Gaseous Explosions - Industrial & Engineering Chemistry (ACS

Ind. Eng. Chem. , 1929, 21 (12), pp 1261–1268. DOI: 10.1021/ie50240a032. Publication Date: December 1929. Note: In lieu of an abstract, this is the ...
6 downloads 0 Views 1MB Size
INDUSTRIAL AND ENGINEERING CHEMISTRY

December, 1929

1261

Gaseous Explosions VIII-Effect

of Tetraethyl Lead, Hot Surfaces, and Spark Ignition on Flame and Pressure Propagation’ Mott Souders, Jr., and Geo. Granger Brown UNIVERSITY OF MICHIGAN, ANN ARBOR, MICR.

The effect of tetraethyl lead, both in the vapor phase 111R R o R s u p P oRT-The and thermally decomposed, on flame-speeds and rate of pressure-element mirror suplier paper (2) the rate rise of pressure following ignition, was determined for of rise of pressure and p o r t was redesigned to reexplosive mixtures of benzene, pentane, isohexane, and duce friction and looseness the auto-ignition temperature heptane in air. Tetraethyl lead vapor was ineffective of the explosive mixture apin bearings. The steel spring in retarding combustion until decomposed by the peared to indicate the relaoriginally serving as a bearburning mixture, whereas decomposed tetraethyl lead, tive knocking tendencies of ing support for the mirror introduced before firing the charge, retarded both different fuels in i n t e r n a l axis was replaced by an adflame-speed and rate of rise of pressure. justable pivot and jewel arc o m b u s t i o n engines. The A hot surface was introduced into the bomb to secure rangement, shown in Figure effect of tetraethyl lead upon auto-ignition of the charge ahead of the advancing 1. T o r e d u c e friction and rate of rise of pressure was flame. This auto-ignition produced an unusually high p l a y s t i l l further, the Vreported a month ago ( 3 ) . rate of rise of pressure. n o t c h e d brass wing of the Last year (12)an apparatus Decomposed tetraethyl lead prevented or delayed the mirror mounting, into which was described in which flame auto-ignition and retarded the resulting combustion. was fitted the slotted presand pressure p r o p a g a t i o n High-frequency pressure waves ordinarily present in s u r e r o d , was replaced by m i g h t be investigated. At the explosions were eliminated by decreasing the numtwo small p a r a l l e l r o d s , that time the pressure waves ber of sparks in the igniting discharge. The effect of s l i g h t l y curved downward, frequently observed in gasethese waves on the combustion and on the initiation of a n d i n s e r t e d between and ous explosions were shown to a violent “shock wave” was determined. b e a r i n g upon similar rods be initiated in the inflamed which were fastened to the m i x t u r e behind the flame front. The influence of such pressure waves 011 combus- pressure rod by an adjustable screw. Thus curved bearing tion has been indicated by Dixon ( 5 ) . Morgan ( 1 7 ) and, surfaces, adjustable to take up play, were used throughout. REFLECTING SURFACE-By employing atomic vaporization more recently, Maxwell and Wheeler (15) have suggested that t,hese pressure waves are definitely related to engine in a vacuum tube, the concave mirrors were platinum-plated and the difficulties of silvering were eliminated. The reknock. Auto-ignition by adiabatic compression (1, 6 , 20) and ig- flecting surface thus produced was more uniform and pernition of the compressed explosive mixture ahead of an ad- manent than the silver plate previously employed. HOTSPOT-A hot spot was introduced into the end of the vancing piston (6, 8) or flame front (23) a t high initial temperature and pressure have been studied photographically. bomb opposite the ignition end. This hot spot consisted of a But heretofore no attempt has been made to obtain simul- cylindrical brass cup into the bottom of which was screwed taneous flame photographs and pressure records of auto- a small brass tube wound with chrome1 wire embedded in ignition induced by a “hot spot” ahead of an advancing alunduni cement. The open end of the cup was joined to flame front initiated by a spark, Moreover, in view of the the threaded metal part of an ordinary spark plug and aluninconsistencies in the literature, a study of the action of dum cement was packed around the brass tube projecting tetraethyl lead and its decomposition products on auto- through this plug. (Figure 2) An insulated thermocouple ignition by a hot spot, on flame speeds, and on rate of rise was inserted in the tube so that the junction was in contact of pressure seemed desirable. Finally, the effect, origin, and with the hottom of the cup. When in place the bottom of possible control of the high-frequency waves usually present the brass cup was flush with the end plate of the bomb. in gaseous explosions initiated by a spark could be investi- Temperatures up to 900” C. could be attained with a few gated in conjunction with the studies on auto-ignition and minutes’ heating. An electric fan served to prevent excessive the effect of tetraethyl lead by use of the apparatus de- heating of the bomb by the hot spot. When the heating elescribed by Hunn and Brown (12). ment failed (after 5 to 10 hours’ service), the cement was removed, a new heating element was inserted, and the plug Apparatus and cup were packed with cement as before. CHARGIXG PIPET-L4 1-cc. Luer tuberculin syringe graduThe apparatus used was essentially the same as that de- ated in 0.01 cc. was used for charging the fuel. A long scribed by Hunn and Brown ( l a ) , except for the following needle with a small opening served to spray the fuel into the improvements and additions: combustion chamber, promoting vaporization and distriPRESSURE ELEMEKT LAMPS-The original pressure-element bution. The needle passed through a cork fitted into the lamp bulbs were found to be too short-lived. Af’ter some charging opening to prevent loss from back pressure when experiment and upon the suggestion of H. W. Corzine, of the charge was injected. the Incandescent Lamp Department of the General Electric Calibration Company, type C-8, g-volt, 18-amperel projection lamps were adopted. These have been very satisfactory. The pressure elements were calibrated, by means of gas 1 Presented before the Division of Gas and Fuel Chemistry at the 78th precsure in the bomb, against a Bourdon-type gage which had Meeting of the American Chemical Society, Minneapolis, Minn., September been previously checked against a dead-weight tester. 9 to 13, 1929.

