extrusion profile geometry measurement methods and case studies.
There are many variables that affect the extrusion size and shape quality of the profile. First of all, the physical properties of rubber compounds vary depending on the elastic properties of natural rubber, the size and structure of carbon black particles, and before reaching the extruder, the uniformity of the dispersion of the mixture and the polymerization history of the compound, and other factors. The consistency of rubber flowing through the mold varies with the thermal history applied by the extruder and the adhesive properties produced by the composite under mold pressure and temperature conditions. These attributes, combined with flow- The limit geometry of the mold, once the rubber exits the mold, it will lead to a change in the degree of expansion of the mold (ref. 1). When the extrusion is carried out through downstream operations, changes in heating and cooling affect the final geometry. The driver, the cutter, and the strip further change the line tension and cause distortion of the profile geometry. In addition, the feed interruption during the batch change will result in a change in the mold pressure, thus affecting the flow rate. Historically, complex profiles are regularly checked using a 10x optical comparator. This requires the operator to cut the rubber sheet, place it on the magnifying glass, and visually compare the sample to the 10x specification on the transparent film. Thin strips of rubber profiles are easily distorted by the operator, and these systems are considered to have high operatorsto- Operator changes. Studies have shown that the results of 10-fold optical comparison measurements are comparablecontact on- Line Measurement (ref. 2). To close- In order for a line statistical check, the sample of the test must represent the entire population. This study shows that it is easy to miss important processes with regular inspections. Sometimes online vianon monitors simple profiles continuously Contact optical width measuring instrument. Traditionally, when presented under a laser scan point, these methods are based on the principle of measuring the shadow of the extrusion projection (ref. 3). The challenge is that many profiles have key features that cannot be identified with ashadow measurement principles. In addition, the movement of the profile relative to the sensor introduces measurement errors. For shadow measurement, there is also a problem with the profile with multiple compounds, inserts, enhancements, and trivia. [ Figure 1 slightly] This article is introduced in two parts. The first is the on-method. This paper describes how the new laser profile triangulation technology can be used in the online profile measurement system. The second part presents several case studies showing common problems in Profile extrusion and how these problems are identified Method of line measurement. Laser Triangulation is a method of distance measurement. As shown in Figure 1, the laser light source projects a point on the measuring surface from d1. The laser is reflected through the lens onto the detector at the dl position. When the measuring surface is located closer to the DO light source, the reflected light falls on the detector at the dO position. When the measuring surface is located further away from the light source in D2, the reflected light falls on the detector in D2 position. The scope of D2- You can scale the detector to the range of the detector d2- Therefore, the detector output is related to the actual distance between the sensor and the measured surface. Earlier laser triangulation sensors used similar devices, such as arsenic and aluminum, to detect the change in position of the reflected laser. These were listed in the 1970 s and are called location sensitive detectors (PSD). The PSDs output a variation voltage or current proportional to the position of the light on the detector. Charging Equipment (CCD) It is now widely used as a detector. CCDs is a light-sensitive micro-chip used in digital cameras. For example, a CCD detector may contain 1,024 pixels, resulting in a detector resolution of 1,024 parts throughout the measurement range. [ Figure 2: Outline triangle measurement in the middle 1990 s, Bytewise measurement system and other companies have developed a contour triangulation sensor that replaces the laser line source of the laser point light source and replaces the double- The size CCD array replaces a single Axis pixel array. As shown in figure 2. Profile triangle sensor using a megapixel detector (1,024 x1,024) Now you can get fixed data equivalent to 1,024. Point laser triangle sensor. These are called \"film\" laser sensors. These sensors can obtain a single profile of 1,024 points at a frequency of 15 samples per second, which applies- Rubber extrusion measurement online. Pay attention to complimentarymetal- Oxide Semiconductor (CMOS) The detector is used for ccddettors that require a higher frequency. In similar applications, CMOS detectors meet the requirements of 2,000 to 8,000Hz. [ Figure 3 slightly][ Figure 4 slightly][ Figure 5 Slightly]Multiple- The sensor system