Document jye8jg3M1YaeBMx7ZGqa5OgKp

FILE NAME: Military Specifications (MS) DATE: 1936 DOC#: MS007 DOCUMENT DESCRIPTION: Journal Article - Heat Insulation - The Society of Naval Architects and Marine Engineers - Transactions THE SOCIETY OE 'NAVAL ARCHITECTS AND MARINE ENGINEERS TRANSACTIONS Volume 44 1936 Published in 1937 by The Society of Naval Architects and Marine Engineers 29 West 39th Street, New York City HEAT INSULATION B y C.u t a i y O h.mo.yd L. C ox, l T. S . X ., V isitor* The present-day high-pressure steam plants, with temperatures in excess of 500 degrees F., have created a new prol)leni in heat insulation. The old standby for many years, 85 per cent hydrated magnesia carbonate insulation, is not suitable as it calcines and decomposes at tempera tures in excess of 500 degrees F. An extensive amount of research has been conducted by various manufacturers in an effort to develop and pro duce insulating materials for the higher tem peratures, which will meet the Navy Depart ment requirements as to conductivity, structural strength, weight, etc. In general im single ma terial will meet all of the requirements, with the result that different types are being used. This paper discusses heat-insulating materials from the standpoint of the requirements of the naval service, service experience and the results of tests conducted at the D. S. Na\al Engineering Experiment Station. the II. S. Naval Engineering Experiment Sta tion. These tests are conducted to determine the suitability of the materials for use on ships of the United States Navy. The test apparatus and procedure were developed with the view of simulating service conditions. The mere fact that a material satisfactorily passes all of the tests cannot, in itself, be accepted as conclusive evidence that it will be satisfactory under actual service conditions. This is especially true of the newrtypes or brands. It is believed, however, that the tests do show which materials are un suitable and which materials give promise of being suitable. Tin1 tests are conducted to determine the fol lowing characteristics : (a) Conductivity. (h) Durability. (r) Ease of application. (d) Density. G txkkal Heat-insulating material may be defined as that substance which prevents or minimizes heat losses. Its use on the external surfaces of main and auxiliary machinery, boilers, evaporators, heaters, connecting piping, etc., under operating conditions, is necessary for several reasons: (a) To promote efficiency of plant operation by minimizing heat losses; (b) To prevent abnormal or unduly uncom fortable temperatures in the machinery spaces, especially in the closely confined spaces cm board ship; (c) To prevent injury to personnel through contact with hot surfaces; and (d) To prevent or minimize fin1 hazards. The various types of heat-insulating materials mentioned in this paper are discussed relative to the requirements of the naval service as speci fied by the various Navy Department specifica tions. Tile data presented were obtained pri marily from the results of tests conducted at * I>irec`tor IX S, Naval EnplnrerlnK K\|icrmuint Station. Annapolis Md Conductivity is of primary importance. If is a measure of the effectiveness of the material to prevent heat losses. The lower the con ductivity the better the insulating value. Next in importance is durability--a quality or characteristic denoting ability of the material to withstand the operating conditions of heat, shock, vibration, abrasion, moisture, etc. Oper ating conditions on board ship, as regards shock, vibration and moisture, are considered to be more severe than in shore plants. Such con ditions require materials of greater durability. Ease of application, for botli the initial instal lation and for repairs, involves cost, time and degree of skill required of the workmen Density is of importance especially on ship board installations where space and weight are at a premium. T kst M ethods The tests conducted at the Experiment Sta tion may he summarized briefly as follows: Conductivity Test. For pipe covering it is determined on the pipe test apparatus (Fig. 1) which consists of a 36-inch section of 3-inch 470 IIKAT IXSI'UATIOX 471 .standard wrought-iron pipe. The covering under test is applied to this pipe. Inside of this pipe there is an electrically insulated 2-ineh iron pipe around which are wound three heating coils. The center coil Covers the 30-inch middle section--each end coil covers a 4-inch end sec tion. The 2'-inch pipe is centered in the 3-inch pipe and the ends are elosed with insulating disks. The heat supplied to each coil is regulated by means of a variable resistance. The input to each eoil is adjusted until the temperature of the pipe over the end coils is the same as over the central coil. Under these conditions, cor rections for end losses are eliminated, and the heat supplied the central eoil is dissipated radially through the central 30-inch section of the covering. Tire temperatures of the inner and outer surfaces of the covering are measured by means of iron constant an thermo-couples. The input to the central eoil is computed from voltmeter and ammeter readings. Runs are made at several different pipe temperatures. The resultant heat loss curve represents the British thermal units lost through the cover, per square foot of pipe surface, par hour, per de gree F. temperature difference (pipe to room) for the particular thickness of the material under test. The conductivity of the pipe covering material is calculated from the following formula: where II zzz he at lu-.. H iitish th e rm a l units, per -quure toot of pipe bin face (30-inch central section only eonMilcied) per hour, k -- heat conductivity of material. 7'i -- inner mo face te m p e ra tu r e of m ate ria l, degrees F. 7'j r= outer surface te m p e ra tu r e of m a t e r i a l, degrees F. n = inside radius of material, inches, is outside radius of m aterial, inches. Using this conductivity, the losses to be ex pected from all sizes of pipes aiuFall thicknesses of material al any pipe temperature may be calculated. For flat sections, such as blocks and insulating cements, the conductivity is determined in the flat plate testing apparatus shown in Fig. 2. The apparatus consists of a round heater plate, 8 inches in diameter, to eaeh side of whie.h is applied a 1-inch thick, 8-inch diameter disk of the material to be tested. Additional insulation is supplied as shown in Fig. 2. The entire assemblage is surrounded with pulverized diatomaeeous earth. Fleetrieal energy is supplied the heater plate by two coils--one eoil heats the central G-ineh diameter area and the other coil the 1-ineh rim. Thermocouples are installed to indicate Ihe temperatures at different points across the face of the date and at the outer face of the test disks. When the temperature across the face of tlie heater plate is constant at all points, it is assumed that all of the heat sup plied to the central heating eoil is being dis sipated through Ihe test disks. However, in order to minimize the effect of losses through the edges of the test disks themselves, only the central 4-ineh area of the test disks and plate are considered in calculating the conductivity. Wiring Diagram of T o t i n g .Apiuintus 4%ft. No.TA Res.Wire LrL*_Pe_rl!lCj?_____ 36" /7 0 ft.No.!4 Res.Wire / 4Turns per inch k ------3 - --------------- ^ ---------- 30'i-------- /*)<------------- 3 - j 8 "3 2 Machine ' Screw 1H; 1| ., . 7T A sbestos P ap er So aked in A lu n d u m C e m e n r Fit;. ].- - P ick ( iivi.iunu Co m u r i m r v T r s n v ; Ar i c v mv i v s Asbestos Lumber Iron Plate 472 HEAT iXSl'EATlON F i o . 2.-- F i .a t P l a t e C o n d u c t i v i t y A p p a r a t u s The conductivity is calculated from the follow ing formula: " h e r e 1 -- lioat loss, B r itis h therm al u n it- per square foot of sin face (c en tral 4-inch section only eonsUlered) per hour. ' t --h eat c o m l u c t h i t y of m ater ial, r e - i n n e r surface tem p eratu re of material, d'T re es F. T s - o u t e r surface tem perature of material, decrees F. x --thickness ot m aterial Inches. Conductivity values are obtained for different temperatures. Conductivity is plotted against the mean temperature of the high and lowtemperature faces of the material. As a general rule numerical limiting values as to conductivity are not specified. To he acceptable in this respect, the conductivity must compare favorably with other approved types. Shock and Vibration Test, A section, in pipe covering form, is mounted on the test apparatus (Fig. 3). The pipe, with the material applied, is installed in a horizontal position on the vibra tion machine. In this position the material is subjected to 700 vibrations per minute through an arc of 15 minutes with a radius of 30 inches, for a period of at least 96 hours. The material is weighed before and after the test. Excessive loss of weight (about 5 per eent), sagging, loose ness, crumbling, etc., are causes for rejection. F i o , 3.