Document 4vYzL5QL7JXa7Y9wXEYrkrBYQ

DownloadSearchable PDFRandom document
Fire-Retardant Coatings for Fabric-Covered Aircraft S. G. WEISSBERG AND G. M. KLINE National Bureau of Standards, Washington 25, D. C. H. L. HANSBERRY Civil Aeronautics Administration Experimental Station, Indianapolis, Ind. Simulated power-plant fire tests of doped fabrics coated . MATERIALS INVESTIGATED with fire-retardant coatings, conducted under wind-tunnel conditions, demonstrate that it is possible to increase the critical time interval between the instant of first contact of fire with fabric and the instant of fabric destruction, from the present value of 2 seconds with cellulose nitrate dope or 6 seconds with cellulose acetate butyrate dope to 12 seconds with a fire-retardant coating applied over cellu lose acetate butyrate dope. A laboratory method for the quick evaluation of relative performance of fire-retardant coatings is described; this test will facilitate the further development of fire-retardant coatings. The protective Table I describes the materials investigated. Materials which have been most successfully used in fire-retardant coatings are (a) film-forming substances' which, on pvrolysis, give off large quantities of noncombustible gases and (b) pigments which func tion by the production of noncombustible gases or by the forma tion of a protective glaze which excludes oxygen (), Since most of the film-forming materials must be plasticized, it is desirable to incorporate as much fire retardancy into the film as possible by using fire-retardant plasticizers; a number were included in this investigation. In addition, several commercial preparations recommended by manufacturers as flame-resistant coatings were evaluated for comparison. action of fire-retardant coatings on cellulose nitrate dope is too small to be of any value. The results of outdoor ex DOPED FABRIC TEST PANELS posure tests reveal several fire-retardant coating systems Wood frames were used for those doped fabric assemblies which with good weathering characteristics which are recom were to be tested in the full-scale wind-tunnel fire tests at Indian mended for further development. The use of a mixture of boric acid and borax as fabric impregnant, while effective as a fire retardant, has a deleterious effect on adhesion of apolis, and metal frames for panels to be placed on outdoor expo sure in Washington, D. C. Grade A airplane fabric, weighing 4 ounces por square yard, was stretched over the 15-inch square frames which had openings 12 inches square. The fabric was then dope to fabric which is apparent on outdoor exposure. coated with 10% by weight of a mixture of 3 parts of boric acid and 7 parts of borax, applied as a 7% aqueous solution. The pur pose was to minimize the tendency for the flame to be propagated under the fabric in the still-air burning test. Four coats of clear AIRCRAFT power-plant fire tests conducted by the Civil l Aeronautics Administration at the National Bureau of cellulose acetate butyrate dope, conforming to Army-Navy Aero nautical Specification AN-D-1, and two coats of pigmented cellu lose acetate butyrate dope, conforming to Specification AN-D-2, Standards in 1943 emphasized the need for providing fire-retardwere applied to the fabric. Dope was applied according to the ant coatings for doped fabric surfaces. Although complete fire proofing is not feasible or perhaps not even possible, substantial protection can be afforded by providing a coating which will de procedure defined by Navy Aeronautical Specification SR-70e. The total weight of dope applied was about 4.5 ounces per square yard. Fire-retardant coats were sprayed on this basic six-coat system, the number of fire-retardant coats depending upon the lay destruction of the fabric, after the outbreak of a fire, beyond tests to which the panel was to be subjected. The weight of each the time interval necessary for extinguishment. Therefore, a fire-retardant coat was approximately 2 ounces per square yard. joint program of the Civil Aeronautics Administration and Na tional Bureau of Standards looking toward the development of fire-retardant coatings for doped aircraft fabrics was set up. Fabrics doped with cellulose acetate butyrate do not present the fire hazard encountered with fabrics doped with cellulose nitrate. However, even the slow-burning butyrate dope would not be satisfactory in power-plant fires in which the doped fabric is likely to be in continuous contact with flame for several seconds. No fire-retardant film-forming materials are known which tauten airplane fabrics as effectively and as permanently as do dopes based on cellulose derivatives. Furthermore, the addition of fire-retardant resins to cellulose derivatives has little effect in increasing their fire retardance but has a