1、% T9J$F n -ii/ 7948NATIONAL ADVISORY COMMITTEE,FOR AERONAUTICSTECHNICAL NOTENO. 1637EFFECT OF PREIGNITION ON CYLINDER TEMPERATURESPRESSURES, POWER OUTPUT, AND PISTON FAILURESBy Lester C. Corrington and William F. FisherFlight Propulsion Research LaboratoryCleveland, Ohio.,YjiiJJi-J7WashingtonJune 19
2、48Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-lllllll!lllllumflM31176014222781NATIONAL ADVTS3RY COMMIITEEFOR AERONAUI!ICSTechnical Note No. 1637mm OF FREIGNITION onTmrmm TmmATURE3PRE3SURE3, POWER OUTPUT, AND KL5TON FAILURESBy Lester C. Corrington
3、 and William F. FisherSUMMARYAn investigationusing a cylinder of a V-type liquid-cooledengine was conducted to observe the behavior of the cylinder whenoperated under preignition conditions. The effects of the follow-ing variables were investigated: fuel-air ratfo, power output,aromatic content of f
4、uel, engine speed, mixture temperature, andpreignition source. The power outputs at which preignition wouldcause complete piston failure for the selected engine operatingconditions and the types of failure encountered when using variousvalues of clearance between the piston and the cylinder barrel w
5、eredetermined.The results indicate that in the engine investigated preigni-tion at high lower levels led to baclc?iringinto the inductionsystem under most condition=. Preignition caused cylinder-headtemperatures to increaseup to 200 F at rates up to 30 F persecond, caused maximum cylinder pressures
6、to increase up to30 percent, and caused power output to be drastically reduced.Runs to destruction indicated that preignition caused overheatingthat resulted in overexpansion of the piston, seizure of the pistonin the cylinder barrel, and consequentmelting of the side of thepiston. By increasingthe
7、clesmince between the piston end thecylinder barrel it was possible to increase the power level atwhich piston failure occurred because of preignition.INTROIXJCTION2renition difficulties experienced by the military servicesand engine manufacturers in recent years have indicated the needfor informati
8、on on the harmful effects of preignition and possibleremedies. During an investigationof a V-type liquid-cooled engineat the NACA Cleveland laboratory, several cases of piston burningand destructive backfiring were encouiitered.Efforts to determinethe cause of the failures showed that preignition wa
9、s responsible,.for a large number of them. Further investigationsshowed thatProvided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-2 NACA TN Oa 1637preignition in the engine was traceable to overheated spark plugsand to overheated exhaust valves. Msmuch as
10、the preignitionproblem appeared to be of great tiportance,an investigationwasundertakenat the Cleveland laboratory to obtain information on thebehavior of an engine during preignition operation.With exhaust spark plugs and exhaust valves uses as ignitionsources preignitionwas excites at different fu
11、el-air ratios,power levels, engine speeds,mixture temperatures,ana with fuelscontainingvarious percentages of aromatics. Obsemations weremade of cylinder-headtemperatures,preignition advance, enginepower, and change in peak cylinder pressure as the engine wasallowed to enter into preignition and rem
12、ain operating unaer pre-ignition conditions. Because of the unstable nature of preignition,a motion-picture camera was used to record the data.Several piston-destructiontests were run to cteterminethetype of piston failure associated with preignition and to deter-mine the effect of a change of clear
13、ance between the piston and thecylinder barrel on these failures.APPARATUSEngine setup. - A single-cylinderengine (bore, 5* in., stroke,6 in.) was used in the investigation. The setup consi=ted of aV-type liquicl-cooledmulticylinder engine block mounted on aCUE crankcase in such a manner that any on
14、e of the cylinaers couldbe used. (Seefig. 1.) One cylinaerwas used for all of theinvestigationexcept the destruction tests which were conducted ondifferent cylinders in the same bank. The inauction system, themain part of which is diagrammaticallyshown in figure 2, includeda pressure-regulatingvalve
15、, a flow-measuring orifice installedaccording to A.S.M.E. standards,a surge tank, a vaporization tank,and an intake elbow designed to simulate the intake elbows on themulticylinder engine.Fuel was injected into the upstream end of the vaporizationtank in which seven inclfnedbaffles promoted vaporiza
16、tion andmixing with the air. The engine efiust passed through a shortwater-jacketedpipe into a large water-cooledmuffler. The pres-sure in the muffler was maintained constantat atmospheric pressure,H inch of mercury.The cylinderwas cooled byamlxture of 30-percent ethyleneglycol and 70-percentwater (
17、by volume) circulate at the rate of120 gallons per minute ina pressurized system. Navy 1120 lubri-cating oil was used throughout the program; two oil Jets located.Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-. NACA NO o 1637 3* on the maJor and mi
18、nor thrust sides of the cylinder and directedtoward the under side of the piston crown provided lubrication andpiston cooling. The flow through each jet was about 4 pounds perm minute for most of the tests. This flow value was arbitrarilywa chosen to insure sufficient cooling and lubrication and the
19、rebyeliminate any failure that might %e caused by insufficient oil flowTO the piston. The flow per jet waa reduced to about 2 pounds perminute for the piston-destructiontests. This lower rate of oilflow probably more nearly simulates the multicylinder conditionsfor this engine.The engine power was a
