I” i y ‘I» I 1it! 11 I ll H | }\ ‘ , ‘ { l I‘ It (1 HI I H! il (9—: '—1 I 0.4 _‘CDN—* THE EFFECTS OF CO-E‘SDETEONFNG. TO? LOA‘DWG, AND {MPACT ON RELEASE TORQUE Thesis for the Dagrae a§ M. S. 'M’iCHIGAN STAT-E UNWERSW‘! :Ronaid W’aifer Hmiszny 1964 WHHWIIWMIHIEMIIHHIHII Wu WWII ” 3129 01107 0590 ABSTRACT THE EFFECTS OF CONDITIORING. TOP LOADING, AND IMPACT ON RELEASE TORQUE by Ronald Walter Horiszny This investigation was undertaken to determine the effects of three primary variables on the release torque of screw caps on glass bottles. An extensive literature search indicated that very little was known about the effects of different variables on release torque. Interviews. however, determined that a great deal more was known than has been made public on the subject. The three variables tested were storage conditions, top loading, and impact on the top of the container. Storage conditions of h0°F., ambient humidity; ambient room temperature and humidity; and 10003.. 90 to 95% relative humidity were tested for both five and ten day storage periods. Immediate release torque was also checked. The other two primary variables investigated were the effects of a 200 pound tap load and the effects of both 1 foot-pound and 2 foot-pound impact shocks. The test results indicate that. in general, storage at any of the three conditions causes metal caps to lose torque. Storage at ROOF. or at room temperature increases the release torque of phenol- ic caps. but high temperature and humidity tend to decrease it. TOp loading and impact were also found to decrease the torque retention of screw caps. ii THE EFFECTS OF CONDITIONING. TOP LOADING. AND IMPACT ON RELEASE TORQUE By Ronald Walter Horissny A THESIS Submitted to the College of Agriculture Michigan State University of Agriculture and Applied Science in partial fulfillment of the requirements for the degree of MASTER OF’SCIINCE Department Of forest Products School of Packaging 1961+ ACKNOWLEDGEMENTS The author extends his deep appreciation to Mr. B.J. Schnorr of Eli Lilly and Company and to Mr. 03V. Mumfbrd and.Hr. 3.x. Flaekamp of Owens-Illinois Glass Company for their suggestions. and to their companies for providing materials for testing. Thanks are also due to everyone of the Packaging Department for their interest and encouragement. with special thanks to Dr. J.V. Goff. major professor. and Mr. D.L. Olsson for their aid and guidance as members of the graduate committee. Thanks are also extended to Dr. S.E. Bryan of the Management Department for his membership. as minor professor. on the graduate committee. . iii TABLE OF CONTENTS ABSTRACT ACKNOWLEDGEMENTS LIST OF TABLES LIST OF FIGURES INTRODUCTION BACKGROUND EXPERIMENTAL PROCEDURE ANALYSIS OF DATA CONCLUSIONS SUGGESTIONS FOR FURTHER STUDY LIST OF REFERENCES iv Page ii iii vi 11+ 1&2 1+5 1+6 Table Table Table Table Table Table Table Table Table Table Table Table 10 11 12 LIST OF TABLES Release Torque Results, Condition 1 (30°F.. Ambient Humidity). Size 20-MO5 Bottles, Application Torque - 10 inch-pounds . . . Release Torque Results. Condition 1 (NO°F.. Ambient Humidity). Size asauos Bottles. Application Torque - 15 inch-pounds . . . Release Torque Results, Condition 2 (Room Ten erature. Ambient Humidity). Size 2 5 Bottles. Application Torque - 10 inc h-poundl e e e e e e e e e e e Release Torque Results, Condition 2 (Room Temperature. Ambient Humidity). Size as-uos Bottles. Application Torque - 15 inch pounds . . . . . . . . . . . . Release Torque Results, Condition 3 (100°r., 90-95%.Relative Humidity), Size 20-h05 Bottles,Application Torque - 10 inchspounds . Release Torque Results, Condition 3 (lOO°F.. 90-95% Relative Humidity). Size 28-305 Bottles. Application Torque - l5 inchppounds. Release Torque Results, Immediate Removal under Ambient Conditions . . . . . . n Comparison of R Factors of Immediate Release and Storage Results. . . . . . Release Torque Results. 200 pound Top Load Test . . . . . . . . . . . . . Comparison of R Ractors of Top Load and Storage Results. . . . . . . . . . Release Torque Results. Impact Tests on net“ Cap. 0 O O O O O O O O O 0 Comparisons of R Factors of Impact and Storage Results. . . . . . . . . . 16 18 20 22 2t 26 29 31 31* 38 no Figure Figure Figure Figure LIST OF FIGURES Sample screw caps and bottles . Owens-Illinois Torque Tester ready for testing . . . . . Rational Forge and Ordnance Compression Tester ready to top load a sample Impact Apparatus ready for testing vi 11 12 INTRODUCTION The purpose of this study was to determine the effect of certain storage conditions. tsp loading, and impact to the container top. on the release torque of screw caps on glass bottles. Interviews (1.2.3) with people in the field of packaging indicated that these areas of testing, especially the top loading and impact tests. would be the most interesting to the glass and closure industry. The reason for this situation is that not much work has been done in these two areas, and none of it has, as yet. been published. In fact. very little has been published concerning storage conditions and release torque either. Three storage conditions were used, each one for both five- and tenrday periods. The conditions were no°r. with ambient humidity. approximate]: 70°F. with ambient humidity. and 100°r. with 90 to 95% relative humidity. The top load checked.was 200 pounds and the im- pacts used wore equal to l foot-pound and 2 foot-pounds. An immediate release torque experiment was performed also, and the above tests' comparisons with it and with each other are analysed in this thesis. BACKGROUND Not a great deal has been published on the subject of release torque. Most of the articles turned up by a literature search were of a practical or production nature rather than reports of experimental findings. Probably the main reason for this is that most of the re- search in this area has been done by the glass and closure manufac- turers, and the highly competitive packaging field causes them to closely guard their findings. Also, the articles tended to be general rather than specific - so two of them are reviewed here to present a background on release tor- que. In providing this background material, information from articles (M) and (5) was combined and revised to present a more complete picture. first of all, it is pointed out that while torque for packaging is the force required to apply a cap to a container or to remove it. torque to a consumer is merely the effort needed to get a top off. This means that a packager must select an application torque which will insure the desired amount of protection to the product. but won't make it too difficult for the consumer to open the container. The articles agree that the main factors affecting torque are the product. the cap. the cap liner, the container, time cycles. and storage conditions. 