A New Ablative Elastomeric Insulation

7/28/2019 A New Ablative Elastomeric Insulation

1/6

233

A NEW ABLATIVE ELASTOMERIC INSULATION

J R PatersonUniversity of Manchester Institute of Science & Technology (UMIST), United Kingdom

INTRODUCTIONA new ablative elastomeric insulation has been inventedprimarily for insulating cables so that they can functionin a nuclear environment, Paterson (1). Such cableswould be subjected through their lifetimes to nuclearirradiation aswell as thermal ageing and superimposedon th i s background may be higher levels of nuclearirradiation which would occur in an accident situation.In a loss of coolant accident, such as would occur in apressure water reactor, fire will follow and in additionthe insulation will be subjected to cycles of highpressure steam at high temperatures. It is essential thatthe cables continue to supply power and servecommunication circuits during the emergency and forsome time atlerwards. This means that, asdistinct fromlow fire hazard cables, the new insulation must continueto function under actual fire conditions. It has beendesigned to function at lo00 to 1100C when subjectedto an extemal heat flux density of 50 to 200 kWlm. Ithas virtually no toxic or corrosive gases.Fire is the common hazard. It can occur in a numberof places, such as operating rooms, engine and controlrooms or on oil production facilities. As the researchproceeded, it was realised that because of its resistanceto fire, the new elastomers would also be used veryeffectively in protecting structural members, walls andbulkheads from collapse and thus preventing personnelfrom being trapped in addition to protecting them frominjury due to inhaling toxic or corrosive fumes. It canalso withstand shock, vibration and resistance to impactso that it also offers protection from falling objects.The paper will concentrateon he effects of fire and willbriefly describe some of the items of equipment thathave been constructed. The development of the ablativeelastomer makes possible the design and manufacture ofbushings and cables which will allow electrical suppliesto pass through a compartment which is on tire to anadjacent compartment.DESCRIPTION OF ELASTOMERAn addition reaction polydimethylsiloxanewas used asthe base elastomer with a platinumcatalyst and ethyleniccross-links. This system was selected because it wasfound to have the best thermal and hydrolytic stabilityand reversion resistance. The elastomer has beenchecked by means of gamma spectrometry to eliminate

as far as possible all elements that could be convertedinto radionuclides. It contains additional fillers ofsilicon carbide and red iron oxide. Silicon carbide wasusedbecause it has a very low neutron absorption cross-section and the red iron oxide because it will react withthe hydrocarbons given off during buming to form asilicate layer, possibly after the reaction:-

with02F,0+C,H,,Si,01 - 2F,Si0,+4COl+6H10The red iron oxide absorbs a large amount of heat.Other fillers can be used, such as aluminosilicatemicrospheres or pyrophyllite rock, particularly forpurely tire protection. The elastomer can be used aloneor made up as a composite with amorphous silica glasscloth or it can be coated onto glass cloth and used forfire curtains.FIRE HAZARDSAs a result of a fire, failure of a structural elementextemal to a building may occur, such as a supportingleg for an offshore petroleum platform or a fire mightoccur in a compartment. Drysdale (2) has pointed outthat there are three stages in the growth of a fire in acompartment:1. The growth or pre-flashover stage. In this stage,the average temperature is low and the fire islocalised in the vicinity of its origin.2. The fully developed or post-flashover fire, duringwhich all combustible items in the compartment areinvolved and flames appear to fill the wholecompartment. It is during this stage that structuraldamage can occur and flame may spread to othercompartments, even to other adjacent buildings.There is a threat to occupants remaining in otherparts of the building.3. The decay period classified as the stage where the

average temperature has fallen to 80% of its peakvalue.

Electrical Safety in Hazardous Environ ments, 1S21 Apri l 1994, Conference Publication No. 390,@IEE, 1994


2/6

234

If a person has not been evacuated from a compartmentbefore flashoveroccurs,he is unlikely to survive. Timemust be allowed for evacuation from ignition to whenthe fire renders impossible the occupation of thecompartment. Marchant (3) has classified this time asbeing made up of the time for a series of events tooccur as follows:-t + re + t" 5 '*, (2)