-- S h o c k a n d V i b r a t i o n M a c h i n e f o k T e s t i n o H e a t - I n s u l a t i n g P i n : C o v e r i n g HEAT INSULATION 473 Hardness. Hardness is determined by a plastometer. The degrees of hardness of the various insulating materials, in 1/100 millimeter using a Vs-inch ball, are specified by the Navy Department specifications. Resistance to Abrasion. The structural strength is determined by an abrasion test (Fig. 4). In this test twelve 1-inch cube speci mens of the material and twenty-four H-ineh oak cubes (specific gravity of wood 0.63). are placed in an oak box having inside dimensions iy., inches by 7% inches by 7:/i inches. The oak box is then closed and it is rotated about its own axis at a speed of 60 revolutions per minute for two 10-minute periods. At the end of each 10-minute period, the eubes of the material under test are removed from the box and the per emit loss in weight, due to pulverization and breakage, is determined. The maximum allowable loss in weight is 40 per cent after the first 10-minute run and 60 per cent after the second 10-minnte run. Physical Changes Under Heat. Specimens of the material are subjected to a soaking heat, at various temperatures, in an electrically heated oven for a period of six hours. Changes in hardness, loss of weight, and linear shrinkage, are recorded. The size of the test specimens, oven temperature and allowable limits of change, are specified by the applicable specifications. Moisture Absorption. Specimens are sub jected to an atmosphere of 90 per cent relative humidity at 120 degrees F. dry bulb for six hours. The increase in weight is noted and the per cent absorbed by volume is calculated. Moistuie absorption is limited to 3 per cent by volume. Watn- Absorption. This test is not required by the specifications, but is conducted to obtain information which might be of future value. Specimens similar to those employed in the moisture absorption test are immersed in water at 80 degrees F. for six hours. The increase in weight is recorded. Chemical Analysis. A complete chemical unalvsis is made n accordance with accepted laboratory methods. Weight. Expressed in pounds per cubic foot. The applicable specifications limit the weight. Adhision Tisl (Insulating G.cments). The adhesive properties of insulating cements are determined on the cement adhesion test appa ratus (Fig. 3). This apparatus consist of a steel plate and a steel ring, each lO'/t inches in diameter and f^-inch thick, centered and held in parallel planes 2 inches apart by four studs set in the ring. The inside diameter of the steel ring is 6'/i inches. The material under test is held between the ring and the plate. A steel disk, 6 inches in diameter and %-inch thick, completes the surface in the plane of the ring. The adhesion of the material to this disk is measured. 474 11 F A T l N S F F A T K I X The insulating cement is applied to the appa ratus as follows: The steel ring is laid on a flat, horizontal plate. The steel disk is laid in the center of the ring, its center coneentrie with that of the ring. The continuous flat surface thus presented is covered with a 2-iueli troweled layer of the insulating cement. King, disk, sup porting plate and cement are dried at 300 de grees F. to constant weight in an electric oven. The 101i-inch steel plate is then secured in place by the four studs. Ilooks are screwed into centered holes on the outside surface of plate and disk. A measured load is applied to the hook in the steel disk while the specimen is suspended from the hook in the steel plate. Adhesive property is expressed in pounds of load (to cause rupture of bond) per square inch of sample. The aver age of three determinations is used. In the eases of new' type materials, of which no previous data or information are available, the lest procedure is somewhat altered--the ma terials being subjected to test for a longer period of time. If any doubt still exists as to tbe suitability of the material, the approval is with held until it lias successfully withstood service conditions for a limited period. C lassification' The Navy Department general specifications for machinery specify heat-insulating materials as follows: (1) For insulating steam, hot water, and hot fuel-oil containing surfaces, and in general to prevonl loss of heat by radiation, ilm standard insulating materials shall be " So per cent mag nesia" tXavy Department Specifications 32M2) wherever the temperature of the insulated sur face will not exceed 500 degrees F. (2) Where t^e temperature of the insulated surface will be above 500 degrees F. and will not exceed 850 degrees F., the insulating ma terial shall be high-temperat ure insulation (Navy Department Specifications 32P3). (3) For"the insulation of combustion surfaces in general, where the normal full-power oper ating temperature will exceed 500 degrees F., and will not be in excess of 1500 degrees F., higli-temperature insulation (Nnvv Department Specifications 32P3) shall he used. (4) In lieu of the materials specified in para graphs 1 and 2 above, ``crinkled" aluminum foil (Navy Department Specifications 47A5, grade D) or rock wool (Navy Department Speci fications 321*5) may be used, .subject to all of the following conditions: (a) Where weight saving is a consideration and where definite saving in weight will result. (b) Where the temperature does not exceed 850 degrees F. (c) When specifically approved by the Bureau. (5) In addition to the optional materials listed in paragraph 4, preceding, where irregular sur faces are to be covered, asbestos cloth pads of convenient size may be used when specifically approved by the Bureau. The asbestos cloth '-hall be in accordance with Navy Department Specifications 32C11, type B. The pads shall be filled with magnesia asbestos cement or pul verized magnesia block (Navy Department Specifications 32M2) where the temperature will not exceed 500 degrees F. and asbestos fiber U Mirysotile) or (Amosite), or rock wool fiber (Navy Department Specifications 32P5) where the temperature will not exceed 750 degrees F. The pads shall be sewed and quilted with fine, wire or with asbestos twine in such a maimer as to provide a uniform thickness of insulation. For many years 85 per cent magnesia insula tion predominated the field of heat insulating materials. The ease with which this material could be applied in its various forms, its low conductivity, its relatively low cost, its compara tive lightness in weight ami durability, were factors for the retention of its use on board ship. The temperature range at which this material could be used, however, was limited. At tem peratures in excess of 500 degrees F., it calcines and decomposes. Tin* use of high steam pressures and tempera tures in excess of 500 degrees F. has presented a new problem in heat insulation. As 85 per rent magnesia is unsuitable, it became necessary to investigate, develop and employ other types of heat-insulating materials. These materials were required to have highly efficient insulating properties and physical properties which would not depreciate it the higher temperatures. Extensive investigations and tests have been conducted on these materials. No one type lias been found to possess all the required proper ties, hence this has resulted in adopting and employing a group of materials which may be classed as follows: ' (a) Sectional and segmental molded insula tion. (b) Fabricated felted rock wool. IIKAT IXSl'LATION 475 () Ahnniiimn foil. Ul) Insulating cement. (e) Asbestos pad and blanket. (f) Asbestos tape. D isc issio n Sectional and scguuntal moldtd insulation materials are generally composed of large per centages of diatomaeeous earth and magnesium or ealeinm carbonate bonded together with small percentages of asbestos fibers. 