3 Product considerations involved. besides whether the product reacts with the cap or liner. are whether it needs to be kept within certain moisture content ranges. or Just in the container. Important cap factors include the material it is made from. its dimensions. and its thread design. The item about liners which is most important to torque is the resiliency of the liner material. for generally. the more resilient the liner. the less application torque is needed to preperly maintain an effective seal. Thread design. diameter. and smoothness of finish are the essential traits of the container that influence torque. Other factors mentioned as influencing release torque are lub- ricants used on the cap. product being spilled on the bottle threads. pressure on the top of the cap during application. and compression set of liners. EXPERIMENTAL PROCEDURE SAMPLE CONTAINERS The bottles used in the testing were of the same volume (30 cubic centimeters). but of two different finishes: 20-n05 and 28-n05. Finish refers to the size and type of closure means. The means in this case are screw threads - one with a 20 millimeter maze imum thread dimension and the other with one of 28 millimeters. Both types of bottle were amber glass. The 20-u05 bottle had an overall height of 2-7/8 inches and an outside diameter of 1-3/8 inches. The 28% bottle was 2% inches high and had an outside diameter of 1% inches (see Figure l). The screw caps used were caps manufactured to fit the bottles used. Both tincplated steel and black phenolic caps were tested. The metal caps were white enamel coated on the outside. All caps tested were lined with 0.035 inch pulp/vinylite lubricant finish liners. See Figure 1. also. for a.visual comparison of the screw caps. All of the bottles and screw cape used were manufactured by Owens-Illinois. and were supplied for testing by them. or by Eli Lilly and Company. CONDITIONING The bottles and caps were stored together in the same area of a room for at least two weeks prior to filling and testing. in an ”m1 Saple screw caps and bottles 6 attempt to eliminate any variations that might be caused by unequal storage conditions. There was no attempt to control the conditions in the room. All containers tested. were tested when full. The "product" used in each case was 30 cubic centimeters of distilled water at room temperature. TEST METHOD The actual test method used was the measurement of release torque on an Owens-Illinois Torque Tester (see Figure 2). This device was also used to regulate the torque used to apply the caps. The application and release torques of screw caps were measured in inchrpounds and the torque tester used read up to 25 inch-pounds for each. To apply a cap. the bottle was placed between four upright pegs and secured by turning the knob at the right of the dish. and the cap placed on the bottle. Assuming a 15 inch-pound application torque is desired. the cap is grasped taking care not to touch the glass. and turned clockwise until the pointer on the dial reaches 15. Ho downward pressure should be put on the cap when applying or removing it. Release torque is measured in the same manner except that the cap is turned counter-clockwise and the pointer must be watched carefully in order to note the highest point reached. In this in- vestigation. readings were taken to the nearest half inch-pound. although the dial is calibrated only in whole numbers. IIGURI 2 Owens-Illinois Torque Tester ready for testing 8 In the case of the 20 millimeter caps. a pair of pliers with rubber covered Jaws (to prevent damage to the caps) was used to grasp the cap during application. as they were too small to hold by hand. Release torque was low enough so that pliers weren't necessary. and the 28 millhmeter caps were large enough so that their application and release could both be done by hand. VARIABLES TESTED Three storage conditions were tested: Condition 1 was h0°F. and ambient humidity. Condition 2 was room temperature (approximately 70°r.) and ambient humidity. and Condition 3 was 100%. and 90 to 95% relative humidity. Condition 3 was chosen to represent an ex- treme condition. and it is cited in Package Eggineering (6) as such. The ambient humidities of Conditions I and 2 should experience approximately the same variations. because the refrigerator used for Condition 1 was in the room used for Condition 2. This room was the same one that the bottles and caps were stored in prior to testing. Condition 3 was obtained in a.Yapor-tenp Relative Humidity Chamber. and the humidity did vary between 90% and 95% although it was set for about 93%. Bottles were filled in groups of ten, each group having caps applied and being placed under its particular storage condition before the next group was filled. Ten each of the metal and black phenolic capped bottles were 9 stored for five days under each of the three conditions. Ten of each were also stored for ten days under each condition. The above storage lots were performed for both the 20-h05 and the 28-305 size bottles. The 20 millimeter size caps were put on with an application torque of 10 inchepounds. and the 28 millimeter caps with 15 inch-pounds. These application torques were chosen because they are mid-points of application torque ranges recommended in the manual for torque testers (6) for caps of these particular sizes. The range suggested for 20 millimeter caps is 8 to 12 inch-pounds. and that for 28 millimeter caps is 12 to 18 inchrpounds. A new bottle and a new cap were used for each sample. with no cap or bottle being used twice. As the groups of bottles were put into their respective storage conditions. the time of day was noted so that they could be conditioned as nearly as possible to exactly a five or ten day