laminar flame speeds of me thane -air mixture s under reduced...

17
COMBUSTION AND FLAME 76: 375-391 (1989) 375 Laminar Flame Speeds of Methane-Air Mixtures Under Reduced and Elevated Pressures F. N. EGOLFOPOULOS, P. CHO, and C. K. LAW* Department of MechancialEngineering, University of California, Davis, CA 95616 Using the counterflow methodology, the laminar flame speeds of methane-air mixtures have been accurately determined ove r the pressure range of 0.25-3 a tm a nd ove r extensive lean-to-rich concentration ranges. Theseflame speeds are then compared with the numerically calculated valuesobtained by using various published kinetic schemes of either the C1 mechanism or the full C2 mechanism. Two such schemes show very close agreement with the experimental data. However, availableinforma tioncannot furthe r differentiate the relative superioritybetween them for flame speed calculations, e s p e c ia lly the importance of C2 reactions for moderately rich situations. Two reduced mechanisms a re also deduced through sensitivityanalysisa nd a re expected to be usefulfor flame speed calculations a nd approximate flame structure studies. INTRODUCTION The laminar flame speed S ° is an important physicochemical parameter of a combustiblemix- ture because it contains the basic information regarding its diffusivity, exothermicity, a nd reac- tivity. However, in spite of the extensive effort expended to accurately determine their values, e s p e c ia lly those of the conventional hydrocarbon- air mixtures, wide systematic spreads in the reported experimental data still exist even though in many instances the experiments appear to have been carefully executed [1]. Figure 1, partly reproduced from Ref. 1 and then updated to include some recent data [2, 3], clearly demon- strates this point for the laminar flame speedsof methane-air mixtures as a function of the equiva- lence ratio ~b. Recently it has been suggested that these sys- te ma tic dis crepancies are probably caused by the coupled effects of flame stretch and preferential * P re s e n t addresses: (P. C.) Department of Mechanical Engi- neering,Michiga n Technological Unive rs ity, Houghton,MI 49331. (C. K. L.) Department of Mechanical & Aerospace Engineering, Princeton University, Princeton, NJ 08544. Copyright © 1989 by The Combustion Institute Published by Elsevier SciencePublishingCo., Inc. 655 Avenue of the Americas, New York, NY I0010 diffusion [4, 5]. Specifically, it has been demon- strated both theoretically and experimentally that the flame responsecan be qualitatively reversed when the flame stretch changesfrom positive to negative, as in the cases of expanding spherical flame versus the Bunsen flame, and when the mixture's effective Lewis number crossesa criti- cal value typically aroundunity, as in the casesof lean methane-airand rich propane-air mixtures versus rich methane-air and lean propane-air mixtures. A particularly seriousimplication of the stretch- induced flame responseis the potential falsifica- tion of the kinetics information determined or validated through comparison between the numeri- cally calculated and experimentally determined results [6, 7]. It is clear that if considerable uncertainty a nd/or ina ccuracy exist in the experi- mental data, their usefulness and fidelity for kinetics studies can be seriouslycompromised. In view of the above consideration, La w a nd co- workers [8-10] have proposed a counterflow- based methodology through which stretch effects can be systematically subtracted out such that S ° can be unambiguously determined. The laminar flame speeds of methane-air and propane-air 001~21801~I$03.50

Upload: others

Post on 06-Oct-2020

2 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Laminar Flame Speeds of Me thane -Air Mixture s Under Reduced …ronney.usc.edu/AME513b/Lecture4/References/... · 2020. 2. 28. · Laminar Flame Speeds of Me thane -Air Mixture s

C O MB US T IO N A N D FLA ME 76: 375-391 (1989) 375

Laminar Flame S pe e ds o f Me tha ne -Air Mixture s Unde r Re duce d and Ele vate d Pre s s ure s

F. N. EGOLFOP OULOS , P . CHO, and C. K. LAW*

De partm e nt o f Mechancial Engineering, Univers ity o f California, Davis , CA 95616

Us ing the counte rflow me thodology, the la mina r fla me s pe e ds of me tha ne -a ir mixture s ha ve be e n a ccura te ly de te rmine d ove r the pre s s ure ra nge o f 0 .25-3 a tm a nd ove r e xte ns ive le a n-to-rich conce ntra tion ra nge s . The s e fla me s pe e ds a re the n compa re d with the nume rica lly ca lcula te d va lue s obta ine d by us ing va rious publis he d kine tic s che me s of e ithe r the C1 me cha nis m or the full C2 me cha nis m. Two s uch s che me s s how ve ry clos e a gre e me nt with the e xpe rime nta l da ta . Howe ve r, a va ila ble informa tion ca nnot furthe r diffe re ntia te the re la tive s upe riority be twe e n the m for fla me s pe e d ca lcula tions , e s pe cia lly the importa nce of C2 re a ctions for mode ra te ly rich s itua tions . Two re duce d me cha nis ms a re a ls o de duce d through s e ns itivity a na lys is a nd a re e xpe cte d to be us e ful for fla me s pe e d ca lcula tions a nd a pproxima te fla me s tructure s tudie s .

INTR O DUC TIO N

The la mina r fla me s pe e d S ° is a n importa nt phys icoche mica l pa ra me te r o f a combus tible mix- ture be ca us e it conta ins the ba s ic informa tion re ga rding its d iffus ivity, e xothe rmicity, a nd re a c- tivity. Howe ve r, in s pite o f the e xte ns ive e ffort e xpe nde d to a ccura te ly de te rmine the ir va lue s , e s pe cia lly thos e o f the conve ntiona l hydroca rbon- a ir mixture s , wide s ys te ma tic s pre a ds in the re porte d e xpe rime nta l da ta s till e xis t e ve n though in ma ny ins ta nce s the e xpe rime nts a ppe a r to ha ve be e n ca re fully e xe cute d [1]. F igure 1, pa rtly re produce d from Re f. 1 a nd the n upda te d to include s ome re ce nt da ta [2, 3], c le a rly de mon- s tra te s this point for the la mina r fla me s pe e ds o f me tha ne -a ir mixture s a s a function o f the e quiva - le nce ra tio ~b.

Re ce ntly it ha s be e n s ugge s te d tha t the s e s ys - te ma tic dis cre pa ncie s a re proba bly ca us e d by the couple d e ffe cts o f fla me s tre tch a nd pre fe re ntia l

* P re s e nt a ddre s s e s : (P . C.) De pa rtme nt o f Me cha nica l Engi- ne e ring, Michiga n Te chnologica l Unive rs ity, Houghton, MI 49331. (C. K. L.) De pa rtme nt of Me cha nica l & Ae ros pa ce Engine e ring, P rince ton Unive rs ity, P rince ton, NJ 08544.

Copyright © 1989 by The Combus tion Ins titute P ublis he d by Els e vie r S cie nce P ublis hing Co., Inc. 655 Ave nue o f the Ame rica s , Ne w York, NY I0010

diffus ion [4, 5]. S pe cifica lly, it ha s be e n de mon- s tra te d both the ore tica lly a nd e xpe rime nta lly tha t the fla me re s pons e ca n be qua lita tive ly re ve rs e d whe n the fla me s tre tch cha nge s from pos itive to ne ga tive , a s in the ca s e s o f e xpa nding s phe rica l fla me ve rs us the Buns e n fla me , a nd whe n the mixture 's e ffe ctive Le wis numbe r cros s e s a criti- ca l va lue typica lly a round unity, a s in the ca s e s o f le a n me tha ne -a ir a nd rich propa ne -a ir mixture s ve rs us rich me tha ne -a ir a nd le a n propa ne -a ir mixture s .

A pa rticula rly s e rious implica tion o f the s tre tch- induce d fla me re s pons e is the pote ntia l fa ls ifica - tion o f the kine tics informa tion de te rmine d or va lida te d through compa ris on be twe e n the nume ri- ca lly ca lcula te d a nd e xpe rime nta lly de te rmine d re s ults [6, 7]. It is cle a r tha t if cons ide ra ble unce rta inty a nd/or ina ccura cy e xis t in the e xpe ri- me nta l da ta , the ir us e fulne s s a nd fide lity for kine tics s tudie s ca n be s e rious ly compromis e d.