'Phis class of material is generally more expensive, is heavier and does not have as high heat-insulating eflicieneies as other types. It possesses, however, high heat-resisting properties and it is suitable for use at temperatures up to 17)00 degrees F, Consequently it is employed for the insulation of pipe and combustion surfaces in general, where the normal full-power operating tempera ture will exceed 500 degrees F., but will not exceed 1500 degrees F. Whenever practicable, especially in pipe covering form, the material is furnished as an inner layer in combination with an outer layer of 85 per cent magnesia. Navy Department general specifications for machinery specify the thickness of the layers that are to be em ployed for different pipe sizes and temperatures. The double-layer type material reacts to the manufacturers' advantage because 85 per cent magnesia is cheaper than diatomaeeous earth. It is advantageous to the Navy because the com bination weighs less than a complete high-tem perature material. The high-temperature mate rials in most eases are not as good heat insula tors as 85 per cent magnesia, lienee more heat protection is obtained from the combination than from a complete high-temperature material. Manufacturers of the diatomaeeous earth molded type high-temperature materials have strived to improve their product. Marked im provement has been obtained through the use of calcined diatomaeeous earth. This has resulted in a more efficient lighter weight material, pos sessing adequate and ample physical properties. Fabricated rock wool materials arc composed of rock wool fibers, felted without the use of added binders and secured to inner and outer face metal supporting members. The rode wool fibers are made from natural stone, composed of ealeinm, magnesium and aluminum oxides and siliea. The rod; is melted and while ill the molten state it is blown with a steam jet. A fibrous formation is obtained. The grade of liber thus formed is dependent upon the process of manufacture. Drastic Navy Department leatlet specifications have been prepared in order to obtain a high quality material, with a low sulphur content and containing a minimum amount of undesirable shot or glassy globules. The metal supporting members of the material on the outer face are screen wire or 1-inch mesh wire covering, and on the inner face are I'/e-ineh wide strips of screen wire, spaeed I 11 to 2 inches apart and running longitudinally the length of the material. The securing and the fastening of metal members together and to the rock wool sect ion of the material is performed by individual, small steel wires, fastened approxi mately 0 inches apart along the metal strips, passing vertically through the rock wool and fastened to the outer .screen wire covering. Hock wool fuses at about 1200 degrees F., hence it is not used for temperatures over 1000 degrees F. It ranks high as regards its heatinsulating properties. Its initial cost is lower than that of most of the other types. Its weight is approximately the same as 85 per cent magnesia. When used on pipe surfaces it is wrapped around in a single layer and the abutting edges of the outer screen wire covering are secured by lacing or some type of elip. It is more dif ficult to apply to pipe surfaces than the molded type materials. A tight snug fit is imperative in order to prevent damage from vibration. For the purpose of saving of weight, experi ments and tests are now under way using this material as ;m inner layer in the combination type. Aluminum foil insulation was used in Ger many before it was introduced in this country. At first tin' idea of using a metal as an insulat ing material to retard the flow of heat seemed foolhardy. Nevertheless, very thin, pliable sheets of aluminum foil approximately 0.0003 inch thick, can he applied in such a manner as to form a rather efficient insulating material. The heat-insulating efficiency will vary widely with the manner of application of the foil and with the degree of expert ness with which the applica tions are made. Tn its most effective form, it is applied plain, in successive layers made of individual sections of foil separated from each other by strips or frames made from heatresisting materials, such as asbestos wood or built-up layers of asbestos paper. The cost of applying the material in this manner is too groat for the material to lx1 considered a prac tical insulation. Practical applications consist, of built-up layers of crumpled foil in which no 47 G HEAT I N F LATION spacer nr supporting sections an1 iw'd. In this form the foil is first crumpled hy hand and thou applied in layers which arc held apart approxi mately :is-ineh, solely by crmnpliinj. For its insulating value in practical installations, it depends on the reluctance of a bright surface to absorb and radiate heat and, in part, on the opposition to convection current presented by the small air cells formed between adjoining layers of crumpled foil. Any number of layers of foil may be applied, depending on the beat protection desired and on the allowable bulk of installation. In view of the fragile nature of the foil, each application, whether on pipe or flat surfaces, must be protected with an outer metal sheathing. Special construction therefore must be provided to support the sheathing. Aluminum foil installations using uncrum pled foil and spacer sections, as described in the preceding paragraph, are very efficient. As previously stated, the cost of application is pro hibitive. It is, therefore, not considered prac tical and its properties should not be used for comparison with those of the practical type materials. Practical applications, employing layers of the crumpled foil, generally are found to have heat-insulating properties which are not as good as those of rock wool and only slightly better than those of the molded type materials. The big advantage claimed for aluminum foil insulation over other types is its lightness in weight. This advantage is much reduced, if not entirely lost, by the added weight of the sheet metal protectors. "With the aluminum foiltype of insulation higher outer surface tempera tures are encountered than in the ease of rock wool or molded type of insulation. 'This is due to the reluctance of the metal cover to radiate the heat transmitted to it. There have been some cases where the outer metal eover of the insulation on Steam lines passing through con fined areas was of such high temperature that it had to be covered with other types of insulat ing materials for protection of the personnel. Installations of the material in horizontal layers gave unsatisfactory service. The layers of the foil sagged or closed in to the lowest level. The melting point of the foil is approximately 1200 degrees F. lienee. Navy Department gen eral specifications for machinery limit the use of the material to temperatures not exceeding 800 degrees F. The cost of aluminum foil its e lf is relatively low. but the labor cost of applying it and the patentee royalties are high. lienee, for cor responding thickness of insulation, the total initial cost of the material is higher than that of other type materials. Insiilat iihj cfmint.s are composed of widely variant materials. As a result, they vary mark edly in their heat-insulating and physical prop erties. In general they may be divided into the following groups, according to their basic constituents: (a) Diatomaeeous earth cements. (b) Ilock wool cements. (c) Exfoliated mica cements. All cements vary greatly among themselves as regards conductivity and weight, and in general are heavier and have a higher conductivity than other type insulating materials. They are prin cipally used on uneven or irregularly shaped surfaces. The primary characteristics to be considered in heat insulating cements are the following: (a) Thermal conductivity. (b) Ease of application and adhesive proper ties. (c) Coverage capacity or cost per unit volume after applied. (d) Refractoriness or the ability to withstand the action of heat at the temperature to which it is to be subjected. (e) Shrinkage, wet to dry. The diatomaeeous earth type is composed of pulverulent diatomaeeous earth, clay and asbes tos fibers. This type, after applied, molded and dried, in general has the most efficient insulat ing properties. It is. however, difficult to apply and has poor adhesive properties. It will not adhere to hot surfaces nor readily adhere to lower or downward-facing cold surfaces, unless reinforcing wire is used. Its coverage capacity in general is lower than that of the other type insulating cements. After applied, molded and dried, its weight, structural strength and other physical properties approach those of the molded type insulating materials. The temperature range for which this type of material is suitable is in the realm of that of the molded type in sulating materials. The rock wool type is composed of nodulated rock wool fibers, clay and asbestos fibers. This type generally can be easily applied and will adhere to any clean, unheated or heated surface. It can be applied to the surface of piping with out requiring the use of reinforcing wires. The adherence of this material to heated or hot surfaces is unique. 1IKAT lXSTbATIOX 477 T a ili; I.- R a n u e o f P h y s ic a l C h a u 'if iist ic s 8 o<:,\ Diuturna- Magnesia eeou.s F.nrth M ar ermi Molded Insulation {.Pipe and Pluck ) MInoqludieedtimi (Pipe und Block W eight, lb. per <u ft . . ... 