period. No group deviated from its schedule by as much as 30 minutes. At the end of the storage period the bottles were removed by groups. and their release torque measured and recorded. An immediate release torque test was also performed for each of the four caps (metal and phenolic 20 millimeter. and metal and phenolic 28 millimeter). In this test. ten bottles were filled as a group and then they were capped with an application torque of either 10 or 15 inchrpounds. depending on their sise. Immediately after the tenth.bottle was capped. the first bottle was checked for release torque. The nine remaining bottles were then checked in the same order in which.they’had been filled. Not more than five 10 minutes elapsed between the time the first bottle was capped and the time that the tenth bottle was Opened and its release torque recorded. The second variable investigated was top loading. For this test. filled. capped bottles were top loaded to 20035 pounds in a National Forge and Ordnance Compression Tester (see Figure 3) at a platen speed of one-tenth of an inch per minute. The 200 pound load was suggested in the interview with Owens-Illinois personnel (3) as a realistic load. because the stacking of pallet loads six or seven high in warehouses may create loads considerably higher than 200 pounds on the lower bottles. The first step in this test was to fill a group of ten bottles with 30 cubic centimeters of distilled water each. These ten bottles were then capped with an application torque of either 10 inchepounds in the case of the 20 millimeter metal caps. or 15 inch- pounds for the 28 millimeter metal caps. After the tenth bottle was capped. the first bottle was tap loaded and then immediately taken out of the compression tester and checked for release torque. The top loading and checking of bottles two through ten in the same manner. followed directly. These steps were repeated on a.second group of ten for each cap. to give a total of 20 trials for each of the four caps. The third variable investigated was impact to the tap of the bottle. This test was also suggested in the Owens-Illinois interview (3) mentioned above. The impact appartus shown in Figure h is simple. Its purpose was merely to direct the steel ball used to impart the impact so that it would hit the cap directly in the center. The ball noun: 3 lotionsl lbrge and Ordnance Compression Tester ready to top load a sample 11 11003] ’4 Impact apparatus ready i'or testing 13 fell on a steel plate 3/32 of an inch thick so as to distribute the impact over the entire cap rather than Just at the point of impact. The samples rested on a 1 inch thick steel base plate while under- going the impact. Because preliminary testing indicated that the phenolic caps cracked when receiving 2 foot-pounds of impact, only metal caps were tested in the impact studies. These were subjected to both 1 foot- pound and 2 foot-pound impacts, but not on the same container. The steel ball used had a diameter of 1 13/16 inches, and weighed in ounces. It was dropped. by hand. from a height of 1 1/7 feet. or 13 23/32 inches. above the steel plate which rested directly on the cap. to impart approximately a l foot-pound impact. The 2 foot-pound impact was approximated by drOpping the ball from a height of 2 2/7 feet. or 27 7/16 inches, above the plate. The procedure was identical to that of tap loading in that bottles were filled in groups of ten. the cape were applied with the prOper torque. and then after all were capped they were subjected to impact and opened in order. .Again twenty samples were run for each cap. All cap applications, release torque testing. tap loading. and impacting were done under ambient conditions. the same as those of Condition 2. ANALYSIS OF DATA In addition to the individual test results. the following tables present the group's range (r). average release torque (E). standard deviation (s). and its average release torque expressed as a decimal fraction of application torque (R). For example. a.figure of .25A would indicate that the average release torque of the group equaled 25 percent of the application torque. The standard deviation as used in this thesis. is merely to give an indication of the width or spread of a group's release torque values and thereby give an idea of their consistency. The higher the standard deviation.value. the lower the consistency of results of that particular group. The standard deviation was calculated by the formula: 8-: NEEXZ - C22)2 '1 N (II-1T In this formula. 's' stands for standard deviation. ”x" stands for release torque. and ”N“ stands for the number of samples in the group. See reference (8). The 'R.factor' is found by merely dividing the average release torque for a group by the application torque with which the caps were put onto the members of the group. The following pages contain the tables of the experimental data gathered. and comments which summarize the data. In the comments. only the cap size and material are referred to because all liners in m 15 the experiment are the same. and the caps determined the bottle and application torque used (see EXPERIMENTAL PROCEDURE). while reviewing the following tables and commentary. it should be remembered that in addition to experimental error. both in per- forming tests and in reading the Torque Tester gauge, material variations may cause differences between individual sample results. For instance. a good seal depends on the cap liner being held firmly against the sealing surface of the bottle. and the holding firmness depends on the application torque. The application torque. in turn. is determined by the friction developed between the cap and bottle threads. Poor glass surface or bottles from worn molds may increase the friction and absorb an undue prOportion of the application torque. thereby weakening the seal and affecting the release torque (9). In addition to the items mentioned in this example. the factors listed in the BACKGROUND section above play a.part in determining the amount of release torque retained. The June. 1962 edition of Modern Pack- aging contains an article exploring the probabilities of looseness of fit for bottles and caps (10). TABLE 1 RELEASE TORQUE RESULTS. CONDITION 1 (u0°r.. AMBIENT HUMIDITY). SIZE 20-h05 BOTTLES. APPLICATION TORQUE - 10 INCHFPOUNDS. METAL CAPS PHENOLIC CAPS ‘__ sgogfigs , sdgaigg sgogfigs signifii ‘ 1. u.o in-lbs. I 1. 