Clothing Test Methods, Edited by L.H. Newburgh (Physiological Tests) and Milton Harris (Physical Tests) of Subcommittee on Clothing of the National Research Council (U.S.A.) (1945)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

66 CLOTHING TEST METHODS equilibrium. It should, be remembered, however, that although the number of clo required increases more rapidly with progressively lower temperatures; neverthe- . less, a far greater amount of energy must be supplied to make up a small clothing deficit when the temperatures are near those of the body. The two dotted lines on Graph No. 4 represent the amount of additional energy needed to keep in equi- librium if the clothing is deficient by \ clo and 1 clo respectively. At low temperatures, to*cite an extreme example, a deficiency of ^ clo may mean that a man need raise his activity by only 10 Kg. Cals/M2/hr. (an amount which is easily achieved by muscular exercise or slight heightening of activity). If he is i clo deficient at high temperatures, he may need to increase his activity by as much as 40 Kg. Cals/M2/hr. This apparently anomalous occurrence is due •to percentage difference in insulation and holds true for small deficiencies of clothing and low activities only. This does not mean that it is safe to issue inadequate clothing, but it does mean that adequate clothing will be safe when the temperature falls a certain amount below average. For temperatures below 14°F., either the time of exposure must be stated or the standard activity of the body must be increased. Graph No. 4 shows the amount of increase necessary with the 4 clo uniform. Graph No. 2 shows the rela- tionship between severity of exposure and time limitations for various clo values. It is assumed that the body may lose only 80 Kg. Cals/M2 before reaching a state of unbearable cold. This figure is somewhat arbitrary but is chosen to represent a suitable average since in combinations of clothing which are poorly balanced, the breakdown may occur sooner due to local cooling; while in those which give equal distribution throughout, the length of time during which he can withstand the cold may be extended for a considerable period beyond that indicated. 2. Adequate Protection for Sleeping Men The following definition of adequate protection for sleeping men is proposed: Sleeping gear should be considered adequate if it provides eight hours comfortable sleep at the average minimum temperature of the environment. This is on the assumption that heat production will be 40 Kg. Cals/M2/Hr. and that total h.eat debt will not exceed 40 Kg. Cals/M2. The debt value of 40 Kilogram Calories has been recommended by the Harvard Fatigue Laboratory in Report No. 19 and substantiated from records of comfort and discomfort made by the Climatic Research Laboratory at Lawrence, Massachusetts. It is estimated to be the amount which can be lost before a sleeping man will become restless in an ef- ' fort to build up a higher metabolic rate. The conditions for applying this standard of adequacy are the same as for general activities. The sleeping gear would be more than adequate on the warmest nights; it would, however, provide a shorter, but still reasonable, night's rest when the temperature dropped considera- bly below the average minimum. Inefficient cold spots or drafts due to badly designed sleeping gear may substantially decrease the adequacy of the bag even though the insulation as de- termined by clo values would indicate comfort for most of the body. Graph No. 3 shows the relationship of"clo 'values, temperature,, and time for the protection of a sleeping man. 3. The Effect of Moisture on Clothing Protection The definitions of adequacy as given are theoretically applicable irrespec- tive of the dryness or dampness of the material. That is, if the man is able to remain in equilibrium indefinitely while standing or if he is able to sleep

DEFINITION OF ADEQUACY OF CLOTHING IN COLD CLIMATES 67 comfortably for eight hours, the clothing and equipment should be adequate under any normal conditions. This is not true, however, of any given item or combina- tions of items which has as yet been constructed. The charts and subsequent dis- cussion of clo values apply to dry clothing only. At the present time, no stand- ard of the decrease in adequate protection afforded by damp clothing can be analyzed. Preliminary tests show that sleeping bags preconditioned at a high relative humidity afforded rest for only half the time which was provided by sleep~ ing bags preconditioned at a low relative humidity. Furthermore, the addition of a minimum amount of sweat to an arctic uniform over a ten day period appears to reduce the value of the equipment by about one clo. It is obvious that evaporative heat loss substantially decreases the value of the insulation. These are, however, only initial studies in a wide field. The humid and wet climatic zones have been mapped, but more research is necessary before the corresponding adequate clo values can be determined. Adequacy for ambient air temperature close to and higher than body temperatures will be discussed in a later paper.

68 CLOTHING TEST METHODS A >> fc -P O O bp TH d fH a d a o <4H C > 0 cd OT T-\ 3 *rH 03 TH cd *rH Q-l o oo r^oo ^-H ojtfN m m t— o\ co H -P fH 0 1 It^^J-rHOChOAO^OrH OJKN-^-in 0 O C £> P t--rHHrH rHrH rHrHrHrH 0 ,0 TH TH 0 0) Cd rH O fe -P S TH rH 0 3 0 S a, O 3 0 r-tlC\l rlfc\l CO rH rH O H|« H rHOJ CVJKNr<>KN r^NI^KNKN H 0 cd FH -P O O O 60 -H d I~H *}-i fj ** Qi J-^ ^^ *rH *rH o jn a d a o CQ cd co T-I p tH •^ 03 iH cd tH Cn ooj oin-^-KNHO ONCT\ c— cr\t^-o\ EH 03 H -P- fn 0 1 mo CT\OO COCO 0000 OOO^ OrHOJKN W 0 O d fi Q rH H rH rH rH H . S O rO TH -H 0 0) Cd rH O P S -P S -H rH K 030 5 S D, M" pr"] Q\ t^ O 3 O H|OI rHrH OJOJ l^rA rAKN KAKNKSrA "^ H rH <J 0n) *' to 0 fn -H H O bO > O ,0 (Vn fj H 6 o) co >> -H a o EH W r< 8 d 3 3 cd 03 -H -H 0 0 S O CO -H a) fn -P 3 OJ OJVO t— 00 CT\O OH CT\OO VOin-^KN g IS) 03 i-H -P f> Cd rH inMOVOVOVOVOt--t^t^t— CO CTvOHCVJ »S (D O d TH 3 cd W 0 fl vH rH 0* > EH 0 cd -P -H 0 rH rH. H O g 5 £5 -P cd 3 ^ c S 0 S 0* <! g 0 0 m fc -p 0 o a) 0 TH » 0* rH rH O 0 0 Cd 303 o H|W H HOJ CVJKN r^^- ^- ^- ^-^-•^^- 1 i m • fTi [V. fV. •H K &,&, fefr, fefe fefc -fe O&<(OJ CrH°O O 0 oo oo oo oo (1,0 r^ojr<A-=i' &, Q 0 W) (^ t~-00 O>O rHOJ r<>^J-O^l- rHI 1 1 O d 3 pti t~- voinin-^-KSojrHini i o 1 •rl -P O 1 1 1 ^t +J cd v£> I 1 1 1 1 1 1 I 1 I i i _p a/11 CO ... 5 0 TH ft -p S j a S 0OO OO OO OO OO OKNOJrH rH CO > ^ot~-ooo\ OH ojKN-=i-ir> ^-HOJKN 0 1 0 EH oooN-voinin^i-r^ojH 1111 PQ K d O rrj 'j (SI •d >> H Q In O r»S r»» r'S r»* H H O (>4FHFHFHF<O0 O "1 |nP> « Q P 0 BT3 g P 0 H ct) 0 •PfiHOHfn-PO 4J c £3 K Oed-HO00K H W£eS00>W £3 •

DEFINITION OF ADEQUACY OF CLOTHING IN COLD CLIMATES 69 ,888888 i-u 2 a. UK <t-SCEUJ u. B82SSSSS

70 CLOTHING TEST METHODS STANDING ACTIVITY— 75 K GRAPH NO. 2

DEFINITION OF ADEQUACY OF CLOTHING IN COLD CLIMATES 71 ACTIVITY - 40 KG CALS./M'/hr SLEEPING HEAT LOSS- 40 KG. CALS./I/ TW IDITIO* - !•(• lt4* GRAPH NO. 3

72 CLOTHING TEST METHODS Kg. cols./ME/hr. 10 20 30 40 50 60 70 80 90 100 110 120 130 140 GRAPH NO.

PART II Physical Teat Methods

CONTENTS Part II I. Introduction II. Testa for Physical Properties 1. Tensile strength ....'. 2. Tear strength 3. Abrasion testing 4. Shrinkage of wool hosiery; eff.ects of shrinkproofing treatments 5. Flexural rigidity 6. Flexural fatigue 7. Flame proofing 8. Flash resistance 9. Frictional properties 10. Water repellency 11. Buoyancy 12. Thickness, compressibility, and compressional resilience . . . . 13- Effective pressures in clothing 14. Thermal insulation 15- Resistance to diffusion of water vapor 16. Rate of drying 17- Air permeability I. INTRODUCTION This outline and discussion of textile testing methods has two purposes: (1) to promote the correlation of physical testing with the ultimate objective of clothing, the protection and service of the wearer; (2) to aid workers in sepa- rated laboratories to keep more closely in line in our work for the services, per- mitting more ready comparison and utilization of one another's work. It should be understood by anyone reading these test methods that we do not consider them final in the sense that each and every test method being used is regarded as completely satisfactory. In order to make some sort of evaluation and permit comparative . testing to go ahead without delay, we adopted certain procedures which appeared to be as good as any that we were able to perform at the moment, although at the same time we recognized deficiencies and undesirable features which call for eventual improvement. 75