16 18 Hurt ness by plusu.m eter, 1 1 0 0 irm . 55-11>2 18 6-27 5 30 95 Uesistunce 10 abrasion (structural strength r Loss in weight, per cent after l&t 10 min. r u n ......................................... 21 7-31 0 Loss m weight, per cent after 2nd 10 nnn. ru n ............... ................ 30 3-19 3 n 5 40 0 21 J 60 0 H ardness nfler soaking heats* After heating 6 hrs. at 750 deg F . . A fter heating 6 hrs. at 10OO deg. F . A fter heating 6 hrs. m 1200 deg. F . A fter h eating 6 hrs. ai 1500 deg F A fter heating 6 hrs. a t 1800 deg F , 57 93 49 -95 5 1 1-109 to 112 5 31 154 C hange in weight du e to .taking heat P er cent loss in w t, after healin g 0 hr?. a t 760 d eg. F ..................................... 3 3-1 0 P e r c e n t loss in w t. a fte r heal ing ft hrs*. at 1000 d eg. F ............................ 3 4 -5 0 P e r c e n t loss in w t. a fte r h o o tin g 0 hr?,. a t 1200 d eg. F ..................................... A 0-19 0 P er ren t loss in w t. after heat mg 6 hrs. a t 1500 deg. F ..................................... 5 1-20 0 P e r c e n t loss in w t. after h e a tin g 6 hrs. a t 1800 deg. F . . ............................ 7 1-21 0 C hange in linear dim ensions due to soaking heats: P e r ce n t sh rin k ag e in le n g th after heating 0 hrs. a t 750 deg. F . . . . 1000 deg. F .......... 0 0-2 0 0 0- 2 0 1200 deg. F ___ 0 0-2 0 3500 deg. F ___ 0 9-5 0 1800 deg, F ......... 2 3-7 5 Moisture absorption: Moisture absorbed, per cent by w e ig h t................................................. 3 3-8 3 M o is tu re a b s o rb e d , p e r c e n t b> v o l u m e .............................................. 0 S-3 0 W ater absorption: W ater absorbed, per cent In weight 144 0-251 0 W ater absorbed, per cent by volum e 70 0-78 0 W ater retained after draining 20 hrs. per cent by: V o lu m e ..................................................... 57 0 -6 8 0 W e ig h t.......................................... 4 . 117 0-212 0 C o v e ra g e c a p a c ity . s<j. ft., 1 ' th ick per 300 lb m aterial. . . . . . V olum etric shrinkage (wet to dry',, fu r cent .......................................... . . . . A dhesive p roperties, lb.per so in. . . Tensile strength, lb. per in wldth .. C hange in tensile strength- P e r c e n t Idss m ten sile s tr e n g th a fte r heating 0 hrs. at 450 deg F 650 deg F . . 750 deg F. C henge m weight P e r ce n t lost in w t a fte r h e a tin g 0 hr. at 450 deg. F . . ... . . . 650 deg F , . . ... 750 deg F ........... ................... .............. * As applied t V> eight m lb p er H'O im . ft liof k Wool Blanket 11 'ul.'Uion *15 ti 19 0 ___ . . . . . 6 7-12 9 0.8-2 4 31 0 -6 7 6 0 4 0-74 0 0 0-5.3 0 0 0-499 0 I ii.-iil.Aim; ( . lueiits Diutomaceous Larth 28 0- 86 , 41 0-51 0 MHott.oWl 2*1 1- 86 0 60 5--121 KMoliated M ica 17 0-18 0 159 0-1S 5.0 6 0 20 0 11 4 10 s 19 n -50 0 20 5 18 1) 21 5 53 0 2 1 5 50 ll 19 0 75 0 22 0 35 25 0-30 n 45 0- 57. 45 0 -5 0 .0 40 2 115 0 111 0-151.0 .30 0- 130 0 131 0-138 0 35 0 130 0 130 0-136 1) 31 s 118 0 120 0-130 0 21 (1- 101 0 To. Soft 3 0-7 0 5 0-110 0 4-15 0 10 5 -1 S 0 12 5 -2 3 5 0 7-2 X l Ot S 1 2 -5 1 1 b-0 1 2 4-0 0 Zero 1 0-1 3 2 3-2 6 134 7 6 6-6 9 f0).0000--00 3t,18 0 6 3 -0 01 0 03-4 36 2 0 0 -5 70 Zero 0 <K> 0 03 0 31-1 25 0 6 3-2 52 1 20-1 65 Zero Zero Zero Zero 1 1-1 4 41 0 -5 8 0 18 2-25 7 Zero 45 0 69 0 H 4-20 0 2 1 -12 1 75 0--84.0 9 8-12 5 9 0-11.2 A sbestos B lanket and Pad D suintion 1 3 .0 -1 4 .2 7.2-17.5 1.8-4 2 284.0-332 0 7 1 .0 -8 4 .2 5 5 .0 -6 5 .0 224 0-256 0 Insulating Tape 13 8 - 2 1 .O t 162-208 0-32 0-41 0-43 1 2-2 6 2.4-5 5 2 6-7.1 Table 2. -Xa w Dkpabtmknt Kim-cii u a'jions Title t'ijn'cxiicntion Genera) Specification- for Macliim-n ; (Heat insulation and L a t i n o for P ip in g and M a c h in e iy ) .................... S u h -w tio n S39-1 Cloth, Asbestos ............................................. 32C11 Ma<:nf'Ma: Block. Cement and Pipe Covering . . ........ 32M20 Insulation, High Tompei aturc: ]lmk. Cement ami Pipe Covering (Diatomaecous K.ulh} ................................ .321*3.1 and 321*31] Rock Woo); Pipe Cmeiing, Blanket, Cement and F i b e r .................................. 3 2 Poli Aluminum; F o i l ....................... . 47Aoa Cem ents; G e n e r a l .................................. In pi pparatnm Pad* and B la n k e t- ; A-lie?tu-> In prepaiuttoti T a p e ; I n-iilat inn, TIkm uial . Tentative Spcfifh atinn- onlv When applied ami dried, the material forms a fdose bond with the surface to which it is applied. It does not crack, flake or crumble off. It becomes tough and strong, and in general not as brittle as the diatomaceous earth type. It is not as easily damaged. The coverage capacity of this type is generally higher than that of the diatomaceous earth type. Its weight and its heat-conductivity properties are higher than those of the diatomaceous earth type. The tem perature range for wliiedi this material is suit able is approximately the same as that of the diatomaceous earth type. The exfoliated mica type is comparatively new as regards its nse in the naval service. Informa tion relative to this material is limited as few ti&ii-l-'att'1.wij>,,*,,< 478 IIKAT I X s n .A T IO X Fh.. ii - C iimu i in in i: \ m,i s in \' \uiiu T\ ri s oi 1(i \ i-1\-vf \ 11 \o M\ ii.ki \ i > brands arc available. The material is composed of particles of exfoliated mica, clay and asbestos fibers. The exfoliating of the mica particles is performed under heat at temperature of approximately 2000 degrees !*'. Tim particles in this form, because of the abundance of air voids, have very cfticient heat-insulating properties. However, they possess very low structural strength and, in the mixing of the cement, the exfoliated mica particles become compressed, crushed and broken. Therefore, its heat-insulat ing properties ar% lower than those of the diatomaceoiis earth and rock wool types. As regards ease of application and adhesion, it compares favorably with rock wool. The salient properties, in comparison with other types, are its high coverage capacity and lightness in weight. After applied, molded and dried, the structural strength and the temperature range for which it is suitable are approximately the same as those of the other types of cements. However, it is relatively soft and is easily damaged. A<tl>cntnx ptuls are composed (if men asbestos cloth in envelope, form, tilled with Amosite or Chrysotile asbestos filters and tufted at close intervals with wire-inserted'asbesto.s mini. 4 he woven asbestos cloth covering is cut mid tailored so that the material in its finished form will conform with the curvature or shape of the surface to which it. is to be applied. The edges and ends of the covering tire sewed with asbestos cord or yarn. ,,Metal hieing hooks sewed or secured to the outer surface of each section of the material provide means for securing and holding it in place after it lias been applied. Asbi.stos bltuihls are composed oi Anmsite asbestos fibers made up into the form of rovings. Tin"rovings are located side by side and fastened together by wire-inserted asbestos yarns so as to form mat or blanket sections. The wire-inserted yarns in each section arc located approximately ;ti-inch apart and run transversely and alter nately mer and under the asbestos rovings. It is manufactured in the form of a felt, approximately l-im-h thick, 40 inches wide, and in any desired length. In its application, the material is built up in layers, the number of which depends upon the insulating ctTe.