AD-A02 l08 ADUA UIV (ITALY) IST 01 IMPIANTI CHIMICI F/6 21/2

ISTRUCTURE OF H2-02 FLAMES UNDER PRESSURE. (U)

UNLSIIDMAY 81 1 SORGATO DA-ERO 75-6-015

00 !EVE.L100

~AD

U; STRUCTURE OF H2-02 FLAMES UNDER PRESSURE

Final Report

by DF1. Sorgato ELEC TE

JUL 2 811981uMay 1981

EUROPEAN RESEARCH OFFICE

United States Army

London -England

GRANT NUMBER DA-ERO - 75 - G - 015

University of Padova - letituto di Impianti Chimici -Padova -Italy

Approved for Public Release; distribution unlimited

uAJ

81 7 27 082

UJNCLASSIFIEDSECURITY CLASSIFICATION OF THIS PAGE (f.n D t. Entered)

REPORT DOCUMENTATION PAGE READ INSTRUCTIONSBEFORE C6M&PL&N' FQRW

1. REPORT NUMBER A2. GOVT ACCESSION NO. 3. RCCIPIENT'S CATALOG NUUERI . X/ % f/At. 114, TITLE (and Subtitle. 5. TYPE OF REPORT & PERIOD COVE D

STRUCTURE OF JAM. -0 _NDER PR E.UR FINAL TECHNICAL RPWT-- E , H2 2 PERFORMING ORG REPORT NUMBER

UNIVERSITA DI PftDOVA7. ALTHOR(a) ., . . . CONTRACTOR GRANT NUMBER(&)

PROFESSORI. ORGA-G 015ORGTO 075G21

9. PERFORMIt:G ORGANIZATION NAME AND ADDRESS 10. PROGRAM FLEMENT, PROJECT TASK- AREA & WORK UNIT NUMBERS

UNIVERSITA DI PADOVA , - .FACOLTA D'INGEGNERIA, INSTITUTO DI IMPIANTI CHIM CI35100 PADOVA, ITALY 1T161102BHET-08

1I. CONTROLLING OFFICE NAME AND ADDRESS 12. REPORT DATE

USA RESEARCH, DEVELOPMENT & STANDARDIZATION GP." OCT 74 -- NOV 77

BOX 65 13. NUMER OF PAGES

FPO, NEW YORK, NY 0951014. MONITORING AGENCY NAME & ADDRESS(iiff deJt frim Controlling Office) 15. SECURITY CLASS. (of this repor')

USARDSG / , / , -"BOX 65 / _JI LTPT ..

'NYK Y 5 15a. DECL ASSI FICATION/DOWN GRADINGFPO, NEW YORK(, NY 09510 SCHEDULE

16. DISTRIbUTION STATEMLNT (of Wle Repor )

APPROVED FOR PUBLIC RELEASE; DISTRIBUTION UNLIMITED

17. "ISTRiBUTION STATEMENT (of the ebetrect entefred In Block 20, if different from Report)

10. SUPPLE.. ENTARY NOTES

19. KEY WOPDS (Continue on revorec side if rocersety end Identify by block nu'ibe9r)

25. -,StA~ ~ r~re-re -tdw li a.-e~~ d 1'kntii b) block n~-,

'The ReporL illustrates the work carried out over 5 years, summarI,.1 Ig

and correlating that which has alrcaly been set out in eig .ht tcch, nmi l re'ports;

on tle staLe of progress of the research with the addition of how ( :u1 h he

bevi don sici e.l./.

~ , 7 ~~ .iz rl): in;rn w- I i-jV f.. I'_ 3az7,ot I.Ttj

l" 1 1" 1 f% 1,&

/1)

' I'.

(UTY C .A,, IF I(,A TIC., 4 t~~I '~

Part [cularly described are the reu.c tor and ,;ubs;idilary eqtIiiieent , Lhe

equipment and measuring methods of temperature profiles, cuCeLcationl of ra-

dical. species, combustion velocity and finally a ni.w hybrid-stochastLic ro.tLhod

of flame simulation.

Particular emphasis is' given to how much has newly been done f rem an ex-

perimental and theoretical point of view and to the results obtained in a

field 4flaxne under pressureP, heterofore studied little. Several sets of com-

plete measurements have been taken at various pressures up to 4 atmospheres.