In vie w o f the a bove cons ide ra tion, La w a nd co- worke rs [8-10] ha ve propos e d a counte rflow- ba s e d me thodology through which s tre tch e ffe cts ca n be s ys te ma tica lly s ubtra cte d out s uch tha t S ° ca n be una mbiguous ly de te rmine d. The la mina r fla me s pe e ds o f me tha ne -a ir a nd propa ne -a ir

001~21801~I$03.50

Page 2: Laminar Flame Speeds of Me thane -Air Mixture s Under Reduced …ronney.usc.edu/AME513b/Lecture4/References/... · 2020. 2. 28. · Laminar Flame Speeds of Me thane -Air Mixture s

376 F . N. EGOLFOPOULOS

50

45

40

35

25

20

15

10 i t i = I 0.7 1.0 1.1 1.2 1.4

d~

- i - Andrews/Bradley (1972) -e- Gunther/Janisch (1972)

Sharma et aL (1981) Babkin/Kozachenko (1966)

.o- IijimalTakeno (1986) -~- Yarnaoka/Tsuji (1984)

= ! = ! = 0.8 0.9

CH 4/AIR p= 1 ATM

| I | 1 .3

F ig . 1. Va rio u s re p o rte d e xp e rim e n ta l fla m e s p e e d s S °(0 ) , fo r m e th a n e -a ir m ixtu re s a t p = 1 a tm .

mixtures unde r a tmospheric pressure ha ve thus been de te rmined [9, 10].

The present inves tiga tion continues our s tudy of the s tructure a nd propaga tion o f la mina r premixed flames with two specific objectives . Firs t, we sha ll apply the counte rflow method to de te rmine S ° o f me tha ne -a ir mixtures unde r reduced as well as e leva ted pressures , ranging from 0.25 to 3 a tm. Although these da ta a re o f practica l inte res t in the ir own right, especia lly in te rms of high- pressure combus tion, we note tha t the y a lso ca rry s ignificant kine tics informa tion because pressure not only influences the fre que ncy o f molecula r collis ion but a lso diffe rentia tes the re la tive e ffl- ciencies o f two-body branching reactions versus three-body te rmina tion reactions . Thus the y pro- vide additiona l cons tra ints on the va lida tion o f the

kine tic schemes . This the n leads to the second objective , which is to compare our experimenta l da ta with those de te rmined numerica lly by us ing the kine tic schemes proposed by diffe re nt research groups , us ing the same fla me code a nd transport properties code . It will be shown tha t through such comparisons , a nd because o f the fide lity o f the present da ta , useful ins ights ha ve been ga ined on these kine tic schemes . S implified schemes a re a lso proposed for accura te ca lcula tion of the fla me speeds within the parametric ranges s tudied here in.

The experimenta l me thodology is brie fly speci- fied in the next section, which is followed by presenta tion o f the experimenta l results . These results a re then compared with the numerica lly ca lcula ted va lues in the following section.

Page 3: Laminar Flame Speeds of Me thane -Air Mixture s Under Reduced …ronney.usc.edu/AME513b/Lecture4/References/... · 2020. 2. 28. · Laminar Flame Speeds of Me thane -Air Mixture s

LAMIN AR F LAME S P E E D S

f O-IAMBER SOLENO1D VALVE ~ PRESSURE GAUGE

EXHAUST ~ ?

. . . . . . ROTAMI~II~R FLOW

MIXINQ FLOW ~ VALVE PRESSURE

I ~ ON/O~ I y

NITROGEN INLET

Fig. 2, Schematic of the experimenta l se tup.

377

E XP E R IME NTAL ME THO D O LO G Y

Figure 2 s hows the s che ma tic o f the e xpe rime nta l s e tup a nd the va rious flow va ria ble s a re de fine d in Fig. 3. The e xpe rime nt ba s ica lly involve s the

1.1

1.0

0.9'

0.8

,~ 0.7

d 0.6

0 . 5

0 . 4

es tablishment o f two s ymme trica l, pla na r, ne a rly adiaba tic fla me s in a nozzle -ge ne ra te d counte r- flow, the de te rmina tion o f the axia l ve locity profile a long the ce nte rline o f the flow by us ing la s e r Dopple r ve locime try, a nd the identifica tion

Rames =

K= - / dx

u = S L CH4 /AIR

0.3 • = 0.8 0.2 i t t . t ~ I , ,

0 1 2 3 4 5 6 7 X~ 1111111

Fig. 3. Typica l axia l ve locity profile across a s tagna tion flame, showing the definitions of K and SL.

Page 4: Laminar Flame Speeds of Me thane -Air Mixture s Under Reduced …ronney.usc.edu/AME513b/Lecture4/References/... · 2020. 2. 28. · Laminar Flame Speeds of Me thane -Air Mixture s

378 F. N. E G O LF O P O ULO S

4 0

3 0

2 0

10

O = 1.0

O = 1.3

I~ ' ~ 0 = 0.6

i I i I i I i 0 2 0 0 4 0 0 6 0 0

K, S "1

C H 4 / AIR p = 2 ATM

I 8 0 0

Fig. 4 . Va ria tion o f the fla me s pe e d S L(K) with the s tra in ra te K.

1000

o f the minimum point o f the ve loc ity p ro fd e a s a re ference ups tre a m fla me s pe e d SL corre s ponding to the impos e d s tre tch ra te K, which by de finition [5] is s imply the ve loc ity gra die nt a he a d o f it. Thus by plotting SL ve rs us K, the la mina r fla me s pe e d without s tre tch, S ° , ca n be de te rmine d through line a r e xtra pola tion to K = 0 (Fig. 4) in a ccord- a nce with the ore tica l pre s crip tions [4, 5]. The line a r e xtra pola tion is me a ningful a nd a ccura te only fo r s ma ll va lue s o f the nondime ns iona l s tre tch ra te

DK Ka = ,~ 1, (s o)

whe re Ka is the Ka rlovitz numbe r a nd D a re pre s e nta tive ma s s diffus ivity. Re ce nt s tudie s on fla me e xtinction [5] ha ve s hown tha t Ka inde e d a s s ume s s ma ll va lue s p rio r to e xtinction. It ma y a ls o be e mpha s ize d tha t SL is on ly a re ference ups tre a m fla me s pe e d be ca us e the ga s te mpe ra ture a lre a dy s ta rts to incre a s e s lightly be fo re the minimum point is re a che d. The SL s o de fine d turns out to incre a s e with incre a s ing K. Ho we ve r, SL doe s de ge ne ra te to the true S O in the limit o f va nis hing s tre tch.

The dia me te rs o f the nozzle s us e d in the pre s e nt inve s tiga tion a re 7 a nd 14 mm. The y ha ve high contra ction ra tios a nd a re wa te r-coole d a nd nitro- ge n-s hroude d. The e ntire bu rne r a s s e mbly is

hous e d in a la rge s ta inle s s -s te e l h igh-pre s s ure cha mbe r with continuous ve ntila tion, the re by a l- lowing fo r pre s s ure a djus tme nt. The cros s s e ction o f the te s t portion o f the cha mbe r is 14.5 by 14.5 c m .

Ignition is a chie ve d by a continuous high- e ne rgy s pa rk p roduce d by re tra cta ble e le c trode s . The LDV ope ra te s in the fo rwa rd s ca tte ring mode with 0.3-/~m a lumina pa rtic le s e e ding.