CCORDIKG to an ear-

A

INDUSTRIAL A X D ENGINEERING CHEMISTRY

1262

Photographic Materials

Eastman hypersensitized panchromatic motion-picture film was used for the flame photographs and Eastman recording paper Nos. 1 and 2 for the pressure records. Fuels and Mixtures

The four fuels used in this investigation were benzene, n-heptane, 3-methylpentane, and a mixture containing 68.8 per cent (by weight) n-pentane and 31.2 per cent methylbutane. Their physical properties are given in Table I. Table I-Properties

FUEL

of t h e Fuels Used

B;zNy

ComPosrrroN

d2::

O:%

c. Benzene n-Heptane Isohexane Pentane

c. P. 80 0 c. P. 98 2 3-methylpentane, trace of pentanes 62-64 68.8% n-pentane, 31.2% 2-methylbutane, trace of butanes ,.

.

0,8774 0 6834 0 6678

1,5022 1.3868 1,3778

0.6293

. ...

The benzene was secured from the J. T. Baker Company, chemically pure and thiophene-free. The n-heptane was obtained from the Ethyl Gasoline Corporation and was fractionated further in a standard column. Analysis in a column such as described by Podbielniak (21) showed no detectable amounts of other hydrocarbons. The hexane and the mixture of pentanes were naturalgasoline fractions supplied by the Phillips Petroleum Company. Fractional distillation analyses of the hexane indicated practically pure 3-methylpentane with a trace of pentanes present. Similar analysis of the pentane mixture showed 68.8 per cent (by weight) n-pentane, 31.2 per cent methylbutane, and a trace of lighter hydrocarbons. The tetraethyl lead was obtained through the courtes of the Ethyl Gasoline Corporation, according to whose analysis it contained 98.5 per cent PbEtr. The commercial Ethyl fluid was not used. Explosions of hydrocarbon-air mixtures of these fuels containing about 20 per cent excess fuel give the maximum pressure and rate of inflammation. Because mixtures containing about 35 per cent excess fuel produce more actinic explosions, yet differ but little in pressure rise and flame speeds from the maximum-pressure maximum-speed mixtures (IC), these richer explosive mixtures were used. The initial conditions in all cases were atmospheric pressure and room temperature. C A M MIRROR

