TS 1109 .H76 Copy 1 DEPARTMENT OF COMMERCE Technologic Papers OP THE Bureau of Standards S. W. STRATTON. Director No. 194 A PRELIMINARY STUDY OF TEARING INSTRUMENTS AND TEARING TEST METHODS FOR PAPER TESTING BY PAUL L. HOUSTON, Associate Physicist Bureau of Standards JULY 27, 1921 PRICE. S CENTS Sold only by the Superintendent of Documents, Government Printing Office Washington, D. C. . WASHINGTON GOVERNMENT PRINTING OFFICE . 1921 DEPARTMENT OF COMMERCE Technologic Papers OF THE Bureau of Standards S. W. STRATTON, Director No. 194 A PRELIMINARY STUDY OF TEARING INSTRUMENTS AND TEARING TEST METHODS FOR PAPER TESTING BY PAUL L. HOUSTON, Associate Physicist il Bureau of Standards JULY 27, 1921 PRICE, S CENTS Sold only by the Superintendent of Documents, Government Printing Office Washington, D. C. WASHINGTON GOVERNMENT PRINTING OFFICE 1921 ! V -<"} <^ o"\ ^ LIBRARY OF C0NGKi§% ] SEP231921 / | POOOMENt) !OM Zhlk ^ A PRELIMINARY STUDY OF TEARING INSTRUMENTS AND TEARING TEST METHODS FOR PAPER TESTING By Paul L. Houston ABSTRACT In this technologic paper a study is made of the relative effect of different sizes of test samples on the tearing strength of paper. A great number of samples of commer- cial papers are torn on three different instruments, using different sizes of test samples and also the same sizes of test samples. Data are collected, accordingly, to show that the larger the test sample the greater are the values of tearing strength. The reason for this is brought out as fabric assistance, which is of considerable importance in the textile industry. The three instruments used in this study are a tensile-strength instrument and two types of instruments for determining the tearing strength of paper. These two types of instruments are called type I and type II. Type I is a recording instrument, while type II is a nonrecording instrument. A study is then made of these two types of tearing instruments for the purpose of investigating their accuracy and reliability, so that the results of this investigation may benefit the paper industry. Conclusions are drawn up to show that the type II nonrecording instrument is the more reliable of the two and is within a 5 per cent error on the majority of papers. CONTENTS Page I. Introduction 4 1. Statement of problem 4 II. Discussion of test methods 4 1. Practical tests with different size test samples 4 2. Practical test with same size test samples 6 3. Fabric assistance 6 III. Discussion of a recording instrument 7 1. Description of instrument 7 2. Calibration test 7 3. Performance test 8 IV. Discussion of a nonrecording instrument 10 1. Description of instrument 10 2. Error caused by beam position 10 3. Error due to impact of water 13 4. Performance test 16 V. Conclusions 17 3 4 Technologic Papers of the Bureau of Standards I. INTRODUCTION 1. STATEMENT OF PROBLEM For some time it has been a recognized fact in the paper industry that there is a need for an instrument that will give numerically the tearing strength of paper. Considerable work has been done along this line at different laboratories. The result has been that two types of tearing instruments have been invented and are used to some extent. One of these types is a recording instrument, giving a curve showing the maximum and minimum tearing strength of paper and the results of five initial tears. The other type gives merely the maximum tearing strength by weighing or measuring the load applied to the paper to tear it. Both types have their good and their bad points, which will be brought to light in the following study. II. DISCUSSION OF TEST METHODS 1. PRACTICAL TESTS WITH DIFFERENT SIZE TEST SAMPLES Before taking up a study of the instruments themselves it must be remembered that the tearing strength of paper depends a great deal on the size of the test sample. For instance, a series of bag papers that had previously been tested for weight (in pounds) of the standard size ream 25X40 — 500 and for bursting strength were torn on the tensile-strength instrument described in Bureau's Circular No. 107, pages 15-17, and on the two types of tearing instruments