76 CLOTHING TEST METHODS It is not the purpose of this outline to duplicate the standardizing work which has been promoted over many years by such agencies as the National Bureau of Standards, the A.S.T.M., and the A.A.T.C.C., and Canadian and United States govern- ment procurement agencies, but rather to aid in interpreting certain methods in terms of functional properties of clothing. Hence the actual methods will only be described when they are new, or different in important respects from standard meth- ods, or when experience has indicated that an intial interpretation of the results of standard methods is needed. The reader is referred to the following list of published standard methods, which will be referred to by number throughout the 'outline: 1. Federal Standard Stock Catalog, Section IV, Pt. 5 Textiles; general specifications, test methods. CCC-T-191a, April 23, 1937. Supt. Documents, U.S. Govt. Printing Office, 5 cents. 2. Textiles, testing and reporting (4th Edition, 1944), 45 p. Commercial Standard OS 59-44. Supt. Documents, U.S. Govt. Printing Office, 10 cents. 3. Quartermaster Corps Tentative Specification, Test Methods for Textiles, PQD No. 447, 26 June 1944. 4. Test Methods and Ratings, by A. C. Goodings and P. Larose, A.C.A.M.R. Report No. C-2419, S.P.C. Report No. 100, Feb. 18, 1943. (National Research Council, Canada.) 5. Canadian Government Purchasing Standards Schedule of Methods of Testing Textiles 4-GP-2-1942. 6. ASTM Standards on Textile Materials, Committee D-13, October 1943 (and anually), American Society for Texting Materials, 260 S. Broad Street, Philadel- phia, Pa., price $2.25. 7. Year Book of the American Association of Textile Chemists of Colorists-- annual--from the Secretary AATCC, Lowell Textile Institute, Lowell, Mass. The present outline is the result of consultations and correspondence be- tween the laboratories at Ottowa and Toronto and at Washington, and more than one person has reviewed and contributed suggestions to every topic. The following individuals have participated: Appel, W. D., Nat. Bur. of Standards, Washington. Burton, A. C., Banting and Best Dept. of Medical Research, University of Toronto. Fourt, Lyman, Textile Foundation, Nat. Bur. of Standards, Washington1. Goodings, A. C., Ontario Research Foundation, Toronto. Harris, Milton, Textile Foundation, Nat. Bur. of Standards, Washington1. Kitching, J. A., Banting and Best Dept. of Medical Research, University of Toronto. Labarthe, Jules, Mellon Institute, Pittsburgh. Larose, P., National Research Council Laboratories, Ottawa. Page, Edouard, Banting and Best Dept. of Medical Research, University of Toronto. Schiefer, H. L., Nat. Bur. of Standards, Washington. Smith, A. L., Textile Foundation, National Bur. of Standards, Washington1. Sookne, A. M., Textile Foundation, Nat. Bur. of Standards, Washington1. Webb, Peter, Office of the Quartermaster General, Washington 1. Present address: Milton Harris Associates, 1246 Taylor St., N. W., Washington 11, D. C.

PHYSICAL TEST METHODS 7? II. TESTS FOR PHYSICAL PROPERTIES 1. TENSILE STRENGTH. Standardized methods are now being used in most lab- oratories. These are described in several widely available references (1^, £, 3_, ^, or 6). Where strength data is needed only as part of an over-all evaluation, the less precise but more convenient grab test can be used. Where more precise results on, for example, the influence of a finish on the degradation of cellulose are needed, the strip method based on constant thread count is more desirable. Comment by H. F-. Schiefer; ...the main justification for the strip test is when it is desired to obtain elongation or load-elongation relations. Comment by P. Larose: Regarding tensile strength, it might be well to point out that the determination of strength is carried out not so much to gain an idea of the serviceability of the material, since we have no definite re- lation between strength and serviceability, but it is used rather as an index of quality, in particular when comparing several fabrics of the same type to be used for any one purpose where it is generally assumed that the fabric with the highest breaking strength is the most durable. Breaking strength is therefore included in specifications rather as a control of quality. 2. TEAR STRENGTH. Two methods have been standardized. The one usually used, in both Canadian and United States laboratories, is the tongue test, in which the pull is applied in the direction of the tear. One tear is made, between two tongues or strips of cloth, with the force applied in the tearing direction. Strip tear is often used as the descriptive term for this test. Tongue tear some- times refers to another type of tear in which both sides of the tongue tear rela- tive to the rest of the specimen, but this is not in current use in the labora- tories represented. The details of the single-tear test are given in references ^, 3_, or 6. In the trapezoid test, the pull is applied at right angles to the direction of the tear, as described in reference §_. These tests do not give a measure of what is needed to initiate a tear, but rather measure the force needed to continue the tear, once it has started. Comment by A. C. Goodings: In testing work which we have been carring out for the R.C.A.F., we have been using the tongue test method as a measure of the tear strength. This was a purely arbitrary decision and in most cases the trapezoid test probably duplicates more closely what actually occurs. The results obtained by the trapezoid and tongue methods on the same fabric are widely different as the following data indicate. Tear Strength Tongue Trapezoid Test Test lbs. lbs. Nylon fabric (parachute cloth) 5i 15 Rayon fabric 3i 5 Cotton fabric—tear in warp direction 2 3 Cotton fabric—tear in weft direction 3 6 However, either method, even though the results from the two do not paral- lel one another, will indicate whether a fabric has low resistance to tear or not and permit some sort of appraisa1.

78 ' CLOTHING TEST METHODS Comment by P. Larose: In the case of tearing strength, many of the remarks made for breaking strength also apply here, although tearing strength is not as widely used as breaking strength, since there is no reliable index to indicate the degree of serviceability to be expected from a fabric pos- sessing a certain tearing strength. As for breaking strength, it is of more value in comparing two or more fabrics with one another. 3. ABRASION OR WEAR TESTING. This type of test is in a developmental state, and can not be regarded as satisfying an investigator. What is needed is more cor- relation between the results of actual wear, and the indications of physical tests. In general, this requires that men wear the clothing under controlled conditions, a procedure which is being developed by the Quartermaster Corps at Camp Lee. At the National Bureau of Standards, several types of wear test are used, for special purposes. The carpet wear test machine, developed by H. F. Schiefer and A. S. Best (Research Paper 315, Bureau of Standards J. of Research £>, 927-937 (1931), Supt. Documents, 10 cents) has been used to measure wear on pile fabrics. The Taber Abrader and an abrasion device developed by Industrial Byproducts and Research Corporation, have been used in tests of Army wool hose. The Vyzenbek machine is also in use at the National Bureau of Standards. The following description is taken from "Test Methods and Ratings," by A. C. Goodings and P. Larose (ref. 4). A sample of' the material is abraded on a Wyzenbeek Abrasion Test Meter. The abrading medium is a No. 8 cotton duck. The pressure applied during the test is 3 Ibs. and the tension on the fabric is 6 Ibs. The test sam- ple is subjected to a maximum of 40,000 double rubs or less, dependent upon the resistance of the fabric to the abrasive action. Where a quanti- tative measure of the effect of the abrasion treatment is required, this is determined in" the case of most fabrics by the loss in tensile strength after a designated number of rubs. The criterion for the end point of the test is taken to be a loss in tensile strength of 4o# or over. It is us- ually possible by visual examination at the end of every 10,000 double rubs on the machine to determine whether a sample is likely to have lost appreciably in strength or not, and testing for strength is then carried out or alternatively the abrading action continued. The appearance of a fabric cannot entirely be disregarded, although in some fabrics it may be found that while the action of the abrading machine has disturbed the surface of the cloth and damaged its appearance, the strength of the cloth may not be appreciably impaired. In such cases the amount of rubbing required to cause an undesirable destructive effect on the appearance is to be taken as the measure of the resistance to abrasion. In the case of pile fabrics a more suitable method of measuring resistance to abrasion than the effect on tensile strength is by the wearing down, and in cases plucking, of the pile. Thickness measurements provide a meth- od for determining the abrasive action on the length of the pile.. Such cases require separate consideration. Material is tested in both warp and weft directions. Note: The strength tests before and after the abrasion treatment are made by the Grab Test Method. This necessitates cutting the samples of width greater than that normally used in the Wyzenbeek Machine when exam- ination by visual Inspection is adopted. The purpose of this Is in order