1.5 in-lbs. ‘1. 5.5 inrlbs. #1. 6.5 inrlbs. 2. 1.0 i 2. 2.0 2. 6.0 2. 8.0 3.5 ; 3. 2.0 . 5.5 7.5 n. 9.0 u 1.5 7.0 u. 7.0 5- 3-5 g 5.- 1.5 5. 5.5 5. 7.0 6. 3.0 ;6. 3.5 6. 6.0 6. 6.0 7. u.0 ; 7. 1.0 7. 5.0 7. 7.0 8. 3.5 g 8. 1.5 8. 5.5 8. 7.0 9. 5.0 {9.1.0 9. 6.0 9. 6.0 10. 3.5 E10. 3.0 10. 6.5 10. 6.5 f r 3.0-5.022.0 E r 1.0-3.5=2.5 r 5.0-7.0=2.0 r 6.0-8.0=2.0 ; 3.8 in-lbs. E ; 1.9 inelbs. ; 5.9 in-lbs. ; 6.9 in-lbs. s .51 - s .82 s .58 j s .63, 2 .38A R .191 R .591 § R .69A 16 17 Table 1 shows that for 20 millimeter caps. a 10 day storage period for metal caps decreases the 'R factor"II (release torque ex- pressed as decimal fraction of application torque) to half of the R factor resulting from a 5 day storage period at the same conditions (u0°r.. ambient humidity). Phenolic caps. meanwhile. actually had an increase of release torque in the 10 day storage as compared to the 5 day storage - .59A after 5 days and .69A after 10 days. Comparison of metal and phenolic caps stored for 5 days shows that phenolic caps retained nearly six-tenths of the torque they were applied with. while the metal caps retained slightly less than four-tenths. For a 10 day period. the phenolic caps had an R factor over three times that of the metal caps. The standard deviations of the groups in Table 1 indicate that the data found is quite consistent. TABLE 2 RELEASE TORQUE RESULTS. CONDITION 1 (NCOF.. AMBIENT HUMIDITY), sxzs 28-h05 BOTTLES. APPLICATION mosque - 15 INCHbPOUNDS. METAL c125 PHENOLIC cars 1% .3035. «513123; 53.32311 51.3.22; . inch-pounds “inch-pounds inch-pounds inch-pounds I 1 1. 8.0 1. 6.0 1. 11.0 1. 13.0 2. 5.5 2. 6.5 2. 11.0 2. 13.0 3, u.0 3. 6.5 3. 12.0 3. 1h.5 h. 8.5 h. 6.0 h. 12.5 h. 13.0 5. 6.0 5. 7.0 5. 12.0 5. 1h.0 6. 7.0 6. 5.0 6. 10.0 6. 13.0 7. 5.5 7. 7.0 7. 10.0 7. 13.5 8. 7.0 8. 6.5 8. 10.5 8. 12.5 9. 8.0 9. 6.0 9. 10.5 9. 12.0 10. 7.5 i 10. 7.0 310. 11.0 10. 13.0 r h.0.8.5= u.5 ? r 5.0-7.0: 2.0 r 10.0-12.5=2.5 r 12.0-11.5: 2.5 ; 6.7 inelbs. ‘ ; 6.u in-lbs. ; 11.1 in-lbs. ; 13.2 inelbs. s 1.142 s .63 s .86 s .71 11 .951 «.3 $31. 11 .7141 ‘ R .SSA 19 Except for the 10 day storage of metal caps. the data in Table 2. which concerns 28 millimeter caps at Condition 1. is not as consistent as that in Table 1 according to their standard deviations. However. eliminating the most devious reading. h.0 (item 3-5 day storage. metal caps). raises the release torque average only 0.3 inch-pounds to 7.0 inch-pounds. And while the standard deviation is lowered 0.30 to 1.12. the R factor is only increased 0.02 to .M7A. Because the.R factor changes so slightly. the Table 2 28 millimeter caps are compared in the same manner as the Table 1 caps. but using the corrected value. i.e. without using the “.0 reading noted above. ‘Unlike the 20 millimeter caps with their great loss of release torque between the 5 and 10 day periods. the 28 millimeter caps had an R factor only .OHA less for the 10 day storage than for the 5 day storage. The phenolic caps again increased their torque retention at the longer storage period. this time by 0.1hA as compared to the 20 millimeter caps' 0.10A. Also as in Table l. the phenolic caps retained more torque than the metal caps. for 5 day storage they held .27A.more than the metal caps. and for 10 day storage they'retained over twice as much. Comparing Table 1 and Table 2 shows that the larger caps retained torque more effectively than the smaller ones. The 28 millimeter metal caps at 5 and 10 day storage periods had 3 factors of .OSA and .2“A.more. respectively. than their 20 millimeter counterparts. The 28 millimeter phenolics had .15A and .19A more at 5 and 10 day periods. respectively. than the 20 millimeter phenolics. TABLE 3 RELEASE TORQUE RESULTS. CONDITION 2 (ROOM TEMPERATURE, AMBIENT HUMIDITY). SIZE 20-N05 BOTTLES. APPLICATION TORQUE - 10 INCH-POUNDS. ‘_hITAL GAPS _ _ PHENOLIG CAPS 5 DAY 10 net 5 Di! 10 DAY STORAGE .._.- i--§3c’_3§63_-__._ . 51°35“: STORAGE - ‘ 1. u.o in-lbs. 1. 2.5 inslbs. 1. 6.0 in-le. 1. 5.5 inrlbs. I 2. 3.5 2. 3.5 2. 6.0 2. 6.5 l 3. 5.5 ; . 1.5 ' 3. 6.0 3 6.0 1 u. n.5 u. 2.5 4 n. 6.0 u 5.5 5- 3-0 ' 5. 1*5 J 5. 5-5 5 5-5 6. h.5 6. 3.5 g 6. 5.5 - 6. 5.0 ’1 I 7. 3.0 1 7. 5.5 i 7. 6.0 7 7. 5.5 i 8. 3.5 , 8. 3.0 1 8. 6.5 8. 6.0 : i i‘ E 9. u.5 g 9. 3.0 .79. 5.5 9 6.0 E 10. h.o £10. 1.0 £10. 6.0 '10. 6.0 . g 3 i 1 - 2,. T’ ..I - 9: 7 I r 3.0-5.5:2.5 : r 1.0-5.5:h.5 3 r 5.5-6.5:1.o . r 5.0-6.c =1.0 ‘ - 1 - - x h.o in-lbe. i x 3.1 inelbs. ' x 5.9 inslbs. 1 2 5.8 inplbs. i s .78 g s 1.32 s .32 f s .h3 E n .hOA 5 n .31; 3 .59A ‘3 .58A 1 I mn’*u-v~u~——- - s-..-o._. 20 21 At room tanperature and humidity (Condition 2. Table 3). the 20 millimeter phenolic caps retain their application torque practically equally well for 5 and 10 day storage periods. For the metal caps. the 10 day group has quite a high standard deviation. Eliminating the 14.5 and 5.5 readings. which seem to be out of line. leaves a range of 1.0-3.5. an average of 2.6 inch-pounds. a standard deviation of 0.90. and an R factor of .26L. ‘Using the refined in- formation. the metal caps lose .IMA‘betwsen the fifth and tenth.day of storage. Also using the refined.R.factor. a comparison of the caps stored for 10 days shows the phenolics retaining twice as much torque as the metal caps. After 5 days of storage. phenolic caps kept .19A more torque than the metal caps did. Comparing the metal caps in Table 3 (Condition 2) and those in Table 1 (Condition 1) shows that they both had less release torque after 10 days than after 5 days. The 20 millimeter phenolic caps in those tables had identical B factors for 5 day storage periods. but the Condition 1 phenolic caps at 10 day storage had .09A more release torque than those under Condition 2 for 10 days. TABLE u RELEASE TORQUE RESULTS. CONDITION 2 (ROOM TEMPERATURE. AMBIENT HUMIDITY). SIZE 28-uO5 BOTTLES. APPLICATION TORQUE - 15 INCH-POUNDS. __ mm. 0125 Parsons 0125 5 1m 10 m1 5 m1 ! 10 1111 5200111158 sromr. STORAGE L STORAGE inch-pounds inch-pounds inch-pounds f inch-pounds 1. 5.0 1. 8.5 1. 10.5 i 1. 9.3 2. 6.5 2. 7.5 } 2. 11.0 ;_ 2. 10.0 3. 9.5 3. 6.5 . 3 11.0 . 3 10.0 n. 6.0 u. 7.5 i n. 11.5 1+ 8.5 5. 8.0 5. 6.5 1 5. 10.5 j} 5. 10.0 6. 8.0 6. 8.5 . 6. 11.0 _ 6. 