can combine multiple profile triangulation sensors into a fixed frame to see the external profile of the extrusion (figure 3). Each photographer takes a snapshot at the same time. The configuration file is converted into a common coordinate system and sent to the \"measurement server\" software component that performs the following actions Matches the reference measurement or CAD template. Therefore, the profile can be moved up and down, side-to- Side and rotate not working on the matching routine (ref. 5). The virtual calipers allow to measure the critical dimensions of the profile. Target and tolerance values are associated with each key dimension. The virtual calipers asEthernet message or analog signal provide numerical output. The results of the measurement server can be obtained by the Distributed operator interface. Multiple operator interfaces can be run on the win32 Platform that displays the current measurement results. This allows operators at both ends of the production line to see the current measurement results. Supervisors and engineers located in a facility with network access to the measurement server can also see the measurement results. Cross-presentation of operating interface software Section template ( CAD or reference measurement data) , The profile of the measurement and the virtual clamp (figure 4). In addition, the powerful \"on- The linecomparator view can be enabled, which shows deviations from the current metrics and templates. Create a vector from each measurement point to the nearest template point. This vector is then plotted with the magnification factor, which enables it to visualize in a higher proportion than the original data. The error vector is encoded in green, yellow and red to indicate tolerance consistency. Different tolerances can be established for each segment of the template. The sensor system consists of sensors, power supplies, cables, data acquisition cards, PC and software. Software can be divided into several functions, including Time data acquisition, calibration transformation, parameter calculation, visualization, testing Plan management, data History management, reporting, and external communications as shown in figure 5. Real- The time application takes a single profile, performs calculations to extract measurements, and outputs measurements for visualization, grading decisions, and automatic feedback. This is done at sensor frequency. Resolution and precision resolution are defined as minimum size changes that can be identified. The resolution of the Bytewise profile sensor is 0. The thickness axis is 001mm or 1 micron. The width resolution is 0. 2% of the width of the field of view (FOV). Accuracy is a function of sensor resolution, system design, and analysis methods. Multi-Precision The sensor system changes from 0. The full scale FOV is 05% to 0. 2% view according to system features (ref. 6). For the 150mm wide fov sensor, the empirical accuracy has been established. A multi- Step authentication block (figure 6) For all measurements. The mass block provides 16 thickness parameters and 4 width parameters for a total of 20 measurement parameters per test. For each parameter, all measurements are repeated 75 times. All measurements ( Parameters 20 times 75 times) Repeat in nine positions within the instrument\'s field of view. The position on the horizontal axis is right, middle, left, and the position on the vertical axis is upper, middle, and lower. Measurement error (EOM) It is a method of expressing the ability of the measuring system, including the components of deviations and repetitive changes. EOM contains 99% confidence intervals. [ Figure 6 slightly][ Figure 7 Slightly]EOM = bias (absolute value)+ 2. 515 [sigma] Calculate EOM for each parameter at each location. The neural network result EOMs of each parameter is averaged to obtain the average error of each measurement parameter. The EOM of 16 thickness and four width parameters are averaged together to produce the average thickness EOM and the average width EOM. The average thickness EOM is 0. 023 mm. The thickness measurement is 99% sure within 0. 023mm of the real value. The average width EOM is 0. 14 mm. There is a widthmeasure determined within 99% months. 14mm of the real value. The worst case for the Thenine EOM value is about twice the average (ref. 7). Case study-- The start-up behavior of the new operation will generate a large amount of waste and reduce the production line output. On- Line monitoring is used to reduce startup time by better understanding startup behavior. By observing the real Time Optical comparator, the user can immediately see the size result of the change of the set point of the process and equipment. The example in figure 7 illustrates how four different dimensions form specifications at different time points. Some Profile360 users who focus on reducing the start-up time claim by about 50%. Process confusion due to batch changes this process confusion ( As shown in figure 8) A change happened. The second batch of rubber has different viscous properties, resulting in reduced extrusion expansion. This reduction in mold expansion indicates