ll SURFACE TEMPERATURE MEASUREMENT IN THEN PLASTEC ’FELMS fines“: {or Hm Degree of M. S. R‘HCHEGM SE'ATE UKEVERSETY Brian Laurence Akers 1965 LIBR A R Y Michigan State University I! iii” villi | i'il ' n I.III I ’3 . SURFACE memmns MEASUREII’LEI‘I'I' IN THIN PIASTIC FILMS by Brian Laurence Akern AN ABSTRACT Submitted to Michigan State University in partial fulfillment of the requirements for the degree of FASTER OF SCIENCE IN PACKAGED Deptrtment at Forest Products 1965 ABSTRACT Present day psokaging economics dictates short-dwell. high- speed hoot sealing so it is becoming increasingly important to know the sotual film surfsoe twperetures st the interface. These taupereturee have never been adequately measured. This study develops an emerimentsl procedure for measuring the temperature at the surface of thin plastic films. The procedure employs s vacuum metallized film thermistor es the temperature- sensing device. In addition, the resulting experimental date is used to validate a computer model for determining thin-£11m surfsoe tem- peratures that should provide eoqierimentsl flexibility for future study of the heat sesling cycle. SURFACE TEMPERATURE MEASUREMENT IN THIN PLASTIC FIEMS by Brian Laurence Akers A THESIS Submitted to Michigan State University in pertiel fulfillment of the requirements for the degree of MASTER OF SCIENCE IN PACKAGING Department of Forest Products 1965 To My m. and Family ii ACKNOWLEII‘IE'ENTS The writer wishes to enqaress his sincere thanks to Mr. David Bi'cuse, packaging research associate, whose film thermistor idea was the basis of this study and whose guidance and sound criticism were often relied upon in Ian-suit of this investigation. Special thanks are extended to Dr. James Book of the MSU Mechanical Engineering Department for his help in sdspting the host transfer computer program to the thin-film appliostion in this study. Finally, the writer is indebted to All of the feculty md stat! of the Michigan State University School of Packaging for their comments and criticisms contributing to moral support during every phase of the work. B. L. A. TABLE OF CONTENTS ACMORIEEGB'TENTS............. LIST LIST I. II. III. IV. V. VI. LIST OF TM 0 '0 O 0 O 0' O O I O O O 0 OF FIGURES O O O O C O O O O O O O 0 INTRODUCTION AND PURPOSE . . . . . . . 0 BACKGROUND ON TWERM’URE MEASURfiiENT . . . . THIN-FEM SURFACE TWERATURE MEASURWT EQUIPMENT EXPERIMENTAL PROCEDURE AND ANAIISIS . . . . TEAIPERATURE MEASURH‘XENT MODELS . . . . o . CONCLUSIONS AND REC01'MWDATION3 FOR FUTURE STUDY OFREFWCFS............ APPENDIX 1 - 031611131210” 0 e e e e e e e e 0 APPENDIX 2 - Figures 0 e e e e e e e e e 0 APPENDIX 3 '- Tables 9 e e e e e e e e e e APPENDIX h - “(13018 e e e e e e e e e e e 111 Page 11 iv 10 17 52 55 LIST OF TABLES Table 1 - quwrimental Calibration Date . . . . . . . . 2-thJerimentelTimingDete . . . . . . . . . 3 - Comparative Date: Experimentel, Model 1, Model 2 . APPENDIX 3 - Tables A - l’hermistor Electrical Resistance Change with Increase in Tanperature e e e e e e e e e e e e B - Thermistor Calibration Results with Initial Resistance atSmpleExtrmes... e eeee see C - Results of Added Weight hiring Calibration at 175°F . D - Theme]. Influence of the Aluminms Coating on the Normal Polystyrene Film Tapereture Profile from MOdGl 2 D8.“ e e e e e e e e .e e e e e iv Page '83 57 57 LIST OF FIGURES Figure 1 - AsseMbled Surface Temperature Measuring System for Thin-Plastic Films e e e e e e e e e e 2 - Experimental Calibration Data . . . . . . . 3 - Experimental Timing Data . . . . . . . . . h -’Ehperimental Results of Final Procedure . . . . 5 - General Interior Nodes . . . . . . . . . 6 - Special Interface Nodes . . . . . . . . . 7 - Comparative Data: Ekperimental, Model 1, Model 2 APPENDIX 2 - Figures A - Thermistor in Clamps with'WOod Spacer in Place to NUll Bridge (set tb) e e e e e e e e e B - Thermistcr in Clamps with Unheated Platen in Place for Experimental Timing or Calibration (heated bar in retracted position) . . . . . . . C - Bridge Circuit Design for Surface Temperature Measurement in Thin Plastic Films . . . . . Page 16 30 31 41 #2 #5 53 53 I. INTRODUCTION AND PURPOSE The problems involved in effecting adequate package closures ‘hy heat sealing are well known. Certainly many packagers have experienced the frustration of being unable to seal a film'uhen all of the conditions on the wrapping machine appeared Just right. With present day packaging economics dictating short-dwell, high- speed.heat sealing, it is becoming increasingly desirable to know the actual film temperatures as related to machine temperature, dwell time, and film thickness. The reduction of production time for package closure has become a prime goal. With this goal in mind, one fundamental ares deserves detailed attention: the heat transfer processes in thin plastic films during the heat sealing cycle. Satisfactory seals can not be obtained if the film temperature at the interface is not in the filmls sealing range. An article directly concerned with this goal'was published by Mr. Sheldon Kevesh in the October, l96h edition of PACKAGE EhGINEERING (b). Kavesh.used a digital computer to make a series of calculations to determine the temperature profile of polypropylene film.when it is contacted.by'the hot and cold platens of a heat sealing machine. Although the study advanced a considerable amount of theoretical information concerning the temperature profile of polyprOpylenc, the experimental proof was very limited. This fact pointed up the need for an adequate surface-temperature measuring device for thin plastic films that might support stesh in his theoretical findings. 2 The purpose of this study is to develop an experimental procedure for’measuring the temperature at the surface of thin plastic films. The resultant procedure hopefully'will provide a.means to better understanding the time-temperature relationship involved in the heat sealing cycle. As a further aid to study of this cycle, two computer.models cf the experimental system are developed in the last section of this thesis. The first model is an adaptation of the computer program Kevesh.used in the forementioned article. The second model was ‘written hy Dr. James Beck of Michigan State University and adapted for use in.this thinnfilm surface temperature measurement application. Both models are included because the computer can be used to great advantage once the analyst has achieved a parallelism'between the practical situation and his model. It is normally much easier to manipulate the model to study the characteristics in'uhich one is interested than it is to try to work with the practical system. A thorough understanding of the simulation procedures in.describing the system modeled is not necessary if the forementioned parallelism can 'be shown to exist (7). The author*will attempt to show this parallel- ism.between the experimental data and the computer models. Hopefully, this may promote future systems analysis of similar systems through computer simulation. This thesis project was partially sponsored.under the multi- sponsor research program conducted'hy the School of Packaging at Michigan State University. The thesis goals directly parallel the 3 stated project goals and the results should represent a. significant ‘ contribution to the goals of the anti-sponsor research program. II. BACKGROUND ON TEMPERATURE MEASUREMENT The measurement of surface temperature in thin plastic films has never been adequately accomplished. In any measurement situation, the measuring technique must first take into account what effects the measuring device will have on the system; that is, it must be deter» mined if the measurement itself will disturb the system and to what extent the observed data will differ from that which would be obtained'were the measuring device not in the system. It can be assumed that the more thermally massive the measuring device, the more probable it becomes that it will significantly affect the system. For this reason when using a temperature measuring device, two things must be kept in.mind: (a) What are the possible effects the device will have on the system and (b) to'what extent will they affect the measurement data derived from the device. The most oommcn.methcds of measuring temperature are based on one of four principles: fluid expansion, as in the familiar*mercury in a glass tube; physical or chemical