mentioned above. For convenience, let the record- ing instrument be called type I and the nonrecording instrument type II. The test samples were cut 1 by 8 inches for the tensile- strength machine, 2% by \y% inches for type I instrument, and 4 by 10 inches for type II instrument. All samples were torn the machine direction, beginning at a point just halfway across the width of the test sample, and an average of 10 tests was taken for each result. The results in Table 1 were obtained. It might be well to mention here that all tests in this study were conducted in a constant-temperature and humidity room, where all test samples before they were used were conditioned for two hours at a temperature of 70 F and a relative humidity of 65 per cent. From these data it is to be noted that the larger the test samples are in size the higher are the values of tearing strength. Study of Tearing Instruments TABLE 1.— Relative Effect of Different Size Test Samples on Tearing Strength of Bag Paper, Using Tensile Strength, Type I, and Type II Instruments [Test samples: Tensile strength, 1 by 8 Inches; type I, 2 3 /« by l'/e Inches; type II, 4 by 10 Inches] Weight of the standard- size ream, 25X«0— 500 Bursting strength Tearing strength Bag paper Identification numbers Tensile strength Type I Typen 16 081 Pounds 32 44 52 49 53 61 32 46 77 77 110 147 39 45 62 45 38 51 65 43 51 72 52 76 Points 11 23 18 29 36 36 32 36 54 31 94 125 21 25 33 28 22 33 47 24 28 46 30 41 Grams 30.8 58.3 47.5 86.6 91.6 108.3 50.0 100.0 133.3 97.5 270.8 433.3 38.3 43.3 83.3 79.1 101.6 116.6 165.6 78.3 61.6 136.6 73.3 129.1 Grams 15.75 34.80 38.40 58.80 59.80 63.40 33.50 43.20 90.60 55.50 188.00 256.50 22.40 30.80 62.80 52.80 73.10 74.60 108.00 63.50 39.50 116.50 49.90 88.50 Grams 33.6 16 082 76.4 16 083 68.4 16 084 96.2 16 085 113.2 16 086 128.6 16 087 64.0 16 088 118.0 16 089 179.2 16 090 149.2 16 091 338.4 16 092 471.4 16 093.... 60.6 16 094 73.2 16 095 135.6 16 096 94.6 16 097... 84.0 16 098... . 121.6 16 099 179.2 16 100 85.2 16 101 83.0 16 102.. . 186.6 16 103 96.2 16104 166.8 In order to illustrate this point further, a series of writing papers that had already been tested for weight (in pounds) of the standard size ream 25 X 40 — 500 and for bursting strength was torn on the tensile-strength machine and type II nonrecord- ing instrument. Test samples were cut 1 by 8 inches for the tensile-strength apparatus and 4 by 10 inches for the type II instrument, and an average of 10 tests was taken for each result. The data in Table 2 show the same results as in Table 1. TABLE 2.— Relative Effect of Different Size Test Samples on Tearing Strength of Writing Paper, Using Tensile Strength and Type II Instruments [Test samples: Tensile strength, 1 by 8 inches; type II, 4 by 10 inches] Writing paper identifi- cation numbers Kind of writing paper Weight of standard- size ream 25X40— Bursting strength Tearing strength Tensile Type II 500 strength Pounds Points Grams Grams 86 55 156.9 181.7 46 19 32.4 54.5 24 10 29.0 34.2 55 27 50.2 77.8 44 19 28.5 44.8 10 5 13.8 16.8 98 72 167.2 230.4 Ratio of tensile strength to typen 14 731. 14 732. 14 733. 14 734. 14 735. 14 736. 14 737. Bond Writing. . Copying. . Printing . ....do.... Manifold . Ledger Per cent 86.4 59.5 84.8 64.5 63.6 82.2 72.6 6 Technologic Papers of the Bureau of Standards 2. PRACTICAL TESTS WITH SAME SIZE TEST SAMPLES On the other hand, if the test samples are cut the same size, very good check results will be obtained. Note the following results in Table 3 on the above series of long-fibered bag papers when test samples were cut 1 by 8 inches for both tensile-strength and type II instruments. An average of 10 tests was taken for each result. TABLE 3. — Relative Effect of Same Size Test Samples on Tearing Strength of Bag Paper, Using Tensile Strength and Type II Instruments [All test samples were 1 by 8 inches] Weight of the stand- aid-size ream 25X 40—500 Bursting strength Tearing strength Ratio of tensile Bag paper Identification numbers Tensile strength Type II strength to Typell 16081 Pounds 32 44 52 49 53 61 32 46 77 77 110 147 39 45 62 