serious effect on tautness as shown by Kline (1). Therefore, at the outset of this program it was decided to provide additional fire resistance to doped fabric by applying a fire-retardant surface coating. The essential prop erties required in the surface coating are as follows: (1) fire retardancy, (2) good adhesion to doped surfaces, (3) minimum effect on the tautness of the substrate-doped fabric, (4) applica bility by usual coating techniques, (5) flexibility, (6) resistance to weathering equivalent to that of currently used dopes, (7) re sistance to action of gasoline, oil, and ethylene glycol, and (8) aerodynamic smoothness. EVALUATION OF FIRE RETARDANCE Still-Air Burning Test. It was necessary to select optimum combinations of the fire-retardant materials to serve as a basis for further development. In the initial stages of the investigation the apparatus developed by Brown and described by Kline (I) was used. It consists of two parallel steel clamps supported by a steel frame, the distance between the two clamps being adjust able. The specimen to be tested is fastened between the clamps and ignited at one end. The time required for the flame front to travel over a given distance (5 inches), as measured along the clamps between two marks, is recorded with a stop watch. Speci mens used in these tests were 3 inches wide and 8 inches long, and were taken from panels with two fire-retardant coats. They differed from the specimens used by Kline in that two row's of Vs-inch-diameter holes, with centers l/, inch apart lengthwise and 2 inches apart crosswise, were punched in the specimen. The purpose of these holes was to provide a constant air supply throughout the test. The apparatus is suitable for either hori zontal or vertical burning tests. Moving-Air Burning Test. It was soon found that, as the development led to coatings which were more fire retardant, the 1742 August 1949 INDUSTRIAL AND ENGINEERING CHEMISTRY 1743 Table I. Materials Investigated in Fire-Retardant Coating Development Trade Name Chemioal Classification Manufacturer Film Basks Pliolite Insl-X Parlon 125 Parlon H Parlon X Mathieson rubber 153b Vinylite VYHH Resin 256-27 Resin X-124 Resin X-120 Resin P-120 Mathieson plastic 153a Amercoat 1138 Araecco paint Pyropex dope Skylac dope Rubber hydrochloride Chlorinated rubber Chlorinated rubber Chlorinated neoprene Chlorinated polyiaoprene Butadiene-diohlorostyrene Vinyl chloride acetate Vinyl chloride acetate Vinylidene chloride-acrylo nitrile Vinylidene chloride-vinyl chloride Vinylidene ohloride-acrylo* mtrile Polydichlorostyrene Vinyl chloride-vinyl acetate soln. (45% solids) Fire-retardant paint Ethyloellulose dope Cellulose acetate butyrate dope Goodyear Tire and Rub ber Co. . Insl-X Co., Ino, Hercules Powder Co. (Inc.) Hercules Powder Co. (Inc.) Hercules Powder Co. (Ino.) Mathieson Alkali Works (Inc.) Bakelite Corp. American Resinous Chemicals Corp. Dow Chemical Co. Dow Chemical Co. Dow Chemical Co. Mathieson Alkali Works (Ino.) American Pipe and Con struction Co. Amecco Chemicals, Inc. Sherwin-Williams Co. Monsanto Chemical Co. Modifying Resins Acryloid C-5 Acrylic reain Acryloid B-72 Acrylic resin Acryloid B-75 Acrylic resin Arochem 345 Alkyd resin Arochem 520 Alkyd resin Aroclor 5480 Aropla2 945 Chlorinated diphenyl Alkyd resin Aroplaz 930 Alkyd resin Aropene 700 Phenolic resin Bakelite XR-976 Phenolic resin Bakelite XR-13630 Phenolic resin . Beckasol 31 Alkyd resin Clorafin 70 Lewisol 24 Lewisol 33 Rezyl 869 Chlorinated paraffin Alkyd resin Alkyd resin Alkyd resin Uformite Urea-formaldehyde Resinous Produots and Chemioal Co. Resinous Produots and Chemical Co. Resinous Produots and Chemical Co. U. S. Industrial Chemi cals. Inc. U. S. Industrial Chemi cals, Inc. Monsanto Chemical Co. U. S. Industrial Chemi cals, Ino. U. 8. Industrial Chemi cals,Ino. U. S. Industrial Chemi cals, Inc. Bakelite Corp. Bakelite Corp. Reichhold Chemicals, Inc. Hercules Powder Co, John D. Lewis Co; John D, Lewis Co. Amerioan Cyanamid Co. Resinous Products and Chemical Co. Aroclor 1254 Clorafin 42 Flexol DOP Dow No. 0 Flexol 3GH Halowax 4001 Hercolyn Salol Santicizer B-16 Santicizer M-17 Rezyl 36-5 Plasticizers Chlorinated diphenyl Butyl-2-methyl-2-nitropro- pyl phthalate Camphor Chlorinated paraffin Dibutyl phthalate Dioctyl phthalate Diphenyl phthalate Monsanto Chemical Co, Commercial Solvents Corp. E. I. du Pont de Ne mours & Co.. Inc. Hercules Powder Co. U. S. Industrial Chemi cals, Inc. Carbide and Carbop Chemicals Corp, Eastman Kodak Co. Dow Chemioal Co. Triethylene glycol di-2-ethyl Carbide and Carbon butyrate Chemicals Corp. Chlorinated paraffin Union Carbide and Car bon Corp. Hydrogenated methyl abie- Hercules Powder Co. tate (Inc.) Hexachloroethane Phenyl salicylate Eastman Kodak Co. Butyl phtlydyl butyl glyoo- Monsanto Chemical