Document n9NdNpOjOoG1R8V9w2N1oyd0R

t Abstract .97 blllcy of Vinyl Foam 1A lua Mfc# /no MM 0 Recent studio on the flawablllty of cosmerclal carpeting are discussed within the framework of sauke t-neratlon, toxicity of combustion products and fleas retardancy. A number of different formulations Including PVC have been evalueted. Although there la presently considerable concern about smoke and toxic gaa generation during a tire, the major emphasis must be toward new coocepts In flame retardation. In addition new mechoda of aeaaurlng and evaluating toxic gases are needed. The tremendous need for new flame retardant suterlals which also afford smoke and toxic gas retardation presents a challenge to the vinyl Industry that can not be Ignored. Stricter federal legislation In the area of flawablllty, toxic gas and smoke generation makes this challenge all the more urgent. To meet this challenge, a program dealing with the various aspects of flams retardation la now being formulated. The following exerpts were extracted from an article published In the Plain Dealer (2/9/70) with regards to the Marietta Nursing Home fire. "The carpeting and backing throughout the facility were burned away; It appeared that heavy, dense acrid smoke was generated from the smoldering or burning of this carpeting, creating the deadly atmosphere." "The fire spread very rapidly, and it la the feeling of the Investigators that the carpeting aod rubber becking... contributed to the spread of the fire aod was the cause of the heavy, dense, black smoke." This fire has Initiated a aeries of Senate hearings Into the adequacy of the Flamneble Fabrics Act of 1967. The carpeting In the Marietta Homs (reportedly of Hylon/SBR composition) amt the minimum requirements under the FFA. The fact that all carpeting material will undergo Increased scrutiny with respect to flasmmblllty, toxic comubation products and smoke generation has prompted a number of questions about \'nyl carpet backing. How does it compare with SBt and other amterlals with reapect to the above considerations? 0S77001 BFG08460 -2- la vinyl carpeting potentially taora dangerous than other Materials since It releases a stolchloaetrlc aaK>unt of HC1 during combustion? Is there toxicological synergism with HC1 and carbon monoxide (as with HCN and CO)? How does vinyl carpet backing compare with other materials with respect to smoke generation? What effect does heat generation, during a fire, have on toxic gas and smoke generation? The abjective of this paper la to answer sow of these questions within the framework of recent government sponsored studies and to show the need for further studies In the area of toxicity, flaw retardaocy and smoke generation. This need Is due not only to the fact that there will be future federel legislation In thla area but also due to the fact that such of the work In the past (especially with respect to toxicity) Is totally inadequate for todayb needs. Primary Causes of Death la a Fire At a recant symposium on the flamawblllty of synthetic materials,^) Prof. I. KInhorn, symposium chairman, related that In his discussions with people la the fire prevention business, the causes of death In a fire (la order of decrea a lag Importance) are: 1. aaphlxlatlon due to oxygen depletion 2. effects of superheated air 3. Impairment of visibility by smoke 4. toxic effect of gases 5. hast from the fire and 6. varying degrees of 1*3 Thus In most fires asphlx1stloo due to oxygen depletion (or dense smoke) Is Che major cause of death. BFG08461 0977002 3 This la not to suggest that sank* (alteration and toxic gas evolution play a alnor role in this reapect. Indeed, as onre Material In the home la rendered flame retardant thereby reducing the possibility of torch fires, the combination of amoke and toxic gases becomes of greater Importance with respect to fire deaths. Jaaes Gasklll, In a paper on the analysis of smoke developed during a fire (work carried out at Laurence Radiation Laboratory) has stated^3) that "except for clothing fires, more than half the deatha due to fire are caused by smoke rather than by heat or burning." (Geskell's definition of smoke Includes solid particles, toxic gases and suspended liquids) Although considerable attention Is being given to smoke sod toxic gas generation these days, primary concern and basic research oust deal with new concepts in flame retardation. Although the deaths at Marietta may have been due to a combination of smoke and toxic gases, the ultimate cause of death was due to the flammability of the carpeting. Toxicity of Combustion and