45 38 51 65 43 51 72 52 76 Points 11 23 18 29 36 36 22 35 54 31 94 125 21 2S 33 28 22 33 47 24 28 46 30 41 Grams 20.8 58.3 47.5 86.6 91.6 108.3 50.0 100.0 133.3 97.5 270.8 433.3 38.3 43.3 83.3 79.1 101.6 116.6 166.6 78.3 61.6 136.6 73.3 129.1 Grams 21.0 58.0 49.6 90.2 93.8 113.0 50.0 92.0 142.2 104.8 271.0 442.8 41.2 44.8 90.4 74.2 96.6 113.4 160.0 70.6 66.8 139 6 74.0 129.6 Per cent 99.0 16 082 100.2 16083 95.8 16 084 96.1 16 085 97.7 16 086 96.0 16 087 100.0 16 088 108.6 16 089 93.8 16 090 93.0 16 091 99.7 16 092 97.8 16 093 93.0 16 094 96.6 16 095 92.3 16 096 106.6 16 097 102.6 16 098 102.7 16 099 104.0 16100 110.8 16 101 92.3 16 102 97.8 16103 99.1 16 104 99.7 3. FABRIC ASSISTANCE After studying the action of the paper as it was torn on each instrument it became evident that each fiber in the path of the tear received assistance from all the fibers adjoining as far as the edges of the test samples. Also, it was noted that the larger the test sample was the greater was the amount of assistance that the fibers in the path of the tear received from the adjoining fibers, provided, of course, that the beginning of the tear was always the same distance from the end of the test sample and just halfway across the width of the test sample. In the textile industry this effect is called fabric assistance and is of considerable importance. Bureau of Standards Technologic Paper No. 194 Fig. i. — The recording instrument, type I Bureau of Standards Technologic Paper No. 194 Fig. 3. -The nonrccording instrument, type II Study of Tearing Instruments 7 III. DISCUSSION OF A RECORDING INSTRUMENT 1. DESCRIPTION OF INSTRUMENT Since it is found that it is necessary to cut test samples the same size in order to compare the tearing strength of papers, it is now best to proceed to a study of the instruments themselves. The recording instrument, type I, Fig. 1, is the type of instrument preferred for laboratory use. However, such an instrument should be very delicate, sensitive, and accurate. There should be the least possible amount of friction between the chart and pen point. The recording arm should move freely, with as little friction as possible in the bearing. However, there seems to be no instrument of this caliber on the market to-day. The instrument under study has the appearance of a delicate, sensitive, and accu- rate little machine, but has certain defects which will be discussed later. It is composed of a sliding plate on which the chart rests and a recording arm which moves on a pin-slot bearing and holds a small glass capillary pen with a platinum point. The test sample of paper is cut by means of a die in such a way as to give five initial tears and a curve showing the maximum and minimum tear. Small angular slits in the path of the tear make possible these five initial tears, and a slit down the center of one end of the sample makes it possible for the starting point of the tear to be always the same distance from the edges of the paper. The test sample is placed on the instrument so that one half of the slitted end is fastened to a pin on the sliding plate, while the other half is fastened to a pin on the recording arm. The instrument is motor driven, and as the plate slides to one side the paper is torn and the recording arm is forced down to register on the chart the load in grams necessary to tear the paper. Two different weights may be suspended at three different positions on the projection of the recording arm as factors or multiples of the gram readings on the chart. 2. CALIBRATION TEST Before operating any instrument of this kind it is always best to calibrate it to determine its accuracy. .This was done by hanging dead- weights on the recording arm. With pen point barely touching the chart, there was so much friction in the pin- slot bearing that it was almost impossible to calibrate the instru- ment at all, as the swinging pen point would stop almost anywhere. However, the following correction curves, Fig. 2, were obtained 8 Technologic Papers of the Bureau of Standards for factors, 5, 10, 20, 50, and 100, but they are not reliable for reasons mentioned above. It was impossible to calibrate the in- strument for any factor below 5 . 