle 214 is located.

It can be seen from the above embodiments that one of the advantages of the present invention is that the method for detecting the flatness of the nozzle is a non-contact automatic detection method, thereby avoiding damage to the surface of the nozzle and without damaging the detecting device.

It can be seen from the above embodiments that another advantage of the present invention is that the method for detecting the flatness of the nozzle utilizes an optical three-dimensional range finder and a uniaxial moving platform to obtain a three-dimensional coordinate point group of the nozzle surface of the nozzle, and then The flatness of the nozzle surface is calculated by image processing. Therefore, the detection method can quickly obtain a large number of sampling point groups, and helps to fully describe the shape of the nozzle, and the detection speed is fast. In addition, the detection method can further introduce a noise filtering step to reduce the vibration error caused by the three-dimensional range finder when moving. The influence of the number on the shape of the casting nozzle can further describe the shape of the casting nozzle more accurately.

According to the above embodiments, another advantage of the present invention is that the method for detecting the flatness of the nozzle can quickly detect the manufacturing flaw of the nozzle during the production process of the amorphous alloy ribbon, thereby effectively reducing the process interruption. The incidence and the quality of the amorphous alloy ribbon can be ensured.

While the present invention has been described above by way of example, it is not intended to be construed as a limitation of the scope of the invention. Therefore, the scope of the invention is defined by the scope of the appended claims.

100‧‧‧Test method

102‧‧‧Steps

104‧‧‧Steps

106‧‧‧Steps

108‧‧‧Steps

Claims (8)

A method for detecting the flatness of a nozzle comprises: scanning an surface of one nozzle of a casting nozzle with an optical three-dimensional range finder to obtain a plurality of three-dimensional coordinate points; and performing a logic operation using an arithmetic device And using at least a portion of the three-dimensional coordinate points to fit an equation of a reference plane; using the computing device and using the equation of the reference plane to calculate a distance between each of the three-dimensional coordinate points and the reference plane; The flatness detection value of the surface is calculated by using the computing device and using the distances of the three-dimensional coordinate points from the reference plane and a predetermined index. The method for detecting the flatness of a nozzle according to claim 1, wherein the optical three-dimensional range finder comprises a laser and a three-dimensional range finder, and the three-dimensional range finder is engaged with the laser. The method for detecting the flatness of a nozzle according to claim 1, wherein the step of scanning the surface of the nozzle of the nozzle by using the optical three-dimensional range finder comprises: setting the optical three-dimensional range finder On a single axial moving platform; the optical three-dimensional range finder is disposed in front of the surface of the casting nozzle; and the uniaxially moving platform is moved in a direction parallel to the surface. The method for detecting the flatness of a nozzle according to claim 1, wherein the computing device comprises a computer device. The method for detecting the flatness of a nozzle according to claim 1, wherein the step of performing the logical operation comprises using a linear minimum variance method. The method for detecting the flatness of a nozzle according to claim 1, wherein the step of performing the logical operation comprises: using the computing device and using the three-dimensional coordinate points to fit an equation of a rough reference plane; Point filtering processing, comprising: calculating a distance between each of the three-dimensional coordinate points and the coarse reference plane using the computing device and using the equation of the coarse reference plane; and removing the relationship between the coarse reference plane and the coarse reference plane The three-dimensional coordinate points whose distance exceeds a predetermined value; and the equations for fitting the reference plane by using the remaining three-dimensional coordinate points. The method for detecting the flatness of a casting nozzle according to claim 6, wherein in the step of performing the logical operation processing, before the step of fitting the equation of the reference plane, the method further comprises: using the three-dimensional coordinate points repeatedly The equation for the coarse reference plane is combined with the noise filtering process at least once. The method for detecting the flatness of a nozzle according to claim 1, wherein the preset index is a maximum value, a minimum value, an average value, or a standard deviation.