Degradation Products Attempting to rate swat materials with respect to the toxicity of their costoistlon products is onre or less academic. When any material, ranging from wood to Teflon is burned, the combuatloo gases are toxic. In a recent study(*) at the hatIona1 Bureau of Standards, 141 materials desisted for aircraft Interiors worm evaluated with regard to smoke generation and toxicity of cambustion products. The concentration of 00, HC1 and othar gases ware measured la the smoke chamber sod the reported values are the maximum concentrations observed. Table I contains a summary of these values for a number of materials used la carpeting formulations. Cro*a(2) has reported tba values In ppm of air. However, this author feels chat chase values ar* misleading due to the fact that different weight samples (same dimensions) ware used la the smoke test. Gross has give* the weight of the samples consumed so tha values in Table I are reported la terms of ppm oar g BFG08462 0977003 pic burned. 4 la Chla way, all aacarlala can be Judged on the sene basis l.e., welghc... TABLE I : Concentration (ppm]/g material of HCl, CO and HCN From Flaming (F) and smoldering (S) Rug Materials Composition Pile: Modacryllc Backing: Polyester fiber Pad: Polyester urethane Pile: poly (propylene-butylene) Center: Ceiluloslc Backing: Polyethylene Wool Pile: Wool Back: Polyester Pad: Urethane foam ModacrylIc/Acryllc Polyester urethane foam (carpet underlay) Modacryllc/Acryllc M Plasticised PVC/FVA (Flooring, Klgid) Polyamide CO F 41.0 20.0 20.2 21.1 24.3 88.5 31.4 36.1 43.6 81.5 (ppm)/g sample burned HC1 HCN SF SF S 10.2 7.4 2.5 1.6 2.5 114.0 9.6 19.8 20.7 26.5 13.2 20.5 13.7 93.0 0 0 0 2.6 0 2.5 0 49.2 0 0 <1 <1 0 1.6 2.1 0 <1 4.0 <1 <1 4.3 0 <1 0 0 2.1 4.4 0 2.0 13.3 27.3 0 <1 5.6 <1 4.7 U iO Z IG O Z Although these date appear to show Che relative toxicity of the various materials Hated, chla writer feels that the study la of limited use. For example, since one of the primary aspects of moat fires la oxygen depletion, the effect of this parameter oo the concentration of carbon monoxide is of primary Importance. Yet all of the analyses listed In Table I were carried out on yeterns containing a normal level of oxygen. It cannot be asauend that the ana relative order of CO concentrations appearing In Table l will also occur BFG08463 under condition* of 1 talced oxygen concentration. 5 For exaaple, the difference* In the level* of CO generated under flaalng and aaolderlng condition* vary conalderably. The PVC plaaclaol 1* an Intereating exaaple of the ltalced uae of che data. The high level of hydrogen chloride In the coabuatlon gaaea Indicate* that thla particular foraulatlon for carpet backing la extreaely dangeroua. (Thla la baaed on the fact that the Maxlana Allowable Concentration of CO la 50 ppa, that for HC1 la 5 ppa). Two additional piece* of lnforaatlon invalidate thla aaauaptloa. Ftrat, Grosa^) ha* shown that the concentration of HC1 rapidly decreaaea aa a function of tlae while the concentration of CO renalna conatant. Work at the U. of Michigan^) dealing with the toxicity of Che coabuatlon gaaea froa PVC ha a ahown that when theae gaaea, generated under a number of different condition* are directed Into anlaal testing cages death In each case la related to the carboxyhcaoglobln concentration In the blood. There waa not evidence of aaeaive tissue daaage due to hydrogen chloride nor was there any apparent toxicological synerglaa between HCl and CO. This correlation shows that analytical coabuatlon data In the absence of anlaal testing studies is of little value in deteraining the toxicity of the coabuatlon products froa any aaterlal. In another sense, PVC la superior to the other aaterlals with respect to the evolution of coabuatlon gaaea. PVC offers an Internal warning devise when subjected to burning conditions. The sense of saall la extreaely sensitive to HCl but totally Insensitive to carbon aonoxlds. 