3. PERFORMANCE TEST After attempting' to calibrate this instrument it was thought best to tear on it a number of samples of paper, using the different weights to represent factors. The same bag papers were used as before, and from them three identical sets of test samples (num- bered 1, 2, and 3) were cut 2^ inches long and 1 y& inches wide. Each test sample was torn in the machine direction, and the average of the five initial tears was taken as the tearing strength of the paper. In the following data, Table 4, it is to be noted that different results were obtained for different factors used. The values which are given in this table were directly observed and have not been corrected by use of the calibration curves of Fig. 2. TABLE 4.— Relative Effect of Different Factors on Tearing Strength of Bag Paper, Using Three Identical Sets of Test Samples and Type I Instrument [All test samples were 2% by l' , Inches) Setl Set 2 Set3 Bag paper Identification numbers Factors Tearing strength Factors Tearing strength Factors Tearing strength 16081 2 5 5 5 5 10 5 S 20 20 20 50 5 5 20 5 10 10 20 10 10 20 10 20 Grams 15.7 30.0 29.1 41.7 42.3 64.2 33.5 41.6 90.6 54.6 177.2 256.5 22.4 29.6 62.8 43.1 73.1 74.6 108.0 63.5 39.5 101.2 49.9 96.2 5 10 10 10 10 20 10 10 50 50 50 100 10 10 50 10 20 20 50 20 20 50 20 50 Grams 17.8 34.8 35.3 58.8 59.8 63.4 24.5 43.2 105.5 55.5 188.0 283.0 30.8 30.8 65.5 52.8 61.0 74.0 114.0 37.2 44.6 116.5 41.2 88.5 10 20 20 20 20 50 20 20 100 100 100 Grams 16082 16 083 16084 16 085 16086 16087 16088 38.8 16089 16 090 72.0 189.0 16 091 16092 16093 20 20 100 20 50 22.2 36.7 75.0 46.2 76.0 16094 16095 16 096 16097 16098 16099 100 152.0 16100 16 101 50 100 50 100 56.0 146.0 49.0 101.0 16102 16103 16104 Study of Tearing Instruments 10 15 30 25 30 35 ACTUAL LOAD IN GRAMS 20 40 SO 80 100 120 140 160180. ACiTJAL LOAD IN ORAMS -,.- HT-»r-':!-'"ri:.:T:'"fT T ~~T"'" r ! n — l ?..::i.:; ; .:'j4--''-i---t-n-f 5 10 J.5 30 35 30 35 40 45 ACTUAL LOAD IN GRAMS 50100 150 300 350 300 35040C ACTUAL LOAD IN ORAMS SOO 11^,::,.,. -, 700 tr g6oo| 9 500 1 e- "+00 g 200 100 ■\v/:.: : .; -. !■■■ f\ ^iJKi;:^!^ 200 300 400 500 60070C ACTUAL LOAD IN GRAMS Fig. 2.— Diagrams showing correction curves for different factors (recording instrument, type I) 52142°— 21 2 io Technologic Papers of the Bureau of Standards IV. DISCUSSION OF A NONRECORDING INSTRUMENT 1. DESCRIPTION OF INSTRUMENT There will now be taken up a study of the nonrecording instru- ment, type II, Fig. 3, which is a simplified instrument adapted for mill use. An instrument for mill use should be accurate, at least within 5 per cent, should be foolproof and yet so simple of opera- tion that paper-mill machine tenders can handle it, and should check itself under standard conditions within the variation of the strength of paper itself. The instrument under study seems to come near being an instrument of this caliber when it is used for bag paper or heavy writing paper. It was built to test the tear- ing strength of bag paper in the machine direction, since paper bags usually tear in this direction. It could be used on writing paper to determine the tearing strength in either direction. The instrument consists of a beam balancing on a knife-edge (the two arms of the beam being equal). One end of the beam holds a 300-cm 3 glass into which water as the load may be poured from a 500-cm 3 burette until the paper tears. At the other end of the beam one half of one slitted end of the test sample is clamped, while the other half and other end of the test sample are clamped against a vertical plate opposite. A special die is used for cutting a slit and eyelet hole at each end in the middle of the test sample. The cubic centimeter or gram readings are taken from the burette at the end of each tearing operation. These readings indicate the maximum tearing strength of the paper. Two weights may also be used and placed, if necessary, at definite intervals on the glass side of the beam. Each weight at each position represents a certain load in grams. (It is well to state here that due to the facts that the two arms of the beam are equal, and that there is very little friction because of the knife-edge, the force at the tearing end of the beam is actually equal to the weight of water in the glass.) 