Accession For

NTIS GRA&IDTIQ TABUnannouncedJustificati-

-By- -

-Distribution/Availability Codes

iAvall and/orDist Special

SECURITY CLASSIFICATION OF THIS PAGEI'W?,n Data Ent.,.d)

AD

STRUCTURE OF H2-O2 FLAMES UNDER PRESSURE

Final Report

by

I. Sorgato

May 1981

EUROPEAN RESEARCH OFFICE

United States Army

London England

GRANT NUMBER DA. ERO - 75 - G - 015

University of Padova - Istituto di Impianti Chimici - Padova - Italy

Approved for Public Release; distribution unlimited

STRUCTURAL OF FLAME H 0 UNDER PRESSURE

FINAL REPORT

The contract between the ERO London and Padua University - Institute

of Chemical Plants refers to "Studies of the structure and composition of

a flat hydrogen-oxygen flame under pressure by means of thermocouples and

spectroscopic methods. The experimental results will be mathematically

correlated".

ABSTRACT

The report illustrates the work carried out over 5 years, suimnarizing

and correlating that which has already been set out in eight technical

reports on the state of progress of the research with the addition of how

much has been done since.

Particularly described are: the reactor and subsidiary equipment,the

equipment and measuring methods of temperature profiles, concentration of

radical species, combustion velocity and finally a new hybrid-stochastic

method of flame simulation.

Particular emphasis is given to how much has newly been done from in

experimental and theoretical point of view and to the results obtained in

a field "flame under pressure", heretofore studied littl_,and in which the

results given in literature are often contradictory.

INDEX

1. - EXPER1Mt4]UTAL.

2. - MATH:I4ATICAL TRIATIMRNT.

3. - I'JSUI,TS.

CONCI ,U; (-.I .

I' I, I i , 1 1*', I

/ .. ___.... ... ... . H I;R ,OCOUPLE

RS EREGULATIONAND VENT FOR IGNITION

----------- IX

GAl! [. \1-1 COLN AE

flL \ I

M A r . .X ----------

I 1-

ROD ,'., - ;1 $ ,:,!_ _ ''/:''', :':' I"."_".RIX

LI ILL if , ,,,,ir

.. ... .... ~- , _ , . . .. . L P 1,

GEAR TRANSMISSION (1 150)

THERMOJCOUPLE

SHOLDER

Jig.~ -- A /~';'ii'uu 7A~~f 1.' 7iP~?P i'7f UARTZ'!'2

-4-

U

p

I

C-

V -~Cs

IiHC-

4'

-C'

-5-

CAPILLARY

REACTOR

N2 VENT

S-%WATER SEPEP.A!CR

L-- WATER OUTCRITICAL ORIFICE

MIXER

Fig. 4 Gas fw qstem

SPECTROGRAPH

LENS

MIRROR

FLO'fl~~FLS~ LEES.... *FtP[

? 00OPERATIVE BENCH - BURNER0 0FLO./ MUMTS c-cr0 IDTR

RACK FUr(' IH tll'HIC I)[ VCi-S

J6

1. -EXPERIMENTAL.

4a) Reactor and subsidiary equipment.

The lay-out of the reactor built by us and used in the study of flamies

under pressure is given in Fig.l.

The characteristic parts are:

- the burner with corrugated matrix; which ensures a perfect mixing of the

entering mixture and laminar plug flow in the reactant mixture.

- th cobuston hambr, . = 5.6 cm, a double transparent quartz tube

that allows wrigpressure uf up to 6-7 10 5Pa and the carrying out of

optical and spectroscopic measurements and observations of flame geomeit ry

and on the position of probes and thermocouples.

- the flamiehiolder is a holed double disk in steel and pyrophyllite.

The combustion chamber is equipped with autonatic scanning, Fig. 2 wich

allows on one hand a high spatial resolution in the measurements by means of

a thermocouple introduced in the flame as well as in the optical and spectro

scopic ones,and on the other velocity in the experimental surveys thus

ensuring constancy in the behaviour of the flamec.

Fig. 3 shows the arrangement of the burner and instruments both for

combustion speed, flamne temperature and concentration of chemi cal specieOs.

Fig. 4 and 5, in particular, show inl detail the gas flow system and

the optical instrumientation.

