A Draw-In Sensor for Process Control and Optimization Numpon Mahayotsanun, Jian Cao* and Michael Peshkin Department of Mechanical Engineering, Northwestern University, Evanston, IL 60201, USA *Corresponding author: Tel: 847-467-1032,
[email protected] Abstract. Sheet metal forming is one of the major processes in manufacturing and is broadly used due to its high degree of design flexibility and low cost. In the sheet metal forming process, draw-in (planar movement of a sheet periphery) frequently occurs and is one of the most dominated indicators on the success of a forming process. Currently, monitoring and controlling draw-in during each stamping operation requires either time-consuming setup or a significant die modification. Most devices have been used only in laboratory settings. Our goal is to design a draw-in sensor providing high sensitivity in monitoring; ease of setup, measurement and controlling; and eventually be implemented in industry. Our design is based on the mutual inductance principle, which we considered physical factors affecting the characteristics of the draw-in sensor. Two different configurations, single-transducer and double-transducer of our draw-in sensors have been designed and tested. The results showed good linearity, especially for the double-transducer case. The output of the draw-in sensor was affected by the type of sheet metal, dimension of the transducer, and the distance between the transducer and the testing sheet metal. It was found that the result was insensitive to the waviness of the sheet metal if sheet thickness was thin. The invention, implementation, and integration of the draw-in sensor will have an enormous impact on revolutionizing the control of stamping process, will provide solid ground for process variation and uncertainty studies, and ultimately will affect the design decision process.
INTRODUCTION The sheet metal stamping process is performed by placing a sheet-metal blank over a die cavity and then pushing the metal into the opening with a punch. The blank is held down against the die by a blank holder. As the punch moves further down, the sheet is drawn into the die cavity, which amount depends on the restraining force imposed by a blank holder force and/or draw heads restraining force. Due to the high efficiency of material utilization and the high productivity that stamping enjoys, this process has been widely used in the automotive industry, beverage can industry, appliance and aerospace industry, just to name a few. Various tests and industrial experiences have proven that the amount of draw-in directly affects whether a stamped part will result in wrinkling or tearing, due to excessive or too little draw-in, respectively. Figure 1 illustrates what the draw-in is, which is the amount of blank edge movement. The existing draw-in sensors are mostly made for laboratory use. For example, the LVDT type, which requires significant setup time at industrial setup, becomes too time-consuming or too expensive to use.
FIGURE 1. The original and final blank shapes in an irregular blank forming. A, B, C, and D show the four corners
With the intention of continuously measuring the drawn-in amount of sheet metals in a stamping process, our approach is to develop a draw-in sensor, using the principle of mutual inductance between two loops as seen in Figure 2 [1, 2]. The excitation current flowing in the primary coil induces Electromotive Force (emf) in the secondary coil. The presence of metal (ferrous or non-ferrous) near the coils affects the magnetic field lines, displayed on the right of Figure 2, changing the degree of mutual inductance. Thus, the induced signals in the secondary coil reflect how much of the coil is covered up.
CP778 Volume A, Numisheet 2005, edited by L. M. Smith, F. Pourboghrat, J.-W. Yoon, and T. B. Stoughton © 2005 American Institute of Physics 0-7354-0265-5/05/$22.50
861 Downloaded 06 Aug 2007 to 165.124.160.13. Redistribution subject to AIP license or copyright, see http://proceedings.aip.org/proceedings/cpcr.jsp
cover up the top surface of the transducers as seen in Figure 4c.
FIGURE 2. A schematic of the sensor element using the mutual inductance used in the double-transducer configuration
FIGURE 3. A schematic of the single-transducer configuration
Based on the mutual inductance principle, we developed a draw-in sensor by experimenting with a prototype printed on a conventional circuit board to address the need of an affordable and accurate draw-in sensor [3]. We proposed two configurations of the draw-in sensor measurement, which will be further explained in the next two sessions. Measuring results using these two designs will be presented in each of these two sessions, followed by a general discussion on conclusions and future work. Since the ultimate goal is to incorporate the draw-in sensor in the real manufacturing environment, the development of advanced closed-loop control strategies with the drawin sensor will be a future work of this project [4].
FIGURE 4. a Transducers b. Schematic of the loops in the transducers c. Transducers covered by epoxy
SINGLE-TRANSDUCER CONFIGURATION
Figure 5 shows the experimental setup of the drawin sensor and the simulated stamping tools were made by steel blocks. In the single-transducer configuration, a transducer covered with epoxy was inserted in the blank holder’s slot, which would detect the induced voltage changes caused by the disturbances of the sheet metal located below. The linear position sensor, having 75 mm detection range, was attached to the sheet metal to determine how much the sheet metal had moved and served as a reference. While the sheet metal was being pulled (manually by using pliers) horizontally to the left, the magnetic coupling area would increase, resulting in the change of induced voltages detected by the transducer. Then, both the transducer and the linear position sensor transmitted signals to the draw-in sensor data acquisition board, which amplified the voltage readings and sent them to the computer for further analysis. Therefore, knowing the amount of voltages indicated by the draw-in sensor, the flow of the sheet metal could be obtained.