-t desired. The layers are pliable and can tie readily shaped to conform with the curvature of the surface to which they are applied 1IKAT INSULATION 479 Asbestos pads <md blankds are employed where tlie application of other types of insulat ing materials would be diflicult or impracticable. The materials are frequently and extensively employed for insulating pipe flange joints, plain and corrugated pipe bends and turbine casings. The materials are suitable and used at tempera tures up to 850 degrees F. Their thermal con ductivities are higher than those of the molded type materials. Their weights are comparable to those of applied rock wool materials. Asbestos tapes are of variant construct ion. Almost every manufacturer of this type material has his patented type of construction which he claims to be superior to all other types. The material is generally composed of Amosite or Chrysotile asbestos fibers in the form of rovings, slivers and listings, totally or partly covered with a braided or woven jacket. This type material is generally manufactured Vi-inch and %-ineh in thickness, 2 indies wide, and in rolls 25 and 50 feet long. Tapes are employed in the naval service to insulate hh-inch and smaller size pipe with curves or bends. The material has relativcly poor insulating properties and it is used pri marily to protect personnel and minimize fire hazards. Most materials of this type are un suitable for temperatures in excess of 500 degrees F. Several special types, manufactured exclu sively for naval service use. were found suitable for temperatures up to 750 degrees F. ( 'on mu: vnvu Data As a lumd\ reference, Fig. G and Table 1 give the conductivity range and the range of physical charaderistics of the various types of heat-iusiilai mg materials discussed in this paper. Table 2 lists the Xavv Department leaflet speci fications coMM'ing these materials. Conclusions The various types of insulating materials dis cussed in this paper have been approved for use in the naval service. There are undoubtedly other types which to date have not yet been tested for conformance to naval requirements. Some types, which are not discussed in this paper, have been tested and found to be un suitable for naval use. A number of types have been rejected on account of their relatively high weight. The present weight and space restric tions in naval const ruction create a demand for light and compact materials. Manufacturers are aware of this and are continually striving to produce lighter and lighter heat-insulating materials possessing at the same time the other required characteristics and properties. The author's thanks are due to Lieutenant Commander A. C. -I. Sahalot, U.S. Navy. Super intendent Chemical Laboratory, and to Mr. W. P. Sinclair, Associate Mechanical Engineer, I'.S. Naval Engineering Experiment Station, for aid in preparation ot` this paper. DISCUSSION The P hesident.- I am sorry that Captain Cox could not be here, not that we regret the presentation by Mr. Whitaker, but as an old shipmate of mine when he was in the Design Department of the Bureau of Engineering, I know he would have liked to present his paper in person and to have handled the discussion. lie has introduced the subject of insulating materials in a manner which I am sure will lead to valuable discussion by others. The paper is now before yon, gentlemen, for discussion. Mm E. C. L l o y d ,* Visitor: The author is to he congratulated on his excellent presenta tion covering heat insulating materials, par * Armstrong Cork l `roduets Company, Lancaster, l'a ticularly as applied to temperatures in the range from 850 degrees F. to 1500 degrees F. Special attention should he directed to the fact that tests are carried out on both cylindri cal samples and fiat samples so that the results gained by each method may serve to some degree as a cheek on the other. The combined use of the two types of samples is of particular interest to those who have been concerned with the testing of insulating materials over the years. Another interesting development in test methods is the abrasion' test which follows the procedure of the rattler test used for many years by the paving brick manufacturer to determine the resistance to abrasion of the samples under test. There is one test mentioned in the paper =, Aictl ; A u 1M i * C L L . ^ J ~ . 4MJ JI KAT l X K L L A T I O X on which I would like to ask a question. 1 refer to the moisture absorption test where samples arc subjected to an atmiispliei e oi fib Iirr emit ivlatiu1 humidit.v at 12b drives 1-. <1 1-\ I,n il) l o r s i x h o u r s It is d i l i r ult to s.>e what b e a r in " tliis m o istu re reanni w ould liti'.i' on tin 1 s e r \ i c e of a il iu s u l.it uia; m a t e r i d Used at .700 decrees F. it waiiild lie interesting' tu know just how the uiuxunuin ailowalile remain of '! per cent t>y \oluine was arrived at. Quite frankly, u Would seem that the water absorption as secured, by immersion would furnish more valuable data as to the behavior o f an insnlatnua muteiiul in seiwiee than would tin' moisture absorption, thouyli neither test would seem essential evept ill the case of insulating materials lor use at temperatures below normal; i.e.. on those ordi narily used for low temperature insulation. On payes 47b and 470 the author points out some interestin" features of aluminum toil in sulation as it applies at temperatures up to Sod decrees F. I should like to ask if any tests have been made eoveriny the effect i\ euosS of this type of material after service of one year or more on shipboard. Also, it would be inter esting to Know whether the effect of moisture or sea water on this foil mutmial has been determined by test. There is one class of beat insulation largely Used by the Xavy and on shipboard generally that is not referred to in tlm paper, and 1 should like to know whether the author has tested iiisulaliu" brick material such is is used for the insulation of boiler furnaces at tem peratures up to 2600 decrees F. It would also be of interest to know whether or not tests of mineral wools have been extended to include the lighter and more efficient "lass wools which are now available commercially. Mi;. It. AV. Moi.fhsON',* Visitor: In Captain Cox's paper on heat insulation the fact is hroiiyht out that installation of li"lit-wei"lit materials for "(moral use seems to lie retarded because of their method of application together with application costs. A new process of heal iiisukition lias recently been placed before naval architects and marine emrinoers. This process affords a secure, easily applied method of liyhl-weiyht and efficient, coatings by sprayiny asbestos liber, Research and invention of the neeessnrv median teal phases have shown that eroeidolite asbestos fiber, known in the trade as Cape blue fiber, Kon^lipy & M a tt i o n C o m p a n y , AniMi-r, ] `n prodmes the best results. Asbestos liber is blown from 1 uo/./.lc directed towards the slU'l,me to be covered, tlm liber appearum is a Imill cloud suspended m mid-air. Atomi/cd iHiillmr pi (>pti11ct I Imm i spray u'lin. roiilarlx the fiber and lays it onto the surface. Mechanhal pressing is necessary in order to e\en the surface built up to the required thickness. In tins method all ymvrally known eharacterist ics of asbestos are retained ; namely, ability to withstand li'rnh temperatures, low thermal conductmty, <uul the power to resist the attacks of eliemieal action. 'fins spray im'hod ol insulation differs from tlie ippl lent ion of cements, blankets, or molded insulations, yet combines the best properties of each. Like a cement it ran be applied to iiTeuulai' surfaces L forms a continuous, homogeneous co\criii";tenaciously adheres to llie .surface, and is applied in place. The spray application of asbestos is of par ticular interest in marine work. The aver,acre weights of commonly used insulation materials are as follows; C't'.*IU`Mt tlMtiMl.lU I n- u i.il 1njr t . i p r - IIi^h-tiMiipi'i atuic A-hr-to- U.HikrU Cuik li.Mid(hound fink IS ti> o() ])i)lllhl' pel t'd'iL' 1' hiL M l> i' l p o t l n d ' |>or r u l e r f*>t n"Med blocks !!i in 27 p o u n d - per ciiliir f"it .tnd ju d lH to IS pouml- per cuhic foot ^ to IS pounds poi ruMc foot 0 to 12 pound- per t lib!o foot Sprayed asbestos fiber weighs only 6 to S jiounds per cubic foot, 'fliis liyhtncss in weight is "overned by the fiber used, the binder em ployed, the type of surfaein" required, ami the decree of firmness desired. In the spray application the density of the eoutin" is reduced in proportion to the built-up thickness. Naturally this decrease eorrespondiuyly increases the efficiency as a beat insulator. thermal conductivity of the material compares favorably with the other insulating materials mentioned. The conductivity is particularly yood for the lower refrigeration temperatures. This sprayed material has been successfully applied in marine work for the insul.ition of partitions, ceilings, tops and side walls of cabins and salons, boiler room ceilings to pre vent beat transmission to rooms above, under side of decks, bulkheads and stack up-takes. Anion" shipowners, this material h is been applied by the French Tune on the s'cainsliip Xoniunidi(. Tn America its use lias been intro duced by the Standard Oil Company (if New Jersey and the Oulf Refiniii" Company in their The low 11MAT I NSTI GATI ON 48i oil tankers. In automobile and bus eonstruetion this method lias shown additional advantages by spraying thin members subject to vibration. The firm adhesion of the coating to the metal hacking has a deadening effect. There may he places aboard ship where a simi lar application would be helpful. In all forms the sprayed insulation is fire resistant. While not entirely impermeable, it is highly resistant to the action of water. In general, the fiber-binder combinations resistant to fire are least resistant to water. 