PHYSICAL TEST METHODS 79 that the sample may later be placed in the jaws of the tensile strength testing machine at right angles to the direction in which the rubbing was carried out. Comment by A. C. GoodIngs: Abrasion tests are notoriously poor in the reproduci- bility of results, and whether a No. 3 canvas duck is the best material to use as the standard abrading surface for all fabrics is very much open to question. Comment by W/C Peter Webb: Because of the desirability of predetermining the life span of garments by means of the results of laboratory physical analysis, the Research and Development Branch, O.Q.M.G., set up an N.R.C. project, the objects being as follows: a. To develop methods of measuring the degree of wear suffered by garments subjected to the test combat course. b. To apply the same measuring methods as proved by paragraph "a" above, to salvage clothing. c. To reduce new fabrics by means of laboratory apparatus to representative conditions of wear. Specially controlled tests on the combat course have just been completed; following analysis of the results by the N.R.C. research workers, it is hoped that agreement can be reached on methods for measuring the effects of wear. Correlation of results, obtained on machine testing for resistance to abrasion with those found in actual wear, is the first stage in selecting the proper type of measurement to be made on new fabrics. Comment by P. Larose: At the recent meeting of Committee D-13 it was decided to change the name to "wear testing." The comments of W/C Webb actually apply to wear testing in general and not simply to abrasion testing. Wear is the result of a combination of actions applied to the fabric. It varies with the location of the fabric on the body. In the past there has been too much tendency to look on wear as a single property like breaking strength, whereas in fact it is a combination of properties. For this reason it is questionable whether any single test will ever siffice as a criterion of resistance to wear. The weight to be given to the various factors affect- ing wear, such as strain, flexing, rubbing, etc., would have to be deter- mined. Moreover, the weight assigned to each of those factors will also vary with the location of the fabric on the body or on the use to which the fabric is put. The effect of light exposure, the effect of wetting, laundering, perspiration, all come into play in affecting wear. It would be difficult therefore to devise accelerative tests to give us the answer in a short time. The same experience has been met in connection with light fading experiments. Even the tests conducted at Camp Lee suffer to some extent from the acceleration. For these reasons I consider the search for comment "c" of Peter Webb as a "Will of the Wisp," except possibly for certain special items where the type of wear is well determined by usage. 4. SHRINKAGE OF WOOL HOSIERY AND EVALUATION OF SHRINKPROOFING TREATMENTS, by Milton Harris and Arthur L. Smith. A description of a preliminary form of a device for measuring wool socks was given in Textile Research, 14, 150-151, 1944 (May). This one device fits

80 CLOTHING TEST METHODS socks of different sizes, but size 12 is usually used. The following test methods have been used in a program for evaluating shrinkproofing treatments of Army cush- ion sole socks, at the Textile Foundation laboratory at the National Bureau of Standards. a. Measurement of socks: These measurements were made with a special de- vice designed by the Textile Section of the National Bureau of Standards. This consists of a sock form so constructed that the toe portion slides freely out of the heel portion. A sock is placed on the form so that the heel gore is aligned with a fixed line, clamped in position, and subjected to a 5-lb. load at the toe. The length of the foot at this load is read directly in inches. b. Shrinkage during washing: Washing is done in a standard, commercial, laundry wheel, 24 inches in diameter by 20 inches long, running at 30 to 32 rpm and reversing every five revolutions. Enough water containing 0.2$ sodium carbonate is used to keep the level 2 inches above the bottom of the inside cylinder. Enough soap (Spec. P-S-566) is added to give a running suds.2 Four pounds of socks are washed for 2 hours at l4o°F, given two 5-minute rinses at 120°F, centrifuged, and dried without ten- sion at room temperature. c. Alkali-solubility test; One-gram (-0.1 g) samples, taken from the leg portion of socks, are dried at 105°-110°C to constant weight, W, . Loosely stoppered 38 x 200 mm Pyrex test tubes containing 100 ml. of 0.4# (±0.004) sodium hydroxide are brought to temperature in a water bath maintained at 65° -0.1°C. At 10-minute intervals, a group of four weighed samples is immersed in the alkali, each sample in a separate tube. In order to wet the samples thoroughly, they are stirred once as soon as they are put into the tubes, and once more 10 minutes later. After a group of samples has been in the hydroxide solution for an hour, all four tubes are removed from the bath and the contents of each tube poured, as rapidly as possible, through a separate disc of 100-mesh copper screen soldered to the bottom of a section of brass tubing 1 3/^ inches in diameter. The tubes are rinsed once or twice and the samples are washed on the screens by means of a divided stream of running water. After about 5 minutes of washing the screens are placed on a towel to absorb some of the excess water, after which the samples are removed. The fuzz which separates from the main body of the sample can be balled together by running the finger around the screen several times. The wet samples are either partially dried with a stream of hot air or allowed to dry overnight at room tempera- ture. After this treatment they are again dried to constant weight, Wf. The general formula follows: Wi " Wf x 100 = % alkali solubility When applied to the leg of cushion sole socks, these values are doubled to get them on the basis of all wool. , A running suds ie a suds that comes halfway up the cylinder (always running downwards) when the top of the cylinder is moving away from the operator.

PHYSICAL TEST METHODS 8l d. Stress-strain teat: A wool fiber may be stretched 30# in water, relaxed for 24 hours, and restretched and the load required will be the same in both cases. If the fiber is chemically altered between the first and sec- ond stretching, the loads required will not be the same. This furnishes a sensitive method for estimating the damage done to the wool fiber without the complications involved in yarns or cloth. Equipment for this type of testing, however, is very highly specialized and is applicable to research only. To make this measurement, a single wool fiber is fastened to two glass hooks with a chemically inert adhesive, soaked in distilled water about '24 hours, and stretched and relaxed at a constant rate. The load and recovery curves are determined with a modified chainomatic balance that is automatically controlled through an electronic circuit. (For a complete description of the apparatus, see J. Research Nat'l. Bur. Standards 31, 25 (194?) RP 1546.) A curve is obtained with an untreated fiber. This fiber is then placed in a perforated Saran capsule, sewed to a sock, and treated. A determina- tion of stress-strain characteristics after this treatment gives another line where the ratio of the areas under the lines, i.e., the ratios of the amounts of work done in stretching the fiber, expressed in percent is known as the 30# index. If a fiber is chemically unchanged by treatment, this ratio will be about 99- This measurement affords a very sensitive method of telling when something has happened to the fiber, but it does not tell what chemical reaction occurred, nor does it predict any wearing qualities. These factors must be determined by correlation with other test methods. 5. FLEXURAL RIGIDITY. The rigidity of a fabric is related to the comfort and it i-s therefore important to have the rigidity as low as possible where com- fort is an essential factor. Army motion or leg motion will be greatly impeded by the use of the rigid fabric. A garment made of a rigid fabric will feel more bulky and more heavy. The work consumed in flexing a fabric and recovered in releasing it can be measured by an instrument, the flexometer, developed by H. F. Schiefer'. (Re- search Paper 555, J. Research Nat'1. Bur. Standards 10,, 647-57 (May 1933). Often, for comparative evaluation of clothing fabrics simpler devices for comparing stiff- ness can be used. One, used for measuring the stiffness of coated fabrics, espe- cially at low temperatures, is similar to the compressometer, except that the sam- ple is laid across two parallel bars. Pressure from the compressometer is applied by a third bar halfway between the two supports. Other methods of comparing stiff- ness depend on measuring the chord of the arc formed by a length of fabric sup- ported at its center, or the tendency to hang down over the edge of a table. 6. PLEXURAL FATIGUE. The M.I.T. fold tester has been applied to fabrics. The results are described in "Note on flexural fatigue of textiles," by Herbert P. Schiefer and Paul M. Boyland. J. Research National Bureau of Standards 29, 69-71 (July 1942) RP 1485; and Tex. Res. 1£, No. 11, 2-7 (Sept. 1942); and Rayon Tex. Monthly 2J5 (Nov. 1942) p. 62. The fold tester makes double folds, through an angle of 135°, at 200 folds per minute, vith a suitable tension, such as 1.5 Kg

82 CLOTHING TEST METHODS on the specimen 15 mm wide. There is a question as to whether military clothing made of wool or cotton ever wears out by flexural fatigue. However, flexural fail- ure of nylon and cotton fabrics does appear in jungle boots, and more highly crys- talline fibers such as linen or high tenacity rayon break sooner in this teat than do more extensible fibers. 7- FLAME PROOFING. The A.S.T.M. method D626-41T is recommended. The ex- tent of flashing on the specimen, and the duration of flame and afterglow are noted (Ref.6). From "Test Methods and Ratings," by A. C. Goodings and P. Larose: In general a satisfactory degree of flame proofness is indicated by the following: No flashing on the length of the specimen shall occur. The duration of the afterflame shall not exceed 2 seconds. Afterglow at the edge of the charred area shall not exceed 20 seconds after the cessation of flaming. The charred length shall not exceed 4 inches. Comment by P. Larose: Besides the strip method referred to, we also use what we call the surface burning test, which perhaps duplicates more closely the conditions that would be met with the flame striking the clothing. This is a method developed by the British Standards Institute and one that is described in our schedule of methods of testing textiles (Ref. 5). Slightly different methods are described in reference ^, as used at the Philadel- phia Quartermaster Depot. 8. TESTS FOR FLASH RESISTANCE, by J. A. Kitching. The material to be tested is mounted in metal embroidery hoops 6 inches in diameter. A bar, 1? inches long and pivoted at it3 midpoint, is caused to rotate in a horizontal plane about a veritcal axis by means of a variable speed electric motor. To each end of the bar an embroidery hoop is bolted by its edge so as to lie in a plane tangential to the motion. The hoops are readily removable. Hoops carrying the material to be tested are revolved at various controlled speeds past a lighted gas jet directed at them horizontally. The gas jet consists of a micro burner for which the gas is supplied at controlled pressure (9 cms. on the gauge), the constancy of the flame being checked by means of a radiometer. The details of construction of ."he test equipment are arbitrary, but standard for all tests car- ried out. The siower the speed of the motor, the longer is the material exposed to the flame. After each exposure a new piece of material is mounted. The method of Resting consists in the determination of speeds at which the material will just ignite and just fail to ignite. The midpoint between these two speeds is regarded as the ignition speed, and is expressed as the time taken for a single complete revolution of the hoops at this speed. It will be noted that the revolution time is proportional to the maximal time during which any point on the material, being revolved at that speed, is exposed to the flame. This test is designed to reproduce in some measure the conditions likely to be encountered when a two-engined plane crashes on landing. It is presumed that the crew, in escaping from the plane, are briefly exposed to hot flame; and that if their clothes ignite they will be burned severely, but that if their clothes do not ignite those parts of their bodies which are covered will be suf- ficiently protected. Thus a flashing off of surface nap is unlikely to matter so long as the material does not catch fire. Material from which fire is easily