1

m i

I

u

n

DIAPHRAGM

Figure 1-Mirror

Mounting

Procedure

The operation of the apparatus and the method of aligning the flame and pressure records have been described by Hunn and Brown (12). The new method of charging the fuel and the introduction of decomposed tetraethyl lead, however, require some explanation. After injecting the fuel into the explosion chamber by means of the syringe described above, 5 to 10 minutes were allowed for vaporization and diffusion before exploding the

Yol. 21, Yo. 12

mixture. This period also allowed sufficient time for placing the film and paper on the drums. That the fuel was completely vaporized and throughly diffused is evidenced by identical records obtained with widely differing diffusion periods. Moreover, the theoretical partial pressure of the fuel vapor in the mixture was, in the case of heptane, less than 43 per cent of its vapor pressure a t the charging temperature. With pentane the partial pressure was 5.9 per cent of the vapor pressure, with 3-methyl pentane 12.7 per cent, and with benzene 27.7 per cent. Tetraethyl lead was decomposed by dropping the liquid onto the surface of a glass bulb heated electrically to a red heat. An outlet tube from the bulb was inserted in the bomb and the violence of the decomposition served to THERMOCOUPLE LEADS distribute the decomposition products in the combustion chamber. The decomposed tetraethyl BESTOS PACKING lead was introduced after ATlNG ELEMENT LEADS the diffusion period and SBESTOS PACKING the charge was fired immediately afterward. E l a p s e of s e v e r a l minutes between introd u c t i o n of the decomposed tetraethyl lead and firing of the charge apparently rendered the decomposition products ine f f e c t i v e , although a BRASS i u a E Tyndall effect from the HEATING COILS glowing hot spot indiALUNDUM CEMENT cated that the smoke had not settled out. Undecomposed tetraethyl lead was charged in solution with the fuel. Figure 2-Hot Spot That the tetraethyl lead was vaporized and well distributed is shown by the uniformly increased luminosity of these explosions. Discussion of Results

Extracts from the experimental data are presented in the accompanying tables. For convenience in discussion, tetraethyl lead charged in solution with the fuel will be referred to as “undecomposed” and the products of decomposition blown into the bomb with the charge will be called “decomposed.” FLAME SPEEDSAND RATEOF PRESSURE RISE-In the investigation of the effect of tetraethyl lead on rate of inflammation and rate of rise of pressure (Table 11),the pressure waves were eliminated, by a method described later, in order to exclude their influence on flame and pressure propagation. From photographs (Figures 3 to 11) of the various hydrocarbon-air mixtures exploded under initial conditions of one atmosphere and room temperature, it appears that the flame front passes through the charge with an accelerating velocity until about two-thirds of the charge has been inflamed. At this point the flame speed is checked and the inflammation proceeds a t a nearly uniform rate (referred to the bomb) throughout the remainder of the unburned charge, unless the far end of the bomb pressure waves again accelerate the flame. This approximately uniform velocity of the flame has been used as a basis for comparing the effect of tetraethyl lead on the rate of inflammation. It has also been observed that combustion is not complete in the flame front, as there is an appreciable increase of

I

Figure 3-3.JlSb

Benzene-Air Mixture Alone, under 1 Atmosphere

a t ZJ*

c.

Spark acljurted to reduce waves: no hot spot

pressiire after the flame front st.rikes tile end platt:.