'Where conditions require, it is adaptable to surfacing for decorative effects or appearances, pro tection against abrasion, ease of cleaning, weather proofing, and additional fire protec tion. The insulation possibilities of this ma terial in marine construction and repair work present a large field for future investigation. Practical applications have been made with entirely satisfactory results, for temperatures ranging from minus 10 degrees T. to plus TOO degrees F . ; and under test conditions this insulation has withstood temperatures rising from room temperature to 1700 degrees F. within one hour. Jilt. C. L. Norton-,* Visitor: The paper in general seems to be very well presented and the scope and methods of tests shown seem to be very satisfactory. The pipe covering conductivity testing apparatus is practically a standard setup and is in accordance with accepted practice. The flat plate condneth by apparatus is a good setup which should give accurate values. The tests devised for measur ing resistance, to vibration, abrasion and adhesion appear to be excellent tests and should give very representative figures. The choice of materials covers the low-tem perature insulating field quite thoroughly with the possible exception of insulating brick", which might possibly have been included in this report. The only criticism of this paper which comes to mind is the question of aluminum foil insula tion. This particular material has been the subject of mudi controversy and 1 fool tliat a little more might have been said in its favor in the paragraph under this heading. In general, I believe that this is a very good description of available low-temperature insulat ing materials and should be of great benefit to a person contemplating the inc of such materials. .Mu. Max I'Kii/ri No.'* Visitor: The author is to be highly complimented on his clear and concise description of the methods ot testing various types of heat insulation; also Ins detailed analysis id' the merits and demerits of the various materials to which he refers. We arc particularly interested in the com ments about aluminum foil insulation, and wish to emphasize the fact that the author most assuredly refers to high-pressure steam pipe insulation in all of his statements and not to the use of aluminum foil insulation "for high- temperature surfaces such as boilers, turbines, uptakes, etc.; nor for low-pressure surfaces such as decks, ship sides, cold storage rooms, bulkhead insulation, etc. v In general the methods of applying Alfol, as developed abroad, have been followed by the United States Navy Department in the appli cation of aluminum foil to high-pressure steam pipe lines. This method of application is fre quently less costly than for other insulations when equal insulating effect is required. The author states. " The big advantage claimed for aluminum foil insulation over other types is its lightness in weight. This advantage is much reduced, if not entirely lost, by the added weight of the sheet metal protectors." It is the usual procedure of the Navy Department to metal jacket, exposed high-pressure steam pipe lines, regardless of what type insulation is applied. Materials such as h5 per cent magnesia and high-temperature insulation for pipe lines are frequently jacketed with metal. These metal jackets are of a similar weight to that required for jacketing aluminum foil insulation. The weight of the Alfol insulation is approximately one hundredth of that of 85 per cent magnesia or other molded typos of high-temperature insulation thereby making a much lighter weight complete application. Tins weight sav ing is particularly important on pipes 4 inches and over in diameter. In fact, 34-ineh pipe insulated with U1> inches of Alfol shows a weight saving of approximately 15 pounds per linear foot when compared with 14-iueh pipe insulated with l b indies of high-temperature insulation and 2 inches of 85 per cent magnesia. 'Hie author makes the following statement: " There have been some eases where the outer metal cover of the insulation on steam lines passing through confined areas was of sneh Iiigh temperature that it had to be covered with other types of insulating material for The Jialx'ock & W;im\ Compiitn, X<"\v York. Alfol Insulation Company, Inc., New York. -t.L.^fc.^ ^ ' .wiw x .a a.J^ tfa^ v y . ^ . f n u r tujiUiiN JO'..T^ a M ^ j.>&^,a Ja H i & ^ v ^ A t <au>-4ri>w ,istf^. < *-^<-' .O iA riu r^ c. gwj 11 KAT IXSU l.ATIt )X protecton of the personnel." Tin' impression is lined that this refers onl\ t" All'll iiwilaifi steam pine lines, hut there me many eases where steam pipe lines insulated with other materials ami metal jaeketed ha\e had to he euvered with another jacket <>t l u s u h i t i o n lor the protection of the personnel. \\ < teel that this point should be made clear, because we arc con fident that it was not the author s intention to infer that aluminum insulated pipe lines metal jacketed were the only lint's aboard ship which have had to be covered with an additional layer of insulation in order to lower (he xurtace tem perature to a point where the personnel would not be burned. The data recorded in Table 1 Kange of Physical Characteristics), as compiled by the author, is of utmost interest. Although no figures are given in this table for aluminum foil insulation, it is an established fact that the aluminum foil insulation is not subject to the following factors: Increase in hardness after soaking heats; change in weight due to soaking heats; change in linear dimensions due to soaking heats; moisture absorption; water absorption; water retention. On the other hand, the figures for water absorption as established for diatomaeeous earth molded insulation, rock wool blanket insulation, and asbestos blanket and pad insula tion, are most important. These figures are: P i ATOMACKOUS K a HTU Moisture Absorption : Moisture absorbed. percentage by weight 3.3 to S 3 K<)< K WOOL H laM vKI . 0.7 to 12 As P.Rs ToS H LAX KK1 AM) l*.u> 7.2 to 17.5 Water Absorption : Water absorbed, per centag e by weight 144.0 to 254 0 31 0 to 070 0 2S4 0 to 332.0 When one considersthat the conductivity of water isapproximately ten times greater than an average good insulating material and, furthermore, tha^ the types of insulation re ferred to absorb moisture and absorb water at a relatively high rate of percentage, it is obvious that laboratory test conductivity fig ures do not and cannot apply in actual service regardless of the amount of heat which is within the pipe line upon which such materials are applied. This is because of (hanging tem peratures within the pipe line and within the surrounding area. The possibility of condensation within an insulation is far greater on surfaces such as ship sides, decks, bulkheads, cold storage room compartments, etc., than it is on high-tempera ture steam pipe lines. Therefore it X obviously highly d e s ir a b le to pro\ide insulating material which cannot absorb moisture and vdio-.e insulating value cannot be appreciably atlected by cnmlensatimi. Aluminum foil imulation. being all metal, meets this reipiireiiumt. The author states. " Installations ol the ma terial m horizontal layers gave unsatisfactory service. The layers of the foil sagged or closed into the lowest level." We have never heard of a condition of this kind. M lion aluminum foil is properly crumpled, it is so stilt' in proportion to its weight that any number of layers used in insulation practice would be self-supporting without any settle ment even under severe vibration. There are many applications which have been in use for approximately ten years, in which there has been practically no change in the insulating efleet of this type of construction nor is then' any sign of settlement of the various layers of foil. In closing this discussion we believe it is a matter of public record that a great many ships of the United .States Xavy are using Alfol (Aluminum foil) insulation, and we have re ceived no serious complaints about its perform ance during the five-year period in which it lias been in operation. Furthermore, there are now many more ships under construction in which aluminum foil insulation is being extensively used for many purposes, as the result of its past performance aboard ship. L i e u t e n a n t C o m m a n d er Tt. W. B r e n e r . U.S.X., Visitor, and M r. F. M. M cG reary*, W.