PHYSICAL TEST METHODS 83 extinguished is also likely to be better than material which burns freely when once alight. On the basis of tests already done, a provisional rating has been attempted as follows: Revolution time for ignition speed Rating 0-19 secs. 1 20 - 39 " 2 4o - 69 " 3 70 - 99 " ^ 100 - 150 " 5 The scores for higher ratings can only be determined by tests on less inflammable materials than those so far used. The indications from service experience are that fabrics with a flash resistance of 3 or higher give reasonably good protection. Comment by A. C. Goodinga: The original flash test used was essentially to deter- mine the time which a material could be exposed to a standard flame with- out igniting and continuing to burn. This had limitations. For example, the case of net fabrics and flame-proofed materials posed difficultiea in evaluation considered in the light of whether a man would suffer severe burns. Further tests were developed in which the test fabric was backed by some material such as nylon which would act as an indicator of tempera- ture on the side of the fabric remote from the flame and permit a time evaluation of how long it took for a fabric exposed to a standard flame to reach a certain temperature. This, however, is not the whole story because it obviously is an important difference whether a fabric will ig- nite and blaze or whether it will not. We are given to understand that the R.A.F. consider wool and silk to be bad in respect of flash resistance but we have not been able to show that wool is any worse than cotton in any of our tests. We feel there may be another explanation for the R.A.F. observations, namely, that at the hands and knees where the large number of burns occur, and which happened to be covered with wool or silk, the fabric is tight to the skin. In other words, there is a great deal of difference in flash protection given by a loose covering and one which is tight. However, we are now looking into tests which will more closely stimulate flash than tests which we have been using where a standard flame is employed and time is the variable factor. 9. FRICTIONAL PROPERTIES. An inclined plane method has been used in Canada, c.f. ACAMR Report No. C-2569, "The slipperiness of lining and related fabrics in flying suits," by A. C. Qoodings and C. E. Coke, and the device described by A. A. Mercier, J. of Research Nat'l. Bureau of Standards, !5, 1930 (RP 196). The method adopted was to measure the coefficient of friction between dif- ferent fabric surfaces by determining the sliding angle. A sample of one of the materials to be examined is fixed to a block which is then placed on the second material attached to a board capable of being tilted to any angle. The angle of inclination of the board at which sliding will occur with approximately uniform velocity is measured, and the coefficient of friction is the tangent of this angle. The static friction, or the force required for sliding to begin, is slightly higher

84 CLOTHING TEST METHODS than the sliding friction, but the latter is the preferable measurement to make from a point of vIew of reproducibility of result, avoidance of irregular readings due to local variations in the material, and the probably greater significance of sliding friction over static friction relative to the matter under consideration. The pressure applied between the fabrics was approximately 0.1 Ib. per sq. inch. The measurement is always made between two fabrics whose frictIonal properties when in contact require to be known, in other words, it is not the practice to use a standard surface against which to measure all other fabrics. Another method employs the friction meter developed at the National Bureau of Standards; described in RP 1562, "A friction meter for determining the coeffi- cient of kinetic friction of fabrics," by Edwin A. Dreby. J. Research NBS 31, 237-46 (Oct. 1943). This instrument is also in use at the Ontario Research Foun- dation, replacing the inclined plane. It measures the force transmitted from a moving layer of one fabric to a stationary layer of another. Comment by P. Larose: The frictional properties of fabrics are of importance for such things as lining fabrics where it is a definite advantage to have the outer garment slip readily over the inner clothing. It has also a certain bearing on the comfort of the garments themselves. An outer garment that slips readily over the inner clothing will no doubt feel less bulky and more flexible than one offering appreciable resistance to such movement. On the other hand, slipperiness of the outer surface of a garment may be of advantage in decreasing the changes of snag. 10. WATER REPELLENT FABRICS, by A. M. Sookne. The distinction should be observed between water proof fabrics in which a continuous film of water proof material, such as rubber, is supported by the fibers, and water repellent fabrics, in which the space between the fibers is not closed, but the surface characteristics have been altered to prevent wetting. The following discussion is primarily concerned with water repellent fabrics. Tests for water repellency may be conveniently divided into the following categories: 1. Tests to determine resistance to wetting inherent in the yarns them- selves (spray tests, immersion or absorption tests) 2. Air permeability tests. 3. Tests to measure resistance to penetration by hydrostatic pressure. 4. Tests to measure resistance to penetration by falling drops, streams or sprays. Since it is the primary purpose of water repellent clothing to keep the wearer dry when exposed to rainfall or other water, these tests will be evaluated from that point of viev. Tests in group 1 bear no direct relation to the ability of a fabric to withstand rainfall. They are, however, useful in evaluating the efficacy of the hydrophobia finish given to the fibers. The standard AATCC spray test (ref. 7_) is rapid and reliable, and can be usefully employed for this purpose. A dynamic ab- sorption test is now being standardized by the Office of the Quartermaster General. It is more reliable than the standard AATCC absorption or immersion test, and should also prove helpful in this connection. Air permeability tests (group 2) also bear no direct relationship to resis- tance to rainfall, since they obviously cannot provide an estimate of the efficacy of the finish, the effect of swelling on the pores of the wetted fabric, the shape

PHYSICAL TEST METHODS 85 and size distribution of the pores, etc. However, for any given type of fabric, the air permeabilities of a series of well treated samples (100 spray rating) are fairly closely related to their resistance to rainfal1. Air permeability tests are also useful in providing a rough basis for choosing grey goods to be given water repellent treatments for specific purposes. The SchIefer-Boyland air permeability apparatus developed at the National Bureau of Standards, appears to be the most suitable machine available for routine laboratory use. Of the simple routine tests, the hydrostatic pressure tests (group 3) are perhaps the most informative, since they measure a quantity which is influenced by both fabric construction and finish. Within any group of comparable fabrics, the hydrostatic pressure test provides a good indication of rainfall resistance. This is not the case, however, when fabrics of widely different construction are com- pared. Thus, a thick, resilient fabric will provide much better resistance to rainfall than a thin, flat fabric bearing a comparable finish and having the same resistance to hydrostatic pressure. The Suter tester (Procedure A, ASTM D583-40T) and the AATCC hydrostatic pressure tester (Procedure B, ASTM D583-40T) (see also ref. 7.) are both in wide use in this country. The former is perhaps somewhat less convenient to use, and requires a larger sample. The tests in group 4 are, by their very nature, most closely related to the utility of water repellent fabrics. They have the special advantage that they can be used to evaluate multi-layered combinations, which have been shown in practice to be especially desirable for water repellent rainwear. They are, however, the most difficult to perform, and the most time consuming of the several groups of tests. Of the several tests of type 4 that are available, the Du Pont Rain Tester and the Textile Foundation Drop-Penetration Apparatus appear to be best suited to testing the tightly-woven fabrics required for good resistance to rainfal1. There is a good general relationship between the results of these two devices on tightly- woven fabrics. The Du Pont Rain Tester is simpler to operate and requires less time for measurement; it is more suitable for routine use. The Drop-Penetration Apparatus to be capable of revealing smaller differences between fabrics. The Bundesmann Tester appears to be too complicated for routine use, and is probably not sufficiently drastic to test the better fabrics in a reasonable length of time. Tests of water repel'lency made on the unworn, unlaundered fabric are of questionable value. New fabrics are frequently calendered, which process gives them a high initial water repellency, much of which vanishes after one thorough wetting. In addition, many uncalendered fabrics lose a substantial portion of their initial water repellency after one wetting. It is recommended that water repellent fabrics be evaluated after one or three launderings of the type to which they will be subjected in use. Such a procedure provides a more reliable estimate of the performance of a water repellent fabric over its period of useful service than does testing the new fabric. The Du Pont Bain Tester consists of a horizontal nozzle, the specifications for which appeared in the American Dyestuff Reporter 3_2., p. 88 (1943). A hydro- static head of 2, 4, 6, or 8 feet is used to project the water against the cloth, which Is backed by blotting paper such as that specified in the AATCC Immersion Test. The Textile Foundation drop-penetration apparatus allows 1 drop per second from each of a group of thirty capillaries to fall through a path about 5 feet 8 inches long, in which they are protected from drafts. The drops strike repeatedly at the Sa'me spota on the sample, inclined at 45°. Gauze or blotting paper may be used as backing, and weighed, or an electric signal used to indicate penetration. For the most severe tests, a hard backing may be used, with measurement of the amount of water penetrating in a given time. Drawings and descriptions of the ap- paratus can be obtained from Milton. Harris Associates, 1246 Taylor Street, N. W., Washington, D. C.