Co* late Methyl phthalyl methyl gly- Monsanto Chemical Co. colate Tricresy! phosphate Eastman Kodak Co. Triphenyl phosphate Eastman Kodak Co. Phthalic alkyd soft resin American Cyanamid Co. Pigments Antimony oxide, asbestos fibers, calcium carbonate, magnesium ammonium phosphate, titanium di oxide, zino borate airspeed m.p.h. MEASURED BY HOT WIRE ANEMOMETER MEKER BURNER PLAME NORMALLY S" Exhaust fan Figure 1. Apparatus for Selecting Doped Fabric Coatings by Moving-Air Burning Test still-air burning test was inadequate because the coatings were self-extinguishing. A moving-air burning test was developed which permitted evaluation of the more fire-retardant coatings. Figure 1 is a sketch of the apparatus. A small laboratory centrif ugal exhaust fan draws air through a funnel-shaped entrance port, which is adapted to accommodate a doped fabric test panel. The doped fabric panel is attached to the bottom of the entrance funnel with the fire-retardant coating on the inside. A Meker burner with air supply adjusted to give a quiet blue flame 9 inches high is placed in such a position with respect to the fabric that the current of air causes the flame to lie on the coated surface. Wind velocity in the center of the entrance port is 6.5 miles per hour, measured with a hot-wire anemometer. The time intervals (1) between flame contact and ignition of coating and (2) between, flame contaot and oharring of the fabric as seen from the outside are measured with a stop watch. For a group of 119 tests of pairs of duplicate samples, the standard deviation of an individual measurement was 1.0 second. The panels used in these tests had two fire-retardant coats. Wind-Tunnel Burning Test. A steel wing section of Clark Y airfoil was suspended, as shown in Figure 2, at the outlet end of an open-type wind tunnel. A 12-inch-square opening was pro vided in the center of the span and chord of the lower surface to admit fabric panels. The angle between the plane of the test panel and the axis of the wind tunnel was 0. The wind tunnel was operated so as to provide an air stream of 70 miles per hour. Just in front of the wing was mounted a spray gun to which were supplied approximately 1.5 gallons per minute of 90-octane aviation gasoline and air at approximately 50 pounds per square inch pressure. The gun introduced a uniform spray of atomized gasoline into the air stream. At zero time the spray was ignited by a spark from a high-tension aircraft spark plug. This pro vided an envelope of burning gasoline which passed both over and under the airfoil, and ensured continuous contact of the test specimen with flame during the test. The gasoline flow was regulated by a solenoid-type electric valve. The course of the fire was observed by two men on the ground with stop watches and by a.motion picture camera aimed at a mirror so oriented in the interior of the wing section as to permit observation of the inside surface of the test panel. A large electric clock with a sweep second hand was plaoed in the field of view of the camera to provide an auxiliary time scale. The motion picture camera was operated at a speed of sixteen frames per second. One observer noted the time for the destruction of the outer coating, and the second observer noted the time of fabric failure. The destruction of the outer coating was evidenced by pieces of coating being blown away by the wind stream. Failure of the fabric was evidenced by two criteria: To the outside observer the disin- 1744 INDUSTRIAL AND ENGINEERING CHEMISTRY Vol. 41, No. 8 L 1 .X------------STAINLESS stee ________ ____________________________________________________________ WING SECTION THERMOCOUPLES------- ---------------------------------------TE3T FABRIC TEST PANEL DETAIL WINS TUNNEL OUTLET included panels coated with Navy-spocification cellulose nitrate dope and cellulose acetate butyrate dope, respectively. The number of coats of these two comparison systems was kept constant at six, equivalent to a coating weight of about 4.5 ounces per square yard. Al though the char time in the moving-air burning test increases with increase in the weight of cellulose acetate butyrate dope added, this in crease is not nearly so great per unit: weight of cellulose acetate butyrate dope, as it is per unit weight of fire-retardant lacquer. Most of the fire-retardant lacquers tested in the wind tunnel gave char times in the moving-air burning test of 10 seconds or greater for two coats (4 ounces per square yard) of fire-retardant, coating added to a six-coat (4.5 ounces per square yard) cellu lose acetate butyrate substrate. To achieve a char time of 10 seconds with the addition of cellulose acetate butyrate dope only, it is neces sary to add 15 ounces per square yard to a doped fabric already coated with 4.5 ounces of tegratio-.i of the