change as in certain heat- seneitive materials; the generation of a voltage. as in a thermoe couple; and a resistance change'with temperature, which is the basis for a resistance temperature transducer or thermistor. The following is a brief description of each. Thermometer. The read-out in this type of sensor is direct and visual. The difficulties of using a sensor of this type in measuring surface temperatures in thin plastic film.is apparent and will not be elaborated upon. Egg Sensitive Materials. more is a variety of this kind of sensor on the market. The mechanism of this type or transchlcer is the ability of the material to undergo a physical or chemical change at a fixed temperature. The material is applied to the source either from a crayon or brushed on in solution or suspension form. The solvmt is a volatile type which will evaporate loam-Lg a deposit of the thermally sensitive material on the surface whose temperature is to be measured. A color or consistency change is a common indication whm a particular temperature is sensed. These sensors appear accur- ate to within approximately one per cent of the indicated value in our brief laboratory checks . The primary disadvantage of this type of device when trying to measure the surface tmxperature of thin plastic films is the response time of the material. Since heat transfer is a time dependent quantity, there will exist a twperature leg between the surface of the film and the thermally sensitive material. In order for the themally sensitive material to respond, it first must be heated to its indicated tmperature. Where time is not a factor, this sort of indicator is quite accurate. However. when trying to make short time duration measurements. it is simply not suitable due to the slow response time, determined emerimentally to average approximately ten seconds. e .oocu sole . This is a tmxperature measuring device which might be used for thin film monuments. Depending on the themooouple, response times in the millisecond range are possible. These would be adequate for the measurenents preposed in thin plastic films which are of the order of one to five seconds duration corresponding to the typical heat sealing cycle. The read-out of a thermocouple is adequately displayed through the use of various read-out devices such as oscilloscopes. The primary disadvantage in using a thermocouple lies in knowing what the couple is actually sensing. A temperature change generates a small difference in potential when two dissimilar metals are in intimate contact. When the point of contact is at some different temperature than the input leads to the read-out device, it is possi- ble to detect the small amount of current which is generated in terms of temperature. The actual current generated is a function of the contact surface area whereas the change in'voltmge is a function of the temperature gradient that the entire contact area experiences (9). If this contact area is one mil thick as in a typical heat sealing situation, the voltage generated will be in proportion to the average temperature of the contact area. In a practical sense, a thermocouple one mil in diameter will measure the average temperature across a one mil film.since the pressure generated by the sealing device will imbed the thermocouple into the film. If the heated bar temperatures are BOOOF. and lOOOF., and assuming the surface temperature that is of interest is lOOOF., it is obvious that the recorded temperature will be in error since the thermocouple is no longer positioned on the surface of the film. In attempting to establish a more accurate surface temperature measuring system, the requirements are that the sensor'must be in contact only‘with the surface of interest and.yet have a small enough thermal mass to not greatly affect the temperature of the system. 7 Thermal-ore. In an effort to comply with the above restrictions, the use of a vacuum metallized aluminum coated film to serve as a thermistor is preposed. Such films are readily available from many commercial packaging film sources. The thickness of the almninum coating is approadmtely 0.001 mil or 1/1000 the thickness of the film whose surface temperature is of interest. In regard to these conditions then: what is the temperamre that the coating will see? Hopefully, there will be no thermal gradient across the aluminum coating, and the film must approach the surface temperature at both surfaces of the alumimzm coating. One indicative measure of this possiluilitr is the thermal diffus- ivity of the film. This is the measure of how m the temperature gradient will disappear through the material as opposed to the thermal conductivity which is a measure of how m heat a material will allow to pass through a given thickness and area under the influence of a temperature gradient or differential. Thermal diffusivity, in other words, is a measure of the speed of the thermal ”wave” that is started when a hot surface is brought into sudden contact with a cold material (6). In the preposed situation, the thermal wave through the aluminum should be much greater than through the film whose surface temperature is to be measured. This value for aluminum is approximately 3.27 ft.2/hr. and for the film, polystyrene for mile, it is 3.26 X lO-Bft.2/hr. (Refer Calculation B, Appendix 1.) The relative velocity of the thermal waves in these trio materials in identical situations show that of alum- inum to be about 1000 times that of polystyrene. This rough comparison appears to support the notion that an aluminum coating of 0.001 mil thickness on a one mil thin plastic film'would not significantly change the temperature distribution in the same film without the coating. A factor which may alter the theoretical situation in an exper- imental temperature measuring system is contact resistance. In some theoretical heat transfer calculations, this factor can be ignored but its effects can seriously'alter theoretical results in a prac- tical situation. Contact resistance is s term.defining the nature of the bond formed between the heat source and the fluid or solid specimen which contacts it. If a solid receives heat hy contacting a solid, it is almost impossible to exclude the presence of air from the interface. Contact resistance than defines the average thickness of this air layer at the interface of the two solids (6). In a heat sealing situation, the heat transfer is independent of the pressure (5). A certain.amcunt of pressure is required, however, to hold the filmw together and to approach the intimate contact neces- sary to minimize the imperfect heat transfer brought about by the contact resistance present. Read-out o! the prOpcsed film thermistor necessitates the use of some device to measure the change in electrical resistance with a change in temperature. The most direct and accurate technique is comparison methods based upon an electrical resistance bridge circuit. For some purposes, bridges are used strictly as comparison devices where, once the bridge is nulled, the value of the unknown area may be calculated by a simple ratio equation. In other cases, the bridges serve as a means to obtain a voltage roughly proportional to the deviation from the null condition. For design purposes, the condition for null must be known. To calculate deviation from the null, it is also necessary to know the bridge sensitivity either to the change in resistance or to the tmperature that caused the resistance change (1). Where the application of the bridge is temperature measurement in thin plastic films, the latter use of the bridge circuit fits the situation. III. THIN-FIB“! SURFACE TEI'EPERATURE MEASUREMENT EJJIPMEL‘ET The following is a description of the experimental equipment used in the temperature measurement systan for thin plastic films. Where necessary, the rationale behind the choice of equipment is explained. :13 earn: star. As indicated previously, the choice of the aluminum coated film, in this case one mil polystyrene, was largely based upon its apparently favorable themed. qualities in the preposed temperature-measuring application. The vacuum metallising process by which the coating is applied is known to hold impurities caused by oxidation during the coating process to a minimum besides maintaining a fairly close control on the thickness of the coating. Both of these factors are highly desirable in resistance temperature transducers. Additionally, it was hoped that the aluminum coating would provide s definite and linear relationship between electrical resistance and temperature over the potential heat sealing temperature range. linearity is desirable in that it simplifies the relationship between resistance and temperature, easing calibration. The linearity of the resistance transducer is an inherent function of the material modified only by the immritios present and some mechanical factors (9). Attempts were made to get coated polypropylene or polyethylene but the delivery time was prohibitive. The film used was available at the time of the study and was used primarily because of this availability. 