86 CLOTHING TEST METHODS Comment by P. Larose: It cannot be too strongly emphasized that what a water re- pellent treatment does ia simply reduce a tendency of the fabric to wet out and that any water proofing properties (resistance to Vater penetra- tion) are as much the result of the structural features of the cloth as that of the treatment. Any tests, therefore, that merely indicate the ease of wetting, such as the AATCC spray test are of no value in estimat- ing the rain-proofness of fabrics. 11. BUOYANCY, by A. C. Goodings. The following is the test method as set up and used by the sub-committee on protective clothing, R.C.A.F. Test Method for Evaluating Buoyancy of Kapok and Kapok Substitutes for Fillers in Life-SavIng Equipment. (a) Density of packing. In testing the buoyancy of kapok, the fibrous mass is packed to a density equivalent to 3 lbs. of kapok per cubic foot of occu- pied space. In the examination of materials other than kapok a quantity of fibre is used such that the packing force exerted is equal to that required to compress 3 lbs. of kapok into a volume of 1 cubic foot. This quantity of fibre can readily be found in the following way: A glass cylinder approximately 3i" in diameter and 6" in height is fitted with a circular metal plate of slightly smaller diameter than the corresponding internal dimension of the cylinder. Additional weight is added to the plate such that when a quantity of kapok under the total applied force (approximately 1/3 lb. per sq. inch) is contained in the cylinder to a depth of 4 inches, the density of the mass is equal to 3 lbs. per cubic foot. The weights of other fibres required to fill the same volume under the same load are then readily determined experimentally with the same equipment. In recording the results of tests, the weight of fibre per cubic foot of occupied space should be stated. • *' (b) Determination of buoyancy. A cylindrical bag approximately 3i" in diameter and 4" in height is filled with the fibre to be examined, the weight of fibre per cubic foot to be used having been determined in the manner described under (a) above. The cylindrical bag is made of some open structure textile fab- ric such as mosquito netting or coarse cheese cloth. The bag and contents are then placed inside an open mesh (£") wire cylinder of the same internal dimensions as the bag. To the bottom of the wire cylinder, but a little below its lower face so as not to prevent access of water to that face, is attached a lead sinker of ap- proximately 2 lbs. in weight. The cylinder and contents are submerged in water to a depth of 12" measured from the upper face. The weight in water of sinker, cylin- der and contents is determined at intervals over a 24-hour period, and the buoyancy is calculated by subtracting the observed weight from the submerged weight of the metal cylinder and sinker with no fibre enclosed. . At the end of the 24-hour immersion period the net bag and contents are re- moved and placed in a cylindrical jar of approximately 4-litre capacity fitted with a suitably tight fitting lid. One litre of water is added, and the Jar and con- tents rotated about an axis through the mid-point of the sides. The rotation peri- od is one hour. At the end of this time the bag and fibre content is removed and replaced in the wire cylinder and submerged to a depth of 12" as previously. Read- ings of the buoyant force exerted are again taken over a period of one hour. The buoyant force is to be expressed in terms of pounds per cubic foot of fibre mass.

PHYSICAL TEST METHODS 87 In certain cases It may be found desirable to extend examination of the buoyancy of a material to conditions of packing where the compressional forces are greater or less than the standard detailed above. Complete evaluation of material for use as a filler in life-saving equip- ment should also include an examination of the susceptibility of the material to breakdown in the dry state, with consequent loss in buoyancy. 12. THICKNESS, COMPRESSIBILITY, AND COMPRESSIONAL RESILIENCY. (a) Thickness: Thickness is of course important in the determination of bulk density whenever this is desired, but the main interest in thickness ia prob- ably due to the relation between thickness and thermal insulation. It has been shown that for most textile materials where the bulk density is above a certain minimum figure, that is, where the air spaces trapped by the fibers are quite small, that the thermal insulation is practically proportional to the thickness. This makes it possible to calculate the thermal transmission simply from the thickness. This is mentioned later in speaking of thermal transmission. The measurement of thickness is of course necessary in determining the compressibility of a material. Since the thickness of certain materials such as pile fabrics de- pends on the previous history of the material, it is important that this should be considered when measuring thickness. The standard instrument is the compressometer, developed by H. F. Schiefer at the National Bureau of Standards. It is described in RP 56l, J. Research NBS 10, 705-713 (1933). A new model has been developed recently, with a span of 8 inches. With a heavy spring, this is capable of exerting pressures up to 1 Ib. per in.2 with the 5-inch diameter foot. With a light spring, and the large diameter foot, thicknesses can be measured at pressures as low as 0.005 Ib/in 2. Different materials require different methods of measurement; for comparison with service conditions the thick, compressible materials, such as pile fabrics and loose sleeping bag materials, are better measured by the larger diameter instruments. Thickness at low pressures can also be measured with a microscope, as described in reference 4: "The microscope must be fitted with a scale and vernier which allows the movement of the tube to be determined accurately. The microscope is first focussed on a plane surface placed on the stage of the microscope and sub- sequently focussed onto the surface of the light flat object exerting the required pressure on the top of the pile fabric, due correction being made for the thickness of the particular plate used which may be plastic, glass or other appropriate plane material. This method has the advantage that the instrument with which the measurement is carried out does not touch and therefore does not affect the pressure exerted on the fabric. "It has been found that the size of the plate exerting the pressure is not immaterial, probably due to an 'edge effect.1 The plate adopted tentatively is a circular bakelite plate of 3" diameter." Revised Method for Thickness and Compressibility, Contributed by P. Larose, from Subcommittee on Protective Clothing Reports Nos. 162 and 163. After tests had been carried out on various kinds of fibers, it was con- cluded that the thickness becomes reasonably steady after 5 cycles of compression of a standard type. The relation between the thickness and the logarithm of the pressure is rectilinear, except possibly in the case of very low pressures in which

88 CLOTHING TEST METHODS it ia difficult to make a measurement. Any divergences from the straight line usually occur in the low pressure end, and the thickness at 0.1 Ib. is obtained by extrapolation in that case. The slope of the line is taken as an index of compressibility, but since the two reference points of 0.1 Ib. and 1.0 Ib. pres- sures can be used, the difference of the logs. is one and the measure then of the slope is simply the difference in thickness determined under these two pres- sures. Finally, the slope is divided by the original thickness at 0.1 Ib. pres- sure. The thickness at a pressure of 0.1 Ib/in2 derived by extrapolation of the logarithmic line (and not by direct measurement) has been adopted as an index of thickness of a fabric. This value has been chosen because it had been shown previously that the fabric of,a flying suit during wear is subjected for the most part to pressures of this order of magnitude. b. Compressibility; The new measure of compressibility proposed above by Dr. Larose seems to be the most useful available. A comparison of compressibili- ties, in non-quantitative form, may be presented to the eye by the curve showing the relation between thickness and pressure. A conventional unit for compressi- bility has been used by Dr. Schiefer at the National Bureau of Standards, based on slopes determined from this arithmetical curve at. arbitrary points. (See J. Research NBS 10, 705 (19J>J>) RP 56l, and J. Research NBS 32, 261-284 (1944).) In this, compressibility = (thickness at 0.5 Ib/in2 - thickness at 1.5 Ib/in2)/(thick- ness at 1.0 Ib/in2). Whereas the unit in the new definition is % change in thick- ness over the pressure range 0.1 to 1.0 Ib/in2, the older NBS unit was expressed in contracted form as in2/lb, or in unabridged form as (inches change/inch thick- ness) per (Ib/in2). Comment by P. Larose; As already indicated, thermal transmission depends on thick- ness and therefore a compressible material will have a varying degree of thermal insulation, depending on the pressure imposed. Since certain parts of the clothing are.under pressure during use, it is important to gain an idea of the compressibility of the clothing materials in order to determine the relative merits in respect fo thermal insulation. As in the case of thickness, compressibility will vary somewhat with the previous treatment of the fabric. Some of this effect is eliminated by measuring thickness after a certain number of compression cycles. However, this does not duplicate the effect of matting that takes place during use and which occurs over an extended period of time. This generally results in an increase in resistance to compression. The1 final thickness might be more or less than the original thidkness, depending on the felting ac- tion and whether the material has a high degree.of shrinkage or not. It is impossible to predict absolutely what the state of the material will be after use, unless all the factors that may affect it during use are known. o. Compressional resilience: The compresslonal resilience is computed from the energy recovered and the energy utilized during compression at the Nation- al Bureau of Standards (see RP 56l), whereas the Canadian laboratories use the ratio of the thickness after recovery from compression to the initial thickness. As with compressibility, the main interest arises from thermal properties, so com- parison can-be more directly made in terms of recovered thickness as percent of original thickness. Comment by P. Larose; Any of the factors that may affect resilience have been pretty well covered in speaking of compressibility. We have found that by determining the compressibility after a certain number of preliminary compressions that there is little need for the determination of resilience itself.