where is the elapsed time from ignition o detection ofthe fire; f is the delay between detection of the fire andthe beginning of the escape activity; t is the time tomove to a position of relative safety; 4 is the time fromignition to the fire producing untenable conditions.The first step was to study changes that take place in thematerial during intense. fire conditions and then to applythe knowledge to help protect buildings and personnel.Thermogravimetric analysis, differential thermalanalysis, dielectric measurements, infra-red gas analysisand surface analysis using static fast atom bombardmentsecondary ion mass spectrometry have all been used tostudy the material. At 200C the ablative elastomer willgive off some vapours and behueen 300 and 500Cfurther breakdown will occur with the production of lowmolecular weight siloxanesgiven off as vapours. Above500C complete degradation of the molecule will occurwith burning and oxidation. Water vapours, carbondioxide and some carbon monoxide will be produced;and silicon dioxide will be formed, some of which willremain behind to form a protective ash. The action ofthe red iron oxide in helping to develop silicate layershas already been pointed out. Changes in the thermalconductivity will take place as well as changes in thespecific heat and density. The thermal conductivity willchange from 0.0632 to 0.4847 W/m"K; the specific heatchanges from I.Ok.J/kg"K at 60K to 4.63kJkg"K; andthe density changes from 1600 kg/m3 to 830kg/m3.Changes will occur in the thermal diffusivity andalso inthe emissivityso that some of the incident energy is re-radiated.The rate of surface flame spread will determine howrapidly a fire will develop. The first step is to controlthe growth of flame spread. Flame will spread from thepoint of ignition and the leading edge of the flame willprovide an advancing source of heat which will raise thetemperature at a local point to the ignition temperature.This is illustrated by considering the flame on thesurface as a rectangle. The situation is then analogousto two rectangles at right-angles to each other and theradiant flux density at a distant point can be calculatedfrom:

(3)

where {is a "viewing factor" calculated from viewingfactor algebra; Qc is the rate of heat release in kW; P isthe flame height in metres; and D is the diameter of thefuel bed(base of the flame rectangle).Th e flame height measured in the cone calorimeter andthe thermal imaging device were 125" and 32mm,respectively, Paterson (1). The cone calorimeters andthe thermal imaging device have been described byPaterson, Paul and Warren (4). Even at the higher limitof the flame height, the flame is unlikely to reach andtravel along the ceiling of a compartment so that therewill be no extemal radiation from the heated ceiling orupper walls to raise the temperature of surroundingareas to the fire point. With an incident extemal radiantheat flux density of 75 kW/mz and a peak heat releaserate of 137.7 KW/m2 and calculated "viewing factors"of 0.185 and 0.007, the radiant flux density from theflame at 150mm and 300mm along the surface are7.95kW/m2 and 0 0 3 KW/m2, respectively. Theminimum flux density for ignition is 24kW/m2,so thatignition at the surface will not occur at these points.While the surface is being heated, heat will penetrateinto the material to a depth given by Jut where o s thethermal diffusivity (k/pc where k is the thermal conduct-ivity in W/m"K; p is the density in kgln?; and c is thespecific heat in kJ/kg"K); and t is the time. It can beshown that the velocity of flame spread is inverselyproportional to the thermal inertia (kpc). For a solid, kis roughly proportional to the density of the solid and asthe density of the ablative elastomers is high, it will notspread flame rapidly. It has also been pointed out thatchanges in the thermal diffusivity will occur so that thepassage of heat through the insulation will be sloweddown and the red iron oxide will absorb energy and willcombine with hydrocarbons which havebeen establishedto be given off by infra-red as analysis, to form asilicate-like structure. This structure will prevent theescape of volatiles so that again the flame travel will belimited since the fuel supply to the surface will berestricted.From the above it will be seen that the ablativeelastomer is stable up to 200C and that it will increasethe time to flashover. Hirschler and Shakir (5 ) considerthat the ratio of time to ignit iodpeak rate of heat releaseis proportional to the time to flashover and m ytherefore be the best individual indicator of overall firehazard. The higher the ratio, the less the fire hazard.Table 1 compares the ablative elastomer with othermaterials


3/6


4/6

236

24

( 5 )T-T ,=345 log h e 0(8t+l)where t is time in minutes, T is furnace temperature "Cand To s the initial furnace. temperature "C. Thefumace temperature was 960C and a 60" thick blockused; after 60 minutes, the 'cold' face temperature was70C. In both cases, the 'cold' face temperatures werewell below the acceptance criteria.Cables. Fig.4 is a photograph of lengths of the newcable after fire testing in the cone calorimeter. Thesecables were placed in a simulated duct and were fittedinto the simulated duct with their surfaces slightlyproud. Hirschler and Shakir (5 ) found that the conecalorimeters gave comparable results to a full-size cabletray test such as laid down in I S 0 338. Fig.4 shows thecables removed from the artificial duct. The newablative elastomer was used as core and sheathinsulation. The cables have been subjected to a f luxdensity of 100 kW/m2. The minimum amount ofdamage should be noted. Buming is restricted to theupper surface because of the close proximity of thecables to each other in the duct. There is no damage tothe cores.

310

Smoke. Smoke tests were carried out in the conecalorimeter because. this gives a quantitative value. Itmeasures the specific extinction area (m'/kg) with time.Fig.5 show curves of the variation of specific extinctionarea with time. The relationship between the specificextinction area and time was found to be:-

where x, is peak value in seconds; x is the time frompeak in seconds; y is the specific extinction area inm2k gand k and c, are constants.It will be seen from the figure that different quantitiesof filler modifying the peaks but then falls to 450m'/kgin 1400 seconds and finally dies away in 2600 seconds.Table 3 - Comparison of new cable with other insulation.