11 It was a one mil metallized Dow Chemical Company polystyrene film, Qébldb, approndnately ten years old. The film was felt to be satisfactory for thermistor use since polystyrene is relatively less susceptible to deterioration than sme other types might be. The physical appearance of the film was excellent and the inherent prwsical properties of the film were assumed to be the same as lmotm polystyrene, although no specific information on this particular film could be obtained. The 1' 11m senqales used were eight inches by approximately one inch out on a Model JDC 25 Precision Sample Cutter made by Twins-Albert Instrument Comparw. The length and width of the individual samples was thus closely controlled. These dimensions showed approximately a ZO-ohn resistance per thermistor sample. ;;doe Cirmit. The bridge used was what might be tamed a half symetrical type. The two resistors in the upper half were equal requiring the two in the lower half to be equal at null. The effect of lead wires and variations in the lead wire resistance acted equally in both legs of the bridge so that the null was undisturbed. A variable rheostat was placed in series with the thermistor to null the bridge. The thermistor was made one am of the four-am resistance bridge. The rhecstat was then used to null the bridge by equalizing the resistance variations that might be present Iran one thermistor sample to the next. A practical condition had to be placed on the bridge circuit. This was the self-heating limitation within the transducer, which had to be held to a minimum to avoid errors in the temperature measurements . 12 The high thermal sensitivity of the film made it susceptible to resistance changes due to heating by the bridge current during the balancing procedure. 0n the other hand, the sensitivity of the bridge circuit varied directly with the bridge current so a balance had to be found between the two in final bridge design (1). The primary consideration appeared to be a maximizing of the circuit sensitivity so rough calculations of the heating effect were performed and are shown in Appendix 1, Calculation A (10). The final bridge circuit design was based upon these calculations. Some self heating was present within this final design according to the theoretical calculations but had negligible effect on the tmperature measurement data. The final design of the bridge circuit is shown in Appendix 2, Figure C. The bridge circuit power was supplied by a variable 30 volt - 225 nillianp DC power supply, Model 721A manufactured by Hewlett Packard. The 20 volts used for the bridge circuit provided a low enough voltage that no special precautionary measures on the emer- imental equipment were necessary. It was still high enough that the read-out noise caused by the antenna properties of the system lead wiring, thermistor, etc., were reduced to a workable level. The final system sensitivity achieved was approximately a 2 33111117011: change per one degree F. with 155 millimps bridge current. The resistance change in a typical thermistor was about one ohm in 20 with a loo-degree rise in temperature. This sensitivity seemed satisfactory for following the environmental temperature changes experienced by the film thermistors throughout the experimentation. 13 Typical resistance changes are shown for some randomly chosen film thermistors in Table A, Appendix 3. Care was exercised in selecting the bridge components with sufficient thermal mass and reactances to insure resistive stability. E232 M. The laboraton heat sealer was constructed from a design by Olin Mathieson Chmical Corporation (8). It is specif- ically for laboratory use and provides control of the three variables of sealing: tnuperature, (hell time, and pressure. The «fling unit is constructed in the fern of a small table. 8% inches 117 b inches, with an opening at the center large enough to admit the sealing bar. In its retracted position, the bar is one-half inch below the top of the table. An air piston actuates the upward trsvel of the bar. The sealing profile can be altered by replacing the dasountable profile head with . head ”mam-mg the desired profile. For this series of temperature measurements the profile head used was aluminum with a coating of approadnately one nil thick Teflon. This profile head provides a flat sealing surface one inch wide by 6% inches long. Heat is supplied to the profile head by a 150 watt heater mounted in the heating bar on which the profile head is mounted. The sealing pressure is controlled by dead weight so it is possible to obtain a variety of sealing pressures. The weight used was 16 pounds seven emcee. mounted on a base 1% inches by 6’} inches. For the one inch wide sealing bar, the pressure was calculated to be 0.68 pounds per square inch. Temperamre and dwell time were regulated respectively by means of a tanperature controller and an electric timer both mounted in a 1h separate console. Although the timer was not used during the final recommended measurement procedure, it was used during the initial temperature measurement procedure. For an accurate estimate of the temperature of the surface of the sealing bar, a thermocouple is located as close to the sealing surface as possible. The electric timer is started by means of a microswitch adjusted so that the seal- in; cycle is timed from the instant the sealing bar first touches the test specimen or in this case, the thermistor. The timer, in turn, spans and closes a solenoid valve to the small air piston which moves the sealing bar into position and back down at the end of the sealing cycle. The sealing table also contained the film thermistor clamps. These clamps were Acco Products paper clamps {125, to which banana plug electrical connections were added. The clamps were located in the heat sealer table in insulated positions at either end. These clamps proved to be somewhat troublesome during the ecqaerimentation as the electrical contact the clamps made with the thermistor alum- inum coating was often imperfect. A Simpson Multimeter was used to check the initial resistance of the then’fisixrr samples to insure good electrical contact. Goggles ppc. The final major piece of equipment was a Tektronix Type 561+ Storage Oscilloscope. The scope display was used as a read-out device for the voltage changes detected by the bridge circuit in the themistor as the thermistor resistance changed with increasing tsunaeratures in the heat sealer. 15 A photograph of the assembled temperature measuring eyetan for thin plutio films is shown in Figure l. The small black box in the foreground contains the resistance bridge excepting the film ther- mistor which is located on the heat sealer table suspended between the clmupo. The rheostet control can be noted on this same black box. The large black console in the background is the Model JP temperature control unit by West Instrument Corporation. The teuqaere‘mre dial visible on this console is in Fahrenheit degrees st 5 degree increments with each major division equal to 25°F. m temperature readings throughout the experimental procedure are taken from this dial to interpret the voltage reed-outs from the scope display. This console also contains a. conventional heat sealer timer control. A poweretat control located in the lower left-hand corner is used to control the rate of bar heating in the heat sealer. Notice the themistor in poeition in the clamp: on the heat sealer table. IV. EXPERIMENTAL PROCEDURE AND ANALYSIS In attempting to find an experimental procedure to measure temperature at the surface of thin plastic films, the following conditions appeared to be of paramount importance. First, the system should detect only changes in enviromental