10.5 7. 8.0 7. 7.5 i 7. 10.5 ' 7. 10.0 8. 7.0 8. 8.0 8. 11.0 8. 10.0 9- 7-5 9. 7.5 9. 9.5 g 9. 11.0 10. 9.5 10. 7.0 i 10. 10.5 a 10. 11.5 r 5.0-9.5: h.5 f r 6.5-8.5: 2.0 L r 9.5-11.5= 2.0' r 8.5-ll.5= 3.0 k i 7.5 in-l'bS. 1.: 7.5 in-lbs. ; 10.7 in-lbs. 3 2 10.1 in-lbs. s 1.“) s .71 s .5h s .86 3 .50A 4 R .5011 R .71A R .671 ‘ J l 23 In Table N (Condition 2. 28 millimeter caps). the metal caps under 5 day storage had a rather high standard deviation. Even so. removing the low reading of 5.0 inchrpounds (because there are no readings closer than 1.0 to it) only reduces the standard deviation to 1.20 and raises the R factor to .52A from .50A. This still leaves the metal caps at 5 day storage and those at 10 day storage practically equal. This is also true of the phenolic caps. At both 5 and 10 day storage. phenolic caps retain more torque than metal ones: .19A for 5 days and .17A for 10 days. Under Condition 2. the 28 millimeter caps retain torque more effectively than 20 millimeter caps at all combinations of storage periods and cap materials. Comparing the 28 millimeter caps under Condition 1 (Table 2) with those under Condition 2 (Table h) shows that the metal caps under Condition 2 kept .05A.more torque than those under Condition 1 for 5 day storage (using the corrected values of .M7A.and .52A). and .07A.more for a 10 day period. The phenolic caps are practically equal for 5 day storage. the Condition 1 caps being .03A higher. The Condition 1 caps are even higher for 10 day storage - .21A in this case. TABLE 5 RELEASE TORQUE RESULTS. CONDITION 3 (100°F.. 90-95% RELATIVE HUMIDITY), SIZE 20-305 BOTTLES. APPLICATION TORQHE - IO INCH-POUNDS. METAL CAPS PHENOLIC cars 5 DAY 10 DA! 5 six 10 n1! 1 STORAGE STORAGE STORAGE l STORAGE inch-pounds inch-pounds inch-pounds ; inch-pound: 3 1. u.o 1. 3.5 1. 3.0 i 1. 2.5 5 2. 3.5 2. h.0 2. 2.5 g 2. 2.0 . h.0 3. h.0 3. 2.5 3. 2.5 i u. u.0 u. 3.5 h. 2.5 g u 2.0 5. h.5 5. u.0 5. 0.0” 5. 2.0 6. 3.5 6. u.5 6. 2.0 6 2.0 7. 3.5 7. 5.0 7. 2.5 7. 2.5 8. 3.0 8. u.5 8. 2.5 8. 3.0 9. 3.5 9. h.0 . 9. 2.5 9. 2.0 10. h.0 10. h.0 Q10. 2.5 10. 2.5 r 3.0-n.5a 1.5 r 3.5-n.5a 1.0 Er 2.0—3.0: 1.0 r 2.0-3.0: 1.0 ; 3.8 in-lbs. ; 8.0 in-lbs. g; 2.5 in-lbs. ; 2.3 in-lbs. s .h3 s .33 5s .25 s .35 11 .38A 11 $01 211 .251 f .2311 l’Inspection of this cap after removal proved it to have a faulty liner. so it was not included in the range. average. etc. 21+ 25 Table 5 shows the effects of extreme storage conditions (lOOOF.. 90-95% relative humidity) on 20 millimeter caps. The metal caps kept practically the same amount of torque - the 5 day samples retaining .38A. and the 10 day. .uOA. The phenolic caps were also practically equal - .25A for the 5 day samples compared to .23A for the 10 day sample. Comparing the 5 day storage figures for metal and phenolic caps shows that the metal caps retained more torque (.38) than the phenolic caps (.25A). They also held more torque after 10 days in conditioning (.HOA) than the phenolic caps held (.23A). The metal caps stored for 5 days under Conditions 2 and 3 vary in torque retention by only .02A. the caps under Condition 2 being the larger at .hOA. The 10 day period caused a much larger differ- ential. for using the corrected value the Condition 2 caps have an R factor of .26A. while the Condition 3 caps have one of .NOA. The phenolic caps. meanwhile. show an Opposite trend. They have higher retention factors for Condition 2 storage. The 5 day Condition 2 factor is .59A. while Condition 3's factor is .25A. Ten day storage causes essentially the same situation. the spread being .58A to .23A. TABLE 6 RELEASE TORQUE RESULTS. CONDITION 3 (lOOOF.. 90-95% RELATIVE HUMIDITY). srzs 28-h05 BOTTLES. APPLICATION TORQUE - 15 INCHFPOUNDS. METALfiCsPS PHENOLIC CAPS 5 Day ; 10 DAY 5 DAY 10 DAY STORAGE .11 3103108 STORAGE ; STORAGE 1. 7.5 in-lbsé 1. 8.0 inslbs. 1. 3.5 in-lbs.: 1. 2.5 in-lbs. 2. 8.0' g 2. 8.0 2. u.o i 2. 3.0 3. 7.5 E 3. 7.5 3. 3.5 3 3. 11.0 h. 8.0 u. 8.5 u. 3.5 g h. 2.5 5. 8.0 5. 9.0 5. u.o z 5. 3.0 6. 3.5 6 8.0‘ 6. 11.0 i 6. 0.5 s 7. 5.0* 7. 10.5 * 7. h.0 § 7 0.5 E 8. 8.0 8. 9.5 a 8. 3.5 i 8. h.0 ; 9. 8.0 9. 8.5 E 9. u.0 E 9. 3.0 § 10. 8.0 10. 8.5 !.10. 3.5 E 10. 3.5 ‘ l 1 ‘t r 5.0-8.5: 3.5 r i 7.7 in-lbs. E s .97 s R .SIA R 7.5-10.5: 3.0 tr 3.5-u.0= 0.5 I- 8.6 in-lbs. x 3.8 in-lbs. .88 is .26 .5711 is .2511 r 0.5-h.0= 3.5 E 2.7 in-lbs. s 1.25 R .18A * These caps had rust on their threads when removed from the humidity cabinet. 26 27 Table 6 shows that two groups of caps had fairly high standard deviations. However. eliminating the 5.0 reading which was given.by a cap that was rusty upon removal from the metal cap 5 day storage group. reduces its standard deviation to 0.09 and raises its R factor 0.02 to .53A. In the phenolic cap 10 day storage results. the two 0.5 readings seem abnormally low. Eliminating them gives a standard deviation of 0.59 and.the R factor increases from .18A to .21A. The revised R factors will be used in the following comparisons. The metal caps differ in torque retention by only .ORA for the 5 and 10 day storage periods. The caps differ by .OHA also. with the 5 day caps. at .25A. the larger. The metal caps are a little over twice as effective as the phenolic caps in retaining torque over the 5 day period. and over 2% times as effective for the 10 day period. Camparison of the R factors for the 28 millimeter caps and the 20 millimeter caps after storage under Condition 3 (Tables 5 and 6) shows that fihe phenolic caps are equal after 5 day storage and differ by only .02A after 10 days. The metal caps. though. show better torque retention by the 28 millimeter size than by the 20 millimeter size. The differences being .15A for 5 day storage and .17A for 10 day storage. Although 28 millimeter metal caps stored for 5 days under Condition 2 (Table h) and Condition 3 (Table 6) have practically identical R.factors. .52A.and .53A. respectively; 28 millimeter phenolic caps show a wide variation. the Condition 2 caps having a retention factor nearly three times that of the Condition 3 caps. 28 For 10 day storage. the metal caps under Condition 3 had an.R factor of .57A and those under Condition 2 had one of .50A. As in the 5 day results. the phenolic caps stored for 10 days had a wide variation of release values. this time the Condition 2 value being over three times as great as the Condition 3 value. fletal caps. 28 millimeter. which underwent Condition 3 had higher torque retention than those undergoing Condition 1 (Table 2). They were .06A greater for the 5 day period. and .1MA greater after 10 days. The phenolic caps again show a great spread. Condition 1 caps being almost three times as effective in retaining torque as Cone dition 3 caps for a 5 day storage period. and over four times more effective for a 10 day period. TABLE 7 RELEASE TORQUE RBULTS. IMMEDIATE REMOVAL UNDER AMBIENT CONDITIONS gg, METAL cars 7* PHENOLIC CAPS é 20—h05 SIZE 28-h05 sxzs 20-h05 szzs ‘ 28-M05 5188 ? APP. Tosqns APP. TORQUE APP. TORQUE Arr. TOBQHE g 10 inelbs. 