Toxicity. Repeated testing with Draeger tubes, infra-red analysis and the cone calorimeter have indicatedonly carbon dioxide, carbon monoxide and watervapour. With an external radiant heat f lux density of100kw/m2, the peaks of carbon dioxide and carbonmonoxide are 3.14 and 0.455 kg of gas/kg samplebumt. At the end of six minutes, the level had fallen to1.92 kg/kg and 0.0061 kgkg, respectively. Overalltoxicity index by specification NES 713 is 0.75 whichis very low and which may well be due to the carbondioxide. It must be accepted that certain gases such asCO, will be present.Corrosivity. Specially formulated so that no corrosivegases would be present and subsequent test confirmedthis.BUSHINGSFigure 6 is a bushing which has been constructed foruse in switchgear up to 12kV. The bushing haswithstood 35kV for 1 minute at power frequency and animpulse withstand voltage of 75kV in accordance withIEC 1980, List 2. The bushing is insulated with thenew ablative elastomer. A modified bushing can beconstructed to go through a wall. The new cable can beconnected to the bushing, the connection being sleevedwith the new insulation. and supplies can be maintainedeven though there may be a serious fire in thecompartment. In a separate test, the cable haswithstood 1000 volts between cores while the surfacewas at 1000C.MECHANICAL TESTSThe elastomer will withstand shock in excess of 200g.under impact the depth of penetration is 0.8mm when a0.28kg mass with point ta p e d at 45" is droppedthrough 0.76mm on to a 4mm thick sheet of ablativeelastomer. Truncated projectiles withan mpact velocityof 540mlsecond and a kinetic energy of 416kJ willpenetrate up to 75mm of the elastomer. The cable hasbeen vibrated for an amplitude of 4mm at 16Hz for 1%hours.

17.5mm wallCSP low fire hazard cable11 12.5mm wallXLP/silicone11 (not ablative)

External Irradiance

75 kW h2

I 5 kW/m2

I 5 kW/m'

Time to Ignition Peak Heat ReleaseRateI kW/m2


5/6

237

A s d emperaturerise was detected but there was noevidenceof f lexuralcracking. The insulation is resistantto cut through.NUCLEAR RRADIATIONThe insulationhas been irradiated to greater than 4.17x io7 rads g a " n concomitant with 5.01 xneutrons/cm2.Negligible effects on the mechanical or electricalperformance of the elastomer up to 2.1 x io7 radsgamma concomitant with 2.55 x loi6 eutrons/cm2 wasobserved.CONCLUSIONIt will be seen that a versatile new insulation has beendeveloped. It will function under tire conditions and ithas the advantage of absorbing impact, for example,from fallingcrane blocks and gives offuirtually no toxicand no corrosive gases, and will not pass fire or hotgases if used as duct seals. It will give protection tosteelwork and electrical terminals against corrosion andmore importantly it will protect personnel fromhorrendous injuries. It can be repaired on site. Thematerial lends itself to design for increasing th e fireresistance of structures economically and effectively.ACKNOWLEDGEMENTSThe author is grateful to the Principal, ProfessorHanLins and to the Dean, Professor C B Cooper forpermission to carry out this work at UMIST. Also toDr Paul of RAPRA for assistance with the conecalorimeter and to Mr M Shipp of the Fire ResearchStation.REFERENCES1. Paterson. J.R. 1992. PhD Thesis.2. Drysdale, D. 1985.An Introductionto Firepmamiq, John Wiley & SonsLtd.

3. Marchant. E.W. 1976.5th International FireProtection Seminar, Karlsruhe.4. Paterson, J.R., Warren, L. and P d , K.T. 1992,Dielectric Measurement and Materids m dADDhtiOnS, Conference.5 . Hirscbler, M.M. and Shakir, S. 1992.Conferenceon Flame Retardants.

F i g u r e 1. S a mp le o f e l a s t o m e rb e f b r e f i r e t e s t .140 WAmblent

JF i g u r e 2. T e m p e r a t u r e r i s e of' c o l d ' f a c e w i t h t i m e .


6/6

238

-.7.i g u r c L. I e n g t h s of new cablea f t e r f i r e t e st & L.. _ _

F i g u r e 3. E l a s t o m e r r e m o v e d f ro mf u r n a c e a f t e r 6 0m i n u t e s .

r a d i a n t f l u x d e n s i t yo f 1 0 W/m?

+* ' 11" ' 14003" 700 Secon&

F i g u r e 5 . V a r i a t i o n of s p e c i f i ce x t i n c t i o n a r e a w i t ht i m e .F i g u r e 6 . B u sh in g i n s u l a t e d

w i t h t h e newe l a s t o m e r .

A New Ablative Elastomeric Insulation

Documents