temperatures. Second, he results should be reproducible in a practical manner with relatively unsophisticated equipment and procedures. With a few basic conditions recommended in the werimental environment, the resultant procedure should be versatile in that duplication of the results can be easily attained in any similar laboratory situation. Finally, the experimental data should be related to a computer program suitable for simulation of the temperature measurement system in thin plastic films to promote future stucbr. To meet the foremontioned conditions, two basic emerimental procedures evolved, one from the other. The initial procedure will be briefly explained with the pertinent experimental data that dictated changes in procedure. The final temperature measurement procedure will then be outlined with the morinental results of the procechre sum- marized. General. Erocefitrfi Both procedures involved the use of the aluminum coated poly- styrene strip as a thermistor. Basically, the thermistor was made one arm of a four—arm resistance bridge. This involved clappiz'zg the film thermistor into the clamps on the heat sealer table. The 18 reomirxler of the circuit including the rheostat was located in a circuit box. An ini ial scope diaplay base line was used with bridge power off. This base line on them naillivclt per centimeter scale should be at the bottom scaleline on the scope face so that the upper- most tenperature changes can be recorded without switching to the next higher scope scale. Using any lower scale made the signal noise identified as 60 cycle interference) excessive. A precise balance position was achieved by adjusting the resistance of the variable rheostat in the adjacent am of the bridge. This balance condition, observed on the oscilloscope display, was used as a base line for the subsequent readings. It win subsequently be referred to as ”tb" and represents the null condition of the initial environmental temperature or in this situation, the room temperature. All weflmentation was conducted under non-controlled environmental conditions. The procedure for balancing the bridge deserves flurther clarifica- tion. 'Ihroughout the experimentation, a ten-turn 50 ohm potentiometer wired as a variable rheostat was used. Even with the fine tuning capabilities of this rheostat, at the upper temperature ranges during the experimen- tation, setting tb involved an averaging of the heated bar temperature cycle. If a certain bar temperature was desired, the setting was made on the temperature scale and the control unit then attempted to maintain that temperature with the thermostat. Nevertheless, enough cycling did occur that setting tb often involved considerable judgnent and a definite "feel” for the system. The rheostat control in the bridge circuit gave approximately a ten millivolt change in the scope displq for each of 19 the smallest divisions on the control knob. A ten-turn 25 ohm potentiometer with a 10 ohm resistor wired in series would provide somewhat more control. This problem with the initial setting of tb, milling the bridge, should be noted but it presented no real diffi- culty when the capabilities and limitations of the system in this regard were mom. This setting detemines all subseqzent system accuracy so its importance cannot be overemphasized. The entire bridge balancing procedure must take place with the themistcr in position in the clamps to cmplete the brifie circuit. Since the sealer bar was at some temperature over room temperature depending on the temperature control setting, some method of isolating the thermistor from the heated bar during the balancing process had to be found; the tb base line at null should represent only the initial room temperature environment. This isolation was successfully accom- plished using a wooden spacer out from 2 inch by h inch wood stock. The spacer was checked on a sample film thermistor in bridge balance and there was no significant change in the tb scope diaplay after appromately 1&8 hours with bridge power on. Figure A, Appendix 2, illustrates the thermistor in the balancing position. In addition, it indicates that resistive heating in the thermistor was at a negligible level. Throughout the emerimental procedure, a l/lé inch commercial neoprene insulation was used to insulate the aluminum coating from the metal weight and prevent a short circuit during the timing procedure. men thwas set, the wooden isolation block was removed and the insulation and weight were placed on the thermistor sample. The heat 20 sealing bar'was then raised and a data record taken. This visual scope display'represented the bridge unbalance signal due to thermistor heat- ing and the voltage drop across the thermistor due to the increase in resistance as the temperature increases. In the experimental procedures used, the read-cut from the scope display indicated an increase from the base line. This display appeared more conventional for the measurement purposes although it should.be noted that the display actually represented a decreasing voltage as the resistance of the thermistor increased. Calibration. During calibration, the primary objective was to sense the absolute changes in temperature of the heated jaw as voltage changes on the scope display for later interpretation into corresponding temperatures. With this in mind, all calibration of the thermistor 'was accomplished with the aluminum coating side of the thermistor in direct contact with the Teflon coated heated bar. The timing solenoid was removed from the heat sealer control system and was replaced.hy a simple air valve. A relief valve to lower the heat sealer bar was provided and 15 PSIG air pressure was used throughout the procedures. Iégggg, Although no actual heat sealing was to be accomplished during the experimentation, the temperature of interest throughout the measurement procedures was at the location on the surface of the thermistor that'would represent the interface of two films in the normal heat sealing operation. This, then, was the desired location for the aluminum.coating so throughout the timing measurements of temperature, the aluminum coating was reversed from the calibration 21 procedures. In this way the temperature at any given time could be sensed at the surface of the film where the heat seal would nomally occur. Figure 13, Appendix 2, illustrates the themietor in position for either calibration or timing with the one difference that the position of the slumimm coating varied as pre'vioue'fqr noted. Iritig Procedure In this procedure, the basic approach was that of detemining an optimum sample size with the samples in this case being the film thermistors. An initial sample size of five was arbitrarily chosen. 0-311er :3“. 2., . During calibration each sample was clamped in turn following the general procedures. A data record was made for each sample and data was recorded from the scope display for temperature settings ranging from 100°F. to 225°F. in 25 degree moments. The average of these results was plotted; temperature (from the temperature control setting) versus the scope display reading in millivolts corres- ponding to this temperature. W. wring timing, the same sample reclamping procedure was used as in calibration following the general procedures but in this case, the bar temperature was set first at s constant 200°F. and then at 175°F. The timer and solenoid were used to control the heat sealer her action. The timer range used was 0.5 to 2.5 seconds in increments of 0.25 seconds. The average of these voltage readings was converted to temperatures from the plot of the average calibration data. 22 M333 rsis. The calibration curve indicated a linear relationship existed betseen the electrical resistance and thermal properties of the film thermistor in the temperature range checked. Marv problems were apparent from this initial test procedure. (a) Althwgh the average plot of the samples appeared linear, the variation from this average value appeared significant. There also appeared to be positive correlation between initial resistance as checked by the Simpson voltmeter and the voltage readings. Higher initial resistance resulted in higher voltage readings and vice versa. This indicated that the vacuum metalizing process for coating the film did not maintain as tight control over coating thickness as originally thmght. Variations of up to 8 class