fabric was sudden and complete. At the same time the illumination inside the airfoil suddenly in creased. In Figure 3 prints of successive frames of the motion picture record show the fabric destruction between the first and second frames. Temperatures were recorded at points immediately out side and inside the fabric by thermocouples and quick-act ing recording pyrometers. Fifteen experimental fireretardant coating systems were tested in the wind tunnel. In each system one, two, three, and four coats of fireretardant material were ap plied to a fabric previously doped with a standard Navy six-coat doping scheme, con sisting of cellulose acetate butyrate dope and applied in accordance with Navy Aero nautical Specification 3R-70o. The selection of the materials for this test was based on their performance in the labo ratory moving-air burning tests. Panels were prepared several months in advance, so that the effects of solvent retention were negligible. Table II gives the formula/tiu.is of these five-retardant coat'.. Panels without fire-retard ant coatings were also tested to provide a basis for com parison. This latter group Table II. Compositions and Properties of Fire-Retardant Coatings Tested in Wi.vd-Tcxivet. Fire Tests at Indianapolis Wind-Tunnel Test Coating No. 1 234 Natl. Bur. Standards Formula No. 120 121a 255 469 5 251 67 470 471 8 9 10 . 11 96 434 472 451 12 450 15 16 17 540 556 558 Nonvolatile Composition*7 Film base Vinylite VYHH 26 Paid on 125 20 26 26 Parlon R 20 26 Parlon X 26 27 30 30 30 30 26 Dow X-124 26 Resin 256-27 48 Amercoat 1138 40 Mathicson rubber 's 153b Polydiohlorostyrene 10 Plasticizer Rezyl 36-5 Aroclor 1254 Dibutvl phthalate Dioct.vl phthnlate 26 26 26 26 13 *4 '4 6.5 13 14 26 io 4 27 14 14 14 14 14 8 8 *8 10 16 Dow No. 6 *4 4 0.5 `e Hexachloroethane i2 Dinlvuyl phthalate 6'.5 '4 Tvicresyl phosphate 6.5 Clorafin 42 '4 '4 '4 '4 io 's '8 Pigment Antimony oxide Calcium carbonate 20 20 20 20 20 24 24 20 20 20 `>ri 20 24 24 20 20 20 20 20 20 20 20 20 20 Magnesium ammo 46 nium phosphate Zinc borate 40 Nonvolatile Composition* * (Solvext) Mthvlctlivl ketone 00 *0 00 100 100 100 too ion 100 C dlosolvo neotate 30 30 Butyl ftCTtato 10 20 10 46 20 Et'wl acetate 83 Di icotonn alcohol 50 Aro runic petroleum 60 'naphtha, Tvne. I T Vila- (Neville Co.) T duene 60 60 100 XvlenC 2 j 40 40 Union solvent No. 8 50 Properties Fnlvic d 'st'm/'tirm 8.2 0.4 7 o 8.8 10,2 10.0 8.0 7.8 10.3 ti n" (wind tunnel), SRC . !> Fabric char time (lab. 12 16 11 9 12 10 10 17 10 moving air), see. '> Aor>fl'-",<vt.ed weather P P G VG G G G P P ing c Panel No. in outdoor 655 656 674 621 exposure t'Jst 9.4 11.8 JO.8 10,0 7.0 8.5 12 10 10 11 oP c; P 701 Figures are in per cent of nonv*ilatiles and volatiles, respectively. The solutions were made up to contain 25% nonv^Htiles. 1 ^ , ,, .. ^ ., b Ti lies are for panels with four fire-retardant coats. Figurs 4 gives results of tests with one, two, and three fi e- -Tt \rdant cont*. Code: P poor, G * good, VG = very good. August 1949 INDUSTRIAL AND ENGINEERING CHEMISTRY 1745 Figure 3. Sample of Motion Picture Record of Wind-Tunnel Burning Tests of Fire-Retardant Coatings cellulose acetate butyrate dope per square yard. OTHER PROP ERTIES OF COAT ING SYSTEMS Not only is it desirable to pro vide the maximum in fire retardance in a coating, but it is also essential that the coating have a reasonably good life expect ancy. The require ments for good air plane fabric coat ings are severe. They must be tough enough in all kinds of weather to withstand impacts from pieces of run way gravel as well as normal flying stresses; they must not interfere with the normal fabric tautness; they must be easily applied, show good adhe sion to dope, and be smooth. ACCELERATED WEATHERING TESTS This accelerated weathering by Method 6021 of Federal Speci fication L-P-406a was first used as a guide in the selection of coating formulations. The specimen is exposed to cycles of condensed fog, alternated with ultraviolet radiation from a General Electric S-l sun lamp. This method has proved success ful for laboratory evaluation of the resistance of plastics to nat ural weathering. A standard exposure of 240 hours of this cyclic treatment was used as an initial screening test. A soft alkvd resin, Rezyl 36-5, was found to be particularly effective as a plasticizing resin and superior to the other resins tested in im parting resistance to weathering. Most of the coatings were therefore built around this resin. Test films approximately 10 mils thick were examined with and without pigments, both as free films and as coatings applied to doped fabrics. To be a candidate for outdoor exposure, the test film was required to show sufficient flexibility after the 240 hours of accelerated weathering to permit a 180 fold at room temperature