The various pieces of equipment art, described in the paragr;!phs thatU

concern the control of the progress of the reiac tion or measurement par Li ciiiars.

b) FlwVelocity_and Pressure__Masturements.

Next to theo reactor a cri ticail orifice is se,(t uip which prodUCes a

pres sure drop, so that the pres sure in thec combus tion chamber is main Li iiw

at a constant va Ive . Cri i jcal or i.IiCcS a1re alS I sop Ced inl thle feedling lii

theseV alIlow the 1710 ow e beit y of the s niecomponen~ltS and tlie coiipos it loll

of the mi x.urie to the bniirr tO het'sue The preCssurel 1before 1and aIftLci

is nia;ue y I aanesof the sir -ag ye(Ph1ilips Tlpe Il% (,,,,!))

set- till i l eit I'ie u g oil a thhol ii lic of-i Stee

p '~~The (ii~i: f di 1 ee frd nllt( lill~ 0i 1d th.l A.i at.11ir i u'iit 1

.,1r measUred (ii .?:iieI e pel a ci vis1y i'I t hr I nj r ie ;11(I

pressure values in the combustion chamber. The signals are also visualised

on milliameters set up on the operative bench.

The transducers can visualise variations of few mm of Hg in normal

working conditions up to 10.10 5 Pa.

c) Combustion Velocity.

It is known how this parameter is one of the most important in flame

study. Data in the literature are scarce, in particular for combustion under

pressure. Experiments, if conducted with accuracy, are the most trustworthy

to obtain data on the combustion velocity of the mixture under examination.

The measurements are conducted visualising the flow by means of

suspended particles of MgO. The stroboscopically illuminated particles are

photographed every two seconds with 0.5 seconds exposure time. The camera is

an Asahi-Pentax with a 135 mm F 2.5 telephoto lens. A double enlargement is

obtained and a focal range from 1 to 2 m. The range is focused on the luminous

particles that are found on the axial diameter of the flame.

The velocity is determined by diagramming the height of the particles

against time. The tangent to the curve at a point gives the combustion

velocity at that point.

d) Temperature Profiles.

Tie temperature profile of a flame is that which analyses more completcely

its evolution, visualizing the preheating, reaction and post-reaction zones.

The temperature values along the flame are obtained through weasurcl,,its

by means of thermoiictric probes immersed in the flame and of special diwc:cions

and shape, in order to disturb the gas flow as little as possible.

The probe used is a 25 microns di ame tier chromcl-alumel thermocouplt

covered with a fused mixture of borax and a1lumina to climi iate the catalaLyi

effects. as much as possible. The very snnll diamet-er and tlhe automatic

scanning allow for good spatial resolution.

The instant position of the probe is recorded by the aid of a movjl ''

transduccr which is capablc of m;vig tihe positiol! with .l 10 rli(CIoW;

acc l cy . The cffect ive he;i lt'Ic le t m ic is , few stcoinds,so as to en:,ii

the co;asvt ;11cy of- t hi oper;itiv , \,VQ a iibl Os. Owing to lthe brief t w,' uil lit

it is n,,t J)rIscilI t i ,.C :1 lilTr I peCn re.Co1n dCr, wilt):;k, c otit:Lnit i!a hi !,,r

tha , ,) iof .t ,. :i' , nil lii a1 c,' .' , lfl i S I, H I (If ,1 - r ly , iirc

8

the 7th order which fits the data of the readings and the temperature

obtained is then corrected for radiation losses of the thermocouple junction.

e) Concentration measurements of the OH and H species.

The concentration values of the radical species are an indispensable

complement to the temperature measurement and a useful indication on flame

behaviour, in particular in the reaction zone where the pressure effects can

have a great effect on the reaction kinetics.

The relatively low flame temperatures 1 -0 -N do not allow the appli- Vcation of the emission spectroscopy except in the case of the fundamental

lines of alkaline metals injected in the flame. This technique permits the

determination of the H concentrations. The emission intensity of the sodium

D lines is due only to the chemiluminescence according to the reactions:

H + H + Na - H2 + Na

H +O + Na 1 *0 + Na2

since the thermal emission is irrelevant, because of the low temperatures

of the flame.

On the other hand, absorption spectroscopy is used for Ol concentration

measurements. For such measurements a quartz xenon lamp is used as light

source and a Philips M' 1003 pheto-multiplier as transducer.