had been noticed in initial resistance with the voltmeter. With this apparent correlation, two themistors with opposite extremes of initial resistance were calibrated and timed according to the initial procedure. The results are shown in Table B, Appendix 3. These results indicate that the average plots used in calibration of the five Maples had not eliminated any errors in clamping or technique. In fact, the results sewed to indicate that even small differences in initial sample resistance might affect the data read-«mt significantly. (b) The contact resistance in the system was another area that might possibly have affected the readings. Since this is representative of the thickness of the average air layer in the heat transfer system, it might be reduced by adding more pressure in the tons of additional weights to the system. The results are shown in Table C, Appendix 3, and indicate that additional weight over the normal weight used with the heat sealer would not affect the system data. 23 (c) Another phenomenon'was noted in this initial procedure. It appeared that a sample inadvertently stretched when in position with the weight in place for either calibration or timing experienced some stress that could affect the read-out scope display. To check this, samples were intentionally stressed by'suspending the thermistor in the«flamps while supporting the weight. subsequent readings indicated that this was indeed a factor. The scope display'for the typical sample in this condition indicated lower'voltage than would be expected initially and as the sample heated. the display tends to rise back toward a normal non-stressed reading. The stress and resistance change appeared to act in opposite directions'with tempera- ture change. As the temperature went up, the stresses were relieved within the thermistor and the resistance increase then dominated the scope read-out. By exercising care in clamping the thermistor in the bridge circuit, no stress effect'was noted. (d) The thermistcrs used in the initial procedure showed drastic physical deformation at ZOOOF. after approximately 2.5 seconds. This in turn radically affected the scope display as the electrical conductive ity of the thermistor‘broke down. The thermistors tested at 175°F. were satisfactory after relatively long lengths of time. (a) The initial environmental conditions including temperature and convective air flow affected the bridge balance condition and the tb setting. This had an obvious subsequent effect on the data read-out. Fina; ocedure This procedure is recommended for any duplication efforts of the surface temperature measurement apparatus for thin plastic films. It 24 is based upon the changes dictated by the effects observed in the initial procedure. The most significant change is the use of one thermistor for both calibration and timing. The resultant curves and data will apply only to that particular thermistor, but it is believed that the errors apparently caused by variations in initial resistance probably due to variable aluminum coating thiclmesees will be eliminated using this procedure. Although not used in this final procedure, it :3th be recog- nized that azv further experimentation should be conducted under closely controlled enviromental conditions. Control of temperature and convective air movement should be given special attention. ‘ Calibratim. . 1) The film thermistor should be positioned following the general procedures outlined on pages 17 through 21. 2) The calibration run should start with the heat sealing bar raised. The calibration cycle should begin with the bar at 175°F. 3) Allow the scope display to stabilize and record the scope voltage read-out in millivolts . 1+) hm the power OFF to the heat sealer. 5) Record the scope display readings at 5 degree increments read from the temperature control scale down through 125°F. as the heat sealer bar cools. 6) Tun the heat sealer control unit ON at Just below a 125°F. indication and with the unit powerstat set at a slow rate of 50, reheat the heater bar. 25 7) Record the scope display'during the heating cycle at the same 5 degree increments through l?5°F. The two sets of readings will be the same if the cooling and.hesting rates are equal as desired. If not equal, the average readings can.be used if the differences are minor. 8) Lower the heat sealer bar and remove the sample. 9) The calibration data is complete (total time should be less than 30 minutes). Egaflgg. NOTE: This should be done immediately following the calibration procedure to minimize changes in environmental conditions. 1) The same film.thermistor should.be positioned following the general procedures. The heat sealer timer‘unit is not used. The air valve assembly'used in the calibration procedure is used with no changes necessary; The hot bar temperature remains at 1750F. 2) Using the storage features of the oscilloscope, trigger the pulse using the 0.5 second per cm sweep time on the scope. 3) Immediately raise the heat sealer bar. b) The storage feature of the oscilloscope will store the data read-out through approximately 0.5 to 5.0 seconds which coincides with a typical range of heat sealing times. 5) Lower the sealer bar and remove the sample. 6) The scope voltage data should then be transcribed at 0.25 second intervals to'be interpreted into temperatures from the plot of the calibration data. 7) The timed data is complete (total time should.be less than one minute). 26 Am "veie. This procedure satisfactorily accomplishes temper- ature measurement at the surfsce of thin plastic films. The calibration results for a typical thermistor are shown in Tobie l and plotted in Figure 2. The timing results from the oscilloscope picture of the timing reed-out shown in Figure 3 are tabulsted in Table 2. This data is then plotted in Figure h to show the temperature- time profile of the thermistor. The temperatures in Table 2 were found by converting the timed scope voltages from Figure 3 to temperatures from the calibration curve in Figure 2. Calibration sensitivity appeared to be approx» imately 2 millivolts/degree in the final procedure. The 31 millivolt indication in the timing read-out picture, Figure 3, before the bar was raised at the first vertical scale line should be noted. This 31 millivolt reading reflects the approximate 1.15 degree environmental temperature eaqaerienced by the thermistor before the heat sealer‘bar was raised. It was the temperature experienced by the thermistor when positioned as in Figure B, Appendix.2. The base line or tb setting is shown on the first horizontal scale line of Figure 3. The'minor disturbance noted in Figure 3 after approximately 1.5 seconds is noise detected by the systan when the heated bar temperature control thermostat cycled in attempting to maintain a constant hot bar tanperature. It should be disregarded in interpreting the results of the timing procedure. 27 Note that the signal in Figure 3 appears approximately 1.5 millivoltl wide. No further resolution of this signal was possible using the storage feature of the oscilloscope for read-out of the timing data due to the 60 cycle noise included in the signal. All data for interpretation from Figure 3 was taken from the center of this trace and no attempt'wae made to road the data to less than i'millivclt because of this lack of resolution. Some indication of the accuracy'of the calibration results ‘wae obtained.by totalling the differences between actual calibration readings from Table l and final calibration curve readings from Figure 2. The total.wae averaged and the results indicated that an accuraqy of t 10F. can be expected during the final procedure calibration. TABLE 1 - Experimental Calibration Data ====================E====anII=3IIIHIIIIIIIIIIIIIEEE=:*— 3‘“ Reading (mv) Cooling Heating Temperature (0F.) 1.75 51.0 51.0 170 09.5 18.0 165 148.0 158.0 160 07.0 two 155 05.0 05.0 150 03.0 03.0 105 02.0 u2.0 10.0 no. 5 1.0.0 135 38.0 38-5 130 37.0 37.0 125 35.0 35-0 NOTE: Initial Resistance - 19.5 ohms, 0.05 potentiometer reading TABLE 2 - Experimental Timing Data Time (seconds) Reading (millivolte) Temperature (°F.) 0.1 37.5 132.5 0.2 39.0 137.0 0.3 40.0 100.0 0.4 41.0 103.0 0.5 01.5 100.5 0075 1‘3 .0 116 .0 1.0 44.0 152.0 1.25 45.0 155.0 1.5 45.5 157.0 1.75 “6.5 160.0 2.0 07.0 161.5 2.5 07.5 163.0 3.0 “8.0 160.5 3.5 “8.5 166.0 0.0 49.0 167 .5 moao>aaeez a: m: m: H: on mm mm L p . p — P p u . - a u - .Mpmm coapmhnfiamo Hmpcmeflamdxm t m mpzmflm mma oma mma 02H mza omH mma owa mod Qua .mo l 0 19‘ 1.... I“. 