EXP ERIMENTAL RES ULTS AND DIS CUS S IONS

Figure 5 s hows the me a s ure d S O a s functions o f the e quiva le nce ra tio ~ a nd s ys te m pre s s ure p fo r me tha ne -a ir mixture s . The re s ults s how tha t the fla me s pe e d de cre a s e s with incre a s ing pre s s ure , which is in qua lita tive a gre e me nt with the pre vi- ous ly re porte d be ha vior [11]. The s e re s ults , how- e ve r, a re not a s funda me nta lly s ignifica nt a s the ma s s burning ra te

O _ 0 m L-- puS L,

which a ccounts fo r the de ns ity va ria tion with pre s s ure a nd is the p rope r e ige nva lue fo r la mina r fla me propa ga tion, whe re Pu is the de ns ity o f the unburne d mixture . Thus the da ta o f Fig. 5 a re re plotte d in Fig. 6 a s m ° ve rs us ~b a n d p . It is s e e n tha t m ° incre a s e s with incre a s ing pre s s ure . Be -

Page 5: Laminar Flame Speeds of Me thane -Air Mixture s Under Reduced …ronney.usc.edu/AME513b/Lecture4/References/... · 2020. 2. 28. · Laminar Flame Speeds of Me thane -Air Mixture s

LAMINAR F LAME S P EEDS 379

7 0

5 0

4 0

r J

°3= r,/3 30

2 0

10

4 - :

0 , I , I , I , I i I i 0 . 4 0 . 6 0 . 8 1 . 0 1 . 2 1 . 4 1 , 6

@ Fig. 5. Experimental laminar tiame speeds S°(O, p) for pressures between 0.25 and 3 a tm.

ca us e (m0L) 2 is proportiona l to the ove ra ll re a ction ra te a s e xpe rie nce d by the fla me , the incre a s ing tre nd o f m0L with pre s s ure implie s tha t the ove ra ll re a ction orde r n , a t a give n ~ , is pos itive for a ll s toichiome try a nd pre s s ure ra nge s s tudie d he re in. Furthe rmore , s ince the diffe re nce in m ° is la rge r for ne a r-s toichiome tric fla me s a nd s ma lle r for ne a r-limit fla me s , the s pe cific va lue s o f n s hould a ls o va ry a ccordingly. The de cre a s ing tre nd o f n for s lowe r burning fla me s note d a bove is e xpe cte d to be due to the incre a s ing importa nce o f the thre e - body, te mpe ra ture -ins e ns itive , te rmina tion re a c- tion

H + O2 + M-*HO 2 + M

re la tive to the two-body, te mpe ra ture -s e ns itive , bra nching re a ction

H + O2--'O + OH,

a s dis cus s e d, s a y, in Re fs . 6, 7, a nd 12 a nd a nticipa te d for the pre s e nt inve s tiga tion.

Figure 7 plots the pre s e nt da ta a nd the ba nd cove ring the e xis ting e xpe rime nta l da ta from the lite ra ture for the me tha ne -a ir fla me s pe e ds a t 1 a tm. Figure 8 compa re s the s toichiome tric da ta for va rious pre s s ure s . The s ignifica nt s pre a d in the e xis ting da ta a nd the re by the ne e d for the pre s e nt a dia ba tic, s tre tch-fre e re s ults a re a ga in e mpha - s ize d.

NUME R C IAL R E S ULTS AND DIS CUS S IONS

(R1) In orde r to nume rica lly s imula te the fre e propa ga - tion o f a dia ba tic, s te a dy, one -dime ns iona l pla na r me tha ne -a ir fla me s , we ne e d a nume rica l code for fla me propa ga tion a nd a kine tic s che me de s cribing me tha ne -a ir re a ctions . The s e a re pre s e nte d in the

(R2) following.

Page 6: Laminar Flame Speeds of Me thane -Air Mixture s Under Reduced …ronney.usc.edu/AME513b/Lecture4/References/... · 2020. 2. 28. · Laminar Flame Speeds of Me thane -Air Mixture s

3 8 0 F . N. E G O LF O P O U LO S

0.08

@1

-0- 0.25 ATM • .e- 0.5 ATM

1.0 ATM -e- 2.0 ATM

0.06 3.0 ATM

E Y

0.04 / ~ 4

\, 0.02

0 . 0 0 , I , I , I ' ' 0 . 4 0 . 6 0 . 8 1 .0 1 .2 1 .4 1 .6

@ Fig. 6. Experimental laminar mass burning rates m°(~, p) for pressures between 0.25 and 3 a r m .

50

4 0

~ 3O o . I

/ t "x% / /~Z • % / %

/ % /

2O " - Ba n d o f S e le c te d Lite ra ture Va lue s • P re s e n t Expe rime n ta l Da ta at P re s e n t Nume rica l Da ta

10 ' ' ' ' ' 0.6 0.7 0.8 " 0.9 1.0 1.1

@

C H 4 / AIR p = 1 AT M

l i I .

1.2 1.3 1.4 1.5

Fig. 7. Compa ris on be twe e n pre s e nt e xpe rime nta l a nd nume rica l S °(~) for me tha ne -a ir fla me s , a t p = 1 a tm, with the "b a n d " o f va lue s re porte d in the lite ra ture a nd nume rica l da ta ca lcula ted by us ing the Ke e e t a l.-C~ s che me [15].

Page 7: Laminar Flame Speeds of Me thane -Air Mixture s Under Reduced …ronney.usc.edu/AME513b/Lecture4/References/... · 2020. 2. 28. · Laminar Flame Speeds of Me thane -Air Mixture s

LAMINAR F LAME S P EEDS 381

50

m Garforth/Rallis • • Babkin et al.

• Andrews/Bradley 40 ~ , ~ • lijima/Takeno lit ~ 4 Present Data

~ 30 ~ • ~ • 0%

20 • C H 4 / AIR

• = 1.0 10 , I , I , I ,

0 1 2 3 Pressure , Atm

Fig. 8. C o m p a ris o n be twe e n p re s e n t e xpe rime n ta l S ° fo r s to ich iome tric m e th a n e -a ir fla m e s , with s e le c te d va lue s fro m the lite ra ture , a t d iffe re n t p re s s u re s .

Flame Code

For the fla me code we choos e the one de ve lope d by Ke e a nd co-worke rs [13-16], which is ge ne r- a lly a cce pte d to be a mong the mos t a ccura te a nd e fficie nt. We ha ve , howe ve r, ma de one modifica - tion in the de s cription o f the tra ns port prope rtie s . Tha t is , by a s s uming tra ce s pe cie s diffus ion, Re f. 14 a pproxima te s the ma s s diffus ion coe ffic ie nt o f s pe cie s k a s

1 -Yk Dk-- N

j~k

a nd its ma s s diffus ion ve locity a s

Vk= - D__~ VXk, Xk

whe re Xk a nd Yk a re , re s pe ctive ly, the mola r a nd ma s s fra ctions o f k, Djk the ma s s diffus ivity be twe e n j a nd k, a nd N the tota l numbe r o f s pe cie s .

Be ca us e Eq. 1 is not e xa ct, ma s s cons e rva tion

N

I

is not s a tis fie d. The re fore in orde r to s a tis fy Eq. 3, Vk a s give n by Eq. 2 is corre cte d [14] by a n a mount e qua l to - Z IN Vk Yr.

In obta ining the fina l re s ults in the pre s e nt ca lcula tions , we dire ctly us e the more rigorous re s ult for multicompone nt diffus ion [18]

N

= 1, 2 , . . . , N , (4 ) (1)

whe re diffus ion due to pre s s ure gra die nt, body force , a nd S ore t e ffe c t a re ne gle cte d. The N unknowns Vk ca n be s olve d from Eq. 3 a nd the N - 1 line a r e qua tions give n by Eq. 4. Ove ra ll diffus ive ma s s flux cons e rva tion, Eq. 3, is a uto- ma tica lly s a tis fie d.

(2) In a ctua l ca lcula tions we firs t s olve the proble m by us ing the tra ce s pe cie s formula tion. This s olution is the n us e d to s ta rt the ca lcula tion with the rigorous d iffus ion formula tion. For a s toi- chiome tric me tha ne -a ir fla me a t 1 a tm, us ing a C1 kine tic s che me with 17 s pe cie s a nd 58 re a ctions [15], the CP U time for the comple te double pre cis ion s olution is a pproxima te ly 130 min on a VAX 11/785 compute r. The a dditiona l CP U time

(3) to a llow for the e xa ct d iffus ion formula tion is a bout 15 ra in.