To raise the spatial resolution, cylindrical lenses focusing on the

centre of the flat flame are used. The image on the entrance of the spectrcmetcr

a (Jarrel-Ash) is enlarged twice. The Ol spectrum band (0,0) from 3060 to

3200 A, is taken in the 2nd order to raise the resolving power of the

spectrometer.

2. - MATHEMATICAL TREATMENT.

a) Heat Flow.

The volumetric chemical heat release rate results at any point

d dtq TcT-X j-) (1)

p0 v 0 = mass flow rate of the mixture (po v 0 density and burning

velocity at normal conditions)

c , X = specific heat and local thermal conductivity.p

Equations (1) describes the heat release rate as the gradient of

energy flow due to the mass transfer (convection) and to the heat conduction.

It does not make into account the heat losses which are due, in particular,

to the radiation. The equation has been numerically calculated by Stirling's

method.

The integrated heat rate througouht the flame interval

f q dz (2)

Z = 0

can be compared (apart from the small heat losses) with the heat calculated

from the reaction enthalpy A Hr

Qt = (Vo/V tot) (All/ 2 ) 0v 0 (3)

Vo, Vto t = volumetric flow rate of 02 and total.

A comparison between the two heats allows good agreement witLh the

temperature measurements.

b) Stochastic hybrid model of flame.

In the matic.matical simulation of the burning process of a

monodimens;iona] Ilame stoch: :; tic solution has bcen adopted for the chumicl

process, wh ich ha;, a more cons is tent phy.ical ba:'is tha 1 the commrle; 1y

1ised do(termi ni. tic. It ad;lpts butter to the use of wt computer (avoi l; i ;.

ru't i-OL i1." an1or;) aih( a)J]x.:; tiF oC, or Cal Iut; i j; I ,a! ; 1-0 to L r )]a0Y [

adopted mi-.Odel iar, exact. zind convit;e .

The Sto a( : i S0t; L on ,';c t t: :1 ia J ,od icl:;tieity o . t . I it

has been possible to hybridise it with deterministic solutions for energy

transfer and molecular diffusion.

Volume V, spatially homogeneous and in thermal equilibrium, contains

a mixture of various chemical species which react among themselves in

different ways. They are:

N. = number of particles of the various chemical species S.1

i = 1,2 .......... n.

Rr chemical reactions, r = 1,2 ...........,m.

The probability of the reaction R taking place is:r

m

Pl(r) = Q Q (4)1 r I r

and the probable time is:

m me r( t) Q r exp- (Z Qr t0 (5)

r 1 1

where Qr = Cr f r with fr combinatory function of N.i and c r, stochastic

reaction constant = k q! N q - 1,2,3 for differents, two or threer

equal reacting molecules and g = reaction order.

In order to select the event and the probable time it is necessary

to generate two casual numbers L and L included between 0 and 1 and to1 2choose r so that:

r-l II rS 1 1Q r Qr (6)

and m

t = ( r/ Qr) In (i/L ) (7)

Once r is known the values of N. are updated at time t according to theSstoichiometry of reaction R .

Once the reaction enthalpies and the heat capacities of the vario.to;

components are I nown, we have for every calculation cycle s, and for the

adiabaLic flame:

T = T - (All) r (c Ni/1) s 0,1,2.

and wt l wer;Iva fi, ow velocity will be:

(NT)"S r, v .s s-J( r0

i~r r .... .. - i i -- "

The conservation equation for the species i in the case of "premixed,

stationary, laminar, ideally monodimensional flame" is:

d dn.-I ni (v+vd] = (8)dz dr

with z = axial coordinate, n = mole density number, v and

vd = mean velocity and diffusion.

Introducing the fractional mass flux we have:

p0 v 0 d G = d n. (9)

M. V1.

with p v = mass flux velocity0

G = weight fraction of species i in the flux.M = molecular weight.

Since the variation of n. takes place in a stochastic way, per1

unitary increase, the displacement of one molecule will happen when:

p0 v0 G. > I negative diffusionI (10)

M. v < -1 positive diffusion

For the type of flame adopted the energy conservation equation,

integrated with the boundary conditions z = 0, T = T, T/z 0 offers

AT v p p (T T) Qo p (I)

Az X

with 7 and -X mean specific heat and average conductivity of the mixture.

p

Equation (11) is called upon only for unitary events linked with the

initial reactions between molecules (reactions 9 and 10 of Table 2). When

these reactions which are the most probable in the high part of the flale

are defined by the programme, their stochastic time determines the entity

of heat transfer.