15_in-1bs. 10 in-lbs. l5 invlbs. ‘ 1. 5.5 inelbs. 1. 7.0 inelbl. 1. 5.0 in-lbs.1 1. 7.0 in-lbs. ‘ 2. 5.0 2. 8.0 2. 5.0 2. 8.0 § 3. 8.0 3. 9.0 3. 5.0 i 3. 7.0 § 1. 5.0 u. 9.0 h. 5.0 i u. 7.0 g 5. 8.0 ‘ 5. 8.5 5. h.5 5. 7.5 3 6. 5.0 7 6. 7.5 6. 5.0 6. 7.0 7. 5.0 7. 8.0 7. h.5 7. 7.0 8. 6.0 8. 9.0 8. u.0 8. 7.0 9. 5.0 i 9. 8.0 f 9. u.5 9. 7.5 10. 5.5 i 10. 7.0 :10. 5.0 10. 7.0 i T“ r u.o—6.0. 2.0 ' r 7.0-9.0. 2.0 gr n.0-5.0. 1.0 r 7.0-8.0. 1.0 E 5.0 inelbs. i 8.1 inelbs. ; u.8 inelbs. i 7.2 in-lbs. s .62 s .78 s .35 s .35 s .501 n .5u1 n .hsi a .nsh 29 30 Table 7 shows that all caps tested lost approximately half of the torque they had been put on with. almost as soon as they were on. The 20 millimeter metal cape lost exactly half. having an R factor of .501. The 28 millimeter caps held a little more torque as they had an R.factor of .BMA. Both sizes of phenolic caps had release torqnes of .u8A. Comparing the 20 millimeter caps shows the metal and phenolice to be almost equal. the metal caps retaining only .OZA.more than the phenolic caps. The case is much the same for the 28 millimeter caps. with the R factor of the metal cape being .06L greater than that of the phenolic cape. COMPARISON OF R FACTORS OF IMMEDIATE RELEASE AND STORAGE RESULTS‘ 5 DAY 10 DAY IMMEDIATE RELEASE STORAGE STORAGE 901113121011 1 - 20 mm. mm. cabs .50A .38A .191. 20 mm. PEENOLIC CAPS A811 .591 .69; 28 mm. mm. CAPS .5111 .14711 .1131 28 mm. PHENOLIC caps .1482 ‘ .7101 .88A CONDITION 2 20 mm. mm. CAPS .SOA .uoa .261. 20 mm. Pmouc CAPS Jun .5911 .58A 28 mm. mm. CAPS .5“ .521 .501 28 mm. mono CAPS J+8A .711. .671 CONDITION 3 20 mm. METAL CAPS .5011 .38A #011 20 mm. PHENOLIC CAPS .118; .25». .23; 28 mm. mm. cm .51u .53A .57; 28 mm. PKENOLIC CAPS J+8A .25.». .2u " Corrected R factors are used where applicable. 31 32 Through the comparison of the Condition 1 results with the immediate release torque results (Table 8). it is discovered that 20 millimeter metal caps decline in release torque from .501 for immediate release to .38A after 5 days storage at “GOT. to .19A after 10 days storage at NOOF. The 20 millimeter phenolic caps. however, go in the Opposite direction - from .M8A for immediate release to .59A after 5 days and .695.after 10 days. The same trends are apparent fer 28 millimeter caps. not to as great a degree in the metal cape. but to a greater degree in the phenolics. The metal caps dr0p from .5RA to .M7A (corrected value) to .h3L and the phenolic caps rise from .H8A to .7HA to .88L for immediate release, 5 day, and 10 day storage. respectively. This order of presentation will be used for the rest of the comparisons in this section. A comparison of immediate release results and the results of caps stored under Condition 2 (approximately 70°F.) reveals that the 20 millimeter metal caps again follow the decreasing trend. Their R factor drop from .50A to $05 to .26A (corrected value). As in Condition 1. the 20 millimeter phenolic caps initially rise, from .hSA to .59‘» but then level off to .58A after 10 days of storage. The 28 millimeter phenolics also exhibit this pattern. going from .hSA to .71A to .67At The 28 millimeter metal caps. meanmhile. re- main practically equal, going from .50A to .52A (corrected value) back to .50A. Comparison of Condition 3 (extreme condition) results and immediate release findings gives very different trends. The 20 millimeter metal caps drop from .50A to .38A and level off at .NOA. The phenolic caps drop even more, from .U8A to .251, and then level 33 off at .23A. The 28 millimeter metal caps, unlike the 20 millimeter. rise from .50A to .53A (corrected value) to .57A. The 28 millimeter phenolic caps are very similar to their 20 millimeter counterparts, dropping from .HEA to .25A and leveling off at .21A (corrected value). TABLE 9 RELEASE TORQUE RESULTS, 200 POUND TOP LOAD TEST 3b. METAL CAPS PHENOLIC CAPS 20-u05 SIZE 28-h05 SIZE . 20.n05 SIZE 28-h05 SIZE APP. TORQUE APP. TORQUE ; APP. TORQDE APP. TORQUE 10 in-lb_s_. 11 in-lbe. _. 10 in-lbs. +45 in-lbs. ‘ 1.- 2.5 in-lbs.l 1. u.5 in-lbs. 1. 1.0 in-lbs.§ 1. u.5 in-lbe. , 7 2. 2.0 f 2. 6.0 f 2. 2.5 i 2. u.5 { . 2.0 Q 3. 6.5 3 3. 1.0 3. 5.0 u. 2.0 E h. 6.0 f h. 1.5 u. 5.0 5. 3.0 {5. 5.5 5. 1.5 5. 5.5 3 6. 2.5 :6. n.5 6. 2.0 6. 5.0 E 7. 2.0 7 7. 5.0 7. 0.0 7. u.5 ; 8. 1.0 8. 6.5 8. 1.5 t 8. 6.0 E 9. 2.0 7 9. u.0 ‘ 9. 2.0 7 9. u.0 2 10. 0.0 $10. 3.5 i10. 2.5 § 10. h.0 i 11. 2.5 f11 5.0 11. 2.5 11. u.5 . g 12. 2.0 g12. h.5 12. 1.5 12. u.5 i % 13. 2.5 €13. 5.5 :13. 2.0 13. h.5 g g In. 1.0 jlu. h.5 i1u. 0.5 1h. 8.0 g g 15. 2.5 :15. 3.0 515. 2.5 15. 5.0 7 E 16. 2.0 316. 8.5 ;16. 2.0 16. h.5 g g 17. 2.0 517. 5.0 é17. 2.0 17. 6.0 3 E 18. 3.0 18. 5.5 318. 2.5 18. 5.0 g f 19. 2.0 19. 5.0 :19. 1.0 19. 5.5 2 ' 20. 2.0 20. h.5 .20. 1.5 20. 6.0 : r 0.0-3.0: 3.0 r 3.0-6.5: 3.5 VTF;DO.0-2.5= 2.5 fir h.0-6.0. 2.0 § 2 2.0 in-lbs. ; 5.0 in-lbs. , § 1.7 in-lbs. ; u.9 in-lbs. ? s .70 s .92 s .71 s .65 i ___ 8 .20A . 2 .33A 7 Ii.~ .l7A V R #:33A J 35 In Table 9 we see that the metal (20 millimeter metal) caps re- tained slightly more torque after undergoing the 200 pound top load than the 20 millimeter phenolic caps did, .20A to .17A. The metal and phenolic 28 millimeter caps retained the same fraction - .33A. For both metal and phenolic caps, the 28 millimeter size re- tained more torque than the 20 millimeter size - .13A in the case of metal. and nearly twice as much (.33A to .l7A) in phenolic's case. Comparing top loading results (Table 9) to immediate release results (Table 7) reveals that top loaded samples have lower.R factors in all cases - 20 millimeter, tap loaded metal caps have release torques 2% times smaller (.50L to .20A), and phenolic caps have them almost three times smaller (.N8A to .17A). TABLE 10 COMPARISON OF R FACTORS OF TOP LOAD AND STORAGE RESULTS* TOP 1.019313) 5 DAY STORAGE 10 DAY STORAGE CONDITION 1 20 mm. METAL CAPS .20A .38A .19A 20 mm. PEENOLIC CAPS .17A .59A .69A 28 mm. METAL CAPS .33A .h7A .h3A 28 mm. PHENOLIC CAPS .33A .7MA .88A CONDITION 2 20 mm. METAL CAPS .2011 Am .25.; 20 mm. PHENCLIC CAPS .17A .59A .58A 28 mm. METAL CAPS .33A .52A .50A 28 mm. PHENOLIC CAPS .33A .71A .67A CONDITION J 20 mm. METAL CAPS .