Since the sensor cannot interfere with sheet metal during the stamping process, we designed the transducers such that they would be small enough to be embedded in either a die or a blank holder in the stamping tool. Our first design, called singletransducer configuration, was to put together the primary and secondary coils into one transducer as shown in Figures 3 and 4. The principle of mutual inductance between the two loops was still valid, but the movement of the magnetic fields and the disturbance of the sheet metal would differ from what was shown in Figure 2. Figures 4a represents the fabricated transducers having a few rounds of outer loops and inner loops. The schematic of the loops is shown in Figure 4b. In order to avoid wear on the sensors and prevent sheet metal from sliding into the transducer slot during the forming process, a thin layer of epoxy was used to
862 Downloaded 06 Aug 2007 to 165.124.160.13. Redistribution subject to AIP license or copyright, see http://proceedings.aip.org/proceedings/cpcr.jsp
TABLE 1. Considered parameters in the singletransducer configuration Materials Transducer Widths 0.76 mm AL3003 1.66 mm AL5052 0.64 mm ST1018 1.14 mm ST1018
8.89 mm 12.7 mm 16.5 mm
Draw-in Sensor: 12.7mm Single-TD Testing 0.76mm AL3003 at 0t gap height
Displacement (mm)
60 50 40 30 20 10 0 6.5
7.5
8.5
9.5
10.5
Induced Voltage (V)
FIGURE 5. Experimental setup
FIGURE 7. Repeatability results of the singletransducer configuration Material type: We considered four sheet metals to see the effects of material type to the sensitivity of the sensors. The results are illustrated in Figure 8. Note that the slopes of the aluminum sheets were different from those of steel sheets. In the single-transducer configuration, the permeability of sheet metal influenced the magnitude of the mutual inductance. The steel sheets have high permeability, which allow the magnetic fields to penetrate to a certain skip depth, causing high induced voltages while covering the entire magnetic coupling area. On the other hand, aluminum sheets have low permeability, which distort the magnetic fields and disturb the mutual inductance, causing low induced voltage while covering the entire magnetic coupling area. Between the different aluminum sheets, the slopes were slightly different. There existed the slight difference between the slopes of different steel sheets as well.
FIGURE 6. a. Flat sheet metals b. Shim sheets to create gaps between transducer and sheet The experimented sheet metals are displayed in Figure 6a. The hole on the top right of each sheet was for mounting the linear position sensor. The shim sheets, Figure 6b, were used to create gaps between the sheet metal and the transducer, which will be later discussed in the double-transducer configuration.
Draw-in Sensor: Different Materials Testing 12.7mm Single-TD at 0t Gap Height
We considered two major factors influencing the characteristics of the draw-in sensor as presented in Table 1. Specifically, in the remaining of this session, we will illustrate each factor, and why we want to study this factor and the result. The factors are: material, and transducer width. Each combination of different factors was experimented at least five trials in order to show the repeatability of each test as an example shown in Figure 7. Note that the initial induced voltage was at approximately 6.8V, indicating that the transducer always detected the mutual inductance in the single-transducer configuration even though the magnetic coupling area was completely covered by the sheet metal.