In this way both for molecular diffusion as well as heat transfer a

system of hybridization is chosen adopting the dete rministic equations

only for unitiry events linked to the s.tocli stic approach.

Thus it is possible to cal cula t::

- tO Wt 11111111W.1 0'-

P.VOil S;;

- the i. iiiihe f t t' eve C ' 'oi' '

- the t ;f ' ie ' tlw Ll c1v Ily ;11d I1ht' dp i t; e5

- 12-

3. - RESULTS.

Among the numerous runs described in the 8 reports, two have been

chosen which characterize better the various theoretical and experimental

aspects and on the other hand are sufficient in describing the behaviour

of a flat flame under pressure.

The characteristics are given in Table 1.

TABLE I

Chamber Ingress Flow Qt. Qs.pressure Composition Velocity eq(3) eq (2)

H2 0o2 N1

Pa 10 % cm sec Cal (cm sec) -

1 16.15 5.40 78.45 9.05 2.80 2.55

3 12.40 5.15 82.55 12.85 11.30 10.51

Characteristics of flame at atmospheric pressure and under jressure,

As regards the heat release in general, a discrepancy of almost 10%

is found between the values calculated by equation (3) and those deduced

by the temperature profile equation (2). The agreement is good, taking

into account the the heat due to radiation is not calculated.

The reaction probability, the temperature profiles and the composition

of the two flames have been calculated with the help of a Laben 70

computer.

It is assumed N = 50.000.to t

Of the 38 reactions that can happen in flames It2-O2-N 2 only -0 are

necessary to obtain satisfactory results from tho simulation without making

the calculation too involved.The 10 reactions are given in Table 2 togCther

with the kinctic conslant,. These last are suggested by BAULCII-DRYSPALE -

HOME-'LLOYD (2).

Reaction enthalipies, specific heats and therinal conductivities arc

those given in the table by I'PENNI (3) except t hat for HO,, the val es

given in ref. ( ) have been used.

- 13 -

TABLE 2

*3 -I -1Reaction k cin. (cm mol sec - ) AT,K AH 298(kJ/niole)

141) 02+H - O+OH 2.2010 exp(-8450/T) 700+2200 70.672

2) H2+0 - H+OH 1.8.1010 T exp(-4480/T) 400+2000 8.255

133) 112+OH - E 20+H 2.2.10 exp(-2590/T) 300+2500 -63.304

4) H+1102 * Oi+OH 2.5.1014 exp(950/T) 290+800 -159.979

5) 1120+11 + 11 2+OH 9.3.1013 exp(-10250/T) 30012500 63.304

6) 11+H+N 2 H2 +N2 2.5.10 • /T 300+5300 -435.973

7) 02+H+N2 HO2+N2 1.5-1015 exp(500/T) 300+2000 -197.096

8) H+1O2 H 20+0 5.1012 TO0 5 exp(-2000/T) 300-350 -231.538

9) H2+02 OH+O 8.1014 exp(-22500/T) 298+2000 78.927

10) 112+02 H+HO2 5.5.1013 exp(-29100/T) 298-2000 238.906

The diffusion velocity, owing to the large excess of N2 in the

flame, is approximated by:

N Di ,N2 d N.

d i N. dz (12)

The binary diffusion coefficient is evaluated from the well-known

equation.T3/2 (M_2)(2 Mi 1/2

D = 2.628 0- 3 cm2 sec - 1 (13)2 2 " p 0* (T )

i-i

The values in the constants are given in ref. ( 5) while the

corrections causcd by the presence of polar species arc valued according

to ( 6). In the simulation progranmlc the reduced coll'isional integral Q

(T) is fitted with a 10th degree polynomial expression in (T)-1/2 frow 2

the data in ( 7 );the polynomial coefficients arc given in Table 3

TABLE 3

a = 0.440149 b 1.050917 c = -0.036973

d = 0.818693 c -1.416095 f = 2.1.53107

g= 6.386901 h = -2.91019 i = -7.933206

I -3.8/606 ) 9.390701

The piffo;, ;ae given in the opptdi..