2. ERROR CAUSED BY BEAM POSITION Before using this instrument extensively for tearing it was thought best to study it from the standpoint of physics. For instance, it was decided to determine the error due to changes in moment of force caused by changes in position of the beam during the process of tearing. Also, were there errors caused by the force exerted by the falling stream of water ? Let there first be made a study of the different moments of force caused by different positions of the beam. Note the following diagram of beam, Fig. 4. Study of Tearing Instruments ii Let BC represent the beam and AR represent a pointer which is fastened at the center A of the beam and at right angles to it, and which determines by its position in respect to the graduations below, numbered o, i, and 2, whether the beam is balanced (and in this case horizontal) or is one or two graduations off balance. It is evident that when the beam is one or two graduations off balance the moment of applied force decreases as the number of graduations off balance of the beam increases. In order to find the amount of the error introduced by the changes in moment of force caused by changes in position of the beam, the moment of the applied force was calculated for three different positions of the beam, using 100 g as the weight of water in the glass. Referring to the diagram, Fig. 4, let MF and NG represent the two positions of the beam BC when it is one and two graduations off balance, as indicated by the pointer AR, which at the same time ^N^- // // // / I FlG. 4. — Diagram showing change in effect of weight of water with change in position of beam {nonrecording instrument, type II) takes the respective position A U and A V. By drawing the dotted line US perpendicular to A U at point U and dotted line VT per- pendicular to AF at point V, and by extending line AR to inter- sect dotted line US at S and dotted line VT at T, right-angled triangles A US and AVT are formed. By drawing dotted lines GE and FD from points G and F and perpendicular to A C, right- angled triangles ADF and AEG are formed. Of these triangles mgle D A F = angle SA U, and angle EA G = angle TA V . Distances \R, US, VT, and AC have been measured very carefully on the yp'e II apparatus, as follows: AR = AU = AV= 8.9 cm; US= .48 cm; VT= .96 cm; ^£ = 35.56 cm. 12 Technologic Papers of the Bureau of Standards Referring to triangle A US, US -rjj = tan angle SA U. ( i ) Substituting in (i) 0.48 for US and 8.9 for AU, we get: ^ = tan angle SAU; 8.9 & or, 0.0539 = tan angle SAU. Referring to a table of trigonometric functions, it is found that if tan angle SA U = 0.0539, cos angle SAU = 0.9986. , (2) Since angle SAU = angle DAF, by substituting in (2) cos angle DAF for cos angle SAU, we get: Cos angle DAF = 0.9986. Referring to triangle ADF, AD cos angle DAF = -^ • (3) Substituting in (3) 0.9986 for cos angle DAF, we get: AD -^ = 0.9986; or, AD = AF X 0.9986. (4) Since AF = AC = 35.56, by substituting in (4) 35.56 for AF, Weg6t: AD = 35.56x0.9986; or, 4D = 35.5i. When the beam is in the position represented by MF and the pointer is at the position of graduation, number 1 , the moment of force is represented by the formula: Moment of force = AD x weight of water. (5) Since 100 g were taken as the load in this case, by substituting in (5) 100 for weight of water and 35.51 for AD, we get: Moment of force = 35.51 X 100 = 3551. Referring to diagram, Fig. 4, when the beam is in the position represented by NG and the pointer is at the position of graduation, number 2, the moment of force is represented by the formula: Moment of force =AEx weight of water. (6) By using the triangles AVT and AEG and angles TAV and EAG the determination of AE is exactly the same as for AD and is found to be 35.35. Study of Tearing Instruments 13 Since 100 g were taken as the load in this