80
Displacement (mm)
70
0.64mm ST1018
60 50
0.76mm AL3030
1.14mm ST1018
40 30
1.16mm AL5052
20 10 0 7
7.5
8
8.5
9
9.5
10
Induced Voltage (V)
FIGURE 8. Results of different materials in the single-transducer configuration
863 Downloaded 06 Aug 2007 to 165.124.160.13. Redistribution subject to AIP license or copyright, see http://proceedings.aip.org/proceedings/cpcr.jsp
10.5
Transducer width: Three different transducer widths sheet metals examined in the double-transducer configuration. were considered to see whether the size of the Transducer width:width: Three Three different transducer widths widths sheet sheet metals metals examined in the indouble-transducer Transducer different transducer examined the double-transducer transducers affected theto whet output oftheher thesize draw-in sensor were considered to see the configuration. were considered seeher whet the ofsize of the configuration. and the results are presented Figure It cansensor be transducers affected the output ofinthe draw-in sensor transducers affected the output of the9.draw-in and the results arewider presented in Figure 9. It can be observed thatresults produced higher and the are transducers presented in Figure 9. It can be observed that wider transducers produced highersince induced voltages. results were as produced expected observed that The wider transducers higher induced voltages. were aswere expected since induced voltages. The results asrises expected the magnitude of The the results mutual inductance with since the the magnitude the of mutual inductance rises with the in magnitude the inductance risesthe with the increase theofarea of the mutual secondary coils. increase in the area the secondary coils. coils. increase in theofarea of the secondary
Draw-inDraw-in Sensor:Sensor: Different Transducer Widths Widths Different Transducer TestingTesting 0.76mm0.76mm AL3003AL3003 at 0t Gap at Height 0t Gap Height 80
80
60 50 40 30 20 10 0 6.5
70
Displacement (mm)
Displacement (mm)
70
8.89mm8.89mm
12.7mm 12.7mm
60 50 40
16.5mm
30
16.5mm
20 10 0 6.5
7
7
7.5
7.5
8
8
8.5
8.5
Induced Voltage (V) Induced Voltage (V)
FIGURE 10. Transducers locations in the doubleFIGURE 10. Transducers locations in the doubletransducer configuration transducer configuration
FIGURE 9. Results of different transducer widths widths in FIGURE 9. Results of different transducer in FIGURE 9. Results of different transducer widths in the single-transducer configuration the single-transducer configuration
the single-transducer configuration
According to the effect of the transducer width, we According to the effect of the transducer width, we could increase the induce voltages by increasingwidth, the According to the effect of the transducer could increase the induce voltages by increasingwe the size of the secondary coils. Therefore, the doublecould size increase induce coils. voltagesTherefore, by increasing the of thethe secondary the doubletra nsducer configuration was introduced and discussed size of the secondary coils. transducer configuration wasTherefore, introduced the and doublediscussed in the next session where the secondary coils were transducer introduced and discussed in theconfiguration next session was where the secondary coils were relocated into the outer loops of another transducer. into thewhere outer loops of another transducer. in therelocated next session the secondary coils were
relocated into the outer loops of another transducer.
DOUBLE-TRANSDUCER DOUBLE-TRANSDUCER CONFIGURATION CONFIGURATION DOUBLE-TRANSDUCER
FIGURE 11. Wrinkled sheets FIGURE 11. Wrinkled sheets The double-transducer configuration also used the CONFIGURATION We considered four major factors influencing the The ance double-transducer configuration used the We considered four sensor major factors influencing the mutual induct principle, which was exactly also shown characteristics of the draw-in as presented in mutual induct ance principle, which was e xactly shown characteristics of the draw-in sensor as pre sented in in Figure 2, where the primary and secondary coils The configuration also used the 2. WeWe focused on following factors: in double-transducer Figure 2,separated where the primary andFigure secondary coilsTable Table considered fourthemajor factors influencing the 2. We focused on the following factors: we re located in transducers. 10 transducer width, gap distancesensor between the mutual inductance principle, which was exactlyFigure shown10material, we re located in separated transducers. characteristics of the draw-in as presented in ma terial, transducer width, gap distance between the shows the locations of each transducer in the and sheet metal, and material shape. In the in Figure where the primary and secondary in coilsthetransducer shows2, setup. the locations of each Table 2. We focused on the following factors: transducer and sheet metal, and material shape. In the experimental In addition to thetransducer purpose of remaining of this session, we will illustrate each were experimental located in separated setup. Intransducers. addition to theFigure purpose material, transducer width, gap distance between remaining of tothis session, we and will each the increasing the induced voltages, this configuration was 10of why we want study this factor theillustrate result. increasing the induced voltages, this configuration was showsto the oftheeach transducer in the factor,transducer aimed studylocations the effect of gap distance between factor, why we want study this factor andwas the result. and sheetto metal, and material shape. In the Again, each combination of different factors aimed to setup. study theIneffect of theAlso, gapthe distance between experimental addition to of the transducer and sheet metal. we purpose could Again, ateach of different factors was remaining of combination thistrials session, we will of illustrate each experimented least five and an example the the transducer and sheet metal. Also,sheet we was could increasing theoutput induced voltages, this configuration investigate the of the sensors to wrinkled experimented at least five trials and an example of the factor, why we want to study this factor and the result. repeatability results are shown in Figure 12. Note that thefrequently output ofgot the sensors to wrinkled aimedinvestigate to study the effect of thewrinkles gap distance between metals since sheets during the sheet repeatability results are shown 12. factors Note thatwas Again, eachproduced combination ofin Figure different this configuration better repeatability results metals since sheets frequently got wrinkles during the forming process. The wrinkledAlso, sheet metals this configuration produced better repeatability results the transducer and effects sheet ofmetal. we could compared to those of five the trials single-transducer experimented at least and an example of the formingsince The effects of wrinkled sheet metals are important wrinkling change forming comparedIn addition, to those of the single-transducer investigate theprocess. output of thecan sensors tothe wrinkled sheet configuration. the initial voltages of all repeatability results are shown in Figure 12. Note that are important since wrinkling can change the forming process by providing more material one zone and the the trials configuration. In addition, initial voltages all metals since sheets frequently got inwrinkles during were at zero, indicatingthe that the mutual ofresults this configuration produced better repeatability process by providing more material in one zone and therefore, delaying The the tearing. Onwrinkled the othersheet hand, the trials were attwo zero, indicatingdisappeared that the mutual forming process. effects of metals inductance between transducers compared tothe those of transducers the single-transducer therefore, delaying the tearing. On the other hand, wrinkling presents a difficulty in controlling the inductance between the two disappeared while being fully covered by the testing sheet metal. are important since wrinkling can change the forming wrinkling presents a difficulty in controlling the configuration. addition, thetesting initial voltages springback effectively. Figure 11 shows four wrinkled while being fullyIncovered by the sheet metal. of all process by providing more material in one zone and springback effectively. Figure 11 shows four wrinkled the trials were at zero, indicating that the mutual
therefore, delaying the tearing. On the other hand, wrinkling presents a difficulty in controlling the springback effectively. Figure 11 shows four wrinkled
inductance between the two transducers disappeared while being fully covered by the testing sheet metal.