04!. 4 II. .I!. \ OVA) u. o I l O n. \. . .l . x . If I . .p. ( f y I (D I o II P: _w a 40! l\ m I '1‘ 0: '\ - ‘ | 0 — III..I-III I'lli I. 'Illlll‘l_|l "II‘ trob(|lt 31 O. z m~.m museuom m.m mm.m o.m mw.m m.m mm.m oo.m mw.a m.H mm.a o.H h b _ _ _ p . . p _ _ _ . . a _ q . _ . .mascmooam Hecam mo mpasmom Amocoeflamdxm L 2 mmzmflm _ _ mm.o m.o mm.o — u . _ _ l ow om OOH. 0.: oma omH 03H oma owa Qua .ho v. TE-‘IFERATURE msumcm MODELS The final procedure appears to fulfill the conditions of tamerature measurement and reproducibilit . In addition, two sttanpts were made to simulate the experimental system on the Michigan State University 3600 Control Data Corporation digital camputar. marrow 5/3600 is the language used in the simulations. Computer models can provide a much more flexible method of experimentation than laboratory procedures. However, the model is use- ful in a practical sense only when it duplicates the behavior of the real world system. The testing or validation of a model can be done by making further observations and measurements of the system or by eJqJerimentation. It was this lack of aperimental validity in Kavesh's article that initiated the development of the surface temperature measurement procedure for thin plastic films previously described (it). Using the results of the ceremonial procedure, an attempt was made to validate a computer model that would simulate the experimental system. This would meet the third of the conditions previously outlined and provide a valuable tool for future study of the heat transfer process in heat sealing. Both programs are mathematical models of heat transfer systems which attempt to describe in two different ways with a set of math- matical eoqoressicns the experimental. surface temperature measurement system. These mathematical moaressions have been rewritten according to computer language rules . The mathematical equations pertinent to both models (with comparative data) are listed at the end of this section and will be referred to throughout the following discussion. 33 A complete program listing for both models is included in Appendix 14. The tauperat‘ure range used for comparison is 80°F. to 175°F. Mot-EL ; This model was an adaptation of the program used by Kaveeh (it). The theoretical data derived from its use prompted the experimental investigation reported previously. Discussion. Assume a slab of infinite thickness comes into sudden contact with a hot surface at a twperature Ts' We want to know how far and how fast heat will penetrate this material. We can find the temperature (t) at any depth (3:) and at any time (t) by solving Equation 1. It is possible to use this same equation going from a theoretical infinite thickness in a slab of film to a film of finite thiclmess (L), if the film is properly backed with insulation (6). The temperature- distance distribution for a short period after the heat has been applied will be nearly the same as for the infinite slab. This will occur if the same heat flow is removed from the remote face of the finite film thickness as would ordinarily flow through a plane in an infinite film at the same distance from the hot surface. The finite thickness of film must then be backed with a quantity of insulation that will remove this amount of heat flow. The quantity of insulation required can be found by equating the heat flow at a distance x '- L from the surface of an infinite body to the heat carried away through the insulation on a finite body of thickness (L). Heat flow at a distance x 3 L from the surface is given by Equation 2 (6). 34 Kavesh detemfined that by use of Equation 3, the number of inches of insulation required in a film of finite thickness could be found. This thickness of insulation should then duplicate the temperature profile of the infinite slab of film in films of finite thiclmess for short time periods (6). Equation 1} (which was derived from equation 3, using the data generated by the computer program for heat in a. slab of infinite thickness) was applied to determine the thiclmass of neoprene inculc- tion required for the unheated platen in the camerimental apparatus (A). The calculated thickness was less than the thicknesses corenerciauy svsileble so 1/16 inch was used. However, Kaveeh recognized the fact that other factors such as the type film or the temperature differential might change so the factor of 0.213 used in Equation h is designed to approadmste the maximum thickness of insulation that could possibly be needed to duplicate the infinite slab temperature profile. Knesh calculated this factor based on Model 1 temperatures generated after 1 second. m. iodel 1 inputs are the thermal conductivity, heat capacity and density of the film (polystyrene in this case) . Equation 1 is an indefinite integral. so to be solved by the computer by numerical means. I numerical appronmtion using an error hmction was used just as Kevesh had done (2). £1222 513. The computer print out of this model is in a tabulated form with tmemtures for various times, film thicknesses and heated bar temperatures . 35 Ana‘fl‘ is . The experimental surface tmperature measuranent system employed an excess of insulation over Equation 3 so the werimental tmperaimre profile should be considerably lower than the infinite solid profile from Model 1. Kavesh's inmzlation thickness calculations were based on Model 1 film tanperatures after one second (h) . With increasing time, it is probable that a heat build-up timid occur within a given insulation thickness so it would cease to perform like the infinite solid model. This could also explain the marinental variations from Model 1 data. In addition, no contact resistance is seemed in the infinite solid program calculations so this was another source of variation in comparative results .. This factor should be programmed into the infinite solid model to mks an equitable comparison of data possible . This model was adapted from a program written by Dr. James Beck of the Michigan State University Mechanical Ehgineering Department (11) . Dr. Beck had used the program in marry theoretical heat transfer situations but its use in the simulation of temperature profiles in thin-film mtws had not been attempted previously. Discuseion. The model uses the Crank-Nicholson approximation for the solution of finite difference equations of heat conduction. This method uses the approximation constant, ETA, which equals 0.5. This appro’sdmaticn is an implicit method and requires that all the equticns for the unlmown temperatures at time (m + 1) be solved simultaneously using the known tmmeramres at time (m) . The method 36 has the advantage of being stable under’mcst conditions (3). The difference equations referred to above are developed using the "heat balance" approach.uhere: heat in - heat out 3 rate of increase of internal energy Equations 5, 6 and 7 follow this general format. Equation 5 is the equation for the temperature calculations at the general interior nodes (11). This equation solves for the temperature coefficients at node (n) in Figure 5, where the initial conditions at time (m.- 1) are known. The temperature at point (n) at time (n + l) is the average of the temperatures at either side of point (n) at the previous time. Equation 6 is the special interface equation, in.this case solving for the temperature coefficients at Node 2 in Figure 6 (11). The special interface node calculations include a contact resistance (h) at the interface. The equation is written in the same heat balance form.but in this case the Crank~Nicholson approximation constant (ETA) could not be used for the "heat out" term in the balance equation but had to be replaced by h(T: - T3). This was due to oscillations noted in initial model runs using ETA.when the contact resistance (h) approached infinity (no air layer). The additional approximation constant, LEMBDA, appearing in this equation and equal to 0.75, is used for the interface boundary calculations. To solve for the other interface Node 3 the equation subscripts can‘be reversed, switching subscripts: 2 and 3, l snd.h, 3 and 2. The result is Equation 7, identical to Equation 6 except for the above subscript changes (ll). 