Page 8: Laminar Flame Speeds of Me thane -Air Mixture s Under Reduced …ronney.usc.edu/AME513b/Lecture4/References/... · 2020. 2. 28. · Laminar Flame Speeds of Me thane -Air Mixture s

382 F. N. EGOLFOP OULOS

For the fla me s pe e ds ca lcula te d in the pre s e nt s tudy, the rigorous diffus ivity formula tion yie lds only a ma rgina l diffe re nce from thos e obta ine d by a s s uming tra ce s pe cie s diffus ion. Furthe rmore , the diffe re nce is within the 0 .5 -2 cm/s unce rta inty a s s ocia te d with the e xpe rime nta l da ta . Thus this rigorous formula tion is a ctua lly not ne e de d for the pre s e nt ca lcula tion. It is dis cus s e d he re be ca us e o f its pote ntia l utility for mixture s with h ighe r re a c- ta nt conce ntra tions .

It is importa nt to point out tha t we ha ve us e d the s a me tra ns port prope rtie s code a nd fla me code in the following compa ris ons o f the va rious kine tic s che me s . Thus diffe re nce s in the ca lcula tions a re s ole ly cons e que nce s o f the diffe re nt kine tic s che me s us e d. S ys te ma tic compa ris ons o f this na ture ha ve a ls o be e n conducte d by Coffe e a nd He ime rl [19] for the tra ns port a lgorithms a nd by Coffe e [20] for the kine tic me cha nis ms . The pre s e nt s tudy provide s a dditiona l contributions in te rms o f nume rica l e va lua tions o f the kine tic s che me s ove r e xte ns ive ra nge s o f conce ntra tion a nd pre s s ure a nd compa ris ons o f the ca lcula te d re s ults with the pre s e nt e xpe rime nta l da ta .

Kinetic S che me s

For me tha ne oxida tion, the C2 che mis try which include s the C1 re a ctions is s ufficie nt for comple te mode ling [7, 17, 21, 22]. The C1 pa th is the principa l cha nne l for fue l-le a n fla me s , whe re a s the comple te C2 che mis try is e xpe cte d to be ne e de d to de s cribe fue l-rich fla me s . In the pre s e nt inve s tiga - tion we ha ve te s te d quite a numbe r o f publis he d kine tic s che me s a ga ins t our da ta a s we ll a s a ga ins t e a ch othe r. The s e compa ris ons a re pre s e nte d be low. Be fore doing s o, howe ve r, it ma y be e mpha s ize d a t this point tha t be ca us e the fla me s pe e d is only a bulk prope rty o f the la mina r pre mixe d fla me , a gre e me nt be twe e n the pre s e nt da ta a nd ca lcula te d va lue s a s s uming a ce rta in kine tic s che me is only a ne ce s s a ry but not s uffic- ie nt re quire me nt for the va lida tion o f the s che me . A comple te va lida tion would re quire a gre e me nt in the me a s ure d a nd compute d te mpe ra ture a nd s pe cie s profile s a s we ll. The pre s e nt compa ris on, howe ve r, is s till quite us e ful be ca us e o f the following re a s ons : (1) Be ca us e a gre e me nt in the

la mina r fla me s pe e d is the min imum re quire me nt o f a propos e d s che me , the la ck o f a gre e me nt is the n a cle a r indica tion tha t the s che me re quire s re vis ion. (2) Be ca us e o f the a ccura cy o f the pre s e nt e xpe rime nta l da ta , a tighte r re quire me nt in the de gre e o f a gre e me nt ca n be impos e d, the re by le nding gre a te r confide nce to the va lida tion. (3) Accura te e xpe rime nta l de te rmina tions o f the te m- pe ra ture a nd conce ntra tion profile s for adiabatic uns tre tche d fla me s a re a s ye t una va ila ble . Be fore the s e da ta be come a va ila ble , va lida tion through fla me s pe e d compa ris on is a via ble inte rim s olu- tion.

We now pre s e nt the compa ris on be twe e n our e xpe rime nta l da ta a nd the compute d one s a s s um- ing va rious kine tic s che me s . Mos t o f the s e s che me s a re ta bula te d in the ir re s pe ctive re fe r- e nce s a nd the re fore a re not s e pa ra te ly lis te d he re .

C1 S che me o f Ke e e t ai [15[ Figure 9 compa re s the e xpe rime nta l fla me s pe e ds with ca lcula te d va lue s us ing this s che me , which include s only the C1 me cha nis m with 17 s pe cie s a nd 58 re a ctions . Ca lcula tion wa s te rmina te d whe n conve rge nce be ca me e xce s s ive ly s low, typi- ca lly for high-pre s s ure a nd low-te mpe ra ture fla me s . The e xpe rime nta l fla me s pe e ds a re s hown a s dotte d line s which a re dire ct re productions o f the s olid line s in Fig. 5. It is s e e n tha t clos e a gre e me nt e xis ts , to within the e xpe rime nta l un- ce rta inty, for a ll s itua tions compa re d, including a ll le a n fla me s a s we ll a s rich fla me s up to e quiva - le nce ra tios o f a bout 1.2 to 1.3. Although the clos e a gre e me nt with the rich fla me s ma y be s ome wha t s urpris ing cons ide ring tha t only the C1 me cha nis m is us e d, it ha s be e n s ugge s te d [23] tha t the a dditiona l C2 re a ctions ma y not be too importa nt for the de te rmina tion o f me tha ne -a ir fla me s pe e ds for e quiva le nce ra tio a s high a s 1.3.

C1 S che me o f Pe te rs and Ke e I241 This is a re duce d Cl s che me tha t include s 14 s pe cie s a nd 18 re a ctions . Figure s 10 a nd 11, re s pe ctive ly, compa re the fla me s pe e d a s a func- tion o f e quiva le nce ra tio a t 1 a tm, a nd a s a function o f pre s s ure a t ¢ = 1. It is s e e n tha t this s che me uniformly ove rpre dicts the la mina r fla me s pe e d by a s ubs ta ntia l a mount.

Page 9: Laminar Flame Speeds of Me thane -Air Mixture s Under Reduced …ronney.usc.edu/AME513b/Lecture4/References/... · 2020. 2. 28. · Laminar Flame Speeds of Me thane -Air Mixture s

LAMINAR FLAME S PEEDS 383

70

I

60 ~

50

40

¢J

r,/3 30

20

10

- - - - EXPERIMENT

qUMERICAL (KEE ET AL.-C1) l a

P /~Z , f ,~ / / , f

/ - ~ ' . , # //g

0.25 ATM N 0.5 ATM / \ 1.0 ATM / \ • \ 2.0 ATM ,/ 3.0 ATM / / \ \

/ / ~ - ~ , \ • ¶ '~ / / 1 4 f \ \

/ / - / \ \ = / ,'= \ \

I \ \ / / ~--* \ s / ,s ~ ~ \ / , / / \ " \

/ / ~/ \ \ k \

/ , / / j , ~ - s - . x \ / ,m ,,,/ ~ " \ X \

• "/ / / / / :\ \\ X / / s \ \ \ = \

/ , ~ / \ \ \

0 , I i I i I t I , I i

0,4 0.6 0.8 1,0 1.2 1.4 1.6

Fig. 9. Compa ris on be twe e n e xpe rime nta lly a nd nume rica lly de te rmine d S°L(~, p ) for me tha ne -a ir mixture s , us ing the C, kine tic s che me of Kee e t a l. [15].