864 Downloaded 06 Aug 2007 to 165.124.160.13. Redistribution subject to AIP license or copyright, see http://proceedings.aip.org/proceedings/cpcr.jsp
TABLE 2. Considered parameters in the double-transducer
configuration, the double-transducer configuration
TABLE 2. Considered parameters in the double-transducer configuration, thedifferences double-transducer configuration configuration displayed no in the direction of the curves configuration displayed no differences in the direction of the curves Transducer Materials Material parameters Gap between the aluminum and steel sheets. Since the TABLE 2. Considered in the double-transducer configuration, the double-transducer configuration Materials Material Transducer Gap T ABLE 2. Considered parameters inWidths the double-transducer between the aluminum and insteel sheets. Since the than theeach double-transducer configuration Shapes Distances configuration, configuration displayed no differences the direction of the curves thickness of material was many times higher Shapes Widths Distances configuration no differences in the direction of the curves thickness of each material was many times higher than Between Materials Material Transducer Gap displayed between the aluminum and steel Since the its skin depth, the magnetic fieldssheets. could not penetrate Between Materials Material Gap the aluminum and steel sheets. Since the than its skin depth, magnetic fields couldtimes not penetrate Transducer ShapesTransducer Widths Distances between thickness ofthe each material was many higher Tra nsducer through the sheets to the secondary coils. Thus, the Shapes Widths Distances of each material was many times higher than andBetween Sheet thickness through the sheets to the secondary coils. Thus, the its skin depth, the magnetic fields could not penetrate and Sheet Between induced voltages differed with different sheet metals Tra nsducer its skin depth, thesheets magnetic fields could not penetrate 0.76mm 8.89mm Ot voltages differed with different sheet metalsthe through the towas the secondary coils. Thus, 0.76mm 8.89 mm 0tand Sheet induced Transducer but their slope the same for allsheets the sheets in through the sheets tosign the secondary coils. Thus, the AL3003 but their slope sign was the same for all the in induced voltages differed with different sheet metals AL3003 and Sheet Flat 0.76mm Flat 8.89 mm 0t the voltages double-transducer configuration. 1.66mm 12.7mm It induced differed with different sheet metals the double-transducer configuration. 1.66mm 12.7 1t 0.76mm 8.89 mm 0t but their slope sign was the same for all the sheets in AL3003 AL5052 but their slope sign was theconfiguration. same for all the sheets in Flat AL3003 AL5052 the double-transducer 1.66mm 12.7 mm 1t Flat Transducer width: different transducer 0.64mm the double-transducer configuration. Transducer width: ThreeThree different transducer widthswidths 1.66mm 12.7 mm 1t 2t 0.64mm AL5052 2t ST1018 were considered to see whether the size AL5052 ST1018 were considered to see whether the size of theof the Transducer width: Three different transducer widths 0.64mmWrinkled Wrinkled 16.516.5 mmmm 2t Transducer width: Three different transducer widths 1.14mm 0.64mm transducers affected the output of the draw-in sensor 1.14mm transducers affected the output of the draw-in sensor ST1018 were considered to see whether the size of the 2t 3t 3t Wrinkled 16.5 mm ST1018 ST1018 were considered to presented see the whether the size of14. the ST1018 1.14mm Wrinkled and the results are presented in Figure It can be transducers affected output of the draw-in sensor and the results are in Figure 14. It can be 16.5 mm 3t 1.14mmST1018 transducers affected the output of the draw-in and the results are presented in Figure 14.sensor It can higher be observed that wider transducers produced 3t observed that wider transducers produced higher ST1018 and the results are presented in Figure 14. It can be
observed that transducers produced induced voltages andbehavior this behavior also occurred induced voltages andwider this also occurred in higher the in the
observed thatvoltages wider transducers produced higher Draw-in Sensor: Repetability Results Testing 12.7mm induced and this behavior also Therefore, occurred in the single-transducer configuration. Therefore, wider single-transducer configuration. wider voltages and thisconfiguration. behavior also occurred in thewider Double-TD with Sensor: Flat 1.66mm AL5052Results at 2t Gap Height Draw-in Repetability Testing 12.7mminduced single-transducer Therefore, transducers provided higher induced voltages for both transducers provided higher induced voltages for both Draw-inDouble-TD Sensor: Repetability ResultsAL5052 Testing with Flat 1.66mm at12.7mm 2t Gap Height single-transducer configuration. Therefore, wider transducers provided higher induced voltages for both configurations. configurations. Double-TD with Flat 1.66mm AL5052 at 2t Gap Height
80
transducers provided higher induced voltages for both configurations. configurations.