PHYSICAL TEST METHODS 89 13. EFFECTIVE PRESSURES IN CLOTHING. From ACAMR Report No. C-2556 (Canada), by J. A. Kitching and H. M. Kolesar. Methods: The subject, 'in full flying clothing, was seated on an aircraft seat with his feet resting on the floor. A small rubber bag having an inflatable area of approximately l£ square inches was placed between the suit and the aircraft seat and was connected to an air inlet and to a manometer for kymographic recording. At a given air flow, a constant rise in pressure was recorded until the point was reached at which the bag began to inflate. At this point, the rise in pressure fell off ap- preciably and the manometric pressure at that point was taken to be that imposed on the bag. In order to determine accurately this point of de- flection in rate of pressure rise, a tracing was first made with the bag clamped off. The drum was then spun back to the starting point and a de- termination made; the second tracing coincided with the first one until the bag began to inflate with a consequent tendency for the tracing to flatten off. The apparatus was carefully calibrated with known loads and the linear relationship between load imposed on the bag and manometric readings ascertained. It was not thought worthwhile to calibrate the ap- paratus for loads under 0.1 lb./sq.In. In the lower back region, the compression load is of the order of 0.3 lb. per square inch with the exception of a small area in the mid-line where it reaches 0.7 lb. per square inch. Compression loads on the rest of the back are minimal (less than 0.1 lb. per square inch). The distribution of the compression loads in the seat region, when no parachute harness is worn, varies according to the softness of the sea't. On a hard metal surface, the bulk of the load is concentrated below the points of the ischial bones where pressures ranging from 5 to 16 lbs. per square inch were recorded. At these same points, pressures of 3 lbs. and 2^ lbs. were registered when using a seat pack parachute and a multiple pile cushion respectively. Compression loads elsewhere in the seat region ranged between 0.5 and 1.0 lb. on the hard metal seat, and between 0.5 and 1.6 lb. on the seat pack, and cushion. The distribution was somewhat more uniform in the case of the softer cushion. Other measurements made were as follows: 1. Subject wearing a parachute harness: (a) Under the seat webbing: 4.8 to over 16 lbs. per square inch. Average: 11 lbs. (b) At the top of the shoulder: less than 0.1 lbs. per square inch. (c) At the breast line, under the webbing: less than 0.1 lb. per square inch. 2. Over the knee cap, under the flying suit: 0.5 to 0.6 lb. per square inch. 3. Under the knee, inside the flying suit: less than 0.1 lb. per square inch. 14. THERMAL INSULATION. A guarded hot plate method is used as the funda- mental standard in both the United States and Canada. This is described by R. S. Cleveland in RP 1055, J- Research NBS 19, 675-684 (1937). This method measures

90 CLOTHING TEST METHODS the rate of heat transfer from a black-painted heat source, through the fabric into air confined in a hood. It has been the practice of the Bureau of Standards to re- port the thermal transmission of the whole system, fabric plus air, as "Thermal transmission" usually in units of B.T.U./(°P hr ft2). It is more desirable for clothing work to express the results in terms of thermal resistance, and to separate the intrinsic thermal resistance of the cloth from the thermal resistance of the air layers. It is also desirable to express the results either in Clo units or in metric units, (°C hr m2) Kg cal, since this latter unit is readily used with the unit used in expressing heat production, Kg cal/m2 hr. The multiplicity of units encountered in the literature of thermal trans- mission or insulation measurements is troublesome. The following table of conver- sion factors is by no means complete, but may be helpfu1. In general, transmission units correspond to electrical conductivity units; insulation units are analogous to electrical resistance, and for each unit of one kind there is a corresponding exactly reciprocal unit of the other kind. Units may be converted as follows: Transmission on units: 1 B.T.U./ft2 hr °F = 1.355 cal/m2 sec °C = 4.88 Kg oal/mahr°C 1 cal/m2 sec % = .738 B.T.U./ft2 hr °F = 3.6 Kg cal/m2hr°C 1 Kg cal/m2 hr °C = .278 cal/m2 sec °C = .205 B.T.U./ft2hr°F Resistance units: 1 Clo = 0.88 °F hr ft2/B.T.U. = O.l8 °C hr m2/Kg cal = .648 °C sec m2/cal 1 °F hr ft2/B.T.U. = .738 °C sec mVcal - .205 °C hr m2/Kg cal = 1.138 Clo 1 °C sec m2/cal = 1.355 °F hr ft2/B.T.U. = .278 °C hr m2/Kg cal =1.54 Clo 1 °C hr m2/Kg cal = 4.88 °F hr ft2/B.T.U. = 3-6 °C m2 sec/cal = 5.56 Clo Power units: 1 watt = 3.41 B.T.U./hr = .239 cal/sec = .86 Kg cal/hr The data for deriving intrinsic resistances from the thermal transmission values given by the Bureau of Standards are available in RP 1055, and in a later paper, RP 1589, "A study of the properties of household blankets," by H. F. Schiefer and others, J. Research NBS 32, 261-284 (1944). The thermal transmission of the apparatus containing only air is 1.6 E.T.U./(°F hr ft2). The reciprocal of this, 0.63 °F hr ft2/B.T.U., may be taken as the intrinsic resistance of the air above the fabric. Since 1 Clo =0.88 (°F hr ft2)/B.T.U., the resistance of the air space turns out to be 0.72 Clo, or 0.13^ hr n^/Kg cal, or 0.46°C sec nf/ca1. Hence, to obtain intrinsic resistance from National Bureau of Standards thermal transmission values, one should: (a) take the reciprocal of the thermal trans- mission; and (b) subtract the insulating value of the air included in the meas- urement, in the appropriate units. It is often sufficient to measure the thickness with the compressometer, at a pressure of 0.1 lb/in2, and from this compute the thermal transmission or

PHYSICAL TEST METHODS 91 Clo value. Schiefer (RP 1589) has found for blankets and underwear fabrics that Rj, the intrinsic resistance, is proportioned to tg ]_, the thickness in inches at 0.1 lb/in8: Rj = 3.0 t0.i, in (°F hr ft2)/B.T.U. This holds within 10#, 95 times out of 100. From this it follows that Ic, the in- trinsic Clo value of the clothing material is: Ic = 3.4 t0.l, in Clo This is lower than the approximate value found by Siple, 4 Clo per inch, based on physiological measurements of Clo values and measured thicknesses of clothing layers while being worn. The discrepancy is probably to be attributed to layers of air trapped between layers of clothing. Comments by P. Larose; In my own tests the value of Ic comes out to 3.6 Clo per inch if the value of .72 Clo is used for the insulation of the air. How- ever, in applying it to clothing we have been in the habit of using a higher figure, namely 4.3 Clo which seems to be more in accord with actual experience. Comment by L. Fourt: Comparisons are frequently made between the insulating value of fabrics and that of still air. This is natural in view of the generally accepted idea that the insulating value of fabrics arises from the still air trapped within them. However, the values of thermal conductivity of air, as determined by different investigators, show an average deviation of "(% about the mean (International Critical Tables, V, 213). The precise value, in terms of the I.C.T. mean for air at 0°C, is 7.37 Clo/inch, and the correction for the increase of conductivity (decrease of resistance) is 0.28# per degree centigrade increase of temperature. Hence 7 Clo/inch of ideal still air corresponds precisely to an air temperature of l8°C, and can be regarded as sufficiently accurate for general clothing problems. 15. WATER VAPOR PERMEABILITY OF FABRICS. The presence of clothing reduces air movement in the space which it occupies and between it and the skin. Hence the methods adopted for testing water vapor permeability are based on diffusion conditions, and do not take account of penetration by wind, or movement of water vapor by convection currents. By diffusion conditions we mean ideal still air, in which the only motion is the random kinetic motion of the individual molecules, convection being excluded. It is agreed that the values for the water vapor permeability of fabrics should be given in terms of an equivalent thickness of ideal still air. This would appear to be a very useful method of expressing results because where one is dealing with several layers of clothing with air layers in between, the total resistance of clothing to the passages of water-vapor can be arrived at by the summation of the separate resistances. Conversion of a total resistance to per- meability units such as grams Kg cal evap. cooling ra'lnr. (mm v.p. difference) m hr. (mm v.p. difference) can always be done by applying the diffusion coefficient of water-vapor in air. Methods; (1) from A. C. Goodings, Ontario Research Foundation, Toronto.