-ntor: We have read this paper and are in agree ment with what has been stated therein, with one exception. We believe the following statement, appearing on page 47(S. should he (ptalifinl; " The big advantage claimed for aluminum foil over other types is its lightness in weight. This advan tage is much reduced, if not entirely lost, by the added weight of the sheet metal protector." This statement is true when aluminum foil insulation is applied to pipe and other pres sure containers which are not customarily lagged with sheet metal protectors. P rofessor Gordon B. W ilk es, f Visitor: This paper by Captain Cox illustrates the thorough ness with which the Navy investigates materials that are to be used on its ships. KdKi(n*T ( i n a t o r i a N ) I'niUM S t a t i c N a v v Do piirtinpu r t M a.ss acliu ae tts I n s t i t u t e of Topht tolo^y. (' ahi hririge, Maws. HEAT IXSl'LATIOX 4*:i Tin* thennal coiiductivily apparatus vquiro some comment. From tlu* drawbngs, there is no indication that any attempt has been made to insulate thermally the heater plates irom the guard-ring heater plates. This would necessi tate a much more accurate balance of tem perature between the heater and guard than in the case where one has a definite thermal break, because of the much higher thermal conductivity of metals as compared in insulat ing materials. The design of plate conductivity apparatus usually calls for a groove cut in the metal plates, in an attempt to reduce the cross .sectional area of the '`through metal." In the apparatus at Massachusetts Institute of Technology, all through metal has been eliminated between the heater and the guard. The moisture absorbed by the rock wool blanket insulation is amazingly high and is probably due to a typographical error. Many experimenters have found that rock wool, slag wool and glass wool will only absorb a fewr tenths of 1 per cent of moisture by weight when exposed to high humidities. Under aluminum foil insulation, it is stated that the east of applying '`spaced" foil is " too great for the material to be considered a practical insulation." This may be so for Xavv work but there have been many examples in the past few years where aluminum foil has been applied in this manner at a com parable cost and efficiency to that of the older types of insulation. The insulation of the air space between the studs of a frame wall by spared aluminum foil is an excellent example of good insulation at a cost comparable with mher types of equally efficient insulation. The paper also states, that " installations of 1he material (crumpled aluminum foil) in hori zontal layers gave unsatisfactory service. The layers of the foil sagged or closed in to the lowest level." It would he interesting to know the actual conditions of some of these unsatis factory installations. There arc a great many commercial installations whore crumpled foil, approximately 21 layers to the inch, has been used in a horizontal position over con siderable periods of time with satisfactory results. Many refrigerator ears and trucks, both in this country and diroad, have their roofs insulated with crumpled foil and the results have been satisfactory despite the obvious vibration. Laboratory vibration tests on crumpled aluminum foil in a horizontal position show no indication of settling when it is spaced about 21-; layers per inch. 2 inches thick and with a foil thickness of about 0.0004 inch. T h e PtiKsinuNT: 1 was glad Mr. Lloyd made inquiry as to whether aluminum foil insulation bus .stood up in service, and whether it has been resistant against corrosion. That was very prominently in the minds of those in the Xavy Department who were considering the question of using aluminum foil more than five years ago. at a time which was given by one of the later discussors. I think this type insulation lias been in use for one or two years longer than that. It seems to me thilt the time has been suffi cient for examinations to show how this material has stood up in service and the reason I go into it, in what, may sound like repetition is that there may be .someone here from the Bureau of Construction and Repair, who is familiar with the results of investigations such as are always made at periodic intervals on materials of this nature. If a representative of the Bureau is present and is disposed to speak, he may be able to add information as to the materials which have been tried out and which, as I understand, have been replacing the aluminum foil. One of the points which seemed to be made against the use of aluminum foil insulating material; namely, that it required sheet iron covering, would apply only in certain instances. We were using it very largely on vertical sur faces which were already covered by sheathing of soaie kind. The only weight that we had to add to bold the foil in place was in the form of some clever contrivances which were arranged to bold it up. As I remember, no large additional coverage was necessary for a very large amount of it, in connection with the uses contemplated for it at that time. Mu. If. C. I'.mu.ktt.* Visitor: The author is to be complimented on presenting such a clear and concise description of the Xavy Department s methods of test, of the nature of the various insulating materials in naval use, of the advantages and disadvantages of each, and of their properties, not only physical and thermal but actual application qualities as determined by sendee experience. The writer is in entire agreement with the author of this paper that there is no single insulation material which will meet all the. .1nhii'-M,! it \ t!1* (`ni i-nr.t t ion, X<<\\ York. 484 III-:AT L WS l ' LAT I OX requirements. Each type must be used where its individual characteristics be-t serve the particular conditions ol' service. When new materials are produced, there is quite often the tendency on the part of a producer to believe that the material will supplant everything else in existence. However, the proper field for any insulating material, as this paper shows, is determined only after thorough development and testing in the researeh laboratory, followed by adequate field service tests. Among the various required properties which an insulation must have, the author lists dura bility as second in importance because of the severity of shook, vibration, abrasion, moisture, anti other unusual conditions on shipboard. Also low density is listed as especially im portant. It is appreciated that low density in insulating materials is generally obtained by sacrificing durability and for this reason, information on the relative value of these two properties would materially assist producers in developing insulation for naval service. The Navy test requirements for insulation are very complete. T h e test for resistance to abrasion indicates the ability of a material to withstand handling, shipping and general re sistance to abuse. In reference to the test for determining llm physical change of a material under heat, it has been the experience of the McMillan laboratory, with which the writer is associated, that the full effects of subjecting some insulating materials to soaking tempera tures in an electric oven, are not obtained unless the material is heated over a considerable period of time. This is particularly true of those materials which show crystalline changes under heat. An extreme example of this is molded mono hydrated bauxite block which required nine teen days at IQOO degrees F. before complete shrinkage occurred, the shrinkage having in creased during that period hv -100 per cent over that shown after the first twenty-four hours. On a precaleined diatomaeeous earth block, however, the full shrinkage is obtained in approximately seven days with an increase of only 75 per cent over that shown after the first, twenty-four hours. The service experience with the various insulating materials as enumerated by the author are particularly valuable. They show that the lighter type of materials such as mineral or rock wool and aluminum foil, which in some cases (Fig. fi) show lower conductivity in the laboratory tests than the molded types such as S5 per cent magnesia and diatomaeeous earth, are also subject to greater variation in application than the molded forms. For example, it is stated that the rock wool type is more difficult to apply, a snug fit being imperative to prevent damage from vibration. Also that installations of aluminum foil in horizontal layers gave unsatisfactory service, the layers of foil sagging or closed in to the lowest level. These changes in physical properties of the material undoubtedly change the insulating value. It would be very desirable if tests could be devised to determine the extent of this change in insulating value after the material had been subjected to service condi tions. It would also be of considerable inter est to have data on the effect of moist air, par ticularly salt air, on the metals contained in insulations. Exposed metals on board ship are painted or kept clean and protected as far as possible from oxidation but metals within the insulation, such as wire reinforcement in the rock wool typo covering and the foil proper in aluminum foil covering, might be subject, to rapid deterioraiion. It is well known