80
70 80 60 70 50 60
Draw-in Sensor: Different Transducer W idths Sensor: Different W idths Testing Draw-in W rinkled 1.14mm ST1018Transducer at 1t Gap Height Testing W rinkled 1.14mm ST1018 at 1t Gap Height Draw-in Sensor: Different Transducer W idths 80 Testing W rinkled 1.14mm ST1018 at 1t Gap Height
3040 2030 1020
50 40 30 20 10
010 0 0
0 0
0
0.5 0.5
0.5
1
1
1.5
1.5
2
Voltage (V) 2 1 Induced Induced 1.5 Voltage (V)
2
2.5
2.5
2.5
3 3
Displacement (mm )
4050
60
Displacement (mm ) Displacement (mm )
Displacement (mm)
Displacement (mm) Displacement (mm)
70
3
70
-1
50 40
20
-1
10
0 0
-1 0
0
16.5 mm 16.5 mm 16.5 mm
30
20 30
10
8.89 mm 12.7 mm 8.89 mm 12.7 mm 8.89 mm 12.7 mm
60
10 20
Induced Voltage (V)
FIGURE 12. Repeatability resultsresults of theof doubleFIGURE 12. Repeatability the doubleFIGURE 12. Repeatability results of the doubletransducer configuration FIGURE 12. Repeatability results of the doubletransducer configuration transducer configuration transducer configuration
80
70 80 60 70 50 60 40 50 30 40
0
1
0
1
1
2
2
3
3
Induced Voltage (V) Induced3 Voltage (V) 2 4
4
5
4
5
6
5
6
6
Induced Voltage (V)
Displacement (mm)
Displacement (mm) Displacement (mm)
FIGURE 14. Results of different transducer widths in in FIGURE 14. Results of different transducer widths FIGURE 14. Results of different transducer widths in F IGURE 14. Results of different transducer widths in the double-transducer configuration the double-transducer configuration Draw-in Draw-in Sensor:Sensor: Different Flat Sheets TestingTesting Different Flat Sheets the double-transducer configuration the double-transducer configuration 8.89m m8.89m Double-TD atFlat 0t Gap Height m Double-TD at 0t Gap Height Draw-in Sensor: Different Sheets Testing Gap Gap distance between transducer andand sheet metal: distance between transducer sheet metal: 8.89m m Double-TD at 0t Gap Height 80 80 Gap distance between transducer and sheet metal: To see whether the distance between the transducer To see whether the distance between the transducer Gap distance between transducer and sheet metal: 1.14mm1.14mm 70 7080 To see whether the distance between the transducer and sheet metal affected the sensitivity o f the draw-in and sheet metal affected the sensitivity o f the draw-in To see whether the distance between the transducer ST1018 ST1018 1.14mm 60 6070 and sheet metal affected thegap sensitivity ofbetween the draw-in sensor, four different distances ST1018 sensor, four different gap distances between the and sheet metal affected the sensitivity ofthe the draw-in 50 5060 1.66mm 1.66mm 0.76mm sensor, 0.76mm four different gap distances between the transducer and sheet metal were considered. The gap transducer and sheet metal were considered. The gap 40 50 sensor, four different gap distances between the 40 1.66mmAL5052 AL5052 AL3003 0.76mm AL3003 transducer and sheet metal were considered. The gap distance was increased by adding the shim sheets in 30 distance was increased by adding the shim sheets in AL5052 3040 transducer and sheet metal were considered. The gap AL3003 distance was by and adding the shim sheets inn inn in between the transducers and the results are show 20 between the increased transducers theadding results are show 2030 0.64mm distance was increased by the shim sheets in 0.64 mm between the transducers and the results are show n in 10 Figure 15. Note that 0t represented “no gap between 20 10 ST1 018 Figure 15. Note that 0t represented “no gap between 0.64 ST1mm 018 between the transducers and the results are shown in 0 10 Figure 15.transducers”, Note that 0thaving represented “no gap between the a long shim sheet a shor 0 ST1 0182.5 the transducers”, having a long shim sheet andand a shor t t -0.5 0 0.5 1 1.5 2 3 3.5 Figure 15. Note that Ot represented "no gap between -0.5 0 0.5 1 1.5 2 2.5 3 3.5 0 the transducers”, having aFigure long shim and a shor t “gap shim sheet as shown in Figure 6b. 1t represented Indu ced V oltage (V ) shim sheet as shown inhaving 1tsheet represented “gap -0.5 0 0.5 1 2.5 3 3.5 Indu ced V1.5 oltage (V )2 the transducers", a6b. long shim sheet“gap and a short shim sheet as shown in Figure 6b. 1t represented distance equals to the thickness of the testing sheet Indu ced V oltage (V ) distance equals to the thickness of the testing sheet FIGURE 13. Results of different materials in th e shim sheet as shown in Figure 6b. It represented "gap metal”, which we added another pair of long and short distance equals to the thickness of the testing sheet FIGURE 13. Results of different materials in the metal”, which we added another pair of long and short double-transducer FIGURE 13. Results of differentconfiguration materials