92 CLOTHING TEST METHODS h 4 The method we are employing ia briefly as follows. In the case of a vessel containing water, covered with the fabric being investigated, there is a vapor pressure gradient from the surface of the water (Pi) to the under side of the fab- ric (P 2). There is a further vapor pressure drop Pw from the under (P2) to the upper side (P3) of the fabric and a further gradient from this latter 'T aurface outwards to the controlled conditions of K the atmosphere in which the vessel is exposed (P.*). Let the distance from the water surface to the 1 under side of the fabric be L, the thickness of the fabric be T, and the distance over which the outer vapor pressure gradient extends be H. Fur- ther let the constants for the resistances to moisture permeation be KI, K2 and K3, respectively (K3 and Ki are both for air but the outer air ia subject to disturbance), then the rate of evapora- tion is given by Pi - P2 P2 - P; Rate KjL K2T Pa - P. Kah Pi - K3h is the troublesome factor, but by finding the distance L has to be when the vessel is covered by a single layer of fabric to give the same rate of loss of moisture as when the vessel is covered by a double layer, then Kah is eliminated being the dame in both cases. We have then KJj + KsT + K3h = KiL2+ 2Kj>T + Kah or K2T. = Ki(L!-L2) \ or the fabric is equivalent to an air layer of Li-L2 in thickness. Experimentally we set up 6 veaaels, four of which are with single layers of fabric and varying L. This gives the relation between L and the rate, and if plotted as I/Rate against L the relationship is a straight line for values of L up to about 2 cms. Two experiments (duplicates) are also run at the same time with two thicknesses of fabric with a definite air space in between (this of course being taken into account in the calculation), and the value of L for a single layer giving the same rate of evaporation is read off from the I/Rate against L plot. By using solutions to give different vapor pressures, i.e. different but known values of Pi, it will be seen that in the relation Pi-P2 Rate = P2 can be evaluated and so a value for K arrived at. This is of course simply an inverae of the diffuaion coefficient (if the same units are used). (2) from Lyman Fourt, Textile Foundation, Washington. The total resistance, R, of a system is given in equivalent centimeters of ideal still air, in terras of Q, the grama of water vapor passing, D, the .diffusion coefficient, C, the difference in vapor concentration in grams/cm3, A, the area in cm and t, the time in seconds, by the equation R = ~ A t (1)

PHYSICAL TEST METHODS 93 The diffusion coefficient, D, varies with the absolute temperature, T, and the baro- metric pressure, B (mm Hg), according to the relation (from International Critical Tables ) D = .220 L^,)1'75 760 (2) (:> B With sufficient practical accuracy, for any temperature m°C between 0 and 50°C, this can be replaced by D = (.220 + .00147 m°C) (3) B The concentration difference C, can be obtained from the difference in vapor pres- sure, Ap, and the absolute temperature, T: c .. 18 27J. A£ . 8 -4 A£ (4) C ~ 22,400 760 T - 2.B9 * 10 T The specific resistance of a fabric is defined as the resistance per unit thickness, that is, as the intrinsic resistance divided by specimen thickness. The specific resistance is, therefore, the ratio of the resistance of the fabric to the resistance of ideal still air. Preparing the sample. Five test systems are prepared for each fabric, with 1, 2, 3, 4, and 5 layers of the fabric, respectively. The most simple pro- cedure for multiple layers is to place each layer in direct contact with the next, avoiding wrinkles which would introduce additional air-space resistance. If a space is provided between the layers it must be accounted for in estimating the intrinsic resistance, and should be kept small, to avoid convection disturbances. An experimental test for the absence of convection is that the total resistance should be increased by exactly the amount of any increase in the space between the layers. It is helpful to humidify the specimens before mounting, by keeping them in a closed container with water. This procedure reduces the tendency to sag during the tests. The area under test (17 cm2 has been used) is defined by sealing the fab- ric with wax (50/6 beeswax, 50/6 paraffin) between metal rings. Care must be exer- cised not to apply an excess of wax too close to the opening, to avoid flow into the open space. For calculation of specific resistance, the thickness is required. This is measured with the compressometer at 0.1 lb/in2 after conditioning the fabric "t 65$ relative humidity, 21.1°C. Since the thickness (and also the content of other material such as sizing) depends on the previous treatment of the fabric, this should be stated, i.e., as "manufacturer's finish," "loomstate,"' "washed and ironed," "washed and dried without ironing," or "impregnated with x# of y." The quantities required for calculation by equation (1) can be obtained by either of two methods: A. Absorption cup method. The assembly of fabric and metal rings is sealed to small dishes (tanning dishes, 70 x 50 mm) containing a granular drying agent. The dishes are inverted so that the drying agent is in contact with the fabric, minimizing the path of the water vapor. The inverted dishes are placed on a wide mesh wire rack in a uniform current of conditioned air (70°F, 65$ RH, 200 ft/min has been used). Drierite (anhydrous CaSO*) and Anydrone (MgC104) are satisfactory

94 CLOTHING TEST METHODS drying agents; the MgClC>4 lasts longer, but has a tendency to become wet and' stick to the cloth, unless mixed at frequent intervals. Weighings are made at uniform intervals of \ to 1 hour; the drying agent is mixed at each weighing, since the resistance tends to increase as the layers next to the fabric become saturated. The rate of gain in weight becomes constant after the first half hour, and con- tinues constant for 2 hours or more, B. Evaporation procedure. The layers of cloth are sealed into metal rings in the same manner as in the absorption procedure. However, the opening must be chosen to be the same as the average inside diameter of the straight-sided, cylindrical vessel containing the water (Kimble 50 x 34 mm crystallizing dishes have been used) in order to secure linear diffusion inside the cup. A special ring, with a shoulder recessed to fit the edge of the dish, is an aid to centering. Since glass vessels vary somewhat in diameter, and in contour across the bottom, each should be individually calibrated. This can be done by finding the weight of water required to fill it flush with a flat glass plate lying across the rim. From the volume, and the area of the top, the volume required to fill to any dis- tance from the brim can be computed. A convenient distance, including the thick- ness of the metal ring below the cloth, is 1 cm, and unless otherwise stated, this standard separation is used. In this respect the method differs from that of the A.S.T.M. or T.A.P.P.I., in which a space of 2.5 cm (or greater) is recommended. The dishes are exposed in a uniform air stream, and weighed at intervals of about an hour. The apparent resistance may be high at first, on account of absorption of moisture by the cloth, but a steady rate is soon reached. If the amount of evaporation is sufficient to lower the water level in the interval between weigh- ings, the average spacing should be used in accounting for the air space under- neath the fabric. Variation in the space beneath the cloth causes an equal varia- tion in the total resistance, within a range up to about 2.5 cm. At greater spac- ings, the total resistance does not increase by the full amount of the extra space and in fact increases only slowly with increasing separation, presumably because of convection disturbances. For the purpose of calculating the resistance, it is assumed that evaporation is taking place at room temperature. The actual tempera- ture of the evaporating surface is somewhat below room temperature, and hence the actual concentration difference between saturated vapor at the water surface and the air of the room is less than the assumed, which implies that the actual resis- tance is somewhat smaller than the calculated resistance. Interlaboratory comparison. Each laboratory has applied its methods to each of four fabrics, with the following results: Fabric Thickness Intrinsic Resistance, cm air cm Method 1 Method 2A Method 2B (Evaporation) (Absorption) (Evaporation) Balloon cloth .018 0.04 0.05 0.07 5 oz. poplin, wind .039 0.09 0.09 0.14 resistant Shirley cloth .064 0.23 0.22 0.27 Bedford Cord (Navy .109 0.30 0.37 0.41 "Jungle" cloth) It appears that the intrinsic resistance of a fabric can be determined to within about 0.03 cm.