without cracking. The fold was made by hand, with no attempt to con trol the radius of fold. OUTDOOR WEATHERING TESTS Fire-retardant coatings selected for outdoor exposure were applied as topcoats over panels of doped fabric. Tables III to VI show the compositions and properties of the coating systems tested by outdoor exposure. The solutions were made up to con tain 25% nonvolatiles. (All fabrics were coated with 10% by weight of a mixture of 3 parts of boric acid and 7 parts of borax be fore doping.) Two coats were applied, the total coating weight added being approximately 4 ounces per square yard. The test panels were exposed continuously in Washington on outdoor racks facing south at an angle of 45. Periodic tautness measurements and impact tests were made according to the procedures de scribed by Kline and Reinhart (3). Table III. N.B.S. Panel No. N.B.S, Formula No. Dope Pigmentation0 Vinyl and Vinylidene Chloride Fire-Retardant Coatings Tested in Washington 624 625 674 732 733 62 62 469 549 550 Al W Al Al Al 761 621 697 671 730 731 562 96 96 96 547 548 Al Al Al Al Al Al 760 561 Al Film base Vinylite VYHH Dow resin X-124 Plasticizer Rezyl 36-5 Diphenyl phthalate Dow No. 6 Clorafin 42 Dibutyl phthalate Pigment Calcium carbonate Antimony oxide Stabilizer Phenyl salicylate Urea Basic lead oarbonate 27.2 27.2 2.7 2.7 20,1 20.1 Nonvolatile Composition, Per Cent by Weight 27.2 27.2 2.7 2.7 20.1 20.1 26 26 'i 20 20 4 25.9 23.9 20 26.2 14 isio 14 14 26.2 3.9' 20 3.0 10 10 10 19.9 29,9 29.9 30 19.9 6.3 'o'.s 26.2 26.2 3.9 3.9 19.9 10.9 26.2 26.2 3.9 3.9 19.9 19.9 250 13.9 10 id' 10 20.9 6.3 25^9 13.9 10 io' 10 29.9 6.3 '26 14 10 io 10 30 Methyl ethyl ketone Butyl acetate Volatile Composition (Solvent), Per Cent by Weight 60 60 100 60 00 40 40 40 40 60 100 100 100 100 100 40 100 Fabric char time (moving-air test), sec. Burning characteristicsb Outdoor weathering Initial tautness, mils Tautness range, mils Time to first brittle failure, mo. Total exposure, mo. Condition of panel6 7 A, B, E 106 73/131 10 11 B. D 7 A, B, E 113 69/139 8 11 C, D 9 A, B, E 68 60/98 4 5 F Properties 9 12 A, B, E A. B, E 81 63/125 9 11 G, D 84 60/127 9 11 B, D 17 17 17 8 9 A, B, E A, B, E A, B, E A, B, E A, B, E 80 68/123 5 7 B, D 106 71/142 10 31 A, D 86 68/3^ 14 21 A, D 100 82/133 2 5 D 96 69/137 ' 'i7 A, D 93 61/137 ' 17 A, D 98 75/140 6 U A, O Code: Al -- aluminum. W white. b Code: A m afterglow, B = blistering, E self-extinguishing. c Code: A -- exposuro tost continuing, B -- poor adhesion of topcoat to dope substrate, C <* erosion of topcoat, D -- tautness poor in warm weather, F = experimental hot-application dope substrate, G poor adhesion of dope to fabric, <i S slack. 1746 INDUSTRIAL AND ENGINEERING CHEMISTRY Vol. 41, No. 8 o 'M 03 2 :; ;2S : -- cec t-. -- 03 c c ; ; ' *d c M o --4< ** QOfftO- .---oO*C* :g? : O O*! ** ,(30C c o' c ; =x 53 =O 5i ?3 00'03 ?o M I*.---N1 z; gocP 03 *1 ?3 03 *............. * "* - -- "* 03 Cl c e e c -CO 03 3 s z S toe z z p< c - Cl 03 T1 ................ C -cs _ 03 r>. h. -- h3. ^ 03 m . . 03 t- o : :g f** c . .Tf c 5 00 M r, f, ^ . f4. O'! -w *l cN 03 a ' ' ^ c -- 5j 53 ' 5 $2.0 i OOlft^ <N 0w0 -- 53 " ! ! . ** ; ; ; -- o w ........................... = .= . :: 03 b- b. M OC lit 03 :: g j w 2 '' ^ O& o% . ,9 >a< . -c$ i-1 1C &0cs C -r E :: > . . -t 00 11 Q Q 3 J < 00 0-3 03 b-In. 03 M 3C Lt LC . 3 ^ :S : :2 : . ,c _ o03 o t-N m2 Sc '5 C a (C 03 In. -- 0030 55 -- -r 0^3 M 03 .1 o Z|| !Sl (5.l 5Q. CO.&^- g< zza j c-<i.=2, ISoO *13 J121 - *3 saJ.** ifti&----r"*v* 0,3tt*ag--r'i s:g!^s!ui" I! ^3a iw3 3 S's O u ^ a e&*.S<<S3^asasiHa^y3o&iSS.gsq*o2-za s<-"s$ ssa ,o ' o MM Sjg SC l ** lift sjsf SSeS |N. 5S :2c ..So >`s , I II a tf *? v. '""' *** .i bx-i 2tiig3.E; II31 C II ` g1 ||s !|l i|. l.r. s |I E'w \ ae T3 li|!lI>ec 8a-g S S ga *cV^-oa* The tautness readings shown in the tables represent the deflection in thousandths of an inch produced at the center of a doped fabric test panel by a concentrated load of one pound. Therefore, low readings mean good tautness. A value of 100 is taken as the boundary between satisfactory and unsatisfactory t.aut- ness. The tendency to poor tautness which results from the nontautening fire-retardant coatings can possibly be neutralized in part by a hot ap plication dope, which gives a high degree of tautness, as the substrate. Thus, the tautness of the composite system might be made equivalent to that of fabric doped with standard cellulose acetate butyrate. However, the hot dope systems require further development to correct poor flexi bility before they will be satisfactory. Although the boric acid-borax treatment does act as an effective fire retardant for fabric, there was evidence that the adhesion of cellulose acetate butyrate dope to fabric was lower in panels containing boric acid and