that the uncured elastic modulus of the second compound is low (ref. 1). Therefore, all dimensions are transferred to small values. Note that the dimension value is fairly stable ( Low standard deviation) Before and After batch changes. On- The production line monitoring enables the production line operator to observe the effect of the compound change and to adjust the process conditions for compensation. The unit is micron. [ Figure 8:[ Figure 9 omitted When a batch of compounds reaches the end, the process is interrupted due to the interruption of the feed, the operator compresses the tip of the next batch of compounds together and then feeds the joint into the extruder. This can lead to heavy joints or light joints. In some cases, there may be gaps in the extruder feed. As shown in Figure 9, this type of process destruction will result in a temporary change in the profile size. Traditional The line-cycle method of quality check is unlikely to detect these disturbances. The chart above illustrates some typical confusion caused by feed disruption. On- The line monitor is used to alarm when these events occur so that the operator can intervene. [ Figure 10 slightly] Other process interference thickness Channel C shows the sine interference of the peakto- The month of peakmagndorf. 7mm more than 40 seconds (About 40 m)(figure 10). . Thickness Channel E display peakto- Peak change 0. More than 70 seconds 6mm (About 70 m) Next is a step. Change shift of 0. More than 5mm seconds ( About 100 m). [ Figure 11 omitted] If this happens in the factory that checks the profile geometry- There is a line 2 every 30 minutes. 2% the possibility of distraction was detected in dimension C and dimension a 9. 4% possibility of detecting interference in dimension E The periodic extrusion line exhibits a certain degree of periodicity, or the periodic pattern of the size change is typical, as shown in Figure 11. This example shows 0. 7 mm peak-to- Peak change of cycle for 6 minutes and 20 seconds. There are many different reasons for the periodicity, such as the puller drive control, the puller belt splicing effect, the saturated cycle control, the tool cycle, the thermal cycle of the heating system and the screw beat. Such data can be analyzed to determine the time. The frequency of the cycle, which can then be used to diagnose the extruder production line and downtime Flow process. The reduction of these periodic variation ranges reduces the standard deviation of the size values and results in better Cp and Cpk. Conclusion since its launch more than 100 years ago, the rubber extrusion industry has made progress through several generations of contour geometry measurement technology. Off- More than 50 years ago, we have realized the method of line optical comparison. These techniques enable the user to characterize the dimensional quality of the sample but not the quality of the entire population. Recent progress- Line technology can accurately and continuously describe the dimensional quality of complex contours. These enable users to realize several values. CREATE program: * On- Online measurements can be used to continuously monitor the dimensional quality of any profile and alert the operator in case of any problem. This can identify the problem that periodicoff-cannot be identifiedline checking. * On- During the start-up of any operation, line measurements can be used to help reduce the time required to achieve stability. * On- Production line measurements can be used to quickly evaluate all production lines in the plant to determine the periodic conditions that affect all profiles. Remediation of these issues will benefit all profiles running on each row. * On- Production line measurements can be used to help tool development. The snapshot file can be exported. A dxf point file that can be opened in a mold design CAD application and compared with the intended design. The Asdie trial is in progress Production line measurements can be used to observe the stability of the production process. This information can be used to optimize the setting of the extruder. * On- The data History generated by the line measurement can be used to compare any operation with its historical performance and verify the effect of the quality improvement plan. References (1. )J. Dick, \"Overview of capillary flow measurement required by the rubber industry\", presented at the CS rubber sector meeting in October 2006. (2. ) Gauge R & R Research, 2005, source of confidential automotive companies. (3. )M. Harris, \"On- Line geometry measurement of complex profiles, \"Rubber World, December 2001. (4. )D. Reynolds, \"On- On-line technology of tire tread profile measurementof- \"Light laser method\", paper at 2004 tire technology expo and conference. (5. ) Bytewise measurement system, Profile360 User Guide, 2006. (6. )D. Renault, \"contour sensor technology for tiremanufacturing Applications\", presented at 2004 International Tire Fair and Conference. (7. ) Bytewise measurement system, \"On- Line profile measurement error-of- Survey Research, March 2006. Dennis Reynolds, Bytewise measurement system ( Dreynolds @ by Bytewise @. com)