PHYSICAL TEST METHODS 95 Comment by L. Fourt; Conversion from resistance to permeability. For many clothing problems it is desirable to convert the total resistance of a system into a permeability, to give directly the grams of water evaporated per square meter per hour, or the evaporative cooling in Kg cal/m2 hr. The chief difficulty is that clothing problems seldom present an isothermal system. The concentration difference, C, can be computed correctly for two dif- ferent temperatures by the equation C = (^-|j} 2.89 x 10-4 in which Pi and P2 are in mm Hg, and T! and T2 are absolute temperatures. The diffusion constant also varies with temperature. However, this does not limit the rate of transfer at intermediate points along the thermal gradient, since a reduction in diffusion constant can be compensated by an increase in concentration at every point except the origin at the evaporating surface, where the concentration is the maximum or saturation concentration. Hence the skin temperature, Ti, is the proper temperature for calculating the diffusion constant. For approximate purposes, the calculation can be made in analogy to the isothermal system, using the skin temperature. Neglecting barometric pressure, and letting Ap be the difference in vapor pressure, in mm Hg grams/m2 hr = .125 I"75 -^ n = (8.38 + .0224 m°C) ^ K = (7.98 + .0124 n°F) ^ n and for skin temperature of 35°C (95°F) grams/m2 hr = 9.16 -~ n The error introduced by the isothermal approximation can be corrected by subtracting the term p2(Ti - TE) /Te from fip. The maximum evaporative cooling possible with a given resistance, Emax> can be formed by multiplying by 0.58, the latent heat of water in Kg cal/gram, or for isothermal systems between 0 and 50°C more accurately by the following linear approximations, which include the variation of latent heat with temperature : Emax = *« cal/m2 hr = (5.00 + .0084 m°C) -& = (4.85 + .0047 n°F) & n For 35°C (95 °F) skin temperature this becomes 5.3 Ap/R Application to clothing problems. In practice there are three additional variables in the case of a man wearing clothes. One of these is the ex- tent to which the clothing is wet, since the resistance to evaporation from wet clothing is only that of the outside air. A second is the ex- tent to which evaporation from clothing is effective in body cooling. The

96 CLOTHING TEST METHODS third is the "wetted area" of the body underneath the clothing, defined by Gagge as the fraction, w, which the observed evaporation, EOT-,S, is of the maximum, Emax- Hence E0bs = w Emax = f (5-00 + .0084 m°C) Ap On the assumption that the whole skin is wet (w = l) under the conditions which produced the maximum observed rates of evaporation, Gagge (Am. J. Physiol. 120, 277 (1937)) found, for low air movement, 17 ft/min, that Hobs = Emax = 2.85 Kg cal/m2 .hr. (mm vapor pressure difference), with a skin temperature of 35°C. From this the thickness of the equivalent layer of dead air which was limiting evaporation was = = 1.86 cm 2. op In comparison with this the intrinsic resistances of the fabrics in most clothing are small. In real situations a man compensates for the resistance of his clothing in either of two ways: (l) by increasing the wetted area, or (2) by satu- rating the clothing with sweat. In the second case resistance of the .clothing to passage of water vapor is no longer a factor. However, in conditions of less than maximum stress, the resistance of the clothing may be a factor of importance in comparing different types of clothing, since Winslow, Herrington, and Gagge (Am. J. Hygiene 26, 105 (1937)) have shown that discomfort is strongly correlated with wetted area. 16. RATE OP DRYING OF FABRICS. In studies of the rate of drying of fab- rics, it is important to express the rate in terms related to weight evaporated per unit area, rather than in terms of percent of dry or wet weight, etc. This is necessary because the fundamental processes involved are the transfer of heat into the fabric, and escape of water vapor out of the fabric, these taking place across the "large scale" surface area, that is, the area covered by the cloth. In comparisons of fabrics it is necessary to control the temperature, rela- tive humidity, air movement, manner of exposure of sample, and size and shape of sample . Comment by L. Fourt; In conditions of free exposure to air movements averaging less than 1 mile per hour the rate of drying of a wide variety of fabrics, ranging from broadcloth, knit underwear fabrics, poplin, twill, serge through blankets, is constant and linear with time until the fabric ap- proached equilibrium moisture conditions. The time required to dry, how- ever, varies with the amount of water initially held, which is closely correlated with the thickness. Dr. Goodings has shown that it is even more closely correlated with the volume of air space in the fabric. When only one side is exposed, the rate of drying of a thick fabric such as pile fabric slows down as the process continues, and is approximately proportional to the amount of water which remains. The similarity of drying rates for such diversity of fabrics at low rates of air movement arises from the smoothing effect of the relatively still air layers near the fabric. The equivalent layer of ideal still air amounts to at least 0.3 cm at about 200 ft/min air movement, as judged

PHYSICAL TEST METHODS 97 from measurements of water vapor.permeability. This is sufficient to smooth out irregularities of surface arising from weave. Dr. Robinson's experiments at the University of Indiana have shown that the amount of water required to saturate clothing, is closely related to the physiological stress imposed by the clothing in a hot environment. This is essentially a measure of the extra sweat required to saturate the clothing to secure the greater rate of evaporative cooling which is possi- ble when the resistance of the dry clothing to passage of water vapor is eliminated. 17. AIR PERMEABILITY. Most laboratories are using the machine developed at the National Bureau of Standards, described in RP 1471, "Improved instrument for measuring the air permeability of fabrics," by Herbert F. Schiefer and Paul M. Boyland. J. Research NBS 28, 637-42 (May 1942). In this device the amount of air which can be drawn through the fabric with a difference of pressure equal to i" of water is measured and reported in ft3/(min x ft2), at this arbitrary but standard pressure. "Air permeability" as determined on this device, usually im- plies the standard pressure across the cloth, although other pressures can be used, from 0.01 to 1.0 inch water gauge. The Philadelphia Quartermaster Depot (ref. J>) uses a densometer. With this device one measures the time required for a certain volume of air to pass through a certain area of cloth when pushed by a sinking piston. The densometer was developed primarily for testing paper or the most tightly woven fabrics. Its makers also make a device for directly measuring air flow through cloth, although its range appears to be more limited than the Nation- al Bureau of Standard's type. The results of the densometer measurements are re- ported as "seconds" required for specified volumes to escape through specified areas. In general, "seconds" x "air permeability" = a constant, for a particular set of specificationa. Comment by L. Fourt; For the current PQD specification (ref. 5), using 0.1 sq. in. specimen area, with a 5 oz. piston, and timing the escape of 300 cc of air, the "constant" appears to be about 340, for low permeabilities, correspond- ing to more than 300 seconds, or permeabilities less than 1. For higher permeabilities, a "constant" of 400 is more appropriate. Caution should be exercised that specifications correspond Comment by P. Larose: By plotting the log of "air permeability" againat the log of l/"seconds" a straight line of slope = 1 was obtained, with deviations well within which I would consider experimental error. Comment by L Fourt; Although the range of the Schiefer-Boyland-instrument estends from 0.64 to 700 ft3/ft2 min, there are some fabrics, such as light weight knit underwear and mosquito net that cannot be measured at 0.5 inch pres- sure. For these it is necessary to use. 0.05 or 0.1 inch. However, flow through the fabric is not proportional to pressure across it except for fabrics of low permeability. Plots of the logarithm of the flow against the logarithm of the pressure reveal the existence of two types of flow through fabrics, stream line or turbulent. The streamline flow, observed at low rates of flow and low pressures, is proportional to the pressure (slope of log-log lines = 1), while in turbulent flow the flow does not increase as rapidly as the pressure. An exact

98 CLOTHING TEST METHODS extrapolation can be made by the log-log plot, but for most of the fabrics which cannot be measured at 0.5 inch, no great error is introduced by assuming that the flow is proportional to the square root of the pressure. It should be emphasized, that there is considerable variation from place to place in a single piece of fabric. In tests of eighteen kinds of fab- rics, ranging in permeability .from 1 to 130, the standard error of the air permeability, as measured at five places, was calculated by the small sam- ple method. Expressed as percent of the air permeability, it ranged from 1.5 to 6.5$* with an average of 3'3$. This implies that on the average, a difference of air permeability of at least 10$ is required for statistical significance, defining the limit of significance as the 0.05 level of prob- ability (1 in 20 chance) that a difference that large could arise by ran- dom sampling error. There is even more variation when different lots, from different makers, of nominally similar fabric are considered. As a probably rather extreme example, eighteen different lots of herringbone twill were measured. The air permeabilities ranged from 6.5 to 25, with an average of 12.8, and a standard deviation, calculated by the small sample method, of 3.6. Although the "rating" of air permeability as the flow through the cloth at 0.5 inch water gauge is convenient and widely used, it may give a mislead- ing impression of the amount of air actually moving through fabric under ordinary conditions. To produce a pressure difference equal to 0.5 inch of water by'impact pressure requires a wind of 31-8 miles per hour. The flow through a combination of fabrics can be estimated from the flow through individual fabrics, on the assumption that each fabric has a re- sistance which is measured by the reciprocal of the air permeability. If AI, AS, AS, etc. are the respective air permeabilities, then 1 = .! 1 + I. etc. A combination AI A2 AS The estimate calculated by this equation comes within 10$, on the average of the permeability obtained by direct measurement, for pairs of fabrics. However, in the tests of air permeability the layers are as close to each other as possible, while in clothing there may be more separation, which would tend to make the resistances more nearly additive. Comment by P. Larose; In the method for air permeability, you give a means of calculating the permeability of combinations of fabrics. Might I suggest that although the method described gives fairly good results where only rough approximation is wanted, that mention might be made of the more ac- curate method which takes into account the variation of permeability with pressure. As you have stated, the logarithm of the permeability against the logarithm of the pressure generally gives a straight line. However, the slope of this line is apt to be quite different for different fabrics. It follows from this that in a combination of fabrics the drop of pressure across any one component of the combination can only be determined by tak- ing into consideration the variation of permeability with pressure. This can be readily done by remembering that the same amount of air must go through both fabrics. I have in some cases calculated the permeability of combinations of papers by that method and found that one can get very close results in this way.

National Research Council Subcommittee on C 1 qthj RA 779 . M3 1945 c Natl°nal Research Council U • S . ) . S LI ta c o m m i 1 1 e e o n 779 oN3 1945 c.i Clothing test methods