in the distance equals to another the thickness testing sheet shim sheets create a gap. 2t and 3tthe also used the metal”, which we to added pair of of long and short double-transducer configuration shim sheets to create a gap. 2t and 3t also used the FIGURE 13. Results of configuration different materials in the shim double-transducer same logic described for From the graph, it can metal", we added another pair of the long andbeshort sheets towhich create a gap. 2t1t. and 3t also used same logic described for 1t. From the graph, it can be Material type: We considered four sheet metals to se e double-transducer configuration seen higher gap1t. distances created lower induced sameshim logicthat described for From the graph, italso can be the sheets to create a gap. 2t and 3t used Material type: We considered four sheet metals to se e the type: effectsWe ofconsidered material type the metals sensitivity that higher gap distances created lower induced Material four to sheet to seeof the seen voltages. distances lower induced the effects of and material type are to the sensitivity of th e seenseen that samehigher logicgap described forcreated It. From the graph, it can be sensors the considered results in Figure 13.th As the effects of material type to shown the sensitivity of e see voltages. Material type: We four sheet metals to voltages. sensors and the results are shown in Figure 13. As seen seen that higher gap distances created lower induced inand the the graph, eachare material the sensors showntoaffected in the Figure 13.draw-in As seen the effects of results material type sensitivity ofsensor the inin the graph, each material affected the draw-in sensor voltages. providing different slopesin on each13. curve. the by graph, eachresults material theFigure draw-in sensor sensors and the areaffected shown AsInseenIn byby providing curve. contrast different with theslopes resultson of each the curve. single-transducer providing different slopes on each In in the graph, each materialofaffected the draw-in sensor contrast contrast with with the the results results of the the single-transducer single-transducer
by providing different slopes on each curve. In contrast with the results of the single-transducer
865 Downloaded 06 Aug 2007 to 165.124.160.13. Redistribution subject to AIP license or copyright, see http://proceedings.aip.org/proceedings/cpcr.jsp
Two different configurations, single-transducer and
Draw-in Sensor: Different Gap Heights Testing 8.89mm Double-TD With Flat 1.14mm ST1018
Displacement (mm) Displacement (mm)
80
Draw-in Sensor: Different Gap Heights Testing 8.89mm Double-TD With Flat 1.14mm ST1018
70
1t
60
3t
80 50 70 40
2t 0t 1t
60 30 50 20
3t
2t 0t
40 10 30 0 20 0
0.5
1
10
1.5
2
2.5
3
2
2.5
3
Induced Voltage (V)
0 0
0.5
1
1.5
Voltage (V) heights between FIGURE 15. Results ofInduced gap FIGURE 15. Results ofdifferent different gap heights between transducer and sheet in the double-transducer transducer and sheet in the double-transducer configuration FIGURE 15. Results of different gap heights between configuration transducer and sheet in the double-transducer Material shape: In order to see the effect of material configuration Material In order to see the effect of material shape to shape: the output of the draw-in sensor, we shape to the output of the draw-in sensor, considered four wrinkled sheet metals as shown in we Material shape: In order to see the effect of material considered four wrinkled sheet metals as Figure compared the the results with the flatshown ones. shape 11 to and the output of draw-in sensor, we in Figure 11presents and the with flatflat ones. Figure 16 the results ofresults both wrinkled and considered fourcompared wrinkled sheet metals as the shown in sheet Notethe thatresults theresults results ofwrinkled the flat wrinkled Figure 16 presents of both and flat Figuremetals. 11 and compared the with ones. sheets were slightly those sheet Note that the results of ofthethe wrinkled Figuremetals. 16 presents thedifferent results ofthan both wrinkled and flat sheets for materials, AL5052 and flat sheet metals. Note that the 1.66mm results the wrinkled sheets werethick slightly different thanofthose of the 1.14mm ST1018. However, the those draw-in sensor sheets were slightlymaterials, different than of the flat and sheets for thick 1.66mm AL5052 provided similar results for both flat and wrinkled sheets for thick materials, 1.66mm AL5052 and 1.14mm ST1018. However, the draw-in sensor sheets in the case where thefor sheets thin, as 1.14mm ST1018. However, thewere draw-in sensor provided similar results both flat andsuch wrinkled 0.76mm 0.64mm ST1018. As wrinkled a result, provided similar results for flatwere and sheets inAL3003 the caseand where the both sheets thin, such as the draw-in is insensitive material shape sheets in the sensor case where the sheets to were thin, such as 0.76mm AL3003 and 0.64mm ST1018. As a result, 0.76mm AL3003 and 0.64mm ST1018. smaller As a result, when the sheet thickness is approximately than the sensor isisinsensitive insensitiveto tomaterial material shape the draw-in draw-in sensor shape 1mm. when the sheet thickness is approximately smaller when the sheet thickness is approximately smaller thanthan 1mm. 1mm. Draw-in Sensor: Different Material Shapes Testing