‘5A .38A .MOA 20 mm. PIENOLIC CAPS .17A .25A .23A 28 mm. METAL CAPS .33A .53A .57A 28 mm. PHRNOLIC CAPS .33A .25A .21A ’ Corrected R factors are used where applicable. 36 37 In Table 10 it can be seen that top loading reduces torque to a greater degree than storage at ROOF. for all time - material - size combinations except the 10 day - metal - 20 millimeter caps. which have nearly the same R.factor for both. For the 28 millimeter metal caps. the Condition 1 caps are at least .lCA higher than the t0p loaded caps, for each storage period. The 20 millimeter phenolic caps. meanwhile, have R factors approximately 3% and n times higher for 5 and 10 day storage periods, respectively. than for top loading. The 5 day figure for 28 millimeter phenolic caps is .hlA higher than that for top loading. while the 10 day figure is .SSA higher. Under Condition 2 the 20 millimeter metal caps under 5 day storage double the R factor of those undergoing tOp loading, but fall back to only .06A more after 10 days of storage. As in Condition 1. the 20 millimeter phenolic caps from 5 day storage more than triple the top loaded ones; but in this case, they then level off. The same is true for 28 millimeter phenolic caps; they more than double the tap loaded caps. and then level off. The 28 millimeter metal caps increase by .19A between tOp loaded samples and Condition 2, 5 day samples. and then level off. ‘Using the order of tap loading to 5 day to 10 day storage values, we see that under Condition 3 both 20 and 28 millimeter caps imp crease by about .20A and then level off. The 20 millimeter phenolic caps increase by .08A and level off, but the 28 millimeter caps de- crease steadily. TABLE 11 RELEASE TORQDE RESULTS. IMPACT TESTS ON METAL CAPS l FOOT-POUND IMPACT TEST 2 FOOT-POUND IMPACT TEST 20-1405 SIZE 28-ho5 SIZE 20-h05 SIZE 284105 SIZE APP. TORQUE APP. TORQUE APP. TORQUE APP. TORQUE '10 in-lbs. 15 in-lbs. lO in-lbs. 15 inelbs. 1. 2.5 inrlbsd 1. 5.0 in-lbs. t A 3 1. 1.5 In-Ibsl 1. 1 05 1n‘1b8 e 2. 2.5 2. u.0 E 2. 0.5 1 2. 5.5 I 3 1.5 3. 3.5 i 3. 2.5 . 3.5 3 u. 2.0 n. u.0 g h. 1.0 1 h. 2.5 ; 5 2.0 5. 3.5 g 5. 1.5 2 5. 5.0 g 6. 2.0 6. 6.0 g 6. 0.5 i 6. 2.5 i 7. 2.5 7 5.0 E 7. 1.0 E 7. 3.5 g 8. 2.0 8. 5.0 g 8. 1.5 i 8. 3.0 f 9 2.0 9. 5.0 9. 1.0 E 9. 3.5 E 10. 1.5 10. 6.0 E10. 1.0 E10. 2.5 11. 2.5 11. 3.5 £11. h.0 Ell. 2.5 12. 2.5 12. 5.0 $12. 2.0 712. u.5 i 13. 2.0 13. 8.0 $13. 3.0 §13. 3.5 1 In. 2.0 In. 3.5 51k. 2.0 21k. 3.5 15. 2.5 15. 5.5 f15. 2.5 g15. 3.5 g 16. 3.0 16. u.5 316. 2.0 316. 3.0 i 17. 1.5 . 17. 6.5 i17. 3.0 117. 3.0 i 3 18. 2.5 ; 18. h.0 ?18. 1.0 i:18. 3.0 § 7 19. 2.0 19. 6.0 719. 2.5 :19. n.0 ' 7 20. 1.5 5 20. 5.5 20. 1.5 $20. 2.0 i r 1.5-3.0. 1.5 fT 3.5-6.5. 3.0 r 0.5-b.0- 3.5 Er 1.5-5.5. h.0 § 2 2.1 inelbs. ‘ ; h.8 inelbs. ; 1.8 in-lbs. 3;;3.3 in-lbs. s .1h 8 .97 s .93 is .97 R .21A E .3211 ___...qu .l8A 1R ~EEAW“___-J 38 39 Table ll shows that the 28 millimeter caps retained more tor— que than the 20 millimeter caps after both 1 and 2 foot-pounds of impact. The differnce was only .01». at the higher impact level, but was .llA at the l foot-pound level. Both sizes of caps had lower torque retention after 2 foot-pounds of impact than after 1 foot- pound. The 20 millimeter caps had .03A less, and the 28's had .lOA less. In comparison to immediate release torque. the 20 millimeter metal caps after a l foot-pound impact had almost 2% times less tore que retention. In the case of the 28 millimeter caps, .22A.more release torque was lost by the impact samples than by the immediate release torque samples. Comparing the 2 foot-pound impact results with the immediate release samples shows the same results, but to a.greater degree. The 20 millimeter caps that vere subjected to impact gave release torques approximately 2% times less than immediate release samples. as did the 28 millimeter caps. The l foot-pound impact-tested samples (Table 11) have almost the same R factors as the metal, 200 pound t0p loaded caps (Table 9). This is also true for the 20 millimeter caps which underwent either a 200 pound top load or a 2 foot-pound impact. They varied by .OZA. while the others varied by only .OlA. The situation changes for 28 millimeter. 2 foot-pound impact tested.caps though; they retain .llA less torque than their 200 pound top loaded counterparts. TABLE 12 COMPARISONS OF R FACTORS OP’IMPACT AND STORAGE RESULTS' A. CONDITION 1 20 mm. METAL 28 mm. METAL CONDITION 2 20 mm. METAL 28 mm. METAL CONDITION #3 20 mm. METAL 28 mm. METAL B. CONDITION l; 20 mm. METAL 28 mm. METAL CONDITION 2 20 mm. METAL 28 mm. METAL CONDITION 3_ 20 mm. METAL 28 mm. METAL CAPS CAPS CAPS CAPS CAPS CAPS GAPS CAPS CAPS CAPS CAPS CAPS l PT-LB IMPACT .21A 2 FT—LB IMPACT 5 1m smug: .18L 5 1m STORAGE .53A ‘ Corrected values used where applicable. 10 DA! STORAGE .26A .50A .57A 10 DAY STORAGE . 19A A31 .26A . hon .57A kl Table 12 shows that the l foot-pound impact results compare with the storage condition results much.the same as the 200 pound top load results did. Again. only the 20 millimeter caps stored under Condition 1 for 10 days had a lower 3 factor than the caps which underwent the impact test. In fact. only it and the 20 milli- meter, Condition 2 caps have 10 day storage R.factors considerably less than their 5 day factors. All the others. after rising cons siderably between the l foot-pound and 5 day results, level off in their 10 day storage R factor. Except for the fact that the Condition l. 10 day stored. 