case, by substituting in (6) 100 for weight of water and 35.35 for AE, we get: Moment of force = 35.35 X 100 = 3535. Referring to the diagram, Fig. 4, when the beam is in the hori- zontal position represented by BC and the pointer is at zero graduation, the moment of force is represented by the formula : Moment of force = AC x weight of water. (7) Since AC = 35.56 and since 100 g were taken as the load in this case, by substituting in (7) 100 for weight of water and 35.56 for AC, we get: Moment of force =35.56 x 100 = 3556. Consequently, there are the following moments of force at zero graduation and at graduations numbered 1 and 2 : Moment of force at o = 3556; Moment of force at 1 =3551 ; Moment of force at 2 =3535. From these results it can be seen that the error due to the different moments of force caused by different positions of the beam during the process of tearing is less than 1 per cent, which is very small. Very few testing instruments of greater accuracy than this are built. It is well to state here that due to the fact that at the tearing end of the beam the pull can not be exactly perpendicular to the beam in any position and that the angle of pull will vary with the position of the beam, another error exists. However, this error is very small, for the reason that the vertical plate X Y (Fig. 4) which holds the test specimen of paper is situated so close to the end of the beam that there is very little clearance between the end of the beam and the vertical plate. Conse- quently, due to this fact and due to the fact that the angular displacement of the beam during the process of tearing is never more than io° because the beam is so long, the angle of pull is always very close to the perpendicular. In this connection note the angles of pull, JNA, IMA, and HBA (Fig. 4), which are very close to right angles. H represents the starting point of tear, which is halfway between the end B of the beam in initial position and the lower end Y of the test sample. 3. ERROR DUE TO IMPACT OF WATER. Let there now be taken up the second question : Are there errors caused by the force exerted by the falling stream of water? In practically all cases the paper began to tear at the balancing H Technologic Papers of the Bureau of Standards position of the beam, and the water was shut off immediately by turning of the stopcock which closed the outlet from the burette, and the reading from the burette was taken. Consequently, it was decided to measure the distance from the end of the burette to the surface of the water in the glass at different applied volumes, as indicated by the graduations on the burette. This was done because the different velocities caused by the falling of the water from the burette through different distances to the surface of the water in the glass were the controlling factors in determining the error due to the impact of water. During this operation the beam was held firm in balancing position and a piece of aluminum attached to a thread was used to make the measurements. The following measurements were obtained : Volume applied, cm" 25 75 "5 Distance, cm 16. s 14. 6 13-0 Volume applied, cm 3 J75 225 275 Distance, cm II. 4 ... IO. 2 8.9 Then a small piece of aluminum about the size of a 10-cent piece was attached by means of very small wires to the end of the beam in place of the glass. This was done in such a way that the water from the burette fell directly on the aluminum. The small size of the aluminum prevented any water from remaining on its surface. The forces were then measured experimentally by placing small laboratory weights on the other end of the beam to balance the force of the water from the burette on the alumi- num. The forces were obtained for heads of water in the burette at different graudations (on the burette) corresponding to the above applied volumes (since the applied volumes are deter- mined by the graduations on the burette), and the burette was lowered at the end of each experimental operation so that the distances between burette and aluminum were the same as the above at the respective applied volumes. The same operations were repeated continuously until close checks were obtained. Ten readings were taken at each head of water and an average of the 10 made. The results which were obtained are shown in Table 5. (It might be added here that the force of impact on the aluminum disk is not exactly the same as the force of impact on the water in the glass, because the energy dissipated is not the same in both cases. However, the use of the aluminum disk would seem to give results that are sufficiently accurate for a study of an instrument of this type.) Study of Tearing Instruments TABLE 5.