1)iitti

concerning the effect of tetraethyl lead oil this rate of pressure rise lisvc also hecn obtained. Table 11-Average Rate of ln~ammarlonand Average Mean dP/df Following Complete haammation

7%

Cm./rec.

Seiondr

Lbr.

m./w in./ ILL.

QXNZIINS-AIR MIXTIIRRS

3.51

206 I98 180

0.136 0.137 0.155

WaPT*MQ-*m

2 61

174 175 I52

8

R 5 8

4308 3293 4320

MihlUBIS

0.155

n.156 0.104

7.5 4

2.5

3420 2526 8480

--

Figure 4-Effect of Tetraethyl Lead Vapor on Flume-Speed. 3.51% Benzene-Alr Mixture with 5% Tetraethyl Leud Vapor Spark adjusted to reduce WBVCB: no hot spot

is introdiiceci with the fuel, this decrease often exceeding :Y$ per cent. This cffect is inore pronounced with the paraffin Iiydrocarbons than wit.11 benzene. E,f& of Decoinpa.*ed Tetraetl~!/ZLead. Tile decomposition products of tt?traethyl lead greatly retard the velocity of the flame front in all the hydroearlJon-air mixtures st,udied. (Table 111) In general, deconiposed t,etracthyl lead also deoreascs the rate of rise of pressure following complete inflair~ination,altliorigh this effect is not nit,lioul exceptions. An explanation of t,he apparent ineonsistmcy is given later. . ~ u ~ o - I ~ ~ i ~ ~ o ~the - l "pressure r o m eilrves obtained with explosions of the hydrocarbon-air mixtures under invcstigatioii, it appears that the velocit,y of flanie travel and tire rat.e of rise of pressurc resulting from auto-ignition hy a hot spot ahenrl of the f l a m front initiated by a spark are from three to fire timcs as great as tlic eorri:sponding normal rates at the sitine stage of tiic explosion. Anto-ipiition in this manner rcsiilts in ~ 1 uiiusiially 1 high rate of release of encrg\..

HKXANB.dlR MlXTVRUS

2 8

153 1411

125

0 170 0.173 0.212

7 4.26 5

2550 1805

2030

3545

3 82

2105

2830

5820 2470 2580

3 51

106 184 160

0 142 0.180 0.178

11.6 7.5

8

4800 3300 42.50

2 8

148 146 I26

0.171 0.170 0.200

8.5 4.Q 3.5

2!B2 20!10 2575

a

160 167 144

0.160 0.163 0.191

9.5

8.5

4240 28.50 2018

2 61

8%

5

E&ct uJ Tetrruthyl Leud Ihpor. Although the fiilai, uiiif m n flame velocities of the liydrocarbon-air explosioris with t~etrat~iiiyl lead vapor present. differ froin those of the w. plosions without tetraethyl lead by an amount lcss than the exgiviiiiciital error (about 5 per cent), averaging a large niimbcr of explosioms, iiidicates a consistent tcnclency lol~ard ilecr~asedflarno velocity in the presence of tetraethyl Ica,d vapir. Moreover, tlie axrerage decrease in the final, uniform flniiir rdocity is niucli greater than tlie average increase in t i l e tirne required for total inflaimiation. This indicates that any effect of tetraetiiyl lead vapor is more pronounced in the later stages of blie explosion, where higher temperatures and pressurrs urwail. 1n.a11 cases &e rate of rise of pressure following complete inflammat,ion is greatly lowered when tetraethyl lead vapor