borax. Pew of the coatings tested in this program are as durable or as flexible as the cellulose acetate butyrate dope which is taken as a standard for comparison. The formulas which showed relatively early failure on impact (brittle failure)--i.e., in less than six months--are not recom mended for further consideration. However, the coating systems to be described should be considered for further development by interested lacquer manufacturers. VlNTL AND VlNYLIDENE CHLORIDE Resins Coatings (Table III). Half of the Vinylite VYHH coatings developed poor adhesion to dope upon exposure. Panel 625 had white pig ment in the substrate dope and showed evidence of erosion. Panel 674 was removed because of impact failure after short exposure. Stabilizers, such as urea and basic lead carbonate, were effective in pro longing the coating life as may be seen by comparing panel 761, with out stabilizer, with panels 732 and 733, containing stabilizers. The coatings using the vinylidene chloride-acrylonitrile copolymer (Dow resin X-124) as film-forming base show favorable properties, in general, both from the point of view of fire retardancy and of weathering. The superior performance of panels 730 and 731 in comparison with panel 760 demonstrates the effectiveness of urea and basic lead carbonate as stabilizers. The use of equal parts of calcium carbonate and antimony August 1948 INDUSTBIAL AND ENGINEERING CHEMISTRY 1747 oxide as pigment in formula 96 appears to be a factor in the marked superiority in fire re sistance as compared with coat ings from formulas 647 and 648, in which a larger proportion of antimony oxide was used. This variable needs further explora tion. Chlorinated Rubber Coatings (Table IV). The four chlorinated rubber coat ings which gave best results in the outdoor weathering tests are formulas 2,84,643, and 544. However, the erosion of for mulas 2 and 84, incident to the Table V. Chlorinated Neoprene Fire-Retardant Coatings Tested in Washington N.B.S. Panel No. N.B.S. Formula No. Dope Pigmentation0 628 100 A1 620 100 W 630 07 A1 631 672 97 536 W A1 673 252 A! 698 728 250 545 A1 A1 729 546 A! Film base Parlon R Amerooat 1138 Plastioizer Rezvl 36-5 Diphenyl phthalate Dibu*tyll iphthalate Tr..i.oreessyyl lp. hosphate Dow No. 6 Aroclor 1254 Flexol 3GH Pigment Calcium carbonate Magnesium am monium phosphate Antimony oxide Stabiliser Urea Basio lead carbonate Nonvolatile Composition. Per Cent by Weight 20,0 20.0 20.0 20.0 10.0 30.0 40.0^ 20,0 4.0 20.0 4.0 26.0 4.0 4.0 20.0 4.0 4.0 60.0 3! 0 13.0 7.5 7'5 4,0 4! 0 3! 0 4a, 0 40.0 20.0 20.0 9.0 20.0 20.0 9.0 30.0 25.8 15.0 7.5 7.5 13.9 ioIo loio 20.0 10.0 20.0 30.0 0.3 25.8 13.9 10.0 10.'0 10.0 30.0 o!s use of water-soluble magnesium ammonium phosphate in the pigment, may have been re sponsible for the failure to show brittleness simply because of the reduction of coating thick ness. Formulas 543 and 544 Metbyl ethyl ketone Toluene Trichloroethylene Diacetone alcohol Unionsolvent No. 8 Aromatic petroleum naphtha, Type I Xylene Butyl acetate VOLATIL k Composition (Solvent), Per Cent BY W RIGHT 34.0 33.0 33.0 34.0 33.0 33.0 34.0 33.0 33.0 34.0 33.0 33.0 50.0 50.0 60`.0 20.0 20.0 solo 50.0 100 100 had superior fire retardancy, as compared with formula 2, with respect both to self-extinguish ing and to afterglow. Another promising coating is formula 534, prepared by adding chlori nated rubber to an aluminumpigmented vinyl chloride ace tate lacquer (Amercoat 1138). Preliminary accelerated weathering tests of Amercoat Properties Fabric char time (mov 14 H, 23 23 ing-air teat), sec. Burning characteristics 6 A, B, E A, B, E A, B, E A, B, E 9 12 A, B, E A, B, E 10 12 B, E A. B, E Outdoor weathering Initial tautness, mils Tautness range, mils Time to first brittle failure, mo. Total exposure, mo. Condition of panel6 86 74/121 11 C, E 90 74/118 8 11 C 94 74/121 11 C, E 93 71 58/104 61/102 10 2 11 C, E 5 F 68 65/109 2 5 F 87 87 59/115 07/135 13 9 10 18 G A, E 92 64/133 18 A. E rt Code: A1 = aluminum, W * white. b Code: A afterglow, a - blistering, E - self-extinguishing c Code: A - exposure test continuing, C - er. o_ s_ ion o. _f top..c.oat,.. hot-application dope substrate, O * poor adhesion of dope to fabric. 