3? The choice of the number of space nodes and the size of the time steps is decided on the basis of experience with the type of problm at hand. Nodes can be reduced in a material that is of little interest or where no tmperamreutime gradient appears to cadet. As the time steps and node spacing are simltanemsly reduced. greater accuracy is obtained. Equations 5, 6 and 7 compute the coefficients necessary to complete the finite-difference equations of heat conduction. This is accauplished in SUBROUTINE 0031‘s and the finite-difference equations resulting are solved for the various tarperature-time-node locations throughout the system in SUBROUTINE TRIDI (11). 21323.3. The input data required is the tunnel conductivity, the product of heat capacity and density of each material, and the time steps and nodes desired in each material. The thickness of each material or region in the system is also necessary. The interface location must be specified and this acts as the location for the contact resistance specified and also as the temperature division point of the system Printout. The Model 2 printout consists of the complete temperature profile of the system for various times according to the nodes selected for all system materials. This requires a previous knowledge of node spacing so the temper-emu at the desired location in the systul can be located. Analysis . Using Model 2 data, attempts were made to detemne a contact resistance that would provide close agreement with the 38 momenta. surface tmzperature data. The results can be seen from Table 3 and Figure 7. The data agrees very closely assuming a contact resistance of 700 BTU/hr. - ft.2-°F./ft. equal to an air layer of approximately 0.2 mils calculated Iran Equation 8. The initial portion of the Model 2 temperature profile proved most sensitive to contact resistance changes during the comparison trials so more effort was made to make the Model 2 data and outper- immtal data closely agree in this area. Contact resistance to the nearest 50 units was considered satisfactory for comparison and proof of agreement. The resultant 0.2 mil air layer value sewed "reasonable” for the emerimental system. Any physical measurement of this contact resistance value in the emerimental system was considered unnecessary if not impossible. Figure 7 then indicates that Model 2 can be used for future eXpez-imentation based upon its validation by the eaqzserimental results. It should be noted that the validation is restricted to the eaqoerimental time- temperature range, although there is no reason to believe it would not also be applicable to higher turperatures. The minor differences between Model 2 and the mmerimental data in this figure could be for several reasons. One possible reason is that the average values of the thermal properties for the various materials in the emerimental system used for Model 2 inputs may be in error. Furthermore, variations of these properties with increasing temperatures xrdght have some effect. Probably the main reason is that the difference depends upon the value of contact resistance used in 39 the Model 2 program. Further agreement might have been shown if values to less than the previously indicated 50 units had been checked but this was considered unnecessary for comparative purposes. The maximum difference of 30F. is considered negligible for a practical heat sealing application and for future surface temperature measurement emperimentation involving Model 2. However, the differences emphasize the necessity of accurately describing the practical system in the couputer simulation to minimize possible errors when using Model 2. The experimental validation of this Model for use in thin-film surface temperature measurements is clearly evident in Figure 7 despite these differences. Model 1 Equations EQUATION 1 z = X txfi = TS + (tO - TS) 2 ‘f 2"a“ _Z2 /F' o e dz where: tO is initial body temperature s is hot surface temperature a is thermal diffusivity of the film (a = K/Cpp) Q is time after contact 9 is temperature at distance x, time U K is thermal conductivity C is heat capacity P p is density EQUATION 2 Q -L2/uag _ e A - K(Ts - to) Vflafl where: % is heat flow per area per unit time EQUATION 3 L1 = KAT Q/A where: L1 is insulation thickness in inches K is the insulation thermal conductivity (BTU/hr—ft.2 - °F./ft.) AT is the temperature change (°F.) EQUATION A L1 = 0.213K hi Model 2 Figures and Eguations NOTE: All equations follow the "heat balance” approach format where: heat in — heat out = rate of increase in internal energy FIGURE 5 - General Interior Nodes K m m K m— n Ax) Tn-l _ Tn)+ H FJH I a s s a ,'_. v At is time m is time that temperature is evaluated T is temperature Ax is equal node spacing n is node point is Crank-Nicholson approximation constant (ETA) J is density of material is heat capacity of material 7:00 is thermal conductivity of material AZ Model 2 Figures and Equations continued FIGURE 6 — Special Interface Nodes l 2 3 9 K1 h K2 +Ax+ EQUATION 6 n K1 (TT - T?) + (1 - n) K1 (TT‘l - Tg-l> Ax Ax l l m m —h (T2 - T3) = A (Dc) Ax m m-l m m-l l 21 T2 - T2 + (l—Al) (pc) Axl T1 - Tl At 2 At collecting terms: 2 m 2 m C (Ax ) T1 [2nKl-A2 Cl(Axl) ] - T2 [2nKl + 2hAxl + Al 1 1 ] —-———-—- At At m + T3 [2hAx1] m-l C (Ax )2 m-l C (Ax )2 = T [28K - A 1 ‘1 ] - T [28K = A .1 1 J l l 2 ———————— 2 l l———-——-— At At where: C is 90 B is n - 1.0 Al is boundary approximation constant h is contact resistance A2 is 1.0 — Al is thermal conductivity of the material EQUATION 7 NOTE: Let 2——»3, 3——+2, 1——»u. A1C2(AX1)2 2 + 2hAX2 + At ] m A m T2 [dhAXZJ - T3 [2nK 2 _ A202(Ax2) ] 2 At + TumE2UK 2 2 CA(AX ) A C (Axn) _ m-l A 2 2 m-l 1 2 2 EQUATION 8 h = k/x where: h is contact resistance air k is thermal conductivity air X is thickness of air layer TABLE 3 - Comparative Data: Experimental, Model 1, Model 2 ‘ Tumperaturo (0F.) Time (seconds) Experimental Model 1 Model 2 0.1 132.5 160.29 12h.60 0.2 137.0 160.57 135.h5 0.3 140.5 130.80 0.“ 1&3.5 1&4.25 0.5 105.0 168.96 (0.6) 1&6.75 0 075 1:49 05 15o 075 1.0 152.5 170.32 153.07 1.25 155.5 155.32 1.5 157.0 171.05 (1.h) 156.71 1.75 159.5 171.51 (1.8) 157.85 2.0 161.0 158.69 2.5 163.5 171.84 (2.2) 160.06 3.0 165.0 161.07 3.5 166.5 161.8% “.0 168.0 162.h6 mocooom 0.: Elm m.m mm.m o.m mn.m m.m mm.m oo.m 34 m4 mmé 0% mafia m.o mm.o lb! - n p n p P . . f u n s n p p — . . . . . - q . _ q . . . . . . om Haucoefipmoxm m. lllll'll' m Hoooz o lllll'. H Hocoz OOH . . . _ o: _ . . oma .m Hoooz .H Hoooz .HmucoEHLmme “flmemQEoo L m opzwflm . . oma _ _1 IIIIIIIIuIIIIII'IIIunII.Iu \\ .... 00H. ' | I. 'lll'l'nllll'l"|u\ .ll ONH .mo VI. COHCIUSIONS AND RBCCM'IEEATIONS FOR FUTURE STUDY The linear relationship between the electrical and theme]. properties of the aluminum coated film thermistor as the environ- mental temperature is increased has been shown. The sensitivity of the experimental systan is satisfactory to detect changes in this envircmnental temperature . It can be concluded that temperature monument at the surface or thin plutic filxns can be accomplished momentarily by the use of the final procedure described in this study. The experimental procedure can be easily duplicated using relatively unsophisticated experimental equipment. However, scare limitations on the procedure are apparent when attempting to duplicate maximum a particular set of heat sealing condi- tions. The film involved in the heat sealing process and the alumirmm coated film used as the thermistor must be identical for accurate couparative data. Each film thermistor must be calibrated individually and any subsequent surface time-twpera‘bire measurements must be performed using this same tmperature calibrated thermistor. If extensive data is sought. deterioration of the film thermistor midway through the experimentation is possible so that recalibration of a new themistcr might be necessary with a possible loss at valuable information. These limitations to the experimental flexibility of the final tmperature measurement procedure can be eliminated using a computer model of the surface tmperature measurement system. “7 A comparison of Model 2 data and the momenta]. data showed a clear agreement that indicates Model 2 can be used to simlate any heat seeding process flth none of the limitations of the experimmtal procedure. Satisfactory surface temperature data for thin plastic films can be obtained using Model 2 if care is exercised in describing the practical heat seeding conditions to be simulated by the