.i

Rimre S-Effecf of DeCompoaed Tetrnerhyl Leed on F l a m e Speed. 3.51% Benzene-All Mixture with S% Decomposed T ~ f r a e f h y lLead Spark adjusted to produce violent TIY~S: no hot spot

nrjicritture of the hot spot, apparently, has a greater etrect on t.11~ rat,e of rise of pressure than on t,he velocity of tlie auto-ignition flame. With init,ial conditions of at.mosplmrio pressure and room tcmpcratiire, a red heat was required to secure auto-iynit,ion far enoi~gli,ahead of the advaiicing flnine front tlmt the autoignition flame might iiave sufiieient duration to bo accurately incasuri:d. Iricidcnt,al experiments indicatcd, as would be cxpect,eti, tiint iricrcase in the inilial tempcratr~reor pressure of tlie explosire mixture permitted auto-ignition mit.11 lower tcmperaturrs of the hot spot. All data reported liere are for initial conditioris of atmosplicric pressure and room temperatnr" _X__. Auto-ignition of pentane required a temperature of the

I.VDUAWI~IALB.VD ENGINEERIXG CflE.MISTIZY

1264

hot spot near the maximum attainable with the apparatus. Heptane auto-ignited consistently with a somewhat lower temperature, and 3-methylpentane a t a considerably lower temperature than that reqnired for pentane. Renserie could not be made to auto-ignite under any conditions used in this investigation. of Tetraethyl Lead on Auto-Ignltiun

Table 111-Effecf

10.7'1

2.61

25

241 2412

243 176 106 244 248 107

1111" 233 175 303 8051 3116

%li

222

3112

106

i16:l

25

7706

770 7711

770 770 570 776 780

783 788 7RR 760

770 783

171 172

770

235 286 277 279 276

778 237

20

740 740 760 74.5

3Vi 108 174 ~4.5 247

173 234

2.13

750"

7n5 770 770 765

770 770 770

795 805 764 763 764

760 765

253

2J6

770 800

254

255

805

808

221 222

680

224

680

223

080

680

173

6x5

174

685 680 680 684

225

226 228 A10 R10 273 272 274 213 214 215 191 193 192

Nolie None

Not)c

None 5 . 0 VI

&.one None None None None

None None None None woixe None

0 5V 5.0 V 0 5V

5.00 0.5 D 5.0 D

i.on

n

1.0

N~~~ None

6.0V 5

ov

5.OD 5.0 D None 5.0D 5.0D

8

68

NO auto-ignition

50 59

12

..

63

7 10 8 4

5n 57

6

kk.5

6 4

~~

65

No

None

Slight Siiaht

...

s

None None Slight

8

None Slight

None Slight None

No auto-ignition No auto-ignition 02 2 66 1.5 62 5 00 7 00 4 61 0 N o auto-ignition N o anto-ignition 63 12 04 14 02 12

0.50 6.0 D

52 58 55

11 8

2.5 No auto ignitioo No auto-ignition

5.0D

?ao

None 5.0V

2.0 D 5.0 D

12

62 47 54 58 52

12

4

waVavcn &Tone

Slight waves

.. .. ..

None None None

None None None Waves Fairly viol.ot

1i.O:I

20

7R2

182 184 187

780

None NO"*

780 780

5.0V' 5.0 D

794 790 800 SI2

N~~~ 5.0 D

188

223 224

C10 252 227 225 226

7s0

820 830 835

5.0 n None

N o anto ignition

None NO~C

5.0 D

5.0D

XBPIANR-UP ldrXT"aKl

2.01

wRv-

w*ves

790 770 785 77n

Wxves

Violent Violent

16

Slight

None

auto-ignition

34 0 50 6 no auto-ignition 48 16

None vioicot viairnt Violent Slight

46 48.5

8

Slight

48

6

Violeat

4

Violcot

Temperature of hot spot too low for ro?dstenl auto-ignition. b lawest temperature of hot spot at which connistcnt autoisnition war, ebfnined c V denotes PIIEf.& charged a3 ~ a p o r . d D denotes PbEt. charged decomposed. 1

E$?& of Tetraethyl Lead Vapor. Tetraethyl lead vapor apparently has little or no effect OIL either the hot-spot temperature required for, or the flame and pressure propagation

None Vapor Decomp. NO"