— Forces at Different Heads 15 Readings Forces In grams for heads at graduations of — 25 cm 3 75cm s 125 cm 3 175 om> 225 cm 3 275 cm' 1 4.80 4.80 4.80 4.85 4.90 4.90 4.90 4.95 4.95 4.95 4.30 4.30 4.40 4.40 4.45 4.45 4.40 4.30 4.40 4.40 3.80 3.80 3.80 3.80 3.80 3.80 3.80 3.80 3.80 3.80 3.50 3.50 3.50 3.50 3.50 3.40 3.40 3.40 3.40 3.50 3.30 3.30 3.20 3.20 3.30 3.20 3.20 3.30 3.20 3.20 3.10 2 3.10 3 3.00 4 3.10 5... 3.00 6 3.00 7 3.10 8 3.00 9 . 3.00 10.. 3.10 4.88 4.38 3.80 3.46 3.24 3.05 From the above results a correction curve was drawn, which is presented in Fig. 5. 35 _50_. 75 100 125 150 175.300 335 250 27S VOLUME APPLIED IB 0UBTO CENTIMETERS Fig. 5. — Diagram showing positive corrections at different applied volumes as indicated by head graduations in Table 5 (nonrecording instrument, type II) It can easily be seen from Fig. 5 curve, by dividing the ordi- nates by the corresponding abscissas, that the errors caused by the force exerted by the falling stream of water are as follows: Volume applied, cm 3 Error, per cent 25 x 9-5 2 75 5-84 125 3.04 Volume applied, cm 3 i75- 225. 275- Error, percent ■ 1-97 1.44 1. 10 Practically all the samples of bag paper that were tested on this instrument tore above the 75 g load when the test samples were cut 4 by 10 inches, which is the size recommended and speci- fied by the inventor. This has previously been presented in Table 1. Consequently, the instrument is within the 5 per cent error on the majority of these grades of bag paper. For writing paper, such as manifold, lightweight printings, and lightweight writings, and for all lightweight short-fibered papers, the error is greater than 5 per cent, as is shown in Table 2. The term "light- weight ' here indicates that the weight in pounds of the standard i6 Technologic Papers of the Bureau of Standards size ream 25 X 40 — 500 is less than 50. For practically all writing papers that are heavier than 50 pounds, and for practically all weights of bonds and ledgers, the error would be less than 5 per cent. These errors might be decreased somewhat if the burette were lowered nearer the glass than was the case in the above experiments. However, the decrease would be very small and the results would be comparatively the same. 4. PERFORMANCE TEST After studying this instrument from the standpoint of physics a number of samples of paper were torn on the machine in order to discover whether the instrument would check itself under standard conditions within the variation of the strength of the paper itself. Two sets of test samples, numbered 1 and 2, were prepared from the same bag papers as were used before and torn in the machine direction with very good check results, as are shown in Table 6. All test samples were cut 4 by 10 inches, and an average of 10 tests was taken for each result. From these results it can be seen that it is possible to repeat tests on the same grade of bag paper with this instrument and get check aver- ages, and that errors due to side pull at the tear and to personal errors in stopping the flow of water are relatively very small. TABLE 6.— Relative Effect of Use of Two Identical Sets of Test Samples on Tear- ing Strength of Bag Paper, Using Type II Instrument Bag paper identifi- cation numbers Tearing strength Ratio of set 1 to set 2 Bag paper identifi- cation numbers Tearing strength Ratio of set 1 to set 2 Setl Set 2 Setl Set 2 16 081 Grams 33.6 76.4 68.4 96.2 112.2 128.6 64.0 118.0 179.2 149.2 338.4 471.4 Grams 32.4 74.0 69.0 96.6 110.0 128.6 64.0 117.4 185.6 147.8 335.4 477.0 Per cent 103.7 103.2 99.2 99.6 102.0 100.0 100.0 100.4 96.6 101.0 101.0 99.8 16 093 Grams 60.6 73.2 135.6 94.6 84.0 121.6 179.2 85.2 83.0 186.6 96.2 166.8 Grams 56.8 73.0 132.4 92.2 85.4 122.0 176.8 83.6 83.2 186.8 96.0 167.8 Per cent 106.7 16 082 16 094 16 083 16 095 102.2 16 084 16 096 102.6 16 085 16 097 . 