-14 -

60

40 __ _ _

20

2~ _2 __

p = 3-10-5 pa

20 - _ _ _ _--

0 0,05 0,10 0 $15 0, P

Fig. 6 h Fv,~ t.l'~ ' ci'8'~ ,? "'"

- 15 -

1200

p. '3 Pa p.10 5 -1pa -

T ,OK

I

I

ilYdI 0

4000-

=, 200

0 O, 05 0,1 0 0,~I O,2(

600 -1-

I F

Fig. 7

- 16 -

a) Contribution of every single model reaction.

The calculation programme takes into account the stochastic path in

the form of number of event of the various elementary reactions carried

out by the same programme (Fig.6). Such data have allowed the programme

optimization to reduce notably the troublesome kinetic model discaring

the radical reactions which in our case were practically nil.

From the diagrams of Fig.6 it is possible to observe:

- the reactions between radicals tend to diminish at the approach of

equilibrium. This can be represented by reactions (3) and (5).

3H + OH H0 + H2 ~ -- 2

5

that reach in proximity of equilibrium, intervention values near to 100%.

- reaction (9) between H2 and 02 tends towards zero corresponding with

the ignitions point of the flame.

- reaction (4) shows a maximum at intermediate temperatures, a zone in which

the species H 02 is stable, and which later decreases to zero at

equilibrium.

b) Temperature Profiles.

The temperature profiles are given in Fig.7. A dotted line shows

the experimental ones, a broken line indicates the calculated ones. The

agreement can be regarded as satisfactory taking into account that:11

- the experimental values suffer from approximate measurements and an

imperfect flat shape of the flame.

- in the model the flame is assumed adiabatic and certain secondary reactions

are disgregarded.

From the two profiles of I and 3 atm it is possible to note, in

agreement with the increase of the global reaction velocity with prcc:;urc:

In (rl/r 2 ) = [(q/2) -i1 In (pl/P) w.here q is the global reaction order8

= 2.081 (8), that equilibrium is reached more quickly as the press;ure is

increased and tim reaction xone becomes thinnur.

It is al so noLed (rutis de,;(ril cd in pre, ots report s), that 1il

effect of diff12 1;o1) il '11,)(, ;IcC('I't t d ;'t ] ot.' pr,2 rl h;'ti at iigh

pre s t.. . .. giveu that:h. " global . . . . . ction v... | y is . r.por t to.. ... .

-17-

0,85-- -

0,80 i

0,75

0,20 - I_

0,15 iV'2

0,0 ------

0 O'O'l oyjO o,()z Wh.

F ig o. 6I P Co . W'~t i' J'i'Ir /' (o).

- 18 -

1,0

p = 1.10 5 Pa

..... p = 3.105 Pa

C)

0,8z

. I!OH 10_ _ .. _ _

< 0,6

1-)

II

0I0,2: I

0 0,05 0,10 01, 0,-2

Z CM

o.{ .. .z:..t!3t.z.<j< £ %f ; i,.,()

- 19 -

p (q/2)- ; the diffusion is inversely proportional to pressure, as it

increases, the reactions gain in importance.

c) Concentration Profiles.

The facts previously described find verification also in the

concentration profiles, Fig. 8-9. The increase of the global velocity

leads to a quicker reaching of equilibrium conditions.

The two most important chain carriers, OH and H, show, as intermediate

species, concentration maxima whose position depends on pressure. The

relative curve at OH presents two of them, the first relative to reaction

(4) masimum and the second together with that of H, due to reactions (1)

and (2) maximum.

The presence of a minimum for OH had been noted in the experimental

determination of such species, report 6.

CONCLUSIONS.

The theoretical-experimental study of flames H -O2-N under pressure

has allowed us to identify the influence of the pressure parameter on

fluid dynamics and on the progress of the system reactions in flow.

The small reaction zone of flame under pressure and the low temperature

of flame due to the excess of N2 are favourable conditions in the assumption

of single dimensional flow and therefore ideal for the experimental chech of

the simulation model adopted.

This,a hybrid-stochastic type, offers better opportunities in respect

of the deterministic normally used. In fact besides avoinding "round off

errors", giving stable and accurate solutions, it is capable of indicating

the most important physical phenomena and the most probable reaction

mechanism for the various temperature and pressure conditions of the

flame considered.