4 Figment content included in film base figure. fine cracking of topcoat, F -- experimental 1138 over fabric doped with "" cellulose acetate butyrate were unfavorable. However, use of the mixture gave good results bilizers, show best resistance to weathering. Formula 97 (panels both in accelerated and natural weathering, except for a single, 630 and 631) appeared better in the laboratory burning tests. early brittle failure. Formula 470, almost like formula 97, was excellent in the wind- Chlorinated Neoprene Ooatings (Table V). Formulas tunnel fire tests (wind-tunnel test coating 6 of Table II and 545 and 546, containing urea and basic lead carbonate as Sta Figure 4). The use of equal parts of calcium carbonate and antimony oxide plus one of the stabilizers would probably yield a coating having the im s'6 AVERAGE TIME FOR FAILURE OF PANEL TREATED WITH CELLULOSE ACETATE SUTVRATC DOPE N-N AVERAGE TIME FOR FAILURE OF PANEL TREATED WITH CELLULOSE NITRATE DOPE proved weathering resistance of formulas 545 and 546 and the fire retardancy of formula 97. The chlorinated neoprene formulas gave the smoothest coatings of any of the materials tested. Chlorinated Polyisoprene Coatings (Table VI). Formulas 466 and 540 gave the best all-round performance in the outdoor weathering test, although both panels became slack on one occasion during a prolonged wet spell. The mixed base coating, using chlori z is (TV COATING SYSTEM (SEE TABLE H) nated polyisoprene (Parlon X) and Amercoat 1138, showed an early brittle failure but otherwise gave good results. TOTAL DESTRUCTION EFFECTIVE (E0T) DELAY TIME FAILURE IN OUTER COATING FIRST NOTED NO. OF FIRE RETARDANT COATINOS Figure 4. Burning Times of Doped Fabrics Covered with Fire-Retardant Coatings in Wind-Tunnel Test RESULTS OF WIND-TUNNEL TESTS Table II and Figure 4 present the results of the wind-tunnel burning teBts at Indianapolis. Figure 5 shows the appearance, after the tests, of panels which had been doped with cellulose nitrate and cellulose acetate butyrate, with and without fire-retardant coatings on the dope surfaces. It is noteworthy that the panel doped with cellulose nitrate and coated 1748 INDUSTRIAL AND ENGINEERING CHEMISTRY Vol. 41, No. 8 Table VI. N.B.9. Panel No, N.B.S. Formula No. Dope Pigmentation b Chlorinated Polyisqprene and Miscellaneous Fire-Hetardant Coatings Tested in Washington 670 690 700 701 724 725 720 535 466 467 540 451 451 a A1 A1 A1 A1 W G w '721 708 709 Pyropex W W \V 722 723 Sky lac BB Film base Parlon X Amercoat 1138 Plasticizer Rezyl 36-5 Dibutyl ph.thala.te Aroclor 1254 Flexol DOP Clorafin 42 Pigment Zinc borate Calcium carbonate Antimony oxide 26.0 40.0/ 14.0 10.0 10.0 Nonvolatile Composition, Per Cent py Weight 30.0 30,0 30.0 30.0 30,0 Standard Navy system 14.0 16.0 14.0 is.'o 14.0 ie.'o 14.0 8.0 2o!o 20.0 2o!o 20.0 2o!o 20.0 40.0 14.0 8.0 *8.'0 40.0 Pyropex system Skylac system Methyl ethyl ketone Tollae Xylene 60 .'0 40.0 Volatile Composition (Solvent), Per Cent by Wright 100 100 100 100 100 Fabric char time (moving-air test), sec. Burning characteristics0 .. . Outdoor weathering Initial tautness, mils 113 Tautness range, mils 74/130 Time to first brittle failure, mo. 4 Total exposure, mo. 18 Condition of panel* D 97 84 06/Sd 66/108 14 2 21 A .1.5. Properties 12 A,,B,E 9 B, E 93 109 70/S4 67/141 13 4 21 11 A B, D no 57/143 9 11 B. D 5o 74 71/103 7 11 5 81 70/100 9 11 G o 59 53/S<* 'l9 A 5 ov 59 53/Srf 19 A 80 65/110 10 19 A, D 85 60/120 ... 19 A ,D No fire-retardant coating. 6 Code: A1 aluminum, B - blue-gray, G -- gray, W = white. e Code: A -- afterglow, B -- blistering, E = self-extinguishing, d 8 sa slack. < Code: A exposure test continuing, B poor adhesion of topcoat, D tautness poor in warm weather, G " poor adhesion of dope to fabric. / Pigment oontent included in film base figure, 9 No borio acid-borax mixture on fabric. with a fire-retardant coating was destroyed more quickly and extensively than the panel coated with cellulose acetate butyrate dope only, Coating failure in this investigation is taken as the time required for the coating to be destroyed to such an extent that it is partially swept away by the windstream. Dur ing this interval the temperature on the outside of the panel rises considerably more rapidly than that on the in side. When the coating is completely swept away, the Inside temperature rises sharply. The elapsed time until the temperature rises on the inside is a partial measure of the fire-retardant effectiveness of the coating. The coat ing, in blistering, apparently acts as a blanket and flame deflector. The sharp temperature rise occurs when this blanket is removed so that the doped surface is again in contact with the flame. Figure 6 shows the time-tem perature relations. The times shown by the bar graphs In Figure 4 for total destruction, as evidenced by fire break-through, are taken from the motion picture records; these records are considered to be somewhat more reliable than the data either from the manually operated stop watch, or from, the time-temperature charts. The values of time given in Figure 4 and Table II are the original data for single panels. In general, increased fire protection was given by in creasing the number of fire-retardant coats, with the ex ception of coating system 2 for which no explanation is apparent. The zone of incident flame from the burning gasoline was sharply confined to half