20977005 BFG08464 Smoke Cent radon Purina Coabuitlon One of che Important factor* In any fire la che race of smoke generation. Thla aapect of flaoaaablllty la presently undergoing close scrutiny with respect to Its effects during a fire. Currently work Is In progress In many Industrial laboratories In generating an ASTM smoke test (B.F.C. Is presently cooperating In this effort). There was a considerable amount of discussion at the previously mentioned symposium on flammability with regard to smoke generation. The general concensus at spimlsa la that most synthetic materials are bad In thla respect. Indeed, the correlation was noted that the more flame retardant a material la rendered, the more It will generate smoke when subjected to flaming conditions. This Is one reason why PVC does generate a considerable asaaunt of smoke when burnt. Table II contains a sumery of sonka data obtained for the rug formulations Hated In Table I. Two different values are listed: Ob and D^i. The first value refers to the naxlrum smoke density recorded for the particular material while Ds.16 refers to the time (In minutes) to reach a point of 841 light obscuration; l.e., 84Z of the available light la obscurred by the smoke. This latter value Is of sere Importance In that It reflects the rate at which Hsmoke blindness" takes place la a fire. The table shows chat with respect to assoks generation (rate and degree) PVC Is among the worse under flaming conditions. Under noa-flamlng or smoldering conditions, PVC Is about the sans as tbs other materials. BFG08465 TABLE II : Smoke Generation Proa Typical Carpeting Formulations Composition Pile: Hodacryllc Backing: Polyeater fiber Pad: Polyester Urethane Pile: Poly(propylene-butylene) Center: Cellulosic Backing: Polyethylene Wool Pile: Wool Becking: Polyester Pad: Urethane foaa Hodacrylic/AcryLie Polyester urethane foea (carpet underlay) Hodacrylic/Acryllc MH Plaatlclsed PVC/PVA (Flooring Bigid) Polyaalde Q* P s1 439 375 96 418 123 87 129 206 350 312 SB 77 295 254 355 327 342 169 SI 65 S-16 1P s\ 0.8 1.2 1.8 4.0 1.6 2.1 2.2 2.0 1.2 2.1 1.9 0.6 0.6 1.3 0.7 2.0 0.6 1.2 3.2 4.8 4004.Z-60Z BFG08466 8 The work of Cross hss bssn extended In s aore rtesoc study by Gasklll.^) In this work, the problca of saokc generation Is desit with exclusively. Cesklll hes decerained the "saoke obscuration Indices" (SOI) for a manber of natural and synthetic Mterlals. The SOI value Is of specie Interest In that It Is a function of the aaxlaua saoke density, the rate of saoke generation and the tlae to 841 light obscuration.* Table II-A contains a suaary of S.O.I. values for a nuaber of different aaterlals. It Is again obvious that PVC Is aaoog the worst (neoprene and ethylene-propylene are so bad that SOI values can not be calculated.'). The data also show the saoke difference when a aaterlal la rendered flaae retardant. For exaaple, flaae retardant acrylic has a SOI of 360 while non-flaae retardant acrylic has a value of 10. Tks sasa order occurs under both flawing and saolderlng conditions. A slallar situation also appears In fhm polycarbonate series. Another correlation to ha noted la that for aost of the plastic aaterlals In Table II-A, aore saoke la generated under flawing conditions than under saolderlng conditions. Eased on this work, Gaskill has noted the following conclusions: 1) *Voods sad aost polyaerlc aaterlals pyrolltlcally degrade, yielding saokes that are dense to wary dense. Ventilation will tend to clear the saoke but In aost cases It will not reduce the intensity to the point of satisfactory visibility. 2) Hoods and those polyaerlc aaterlals that burn cleanly will under the Influence of heat and flans yield saokes of soaswhat less density. With fire retardsnee Incorporated into the polyasr....heat and flaaa generally produce very dense saoke and they do thla rapidly. 3) Urethane fossa under flawing or non-flawing exposures generally yield dense saokes and with few exceptions obscuration occurs la a fraction of a alnuta.......... *The higher the nuaber, the greater the saoke density. `T / BFG08467 TABLE II-A : Snoke Obacuraclon Indlcaa For Varloua Matarlala - 9- Planing Hardwood Red Oak White Oak Acrylic (fire retardant) Acrylic Polyethylene PTC (rigid) Polyetyrene Nylon Neoprene (aelf-extlngulahlng) Ethylene-Propylene Silicone rubber Polycarbonate (eelf-extlngulahlng) Polycarbonate (non-aelf-extlngulahlng) Urechene rubber (TDI-eatar) S 1 4 360 10 20 2400 900 1 a 105 a 60 so a - valua coo high co ba calculatad Snolderlng 25 35 80 25 3 65 70 21 1 J. a 90 0 0 5 4) Tha Caw data we have accuaulatad co data on Che apeclflc coaponenta of tanka auggaac only thee, trtkere caapounda auch aa HOI aod CO are preaent, aanka la likely co ba aora Coxlc Chan If cbe cyanide ware abaanC. Work dona aC Duka Dnlveralty lndlcacaa a aynarglatlc eosologlcal efface of chaaa cwo coapounda In coablnaClon. One alghc auapact tha praaanca of 1C1 contained wlch 00 would aoc ba beneficial." Hi--111 11TT Tf Carnet laa FornulaClone Tha above conaldaraelooa (anoka, generation of tonic gaaaa) are laporeanc with reapec C co flanablllty but of print laporcaoca la the lecemal flaaa racardancy of Che notarial. Tha FM of 1967 daala wlch chle paraaacar