16 086 16 098 . . . 99.6 16 087 15 099 101.4 16 088 16 100 16 089 16 101 16 090 16 102 .. 16 091 16 103 100.1 16 092 16 104 99.4 Study of Tearing Instruments 1 7 V. CONCLUSIONS Conclusions to be drawn from this study of type I and type II instruments are that neither one of them has been perfected enough for general commercial use. Type I, the recording in- strument as now manufactured, is neither a delicate, a sensitive, nor an accurate piece of apparatus, since the amount of friction in the pin-slot bearing and the friction between the pen point and the paper chart do not allow careful accurate calibration. Since the test results obtained by using different factors will not check, it would indicate also that there is a defect in the mechanism of the instrument. Type II, the nonrecording instrument, is a fair in- strument for bag paper, since most bag papers tear above the 75 cm 3 or gram mark. However, many recommendations could be made, such as stronger clamps, a better device for cutting test samples an exact size, a better device to control the distance of tear of each sample and to keep the distance the same for all samples, and the elimination of the use of weights on the glass side of the beam, by which the force is immediately applied instead of being gradually applied as in the case of the water. The general idea of both instruments is good. The type I recording instru- ment gives a curve showing the maximum and minimum tearing strength of paper as well as five peaks in the curve showing the results of five initial tears. No fault can be found with a curve that represents the maximum and minimum tearing strength, and the initial tearing strength is just what is wanted for writing papers. However, there is great doubt whether the five peaks in the curve of type I instrument actually represent five initial tears. The fibers very near the edge of an angular slit in the test sample (or perhaps halfway between two angular slits between which the paper is torn) may be stronger than those fibers at the very edge. In such a case the result would be a rising curve and the peak would not represent the initial tear but the tearing strength of the fibers near the edge or halfway between two slits in the paper. During the work on this instrument it was noticed that some peaks in the curves were double- toothed or double-peaked. In this case the first peak probably more nearly represents the initial tear. This may be a very fine distinction, and yet if this recording instrument is going to be used as a laboratory instru- ment it must be accurate to the highest degree. The type II non- recording instrument, on the other hand, gives merely the maxi- mum tearing strength of paper, which is all that is necessary for a 1 8 Technologic Papers of the Bureau of Standards mill test. All tests on the two instruments were made in the machine direction of the paper (the direction in which paper moves on the paper machine) for the reason that better comparative results could be obtained in this way. Most papers are much stronger in the cross direction than they are in the machine direction. Since this is true, if you attempt to tear these papers in the cross direction, the direction of the tear as a rule turns to the machine direction soon after the beginning of the tear. This change in the direction of the tear never occurs when the paper is torn in the machine direction. Since in practically all grades of paper a good tearing strength in the machine direction is just as essential as a good tearing strength in the cross direction, and since better comparative results are obtained by tearing paper in the machine direction, it would seem that all tearing tests should be made in the machine direction. Washington, January 5, 1921. LIBRARY OF CONGRESS