the panel area. This provided a criterion of extent of propagation of the fire to areas not in contact with the original flame. The burnt areas of the fire-resistant panels were confined to those regions immediately in contact with the flame. This was true also of the panels coated with cellulose acetate butyrate dope only, but not of the panels coated with cellulose nitrate dope. In some instances the cellulose nitrate panels were completely consumed. B6 II D Figure 5. Appearance of Panels after Simulated Power-Plant Fire Tests In Wind Tunnel N 3. Cellulose nitrate dope; destruction, time, 2 seconds N 5. Cellulose nitrate dope plus four coats of fire-retardant coating 11 (Table II) | destruction time, 4 seconds B 6. Cellulose aoetate butyrate dope; destruction time, 6 seconds 11 D. Cellulose acetate butyrate dope plus four coats of fire-retardant coating 11 (Table II); destruction time, 12 seconds August 1949 INDUSTRIAL AND ENGINEERING CHEMISTRY 1749 outer coating is designed to have greater mechanical strength after burning, so that it will not be so easily blown away by the wind stream. There appear to be no system atic differences in the results obtained with the several plasticizers investigated. Afterglow, although a significant property in the still-air burning tests because of reigni tion, was not important in either the labo ratory moving-air or wind-tunnel burning test. However, in the wind-tunnel tests large sections of the aluminized coating on a panel, doped with cellulose acetate butyrate containing aluminum pigment, were ignited and blown away as a shower of burning particles downstream. This could conceiv ably constitute a reignition hazard. Since coating systems examined in the wind-tunnel tests were selected from those showing up best in the laboratory tests, the differences in fire-retardant effectiveness among the systems is relatively small. Figure 6. Typical Time-Temperature Record of a Test Panel during WindTunnel Burning Test However, there is no doubt that these coat ings are substantially superior in fire retardancy to a cellulose acetate butyrate system alone. Their superiority over cellu Under the conditions of fire, the cellulose acetate butyrate melted lose nitrate systems is marked. before it ignited. The coating melted also with chlorinated neoprene systems 5 and 6. An almost universal property of the fire-retardant systems was the development of blisters by the generation of gases incident to the pyrolysis of the fire-retardant materials. The only exception was system 11 which did not exhibit the blisters characteristic of the other systems but seemed to be one big bubble; zinc borate ACKNOWLEDGMENT The authors wish to acknowledge the assistance given by P, Dier and J. W. McElwain of the National Bureau of Standards and the staff of the Fire Test Unit of the Civil Aeronautics Ad ministration Experimental Station in supplying some of the data reported here. was the pigment used in this system, whieh showed the best fireretardant properties of any tested. From the point of view of keeping the coating in place for as long a period as possible, the formation of small blisters is better than the formation of one large bubble, which might easily be blown loose by the wind stream. LITERATURE CITED (1) Kline, G. M., J. Research Natl. Bur. Standards, 14, 575 (1935). (2) Ramsbottom, J. E., "Fire Proofing of Fabrics," p. 24, London, His Majesty's Stationery Office, 1947. (3) Reinhart, F. W., and Kline, G. M., Ind. Eng. Chism., 32, 185 (1940). During the progress of the tests, it became apparent that an important limiting factor in protection was the brittleness of the burnt outer coating. Some improvement can be expected if the Received June 16, 1948, Presented before tbe Division of Paint, Varnish, and Plasties Chemistry at the 113th Meeting of the American Chemical Society, Chicago, 111, Correction Factor for Latent Heat Method for Obtaining Haggenmacher Correction Factor without Use of Critical Values R. R. DREISBACH The Dow Chemical Company, Midland, Mich. A method has been developed for calculating tbe Hag- genmacher correction factor for latent heats by means of a modified Antoine equation for vapor densities and the law of rectilinear diameters for liquid densities. heat of vaporization and external work of vaporization can be accurately calculated. The three Haggenmacher equations pertinent to this article are: This method eliminates the necessity of having critical values, and uses instead the boiling point and liquid V, - V, RT\T7~ TP mp '1 pepa (1) density. HAGGENMACHER (S, 4, s, 6) has developed a number of formulas from an equation of state for saturated liquids and the Ciapeyron-Clausius equation by means of which the. latent AH RT8 dp / pap MP df\ 1 " pjpi -AA P(.Va - Vi) RT J M y1 pap pap* (2) O)