60

5O

4O

~ 30

20

10

0 0.4

-- -- EXPERIMENT CH 4 / A I R • PETERS/KEE-C1 • • • p = l ATM • ECL-Cl(19) • • ECL-C1(23)

* ..4r~ - - i t . . / 0 0 "~ l

v\ /

s \ \ / o \ \

/1l' N \ /

I i I I i I

0.6 0.8 1.0 1.2 1.4 1,6

Fig. 10. Compa ris on be twe e n e xpe rime nta lly a nd nume rica lly de te rmine d S°L(~b) for me th- a ne -a ir mixture s a tp = 1 a tm, us ing P e te rs -K•• [24] a nd the pre s e nt ECL-CI(19) a nd ECL- Cj (23) kine tic s che me s .

Page 10: Laminar Flame Speeds of Me thane -Air Mixture s Under Reduced …ronney.usc.edu/AME513b/Lecture4/References/... · 2020. 2. 28. · Laminar Flame Speeds of Me thane -Air Mixture s

384 F . N . EGOLFOP OULOS

° ~

70

60

50

40

30

20

10

- - - EXPERIMENT ,D PETERS/KEE-Cl

\ . • ECL-C1(19} o \ • ECL-Cl(23)

IN ,o. \ ~'

\ •

O, ~ ,O

C H 4 / AIR . @ =1.0 *

0 1 2 3 4 Pressure , A~

Fig. 11. Compa ris on be twe e n e xpe rime nta lly a nd nume rica lly de te rmine d S o for s toichiome - tric me tha ne -a ir mixture s a t d iffe re nt p re s s ure s , us ing P e te rs -Ke e [24] a nd the pre s e n t ECL- Ci (19) a nd ECL-C1 (23) kine tic s che me s .

Pre s e nt C~ S che me s In orde r to e xpla in the la ck o f a gre e me nt for the s che me o f P e te rs a nd Ke e [24], a s e ns itivity a na lys is wa s pe rforme d us ing the C~ s che me o f Ke e e t a l. [15]. The re s ults s how tha t the ra dica l re combina tion re a ction

CH3 + H + M~ C H4 + M,

which is ne gle cte d in the s che me o f P e te rs a nd Ke e [24], is a ctua lly o f le a ding orde r importa nce . Its omis s ion ca n a ffe ct the fla me s pe e d by a s much a s 8 cm/s for s toichiome tric me tha ne -a ir fla me s a t 1 a tm. Figure s 10 a nd 11 s how tha t much improve - me nt re s ults by including this s te p. This s che me o f 14 s pe cie s a nd 19 s te ps , de s igna te d a s ECL- C~ (19), is lis te d a s the re a ctions without a s te ris ks in Ta ble 1.

The s e ns itivity a na lys is furthe r s hows tha t the re a re four a dditiona l re a ctions o f s e conda ry impor- ta nce not include d in ECL-CI(19). By including the m, Figs . 10 a nd 11 s how ve ry clos e a gre e me nt with the e xpe rime nta l da ta . This s che me o f 23 s te ps with 15 s pe cie s , de s igna te d a s ECL-C1 (23), is lis te d a s re a ctions 1 through 23 in Ta ble 1.

Although ECL-C1 (23) s e e ms to provide a clos e r a gre e me nt with the e xpe rime nta l da ta tha n ECL- C~(19), we ca nnot conclude tha t it is a be tte r s che me . This is be ca us e o f the s ma ll unce rta intie s

tha t s till e xis t in our e xpe rime nta l da ta a s we ll a s the C 1 s che me o f Ke e e t a l. [15] ba s e d on which the s e ns itivity a na lys is is pe rforme d. Tha t is , if the e xpe rime nta l fla me s pe e ds a round s toichiome try we re lowe r tha n our re porte d va lue s by, s a y, 2 cm/ s , which is s tre tching our unce rta inty limit but is ne ve rthe le s s a rgua bly pos s ible , the n the e xpe ri- me nta l da ta would mos tly lie e ithe r midwa y be twe e n ca lcula te d va lue s give n by the s e two s che me s or a ctua lly be clos e r to thos e o f the ECL- C1 (19) s che me .

C2 S che me o f Glarborg e t ai. [251 This s che me , de s igna te d a s Gla rborg e t a l.-C2, cons ide rs 26 s pe cie s with 123 re a ctions . Figure 12 s hows tha t a t 1 a tm it pre dicts the le a n fla me s we ll but ove rpre dicts the rich fla me s . Figure 13 s hows tha t the pre dicte d va lue s ha ve lowe r fla me s pe e ds unde r lowe r pre s s ure s but highe r fla me s pe e ds unde r highe r pre s s ure s . This diffe re nce is be lie ve d [23] to be ca us e d by the us e o f a s ome wha t s lowe r ra te for the fa llo ff be ha vior

CH3 + CH3 + M-*C2H6 + M.

Modifica tions a re curre ntly be ing imple me nte d by Mille r [23] to re ctify this e ffe ct.

Figure 12 a ls o s hows the ca lcula te d fla me s pe e ds us ing only the C1 pa rt o f the full C2 s che me , de s igna te d a s Gla rborg e t a l.-C~. Al-

Page 11: Laminar Flame Speeds of Me thane -Air Mixture s Under Reduced …ronney.usc.edu/AME513b/Lecture4/References/... · 2020. 2. 28. · Laminar Flame Speeds of Me thane -Air Mixture s

LAMINAR FLAME S PEEDS

TABLE 1

E C L-C I (19) An d ECL-CI (23) S che me s °

385

Re a c tion .4 /~ Eo

1. CH3 + H + M = C I-h + M 8 .0E26 - 3 , 0 0. 2 . CI-L + H = CH3 + H2 2.2E,4 3 .0 8750.

*3. Cl'Lt + O = CH3 + O H 1.6E6 2 .36 7400. 4. CI'L + O H = CH3 + H2 0 1 .6E6 2.1 2460. 5. C Ha + O - C H 2 0 + H 6 .8E13 0 .0 0.

*6. CHa + O H = CH2 + H2 0 1.5E13 0 .0 5000. *7. C H 2 + O 2 = C O 2 + H + H 1.6E12 0 .0 1000. *8. C H 2 + 02 = CO2 + H2 6 .9E11 0 .0 500.

9. CH~O + H = HC O + H2 2 .1 9 E 8 1.77 3000, 10. C H2 0 + O H = HC O + H2 0 3 .43E9 l. 18 - 477. 11. H C O + H = C O + H 2 4 .00E13 0 ,0 0, 12. H C O + M= H + C O + M 1 .6 0 E I4 0 ,0 14700. 13. HC O + O 2 =HO 2 + C O 3 .3E13 - 0 , 4 0, 14. C O + O H = C O 2 + H 1 .51E7 1,3 - 758. 15. H + O 2 = O H + O 5 .1 3 E l6 - 0 , 8 1 6 16507. 16. O + H 2 = O H + H 1.8E10 1 .0 8826. 17. O H + H2 ~ H2 0 + H 1 .17E9 1.3 3626. 18. O H + O H = H 2 0 + O 6 .0E8 1.3 0. 19. H + O 2 + M= H O 2 + M b 3 .61E17 - 0 . 7 2 0, 20 . H + O H + M= H 2 0 + M c 1 .6E22 - 2 . 0 0. 21. H + H O e ~ 2 O H 1.4E14 0 ,0 1073. 22 . H + H O 2 ~ H 2 + O 2 1.25E13 0 ,0 0. 23 . O H + HO2 = HzO + 02 7 .5E12 0 ,0 0.

o Re a c tion m e c h a n is m ra te coe ffic ie n ts in the fo rm k~ = AT# e xp(-E,/R T) (units a re mo le s , cub ic ce n time te rs , s e conds , Ke lvins a n d ca lorie s /too l, ta ke n fro m Ke e e t a l. [15]). An a s te ris k indica te s the re a c tion is no t inc lude d in the ECL-C~ (19) s c h e m e .

b Th ird -body e ffic ie nc ie s : k(02) = 1 ,2 6 k(Ar), k(N:) = 1 .2 6 k(Ar), k(H:O ) = 1 8 .6 k(Ar), k(C O ) = 2 .1 l k(Ar), k(CO2) = 4 .2 k(Ar), k(H2) = 2 ,8 6 k(Ar).

c Th ird body e ffic ie nc ie s : k(O2) = k(N2) = k(C O ) = k(COz) = k(Hz) = 1, k(H2 0 ) = 5 .0 k(Ar).