computer. The experimental fleadbility offered by I-Iodel 2 should provide the packaging engineer with an additional tool to investigate the heat sealing cycle. Contact resistance in a practical or experimental system is an uncertain area in surface temperature monument. The 0.2 mil air layer in the eyqaefimental system determined using Model 2 data could perhaps be considered representative in a similar systm. However, it should be noted that definition of a contact resistance in a specific heat transfer system in terms of air layer thickness is extremely difficult. Additional work using the final temperature measurement procedure should be accomplished on a plastic film of 1mm thermal properties and a higher melting point. In this way, a wider temperature range to fully encompass the actual heat sealing range normally mommtcred can be investigated. This will extend the known applicability of the tanperature measurement procedure for thin plastic film. Accordingly, Model 2 should be attended in its application. Model 1 may still have valid applicati on in surface temp erature measurement experimentation if reprogrmng is accomplished to make an equitable comparison of data possible for eocpcrimental validation. ’48 Only through a complete mfiemtanding of the elements of the heat sealing cycle can satisfactory production control of package (:1er be assured. Hopemlly, the results of this shady will aid in developing this understanding. 1. 2. 3. 9. 10. LIST OF REMCES “Bridge Techniques in Temperature Measurement," Application W #1, NOVA—NETICS corporation, 1963, p. 3. Cerelaw, H. S. and J. C. Jaeger, Conduction 9; Heat in Solids. Oxford University Press, 2nd Edition 1959, p. 130. Foreythe, G. E. and W. R. Waeow, Waite-Difference Methodg for Partial Differential Erpmtione, John Wiley and Sons, New York, 793, p. U). Keveeh, Sheldon, "Heat Sealers Need Insulation-Polypropylene Study Shows," PACKAGE Engineering, 9: 125-M, October, 1961+. Knveeh, Sheldon and Robert J. Ridgwny, ”Heat Sealing-The Filn'e Opinion vs. The Machine“ Position,” fig W, Packaging Professional Edition, 3:10-15, Winter, 196% Kern, Donald 0., from” Heat Transfer, New York, MoGrew-Hill Book Company, 1956, pp. 9, 11, 61.1.5, Molviillan, Claude and Richard F. Gonzalez, Systems Agni-3:13- 5 Carroter grown 1:2 Boision motile, Homewood, Illinois, Richard D. Irwin, Inc., 1965, pp. 7-8. Ninnemann, K. W., "An Improved Laboratory Heat Sealer," Modern P80k in , 313171.5g Novambor’ 1957. "Resistance Tanperature Transducers,” Application Billetin f2. NOVA-NETICS Corporation, 1963, p. 2. White, Harvey E., Modern College stics, New Jersey, D. VanNoetrand Gunpany, Inc., Third Edition, 1956, p. 52“- Pereonal Conferenoee with Dr. James Beck, Mechanical Engineering Department, Michigan State University, East Lansing, Michigan. APPENDIX 1 - Calculations 51 CALCULATION A - Bridge Circuit Thermistor Theoretical Self— Heating Effect Data: 0.225 cal./gram @ 100°C specific heat of —4 aluminum on thermistor 3'55 X 10 (grams) Solution: 1. 0.225 x (3.55 x 10‘“) = 8.0 x 10'5 calories (Necessary to Raise Thermistor 1°C) _ 2 2 2. H — RI t = 20 X I X 20 = 8 O x 10-5 n.18 M.IE ' I = 2.9 milliamps (Bridge Current to Raise the Thermistor Temperature 1°C in 20 seconds) where: is calories is ohms electrical resistance is amperes of current is time in seconds dH’IJf—E CALCULATION B — Thermal Diffusivity of Polystyrene Solution: 8 5 = 3.26 x lO_3ft.2/hr. ll 0 C‘\ .O O. J:- l\_) a is thermal diffusivity of polystyrene is thermal conductivity of polystyrene C is heat capacity of polystyrene p is density of polystyrene 52 53 FIGURE A - Thermistor in Clamps with Wood Spacer in Place to Null Bridge (set tb). FIGURE B _ Thermistor in Clamps with Unheated Platen in Place for Experimental Timing or Calibration (heated bar in retracted position). FIGURE C - Bridge Circuit Design for Surface Temperature Measurement in Thin Plastic Films. 2R9Vbufls flan/57' one 1——R540 Our where: f? == 250 ohms (5 watts) APPENDIX 3 - Tables 55 TABLE A - Thermistor Electrical Resistance Change with Increase in Temperature Sample f Initial Resistance (75°F.) Final Resistance (200°F.) 2 14.2 1h.9 3 lu.6 15.“ b lu.0 14.8 5 18.3 19.2 TABLE B - Thermistor Calibration Results with Initial Resistance at Sample Extremes ‘Reading_fmilllv61ts}' Ismperature (“F.) Sample 1 - 22 ohms Sample 2 - 14.0 ohms 170 79 6b 165 76 60 160 73 58 155 70 56 150 67 55 1&5 6“ 53 Ibo 61 58 135 59 M 130 57 ‘ El 125 53 38 TABLE C - Results of Added Weight During Calibration at 175°F. (Sample removed and reclamped each time) W Reading (millivolts) Sample # Normal (h 1b. 7 oz.) 6 1b. 5 oz. 8 lb. 5 oz. 18 lb. 5 oz. 1 £15 1:5 as 138 2 b1 40 A1 41 3 49 ms to to NOTE: The minor variations in the readings were caused by the reclamping procedures. TABLE D - Thermal Influence of the Aluminum Coating on the Normal Polystyrene Film.Temperature Profile from Model 2 Data Temperature (0F.) Time (seconds) With Coating Without Coating 1.0 153.18 153 .143 1'5 156070 156071 2.0 158 .69 158 .69 2-5 160.05 160.06 3.0 161.07 161.07 3-5 161 .81» 161.85 4-0 162M 162 .A6 APPENDIX 1‘ - Models 'J0804400790TEMP030AKERSeBRIANeLeGROUP A .FTNOLOXO* ' PROGRAM FILM C MODEL 1 OF EXPERIMENTAL TEMPERATURE MEASUREMENT SYSTEM C PROGRAM TO_CALCULATE FILM SURFACE TEMP FOR NO CONTACT RESISTANCE. WELL C INSULATEDe UNHEATED PLATEN C POLYPROPYLENE PROGRAM HAS ALPHA = 00001573 C POLYSTYRENE PROGRAM HAS ALPHA 3 T0 00001311 'DIMENSION X(10)0 THETA(1110 T(10)0 TEMPIIO) PRINT 300 300 FORMAT (1H1*OTEMP = 80*5X*ALPHA = 00001311*//) PRINT 1 ’ 1 FORMAT (/3HHOT04OX027HFILM TEMP AT GIVEN DISTANCEe/07HSURFAC504X07 1HCONTACT025X041HFROM HOT SURFACE (COLD SURFACE INSULATED)0/04HTEMP 207X04HT1M5040X06H(TEMP19/03HIT108X07H(THETA)0/09HDEGREES F03X03HSE 3C06X05H0000503X05H0001003X05H0001503X05H0002003X05H0002503XO5H0003 4003X05H0003503X05H0004003X05H0004503X05H000501 READ 20 OTEMP 2 FORMAT (1F1000) READ 30 ALPHA C ALPHA = K/C TIMES RHO 3 FORMAT (IF1007) READ 40 (X1110 1:1010) 4 FORMAT (8F1004/2F1004) READ 50 (THETA(J)0 J=10111 5 FORMAT (8F1002/3F1002) READ 60 (TIK10 K=10101 6 FORMAT (7F1000/3F10001 DO 205 K=IOIO DO 204 J=1011 DO 25 1:1010 C bEGIN PROGRAM CALCULATIONS C CALCULATE UPPER LIMIT OF ERFZ Z=X(11/(200*bORTF(ALPHA‘THETA‘J)1) C CALCULATE ERFZ 21 ZZZZI=((((((043063BE-4*Z + 02765672E-J)*Z + 01520143E-31*Z + 09270 1527215-2)*Z + 0422820123E-11*2 + 0705230784E-11*Z + 1001**16 23 ERFC = 100/22221 24 ERFZ = 100-ERFC C CALCULATE SURFACL TLMP 0F FILM 25 TEMP(I) = T(K1 - ERFZ*(T(K1 - OTEMPI IF (J‘l) 20092000202 200 PRINT 2010 TIK)0THETA(J)0(TLMP(I)0 1:1010) 201 FORMAT (/1F90001F1002010F802) GO TO 204 202 PRINT 2030 THETA(J)0(TEMP(1)0 1:1010) 203 FORMAT (9X01F1002010F8021 204 CONTINUE 205 CONTINUE 206 END SCOPE 'LOAD 'RUN030200000M 80 00001311 00005 00010 00015 00020 00025 00030 00035 00040 00045 00050 002 004 006 005 010 020 060 1000 1040 1080 2020 80 100 125 150 175 200 225 250 275 300 60 mpqumzou zqaooaa maq _z~o uuzmamuu_o uqohqamazm» zupm>m urk u m»z_ mozoumm n.¢ u mamhm mz_h wr» z~ mwozqru wt» 20 h~z_4 m2.» m:» u >>> zqaooaa z. am»m Iu npqzo~mma o» amnzaz mmqu >a>>.»a.mm» annu~kawm u 222 u mzo_uma m z~ awkqmr zazmsz< Ip~z o u mzo_oma mo ammxaz u 222 u Azzz._u<_..<~.pqz..zzz.o~_ op_>~punozou Jh~mzmo muz_h >»_uum~ .v.mu. hazaou oo_ ya.ua.zz.n.ku.u»z.~ UUUU UUUUU .ZEZ.PQZ.ZS.Jh.~Zw\0fl Mid ZdIOOIQ MIF Z. AJD< mUQDZ m0 IJQSDZ EDE~XUFW>JOQ J_Z UZO Ihua Q3 hum.m~ SdQOOQQ wIk Zwkw>m HZMZMQquwE mathlmuimk JOOQ¢.BOW. UUUUUUUUU 61 ..oflvzz..o~.JF ..oomvh..om.o..om.u.fiom.m..om.< zo_mzwz_o. mpmou wz_hnoam3m ozw waz.»zou 0 mo_ 0» 09 ~00. waz_hzou ooo~ mzo~h>> u . ooo_.o_m.o_m.>>>u»pvm_ "om .oc.ofiux ..x.». .m» .oo~ hz_aa o u <4m~ flop gov._0b.~oo.o_n44m_. m5 .ooom*»» u wk ooh mozouum 2— mz_» wt» u m» u y mooz F4 maaka_ u .cxx.» om+exxu~axy om.aa~m2ucxx com on 000 u aQ¥¥vP 00” 0m+M¥¥uQ¥¥ 000 u aflxva agmzomufl¥¥ 00— OD N+~Z£mZuQQ~mZ —+_ZZmZuQ—w2 N+N22m2+ ~ZEWZu ZZ NZZmZ+AN¥¥UZZuNZZWZ «MA ZZZ.QPZ_NN¥¥ um" 00 ~+uPZ~unPZ~ Izzmz+._¥¥022uazzmz on" uPZ_.~u~¥¥ cm" 00 ONNZEWZ Ou.ZZmZ mwmaowo LIPLIQI\DPE u ¥ m0 WP—ZD NIP U ooouPP 0.0".N_.—mmm ANH ¢._uN_ ~N— OO Hui—wfi _ u ~Z<0 ~2UPZN PWE~L NIP oZEZoPJom mz_»:oam3m m_I» xa.uu.zz.n.»m.m»z.~ .zzz.h4z.zz.4».~24o.»»._z444.4»m.»o .h.o.u.m.4 zozzou . .0".xa..o_.»4z..o_.zz..o_.4».~ .o_.ua.aomvoo..oavuu..oomvh..om.o..om.u..om.m..om.4 zo.mzmz_o _o_ah mz_PaoamDm ozm zaakma .kzzvh*mmzo+.Fz.h*44zou.zz.o _+»2upzz ~1mx+nwuhz .z_m+.z.¥a*4»wm*o.m.uummzo Z—«Ia£.¥Q*