16.5mm Double-TD at 3t Gap Height 80
Displacement Displacement (mm) (mm)
70
Draw-in Sensor: Different Material Shapes Testing Wrinkled16.5mm 1.66mmDouble-TD at 3t Gap Height
Flat 1.14mm ST1018
AL5052
60 80
Wrinkled 1.66mm Flat 1.66mm Wrinkled 1.14mm Flat 1.14mm AL5052 ST1018 ST1018 AL5052 30 50 Flat&Wrinkled Flat 1.66mm Wrinkled 1.14mm 20 0.76mm AL3003 40 Flat&Wrinkled ST1018 AL5052 10 30 0.64mm ST1018 Flat&Wrinkled 0 20 0.76mm AL3003 Flat&Wrinkled 0 1 2 3 4 5 6 7 8 10 0.64mm ST1018 50 70 40 60
Induced V oltage (V)
0 0
1
2
3
4
5
6
7
8
Two different configurations, single-transducer and double-transducer, of draw-in the draw-in double-transducer, of the sensor sensor were were introduced. The single-transducer configuration introduced. The single-transducer configuration Two different good configurations, single-transducer and produced linearity. However, the permeability of produced good linearity. However, the permeability double-transducer, of the draw-in sensor wereof each material affected the induced voltages such that each material affected the induced voltages such that introduced. The single-transducer configuration the initial voltage while thethemagnetic coupling area the initial voltage whileHowever, the magnetic coupling area produced good linearity. permeability of fully covered was not On zero. the other hand, the was was fully covered was not zero. theOn other hand, the each material affected the induced voltages such that double-transducer produces a “zero” readout double-transducer produces a "zero" the initial voltage always while always the magnetic coupling area readout when the transducers are fully covered. As expected, when the transducers are fully covered. As was fully covered was not zero. On the other hand, theexpected, the sensitivity of the sensorasensor increased as the as the the sensitivity ofdraw-in theproduces draw-in double-transducer always “zero”increased readout width the the transducer increased in both width of increased in both when theoftransducers are transducer fully covered. As expected, configurations. For For the double-transducer the sensitivity of the draw-in sensorthe increased as the configurations. double-transducer configuration, draw-in sensor provided better width of thethe transducer increased inprovided both better configuration, the draw-in sensor repeatability results than thosetheof thedouble-transducer single-transducer configurations. For repeatability results than those of the single-transducer configuration. the Different materials provided different configuration, provided better different configuration. draw-in Differentsensor materials provided induced voltages butthan their slope sign was the same for repeatability results those of the single-transducer induced voltages but their slope sign was the same for all the sheets ofDifferent the same type. Ifprovided the gapdifferent distance configuration. materials all the sheets of the same type. If the gap betweenvoltages the transducer and sheet metal thedistance induced but their slope sign was increased, the same for between the transducer and sheet metal increased, the induced voltages would be type. reduced. Furthermore, the all the sheets of the same If the gap distance induced voltages would be reduced. Furthermore, the draw-in sensor was not affected by the shape of the between the transducer and sheet metal increased, the was affected by the shape sheetdraw-in (flat or sensor wrinkle) when the sheet thickness was of the induced voltages would be not reduced. Furthermore, the sheet (flat orinwrinkle) sheet thickness was thinner than 1mm ouraffected setupwhen with gapshape between the draw-in sensor was not by athe the of the die and theorbinder being approximate 3t.a gap Thiswas is a thinner than 1mm inatour between the sheet (flat wrinkle) when thesetup sheetwith thickness verydie important wrinkling occurs 3t. inthe the thinner than in ourassetup with aoften gap between and1mm thefeature binder being at approximate This is a flange area and webeing wanted our draw-in3t.sensor