20 millimeter caps' R factor does not fall quite below'their 2 foot- pound impact results, the same is true of the 2 foot-pound results as is for the l foot-pound results. The differences between the 5 day and 2 foot-pound 3 factors are a little greater. though, especially for 28 millimeter caps. CONCLUSIONS An important conclusion which may be reached from the data in the preceding section. is that small caps constructed of tin plated steel or black phenolic plastic and having pulp/vinylite liners lose approximately half of their application torque within the first five minutes of being capped. If the caps are not removed immediately. but the caps and bottles are allowed to remain at room conditions for a.period of time. the release torque of metal caps declines steadily. with the smaller caps dropping at a.much higher rate. The release torque of phenolic caps. meanwhile. increases for a few days. and then begins to decline. The larger caps show these changes to a greater degree than the smaller caps. Storage at “0°I. causes the release torque of metal caps to decrease Just as those stored at room conditions. except that the larger caps decrease a little faster in the cooler condition. The refrigeration causes phenolic caps to have a steadily rising release torque for at least 10 days. with the larger caps again having larger increments. The release torque of metal caps stored at extreme conditions (1000!. and 90-95% relative humidity) declined for a few days. and then began to rise. The larger metal caps are more stable than the smaller ones. Extreme conditions cause phenolic caps. regardless of size. to lose torque rapidly for a few days. and then to seemingly level off. In general then. it may be said that storage at any condition he 1*3 causes metal caps to lose release torque; and that storage at any temperature between uo°r. and 70°F. causes phenolic caps to improve on what their immediate release torque would be. while storage at high temperature and humidity causes phenolic caps to lose torque. It must be remembered that the conclusions above are derived from the results of tests on only two sizes of caps. both relatively small. with only one kind of liner. Also. no intermediate tests (between the immediate and 5 day or between the 5 day and 10 day tests) were run; nor were any tests of longer than 10 day storage run. And it is possible that testing other size caps. or different kinds of liners. or more storage periods would alter or completely change the above trends. Top loading decreases the torque retained by caps considerably. the amount depending more on the cap size than on the material. In general. it also decreases torque retention by greater amounts than storage at various conditions does. The exceptions being the smaller metal caps at normal or refrigerated conditions for 10 days or longer. and the larger phenolic caps stored at high temperature and humidity. The conclusions about the effects of top loading are subject to the same limitations as the conditioning results above. plus the fact that more tOp loading weights (both greater and less than 200 pounds) would need to be tested for more inclusive, and still re- liable. results. Impact to the tap of metal caps also causes them to lose more of their release torque than storage causes them to lose. The 20 millimeter caps stored for 10 days at uO°F.. as compared to those nu subjected to l foot-pound of impact. are an exception to this state- ment. The data also indicates that the greater the impact. the greater the amount of torque lost. This effect seems to be greater on the larger caps than on the smaller caps. Again. the reliability of these conclusions are limited by the small number of trials, cap sizes, etc. tested. Increasing the number of all variables mentioned above, and adding more impact levels to the experiment. would lend more conclusiveness to the re- sults. SUGGESTIONS FOR FURTHER STUDY Expanding the tests used in this experiment to remove the limitations mentioned in the CONCLUSIONS section would be advisable. Testing many sizes of caps - made of the various materials available and utilizing more than Just one kind of liner - at both shorter and longer intervals of storage in a wider range of con- ditions would give a much more complete and comprehensive picture of the effects of conditioning on release torque. than this study. limited by both time and availability of materials, was able to present. It would also be beneficial to test a wide range of application torques. Another area that might provide valuable results is a study on the effects of the contents of a bottle on release torque. taking the above cap and conditioning variables into account. Just as in the conditioning tests. it would be wise to increase the cap variations in the physical tests. at the same time adding more tOp loads and impact levels to the procedure. Both reduced and increased loads and impacts would be advisable. Other physical tests - such as shipping. vibration. incline- hmpact. and drOp tests - would also undoubtedly have effects on release torque. and studies of these are recommendable. All of the above tests could also be performed using lug caps or rolled-on caps. in addition to screw caps. 10. LIST OF REFERENCES Eli Lilly and Company. Indianapolis. Indiana. Telephone interview with.R.J. Schnorr. Package Design Chief. February. 196k. Glass Container Manufacturers Institute. East Lansing. Mich- iggfi. Personal interview with Dr. J.G. Turk. March. 19 . Owens-Illinois Glass Company, Toledo. Ohio. Personal inter- view with C.V. Mumford and R.K. Flaskamp, Closure Division. March. 196”. Muraski. T.C.. ”How'Much Torque Is Just Right.‘ Pack e Nngineerigg. Vol. 7. No. 8 (August. 1962). p. E5. Sheffler. B.J.. ”Evaluating Performance of the Screw Cap fer Glass Containers." Glass Packer. Vol. “0, No. 8 (Aug- ust. 1961). p. 32. Curtis. E.A.. ”Simple Methods and Good Judgment Help You Predict Package Shelf Life.“ Package Engineering. Vol. 7. NO. 5 (May, 1961). p.h5$ The Owens-Illinois Torque Tester Manual . Owens-Illinois Glass Company. Toledo. Ohio. Weinberg. G.H.. and Schumaker. J.A.. Statistics: An Intuitive .A roach. Belmont. California: Wadsworth Publishing Company. Inc.). 1962. Hughs. D.A.. ”Demands on the Bottle Today.” Glass Technology. Vol. 2. No. 3 (June. 1961). p. 101. Sheffler. R.J.. I'Screw Cap Problems.” Modern Packaging. Vol. 35. No. 10 (June. 1962). p. 150. #6 ROOM USE ONLY “Luigi {’31: U“ rm M243 ‘56 M-Gsifi-riOGS MAGIC! SEP ‘5‘ iii MAY221999 “ll HICHIGRN STATE UNIV. LIBRQRIES 3129301 1070590