-20-

REFERENCES

(1) D.T. GILLESPIE, Jo~na o6 Physicat ChcnitAty, 81, 2340 (1977).

(2) D.L. BAULCH, D.D. DRYSDALE, D.G. HORNE, A.C. LLOYD "Evaluated

Kinetic Data for High Temperature Reactions",

vol. I - Butterworths, London, 1972.

( 3) S.S. PENNER "Chemistry Problems in Jet Propulsion", Pergamon Press,

New York, 1957.

( 4) D.L. RIHANI, L.K. DORMISWAMY, Ind. Eng. Che n. Futdameata&&, 4,

17 (1965).

( 5) R.A. SVEHLA, NASA TR, R-132, 1962.

L. MONCHICK, E.A. MASON, J. Ohem. Phy6., 35, 1676 (1961).

6( ) M.V. TAYLOR, P.J. PETROZZI "The Performance of High Temperature

System", Gordon & Breach, New York, 1968.

( 7) J.O. HIRSCHFELDER, C.F. CURTISS, R.B. BIRD "Molecular Theory of Gases

and Liquid", J. Wiley & Sons, New York, 1954.

( 8) A.E. POTTER, Jr. "Progress in Combustion Science and Technology"

Vol. 1, Pergamon Press, Oxford, 1960.

Li

I. I -- "

"0 -

T '1 =- pA[ 2 "

1 " , 1 '-;

I . . . . . . ' - ' - '

* r .. . .[4f .° " ° ° 1 '

,- . , ,- .9 ,

-. ' - -, • . ' L

I -- 1 .' 7 - '

f -l y. " . 0 " . .

- P

g I-

1 , -' I ;. ~ ' '

1: - r If7 . ' " "f/

f lir r T -- 2 1< + /1 ,' . ." r i '

f'lI(: - - --. . (-) . ' r- ,,rq /i- '1",-

'C. , ,nn. I 7 ' ) -( :l

-. ~ ' 'I - ' " " 1 "

ri < - " -<, 7 -- " : <

"

r,, I .. ]' = " ' ,.--- i._.ll ix. 0302I , P1 Ia 2l . . . . . . . . .. . . .. . . ... .,.. .

O2~9 I PT F = 01 1'

-# 1'

I -

17 7

*1-1 -

I -, ~ -

C --

C: -

U, 7

C :1

( 7 1(CE-I U CF [Ct A A

7 AZ 41ft ~ p AZ

A' * - - -

1 1 C['~ '~AC - -

Cl

- 1 '-1- K

C *- A

1~hr

U- I - -

a ~- - -,

- ~- - -J'I. - P~'~~~'p -<--

-I *i'--

*

I p I'~

ft I '-.1

'1 -C - .1 -* --- ';~( I

A AZp I - --e 1 * -/ * I' I ~Cr - 1~CI

('$10 (Al 1 i-i

.1 -

7 C

'I ,-

* - ~ ., _ . ,i

-- " . - * -

•I " -, A 4 ' 1 - -

111 I - ( ' '-(A - *

1" I '= 1 1

II -4 -- 4

-

11 - "]*- - - / o

,S. ,

'vi 91T F-f-

I '. -r 1 ..

Ic A!' T . I,,

' I' L ' I (-< - "" r

166(1 LT "i F,' I '

-1= , N . '*

, *

" . 1 . - ' " .-

" 1 : _- -. . -.-

I- - '--I'" 'V

'I. , -.-

-, ] " ,-N

I -: -- I,

r. - -- - .- - -.

" : ' ' ( = ' -1

-' ',r ' , i :

S,1 1 -- . . --rI

1l " - - N,-

-I-'

I ' l ' :

.. . . . . . .I, ... -I Il-, . . . . . . . . . . . , l

' 1. 1 '

' I T I +(

>4-

, ,1 ' -1 _ - '

f "t :: 1== -,4- 1

'"{ 1 . 1 U- Crfl{

( U

-r i " 4rf

,-' - -. . .. i : 4 1=

,r~o,. i r = ,

r-' { .c _- 1-41 . - + ; - "

1'11< ' . )4 ,

,I < ' i r f r

,- IC' 1 1 r

I' 1

tI

O~ATILE9l