60

50

4O

30 o~

20

10

0 L m ~ I _ •

0.4 0.6

- - - - EXPERIMENT • GLARBORG et aL-C2 o GLARBORG et al.-C1 • COFFEE.C1

/ ,11 /

/ 0

\ • ta a \ '~

a N A

CH 4 / AIR p= ! AT M

\\m a\

"x \

I . . _1 . . . . | • I

0.8 1 .O 1.2 1.4 1.6

Fig. 12. C o m p a ris o n be twe e n e xpe rime nta lly a nd nume rica lly de te rmine d S °t(~) fo r me th - a n e -a ir m ixtu re s at p = 1 a tm , u s in g the C2 a nd C~ kine tic s c h e m e s o f Gla rborg e t a l. [25] a nd the C~ kine tic s c h e m e o f Coffe e [20].

Page 12: Laminar Flame Speeds of Me thane -Air Mixture s Under Reduced …ronney.usc.edu/AME513b/Lecture4/References/... · 2020. 2. 28. · Laminar Flame Speeds of Me thane -Air Mixture s

386 F . N . E G O LF O P O ULO S

70

60

50

• 4o

r./3 30

20

10

\ \

• \ \ , , ,q ,

CH 4 / AIR ~ = 1.0

-- - - EXPERIMENT • GLARBORG et al.-C2 m COFFEE-C1

, I t I i I i

0 1 2 3 4 Pressure, A i m

Fig. 13. Comparison between experimentally and numerically determined S ° for stoichiome- tric methane-air mixtures at different pressures, using the C2 kinetic scheme of Glarborg e t al. [25] and the C, kinetic scheme of Coffee [20].

though ide a lly one would e xpe c t tha t the ca lcula te d re s ults s hould a gre e with thos e obta ine d from the C~ s che me o fKe e e t a l. [15] s hown in Fig. 9 , Fig. 12 s hows tha t the y a re un iformly lowe r tha n the pre vious one s o ve r the e ntire ra nge o f s to ichiome - try compa re d.

In o rde r to ide ntify the ca us e fo r s uch a d iffe re nce , Fig. 14 compa re s the individua l re a c- tion ra te s from the two s che me s tha t s how s ome diffe re nce , with kc2 a nd kc l, re s pe ctive ly, de s ig- na ting the s che me s o f Gla rborg e t a l. [25] a nd Ke e e t a l. [15]. F igure 15 s hows the s e ns itivity o f the s e individua l re a ction ra te s on the fla me s pe e d. The compa ris on s hows tha t pe rha ps e xce p t fo r the

C H2 0 + H - ' H C O + H2 0

re a ction, the re is no s ingle re a ction tha t is dra s ti- ca lly d iffe re nt in the two s che me s . Thus the d iffe re nt fla me s pe e ds pre dic te d a re the cumula - tive d iffe re nce s from a numbe r o f re a ctions .

The a bove re s ult illus tra te s the s ubtle ty o f code ca libra tion. Tha t is , in de ve loping kine tic s che me s , it is a fre que nt pra c tice to a djus t s ome kine tic cons ta nts s uch tha t the ca lcula te d re s ults a gre e with the e xpe rime nta l da ta a t one point, typica lly the s ta te o f s to ichiome try. The re fo re if ca libra tion ha s be e n s e pa ra te ly conducte d fo r the C~ a nd the full C2 s che me s , it is the n re a s ona ble to

e xpe ct tha t re s ults from the C1 pa rt o f the (?2 S che me s hould de via te from the ca libra tion point.

C2 S che me o f Coffe e [2011 This s che me cons ide rs 21 s pe cie s with 57 re a c- tions . A de ta ile d compa ris on ha s not be e n con- ducte d, a lthough a s ingle point che ck a t q~ = 1 s hows a ca lcula te d fla me s pe e d o f 45 .7 e ra /s , which is cons ide re d to be high re la tive to the e xpe rime nta l va lue o f 40 cm/s . Figure s 12 a nd 13 s how compa ris ons by us ing the C1 pa rt o f this s che me . Good a gre e me nt is obta ine d up to the s ta te o f s to ichiome try. F o r riche r fla me s Fig. 12 s hows h ighe r ca lcula te d va lue s a t 1 a ttn.

Although both the C1 s che me s o fKe e e t a l. [15] a nd o f Coffe e [20] yie ld good a gre e me nt fo r le a n fla me s , Co ffe e 's s che me give s h ighe r va lue s fo r rich fla me s . A de ta ile d s e ns itivity s tudy s hows tha t this d iffe re nce is ca us e d by the s lightly d iffe re nt ra te s o f the crucia l re a c tion (R2).

We note in pa s s ing tha t ou r ca lcula te d re s ults a re cons is te ntly h ighe r tha n Co ffe e 's publis he d va lue s . Thus his C2 ca lcula tion a ctua lly yie lds a fla me s pe e d o f 40 cm/s a t ~b = 1, which a gre e s with the e xpe rime nta l va lue . Be ca us e both his

The value of Eo/R for the reaction CH3 + H = CH2 + H2 in Ref. 20 should be + 1500 K instead of - 1500 K; this is only a misprint.

Page 13: Laminar Flame Speeds of Me thane -Air Mixture s Under Reduced …ronney.usc.edu/AME513b/Lecture4/References/... · 2020. 2. 28. · Laminar Flame Speeds of Me thane -Air Mixture s

LAMINAR F LAME S P EEDS 387

1.0

0.5

0.0

-0.5

-1.0

-1.5

-2.0

1.0

0.5

0.0

-0.5

-1.0

-1.5

-2.0 1200

• • • • • • • • • • • • • • • • m [] • • •

.ii,,i; ||lllll

• H+OH+M,=H20+M • H+O2+M,.HO2+M x HO2+H=H2+O2 • HO244,,OA+OH • HO~+O-OH*O2

i I m I i I i I i

: : : : : ; : ; : | l l l! .

÷ 4- 4- 4- 4- 4- 4- 4" 4- 4- 4- 4- 4- 4- 4-

t I m I t 1400 1600

T, K

- ~ + H , ÷

1800 2000 2200

F ig . 14 . C o m p a r is o n b e twe e n th e s p e c ific re a c t io n ra te s , k2 a n d k l, o f s e le c te d C i- im p o r ta n t re a c t io n s fr o m th e kin e tic s c h e m e s o f G la r b o r g e t a l. [2 5 ] a n d Ke e e t a l. [1 5 ], re s p e c t ive ly.

tra ns port a lgorithm a nd fla me code a re be lie ve d [26] to be a ccura te , the ca us e o f this dis a gre e me nt is not known a t pre s e nt.

C2 Scheme o f Warnatz 122] This s che me cons ide rs 26 s pe cie s with 74 re a c- tions . Figure s 16 a nd 17 s how compa ris ons for the full (?2 s che me a s we ll a s its Ci compone nt. The compa ris on ca n be cons ide re d s a tis fa ctory for the following re a s ons : (a ) The (2?2 s che me a gre e s we ll ove r the e ntire ra nge o f s toichiome try compute d, with the la rge s t d iffe re nce occurring a round s toi- chiome try. Co) The pre s s ure de pe nde nce o f the ca lcula te d va lue s a gre e we ll with the e xpe rime nta l tre nd. (c) Contra ry to the (:72 me cha nis m o f Gla rborg e t a l. [25], pre dictions o f the le a n fla me s by us ing the Cx compone nt o f Wa rna tz 's C2 s che me a gre e we ll with thos e obta ine d from the

full (?2 s che me a s we ll a s the e xpe rime nta l da ta , indica ting inte rna l cons is te ncy o f the s che me .