exclualvely and any raviaIona of Che ace will aclll rela gate the flaaa racardancy of any aacarlal aa cop priority. Croee In hla atudy of cha carpet aacarlala aeaaurad cha candaocy of each Co apraad flaaa. Thaaa valuaa are reported In eerae of total heac ralaaaad and cha aexlaua race of heat raleaaad (Btu /alo.). Table 111 concalna a auaary of choaa valuta for cha carpet acerlale 11acad tn Tablea 1 and IX. BFG08468 N 600ZZS0 - 10 - Th* rat* of heat ralaaaa data ar th* aoat significant In that this value la a Measure of th* flaac spread of the Material. Noticeably of those tested, PVC has the second lowest rate of all the Materials tested and this of course la due to the Inherent flaae retardancy of PVC. Of the three aeasureaents carried out by Cross, the heat release data la aoat significant. It la on this basis that these Materials are Judged under the Placable Fabrics Act of 1967. Although th* specific test specified by the FPA la not th* flane spread test used by Gross, the end result of the test la essentially the sea*. Conclusion* Th* data presented In this paper show that each of the Materials lla cad In Tables I-lII differ with respect to the toxicity of thalr eoabustlon products, the aaount of evoke generated during eoabustlon and inherent flaaa retardancy. The eoabustlon data la of Halted us* with respect to tonicity due to th* lack of eniaal testing studies. An even aora restricting draw back is that th* study is valid for only one set of conditions...conditions which do not actually reflect a real fir* situation. It has been stated that th* ultiaat* answer to th* problaae of saoka and tonic gee generation lies In the developeant of new concepts in flans retardation. Currently, Goodrich la involved to sons extent in this area. A prograa has been proposed at th* Research Center dealing with the aechanlaa of flaae retardation. Zn addition, 1FG la cosponsoring a study at Stanford Research Institute Involving th* aachanlaas of burning and flaaa retardation la polyethylene, PVC and oxygenated polyaara. Llalced attention will be given to sank* and toxic gas pane rat ion. In addition, Halted work has been carried out at the Oavelopnant Center In two areas: 1) the relationship between a polyaara' pyrolysis products and 203770X0 flanaeblllty, and W -~r BFG08469 TABU 111 : Heat Baleaae and Rata of Heat Release Prom Tvolcal Carpet Material. Pile: Modacryllc Backlog: Polyaatar fiber Pad: Polyaatar urethane Pile: Poly(propylene-butylene) Canter: Celluloelc Backing: Polyethylene Wool Pile: Wool Backing: Polyaatar Pad: Urethane foam Nodacryllc/Acrylic Polyaatar Orethana foam (carpet underlay) Nodacryllc/Acryllc MH Plaatlclaed PVC/PVA (Flooring Blgld) Polyamide Total Heat Release (Btu) 1010 560 not given 940 not given 140 510 800 420 0 Maximum Rate Btu/nln. 480 330 not given 140 240 140 190 330 60 0 The threat of future federal legislation la tha area of fiaua retardancy, smoke aad toxic gas generation ant bo cooatruad at a challenge to cha vinyl laduatry. Additional studlaa complimenting current work at Brecksvllle and f.ft.I. are urgently naadad now If B.P.Goodrich la to uaat tha naada of tomorrow. A raaaarch program daallng with thla overall problan la now being formulated at tha Development Canter. tT l)4 6 0 Z 29765 BFG08470 BIBLIOGRAPHY 1. Polyaer Conference Series, June 16-20, 1969, University of Detroit. 2. Snoke and Gases Produced by Burning Aircraft Interior Materials, D. Gross, et al., N.B.S., AD675513, June 1968. 3. Presented at Polyner Conference Series, June 16-20, 1969, University of Detroit. 4. Cornish, H.H., Arbar, E.L., Arch Environ Health. Vol. 19, July 1969. N BFG08471 2097701 Distribution: Develop--nt Center C. A. Clark L.B. Crider D. P. Knechtgea R.M. Kreager C.H. Lufter R.J. Heyer C. V. Purka N.H. Sherwood D. W. Ward C. L. Wheelock R.J. Wolf Technical Council Cleveland D. H. Hall L. Pilar D.L. Kant H.L. Wuerth Breckaytlle H.L. Dannla J.M. Poulk D.B. Ley Marietta C. Server t. ManefieId A-- rlPQl D. Wright Soon-- Product (Shelton. Coon.) D. Dobey 20977013 BFG08472 I To Subject B. F. Goodrich Chemical Company IiiltiCorr.i^oiiuuicc From March 9, 1970 M.M. O'Hara Correct Lon The report you have received recently entitled "Flarr.-.vibil ity of Vinyl Foam Backing" was written by M.M. O'Mar*. 20977014 BFG08473 29765  rt.iHM'l OATS___ ra INMXER. pnojyoePT. SHEET t wl ________ TITLE AIt - L-2 3* ZtA/YL F.QArt .AAAKf*6- AUTMOR(S) A`MJA.4 M.X. ABSTRACT ____________________ 4~ ACCItMOM NO, u - M i! TfRMS S-M P'MA^A, fl./l. 2A-. i .Z4J.T ftUS.< MMl H7JL AfJc.____ ku>t (F&ittJ if JSL IT POL 11 OVER NO TIMI BFG08474