that metallic aluminum exposed to air and moisture does not remain bright, indicating a change in the surface, so that data on the permanency and ultimate insulating efficiency of these cover ings containing metals would be valuable. Tn addition, I would like to ask the author whether the thermocouples in the pipe covering apparatus are all on one plane, as the drawing might indicate. It has been our experience that unless these are placed in planes DO degrees apart in a spiral direction around the pipe, you do not get a true average temperature of the pipe, since there is a difference between the temperatures on the top, sides, and bottom of the pipe. Another point on which I would like some information is in connection with the conduc tivity curves on page 478. These show a wide variation between the maximum and minimum values for some materials such as aluminum foil, cements, etc., but 85 per cent magnesia and mineral wools are shown to have a very small variation. This agrees with the experience of the McMillan Laboratory with the exception of tests on mineral wool. In many tests on all types of mineral wool we have found a wider variation in conductivity than with most other types of material, excluding cement. The min eral wool curves arc steeper than found by the author, particularly on low density material, 'fhe variation in conductivity has been such that HEAT INSULATION 485 iu many cases the values wen* higher than for 85 per cent Magnesia. In conclusion, it should lie stated that this paper has given a clear picture of the naval requirements for insulation and il is hoped that future papers will follow from time to time as more materials are tested and used in naval service. C aj'TAIN C o x : I am grateful to Mr. Whitaker for the excellent manner in which lie presented the paper and I tun greatly pleased with the spirit in which it 1ms been received and discus sed. Before replying to the discussion, I would like again to point out that the principal object of the paper was to discuss heat-insulating materials from the standpoint of the require ments of the naval service, service experience and the results of tests conducted at the U. S. Naval Engineering Experiment Station. For many obvious reasons, I could not discuss specific brands of materials but bad to confine myself to a discussion of the propert ies of various types; neither could confidential information be divulged, nor could unsuitable types or brands be discussed. If in answering the various ques tions raised I stray from my original purpose, it will be only because the question cannot be answered in any other way. Mr. Lloyd questions the value of the moisture absorption test and further adds that it would appear that more valuable data would be ob tained from water absorption by immersion. Some types of materials are not appreciably affected by moist-air conditions. However, there are other types in which the stability is affected by moist-air. The 3 per cent limit on moisture absorption by volume was arrived at from the results obtained from the tests of the various types and brands of satisfactory materials. The main propulsion plants of naval vessels arc secured over 50 per cent of the time and dur ing the shut-down periods the insulation material is subjected to low-temperature moist-air con ditions. The range of water absorption by the various materials varies considerably. Such a wide range made it impracticable to set definite limits. Although no water absorption limit was established, the materials are required to with stand this test without showing signs of deterio ration. The data on rock wools does not include glass wool. Only one partial test has been made on glass wool and that was some five years ago. The material submitted at that time was not considered satisfactory because the method em ployed for its application was not practicable and the supporting members were combustible. Insulating brick material was not included iu this paper, a- the tests of such material are not conducted at the Engineering Experiment Slat ion. Mr. MollHon disi-uws a type of insulating material which has not been fully tested it the Engineerhur Experiment Station. Some time ago one brand of this type of insulation (sprayed asbestos libers and hinder) was tested for its aceonstical properties. This material took quite soim1 time to apply and the binder supported combustion. For these reasons it was not even considered is a heat-insulating material. Referring to Professor Wifkes' comments about the conductivity test apparatus, the heater plates consist of two alundum disks and are not of metal, as is to be inferred by the comment. This apparatus and the pipe test apparatus are in accordance with the recommendations of the Committee on Heat Transmission, Division of Engineering and Industrial Research, National Research Council, published February 1, 1929. The. data on moisture, absorption of rock wool blanket insulation, ns given by Table 1, are cor rect. These values appear high to Profes sor Wilkes, but it must be remembered that standard-size specimens 4 inches by 4 inches by 2-1/2 inches are used iu the test and that the relation between exposed surface area and weight is high. For large blanket sections, in which the relation between exposed surface area and weight is less, the per cent of mois ture absorbed will also be less. The Station has firsted water-repellent rock wool blankets which absorbed little or no moisture. The author agrees with Mr. Parlett that the full effect of the soaking heat tests cannot, with some materials, he obtained within twentyfour hours. However, the soaking heat test is not fully representative of serviee conditions. A test of only six hours' duration is employed to determine if (lie material is susceptible to change. The thermocouples of the pipe test apparatus are all in the same plane. The Station has not conducted any tests or experiments to determine the best positions for the thermocouples. Ac cording to Mr. R. II. Heilman, staggering the thermocouples in different planes is a useless refinement (Mi chnniral Engineering, Volume 40, No. 10, pages (>05-G0fi). In any case, as previously staled, the apparatus is constructed in accordance with the recommendation of the National Research Council. 4s(i 1i i-:AT I X S l ' L A T l O X J?t'ierr!ii" to tin* conduct n n \ curves for mineral wools, tin* curves shown :n Kiu (> arc f'lily for three approved types wholi arc m ac cordance with Xavy Department Specifications. In regards to aluminum foil insolation, let ns first discuss the shoot metal protectors. <Jalvanizod shoot steel or aluininimi alloy sheet arc required in the naval service to In' used mi the main steam lines and on other lines only when exposed to chafiny. abrasion or to weather. When aluminum foil is used as the heat-insiilatiii" material, the sheathing is ().02d inch nominal thickness; with other types of limit insulation, the sheathin<r is required to he 0.014 inch nomi nal thickness. The author concurs with Mr. Hrietuny's com ment that aluminum foil insulated pipe lines, metal jacketed, are not the only lines on board ship which have had to be covered with an ad ditional insulation in order to lower the surface temperature to a point where the personnel will not be burned. All aluminum foil insulation requires sheet metal protectors, while other t.\pes of lieatii)subitin*i material, especially for many pipe lines, do not require metal protectois. There are a number of steam lines on board ship which, if insulated with aluminum foil, require addi tional insulation o\er the metal protector to prewnit burning of personnel, while these same lines, if insulated with other materials, would not require this additional insulation over the canvas rover. 1 feel that my choice of words in tl <follow imr sentence. " The layers of the foil sayyed or closed in to the lowest level," has been misleadiii". It did not intend to imply that the layers flattened out completely, but that in horizontal layers the optimum void space for obtaining best efficiency was considerably minced. The Xavy continues to use aluminum foil for vertical insulation but found it very difficult to apply properly in a horizontal plane. The insulation was made up in sections and then installed. Dui'iny the in stallation the thickness of the material chamred as much as from one inch for three layers to less than one-half inch for three layers. In reply to Mr. Lloyd's question, no tests of aluminum foil have been made after it lias been in service for a year or more on board ship. Thk Piutsinrvr: We will convey the thanks of the Society to Captain Cox for his very interestinir paper, whi'di we are so yiad to lone from a man of his very outstanding experience in that field.