Present C2 Scheme A s e ns itivity a na lys is o f Wa rna tz 's C2 s che me a llows us to propos e a C2 s che me with 22 s pe cie s a nd 40 re a ctions , a s lis te d in Ta ble 2. Figure s 16 a nd 17 s how tha t the ca lcula te d va lue s a gre e we ll with the e xpe rime nta l da ta . Compa re d with the C2 s che me o f Wa rna tz [22], the ca lcula tion time -for the pre s e nt s che me is re duce d.

C2 Scheme o f We s tbrook [27] This s che me cons ide rs 24 s pe cie s with 74 re a c- tions . The ca lcula te d fla me s pe e d for the s toi- chiome tric me tha ne -a ir fla me a t 1 a t • is 27.5 cm/ s . Cons ide ring this va lue to be low, we did no t

Page 14: Laminar Flame Speeds of Me thane -Air Mixture s Under Reduced …ronney.usc.edu/AME513b/Lecture4/References/... · 2020. 2. 28. · Laminar Flame Speeds of Me thane -Air Mixture s

388 F . N . E G O LF O P O ULO S

CISCHEMEOFKEEETAL. CH4/AIR

O=I.0 p=IATM

HCO-tO--CO2+H I CH20+H=HCO+H2 I

CH4+OH=CH3+H201 HO2+OH=H20¢O21

HO2+O=OH+O2 I

HO2+H=H2-¢O2 I

I CH2+O2=CH20+O

-0.100 -0.075 -0.050 -0.025 0.000 0.025 0.050 Fig. 15. Normalized sensitivity on the flame speeds of the reactions shown in Fig. 14.

continue ca lcula tions a t d iffe re n t conce ntra tions a nd pre s s ure s .

C2 S c h e m e o f P itz and We s tbrook [28]

By e xcluding fro m the lis te d s che me re a c tions involving CH3OH, CH3CHO, CH2CO, a nd HC C O , which a re be lie ve d [22] to be min ima lly importa n t fo r fla me s pe e d ca lcula tions , the re - duce d s che me cons ide rs 24 s pe cie s a nd 82 re a c- tions . The ca lcula te d fla me s pe e d o f s to ich iome tric m e th a n e -a ir mixtu re a t 1 a tm is 23 .0 cm/s , which re duce s to 21 .0 c m /s whe n on ly the C1 pa rt o f the s che me is us e d.

We note tha t the ca lcula te d fla me s pe e d o f s to ichiome tric m e th a n e -a ir mixtu re a t 1 a tm b y We s tb rook [29] is 38 .0 cm/s , which is cons ide ra - b ly h ighe r tha n the va lue s ca lcula te d he re . Be ca us e the s a me kine tic s che me is us e d, the d iffe re nce is be lie ve d [30, 31] to be ca us e d b y the d iffe re nt tra ns port code s us e d in Re fs . 2 7 -2 9 a nd the pre s e nt ca lcula tions .

C O N C LU D IN G R E MAR KS

The pre s e nt s tudy ha s de mons tra te d the impor- ta nce o f a ccura te de te rmina tion o f the la mina r

¢/3

o ~

6O

5O

4O

30

2O

10

0 0.4

- - - - EXPERIMENT CH 4 / A I R a WARNARTZ-C2 p = 1 A T M + WARNARTZ-C1 x ECL-C2(40)

/ ÷ \ ~

+ \+ /

I I I I I I

0.6 0.8 1.0 1.2 1.4 1.6

Fig. 16. Comparison between experimentally and numerically determined S t(O) for meth- ane-a ir mixtures a t p = 1 a tm, us ing the C2 and C1 kinetic schemes of Warnatz [22], and the present ECL-C2(40) kinetic scheme.

Page 15: Laminar Flame Speeds of Me thane -Air Mixture s Under Reduced …ronney.usc.edu/AME513b/Lecture4/References/... · 2020. 2. 28. · Laminar Flame Speeds of Me thane -Air Mixture s

LAMINAR FLAME S PEEDS 389

TABLE 2

ECL-C2 (40) S che me °

Re a ction A /3 Eo

1. H + O 2 + O H + O 1.20E17 - 0 . 9 1650~ 2. O + H2 = OH + H 1.50E07 2.0 7548. 3. OH + H2 = H20 + H 1.00E08 1.6 3296. 4. H + H + M= H 2 + M b 9.70E16 - 0 . 6 0. 5. H + O H + M= H 2 0 + M ~' 2.20E22 - 2 . 0 0. 6. H + O 2 + M= H O 2 + M b 2.00E18 --0 .8 0. 7. H + H O 2 = O H + O H 1.50E14 0 .0 1003. 8. H + HO2 = H2 + 02 2.50E13 0.0 693. 9. OH + HO2= H20 + O2 2.00E13 0 .0 0.

10. C O + H + M= H C O + M b 6.90E14 0 .0 1672. 1 l. C O + O + M= C O 2 + M b 7.10E13 0 .0 -4 5 3 8 . 12. CO + OH = CO2 + H 4.40E06 1.5 - 740.4 13. CI-L + H = CH3 + H2 2.20E04 3.0 8742. 14. CI'L + O = CH3 + OH 1.20E07 2.1 7619. 15. CI-L + OH = CH3+ H20 1.60E06 2.1 2460. 16. C H3 +H=C I-L c 6.00E16 - 1.0 0. 17. CH3 + O = C H2 0 + H 7.00E13 0 .0 0. 18. CH3 + 02 = CH20 + H + O 1.50E13 0 .0 28662. 19. CH20 + H = HCO + H2 2.50E13 0.0 3989. 20. CH20 + OH = HCO + H20 3.00E13 0 .0 1194. 21. H C O + H = C O + H 2 2.00E14 0 .0 0. 22. H C O + O = C O + O H 3.00E13 0 .0 0. 23. H C O + O = C O 2 + H 3.00E13 0 .0 0. 24. HCO + OH = CO + H20 5.00E13 0.0 0. 25. HCO + 02 = CO + HO2 3.00E12 0 .0 0. 26. CH2 + 02 = CO2 + H + H 1.30E13 0 .0 1505. 27. C H + O = C O + H 4.00E13 0 .0 0. 28. CH3 + CH3 = C2H6 c 2.40E14 - 0 .4 0. 29. CH3 + CH3 = C2H5 + H 8.00E14 0 .0 26512. 30. CH3 + CH3 = C21"L + H2 1 .00El6 0.0 32005. 31. C2H6 + H = C2Hs + H2 5.40E02 3.5 5207. 32. C2H6+ OH = C2H5 + H20 6.30E06 2 .0 645. 33. C2H~ = C2I~ + H c 2.00E13 0 .0 39648. 34. C2I'L + OH = C2H3 + H20 3.00E13 0 .0 2986. 35. C 2 H3 +H=C 2 H2 +H2 2.00E13 0 .0 0. 36. C2H3 + O2 = C2H2 + HO2 1.00El2 0 .0 0. 37. C2H3 =C 2 H2 +H c 1.60E14 0 .0 37976. 38. C2H2 + O = CH2 + CO 4.10E08 1.5 1696. 39. C2H2 + O = HCCO + H 4.30E14 0 .0 12110. 40. HC C O + H = C H2 + C O 3.00E13 0.0 0.

a Re a ction me cha nis m ra te coe fficie nts in the fo rm k~ = A T ~ e x p ( - E . / R T ) (units a re mole s , cubic ce ntime te rs , s e conds , Ke lvins a nd ca lorie s /tool, ta ke n from Wa rna tz [22]).

b Third body e fficiencies : k(O2) = 0.4k(H2), k(N2) = 0.4k(H2), k(H20), = 6.5k(H2), k(CO) = 0.75k(H2), k(CO2) = 1.5k(H2).

c High pre s s ure va lue ; corre ction for fa ll-off be ha vior mus t be included.