tois be die and theimportant binder at approximate Thisoccurs a in the very feature as wrinkling often insensitive toarea the wrinkles. very important feature occurs insensor the to be flange and as wewrinkling wanted often our draw-in flange area and to wethe wanted our draw-in sensor to be insensitive wrinkles. To further the interaction between these insensitive to theanalyze wrinkles. factors and the draw-in sensor, more sophisticated To further analyze thea interaction between these study as aanalyze DOE study should be between performed, for To such further the interaction these factors and the draw-in sensor, a more sophisticated factors and draw-in sensor, a more shape sophisticated example, to the study the effect of wrinkle in terms asgap a study DOE study should beSince performed, for study such such as a DOE should be performed, for of astudy normalized and its wavelength. the example, to study the effect of wrinkle shape in terms example, to study effect of wrinkle shape sensor in termsin ultimate goal is totheincorporate the draw-in a normalized gapitsand its wavelength. Since the of gap and wavelength. Since the the a ofnormalized real manufacturing environment, the ultimate goal is to incorporate theexisting draw-in ultimate goal is to the draw-in sensor and insensor in implementation of incorporate draw-in sensor into environment, the the real realmanufacturing the new the tooling will manufacturing be studied. environment, In addition, the implementation of draw-in sensor into existing and implementation of draw-in sensor intocontrol existing and development of advanced closed-loop new tooling will be In beaddition, the strategies with the draw-in sensor will explored. new tooling willstudied. be studied. In addition, the development of advanced closed-loop control control development of advanced closed-loop strategies with the draw-in sensor will be explored. strategies with the draw-in sensor will be explored.
REFERENCES 1. 1. 2.
FIGURE 16. Results of different material shapes in Induced V oltage (V) 2. the double-transducer configuration FIGURE 16. Results of different material shapes in the double-transducer configuration FIGURE 16. Results of different material shapes in 3.
the double-transducer configuration
CONCLUSIONS
3. 4.
4. The draw-in CONCLUSIONS sensor was developed to measure the flow of materials in the stamping process at a real-time CONCLUSIONS sensor was to measure the base The withdraw-in no interference withdeveloped regular stamping press. flow of materials in the stamping process at a real-time sensor was to measure baseThe withdraw-in no interference with developed regular stamping press. the
flow of materials in the stamping process at a real-time base with no interference with regular stamping press.
REFERENCES P. Kopczynski, “Lvdts. Theory & Application”. REFERENCES Sensors 1992, 9: 18-22. P. “Lvdts. TheorySensitivity & Application”. D-j Kopczynski, Choi, C-T Rim; et al., “High Inductive 1. P. 1992, Kopczynski, "Lvdts.Measurement”. Theory & IEEE Application". Sensors 9: 18-22. Sensing System for Position Sensors 9:al., 18-22. D-j Choi, C-T 1992, Rim;and et “High SensitivityTechnology Inductive Instrumentation Measurement Sensing for Position Measurement”. IEEE Inductive 2. D-j System Choi, Rim; et al, "High Sensitivity Conference 2000,C-T 2: 595-599. Instrumentation Measurement Technology Cao, Sensing J., Lee, System J. and and for Peshkin, M., Measurement". Northwestern Position IEEE Conference 2000, 2: 595-599. University. Real-time draw-in and methodsTechnology of Instrumentation and sensors Measurement Cao, J., Lee, J. 2000, and Pat. Peshkin, M., Northwestern fabrication. 2002. U.S. 6,769,280. Conference 2: 595-599. University. Real-time draw-in sensors and methods of N. Krishnan, J. Cao, “Real-time calculation 3. Cao, J., Lee, J. and Peshkin, M., of Northwestern fabrication. 2002. U.S. force Pat. 6,769,280. optimal blank-holder history sheet metal University. Real-time draw-ininsensors and methods of N. Krishnan, “Real-time calculation of forming using J.an Cao, ARMA model”. 2002 Japan US fabrication. 2002. U.S. Pat. 6,769,280. optimal blank-holder forceAutomation, history in sheet metal Symposium on Flexible Hiroshima, 4. N. Krishnan, J. Cao, "Real-time calculation of forming Japan. using an ARMA model”. 2002 Japan US optimal blank-holder force history in sheet metal Symposium on Flexible Automation, Hiroshima, Japan.forming using an ARMA model". 2002 Japan US Symposium on Flexible Automation, Hiroshima, Japan.
866 Downloaded 06 Aug 2007 to 165.124.160.13. Redistribution subject to AIP license or copyright, see http://proceedings.aip.org/proceedings/cpcr.jsp