Page 16: Laminar Flame Speeds of Me thane -Air Mixture s Under Reduced …ronney.usc.edu/AME513b/Lecture4/References/... · 2020. 2. 28. · Laminar Flame Speeds of Me thane -Air Mixture s

390 F . N . E G O LF O P O ULO S

70

60

50 t~

, o © ~ -,,

30

20

10

\ \

\

÷

- - - EXPERIMENT " WARNARTZ-C2 + WARNART'Z-C1 x ECL-C2(40)

CH 4 / AIR + @ = 1.0 +

i l m I m I

I 2 3 Pressure, Arm

Fig. 17. Compa ris on be twe e n e xpe rime nta lly a nd nume rica lly de te rmine d S O for s toichiome - tric me tha ne -a ir mixture s a t diffe re nt pre s s ure s , us ing the C2 a nd CI kine tic s che me s o f Wa rna tz [22], a nd the pre s e nt ECL-C2(40) kine tic s che me .

fla me s pe e ds fo r the (pa rtia l) va lida tion o f kine tic s che me s . It is c le a r tha t critica l compa ris ons o f the va rious kine tic s che me s could not ha ve be e n conducte d without a s s uring tha t on ly s ma ll unce r- ta intie s a re a s s ocia te d with the e xpe rime nta l da ta . Cons e que ntly, ce rta in s che me s ha ve e me rge d to be more pre dic tive tha n othe rs a s fa r a s the fla me s pe e d is conce rne d. S pe cifica lly, the C1 s che me o f Ke e e t a l. [15] a nd the C2 s che me o f Wa rna tz [22] both a ccura te ly re produce ou r e xpe rime nta l da ta ove r the e xte ns ive conce ntra tion a nd pre s s ure ra nge s te s te d.

The C~ s che me o f Ke e e t a l. indica te s tha t re a ctions involving C2 s pe cie s ma y not be impor- ta nt fo r fla me s a s rich a s ~b ~ 1 .2 -1 .3 a nd pre s s ure s a s high a s 3 a tm, while the C2 s che me o f Wa rna tz s hows tha t the y do ha ve s ubs ta ntia l influe nce e ve n fo r the s to ichiome tric , a tmos phe ric fla me . This d iffe re nce ca nnot be e a s ily re s o lve d be ca us e it is pos s ible tha t the C~ s che me o f Ke e e t a l. ca n be e xpa nde d to a n improve d , full C2 s che me tha t inde e d doe s not s how s ignifica nt influe nce o f the C2 re a ctions up to , s a y, ~b ~- 1.3. Furthe r cons tra ints in the compa ris ons be twe e n the e xpe rime nta l da ta a nd the fla me code ne e d to be de ve lope d to re s olve this s ma ll, but importa nt, d iffe re nce .

Fina lly, we note tha t a s fa r a s fla me s pe e d ca lcula tion is conce rne d , the Cl s che me s (Ta ble 1) propos e d he re in a re e ffic ie nt a nd a ccura te . Fur-

the rmore , a n a pproxima te fla me s tructure ca n a ls o be obta ine d by us ing the (272 s che me (Ta ble 2) de ve lope d in this s tudy.

FNE and P C were s upporte d by the A ir Force Office o f S cientific Research unde r the technical m anage m e nt o f Dr. J. M. T is hko ff and CKL was s upporte d by the Office o f Bas ic Ene rgy S ciences o f the De partm e nt o f Ene rgy unde r the technical m anage m e nt o f Dr. J. Welty. Additional com pute r tim e was provide d by the N S F S upe rcom pute r Center at the Univers ity o f California at S an Diego. W e appreciate the technical discuss ions with Drs . T. P. Coffe e , R . J. Kee , J. A . Miller, and C. K. W e s tbrook on this proble m .

REFERENCES

1. Andre ws , G. E., a nd Bra dle y, D., Com bus t. Flam e 19:275 (1972).

2. Yamaoka , I., a nd Ts uji, H., Twe ntie th S y m p o s iu m (Inte rnational) on Com bus tion , The Combus tion Ins ti- tute , P itts burgh, 1984, p. 1883.

3. Iijima , T., a nd Ta ke no, T., Com bus t. Flam e 65:35 (1986).

4. Ma ta lon, M., a nd Ma tkows ky, B. J., J. Flu id Me ch. 124:239 (1982).

5. Law, C. K., Twe n ty -S e cond S ym pos ium (Inte rna- tional) on Com bus tion , The Combus tion Ins titute , P itts burgh, in pre s s .

Page 17: Laminar Flame Speeds of Me thane -Air Mixture s Under Reduced …ronney.usc.edu/AME513b/Lecture4/References/... · 2020. 2. 28. · Laminar Flame Speeds of Me thane -Air Mixture s

LAMIN AR F LAME S P E E D S 391

6. Wes tbrook, C. K., and Drye r, F. L., Eighte e nth S ym pos ium (International) on Com bus tion, The Combustion Ins titute , P ittsburgh, 1981, p. 749.

7. Warnatz , J., Eighte e nth S ym pos ium (International) on Com bus tion, The Combustion Ins titute , P ittsburgh, 1981, p. 369.

8. Wu, C. K., and La w, C. K., Twe ntie th S ym pos ium (International) on Com bus tion, The Combustion Ins ti- tute , P ittsburgh, 1985, p. 1941.

9. Yu, G., La w, C. K., and Wu, C. K., Com bus t. Flam e 63:339 (1986).

10. Law, C. K., Zhu, D. L., and Yu, G., Twe nty-Firs t S ym pos ium (International) on Com bus tion, The Combustion Ins titute , P ittsburgh, 1988, p. 1419.

11. Le wis , B., S e le cte d Com bus tion Proble m s (A G A R D), Butte rworths , London, 1954, p. 177.

12. S trehlow, R. A., Com bus tion Fundam e ntals , Mc- Gra w-Hill, Ne w York, 1984.

13. Ke e , R. J ., Mille r, J . A., and Je ffe rson, T. H., Sandia Re port SANDS0-8003, 1980.

14. Ke e , R. J ., Warna tz, J ., and Mille r, J . A., Sandia Re port SAND83-8209, 1983.

15. Ke e , R. J ., Grca r, J . F ., Smooke , M. D., and Mille r, J . A., Sandia Re port SAND85-8240, 1985.

16. Grca r, J . F ., Ke e , R. J ., Smooke , M. D., and Mille r, J . A ., Twe nty-Firs t S ym pos ium (International) on Com - bus tion, The Combustion Ins titute , P ittsburgh, 1988, p. 1773.

17. Mille r, J . A., Ke e , R. J ., Smooke , M. D., and Grca r, J .

F., Wes te rn S ta tes Section Me e ting of the Combus tion Ins titute , Boulder, Colorado, April 1984.

18. Williams , F. A., Com bus tion The ory, Benjamin-Cum- mins , P a lo Alto, 1985, Appendix D.

19. Coffe e , T. P ., and He ime rl, J . M., Com bus t. Flam e 43:273 (1981).

20. Coffe e , T. P ., Com bus t. Flam e 55:161 (1984). 21. Warna tz, J ., Twe ntie th S ym pos ium (InternationaO on

Com bus tion, The Combustion Ins titute , P ittsburgh, 1984, p. 845.

22. Warna tz, J ., Com bus tion Che m is try (W. C. Gardiner, J r., Ed.), Springer-Verlag, Ne w York, 1984, p. 197.

23. Mille r, J . A., Persona l communica tions . 24. Pe te rs , N., and Ke e , R. J ., Sandia Re port SAND86-

8842, 1986. 25. Glarborg, P ., Mille r, J . A., and Ke e , R. J ., Com bus t.

Flam e 65:177 (1986). 26. Coffe e , T. P ., Persona l communica tions . 27. Wes tbrook, C. K., Com bus t. S ci. Te chnol. 20:5

(1979). 28. P itz, W. J ., and Westbrook, C. K., Com bus t. Flam e

63:113 (1986). 29. Wes throok, C. K., Com bus t. S ci. Te chnol. 23:191

(1980). 30. Wes tbrook, C. K., Persona l communica tion. 31. Coffe e , T. P ., Com bus t. S ci. Te chnol. 43:333 (1985).

R e ce ive d 11 March 1988; revised 28 Augus t 1988