20、bsorbed by an eddy-current hype dynamom-eter and a direct-currentmotoring dynamometer. During negativetorque operation, caused by advanced preignition, the dynamometersupplied power. An electronic speed regulator controlled the speedby varying the loading of these units. A balanced-diaphragmtorquein
21、dicatortith a remote-readingmanometer was used. (See reference 1.)Preignition source. - In most of the rurm,spark plugs ofvarious heat ranges were used = the source of preignition. Oneof these spark piugs was installed in the exhaust side with acolder-operatingplug on the intake side. When excitatio
22、n of pre-ignition by the eZhaust valves was desired, a valve (high chrome-nickel austentic steel, Al-5700) with a badly corroded head wasinstalled in place of one of the Nichrome-coatedvalves normallyused in these tests and cold-operatingspark plugs were installedin both spark-plugholes. Throughout
23、the program the operating temp-eratures were never severe enough to cause scaling of the Nichrome-coated valves to the point where they would excite preignition.Temperature measurements. - The cylinder-head temperature wasmeasured with an iron-constantanthermocouplemounted about three-sixteenthsinch
24、 from the inner wall of the combustion chamber betweenthe two exhaust-valveseats ad connected to a sel.f-lalancingpoten-tiometer. A check was made on the rate of response of the potenti-ometer and it was found to be stificiently high to record the rateof cylinder-head-temperabme rise during preignit
25、ion. Spark-plugelectrode temperatures were measured.with a chromel-alumel thermo-couple mounted in the center electrode of the intake plug aboutone-sixteenth inch from the firing end. This installationwespossible only with certain types of spark plug and for this reasonthe sperk-plug-electrodetemper
26、ature was measured for only a fewof the tests. The design of the exhaust pipe prevented the use ofa thermocoule in the exhaust spmk plug.Electronic equipnent. - A piezoelectric pickup, installed near-the intake spark plug and coupled to an oscilloscope through anProvided by IHSNot for ResaleNo repro
27、duction or networking permitted without license from IHS-,-,-I 4 NACA %NOa 1637 .amplifier,provided pressure-timediagrems. A special timing a71device indicated on the oscilloscopetrace the positions of 180,120j and 60 B.T.C., top center, and 60 A.T.C. In orderto deter-mine the time of ignition,the s
28、park gap of the exhaust spark plugwas connected to serve also as an ionizationgap. When the gapbecame ionized because of the presence of a flsme, a current sup-plied by an electronic instrumentwas permitted to flow through it.By suitable connectionswith this electronic instrument, a verticaJ.line we
29、s superfmsed on the oscilloscopetraoe, which indicated U2the presence of the flame at the spark plug. Inasmuch as this seine .+mspark plug was used as the preignition source in most d the runs,the ionization-gaptrace together with the timing marks made itpossible to detenuine the approxhate ignition
30、 advance during pre-ignition operation.Data-recordingequipnent. - A 16-milltietermotion-picturecamera was set up to photograph simultaneouslythe oscilloscopetraoe and indicationsof engine speed, ttie, engine torque, spark-plug-electrodetemperature, and cylinder-headtemperature. Photo-graphs were tak
31、en at the rate of eight francs per second throughouteach preignition run. Reproductions of three f?xuues,representingthree successive stages in a typical preignition run, are shown infigure 3. .Fuels. - Because aromaticfuels are, in general, raoresuscep-tible to preignition than paeffinic fuels, the
32、 arcanaticcontentwas chosen as the vsrtable in selecting fuels. In order to pre-clude the possibility of knock under all test conditions, the fuelswere blended to have a very high performance number. The followingblends, all leaded to 6.0 ml TEL per gallon,were used to obtainthe data:Percentage coqF
33、uel Cumene Triptane(aromatic)1 0 502 15 353 30 204 50 0Hot-acid octaneswere used in placebecause of acumene blendtent was notIsitionby volumeS-reference/Hot-aoidTfuel octanea- 5050 -50 -50 .-.-of S-reference fuel in fuel 1temporary shortage of S-r#erence fuel. The 30-percent(fuel 3) was used for all
34、 runs in which smmatic con-a variable.Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-NACA TN No. 1637.u-lIPm5For the preignition progrem reported herein the followingbasic operating conditions were used as the basis for comparing,at a variety of eng
35、ine conditions, the behavior of the engine underpreignition operation: fuel-air ratio, 0.070; engine speed,3000 ?pa;mitiure temperature, 175 F; and fuel, 30-percent cumeneblend. The values for each of the variables and the values of theoperating conditions that were held constant are listed in thefo
36、llowing table:Variable APProxi- Fuel- O.unene Sourcemate imep air in fuel of pre-(lb/sq in.) ratio (percent) iit ionFuel-air ratio 300 0.070 30 Spark.099 plugPower output 300 0.070 30 Spark250 plug185Aromatic con- 250 0.070 0 Sparktent of fuel 15 plug3050Engine speed 250 0.070 30 SparkplugMixture te
37、m- 250 0.070 30 Sparkperature plug.Sourceof pre- 300 0a71 070 30 Sparkignition plug,etiaustvalve*IEngine Mixturespeed tempera-(rpn) ture(%)-t-3000 1753000 175,3000175t2753(X)0 175Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-6 MACA TN NO. 1637The f
38、ollowing operating conditionswere maintained constant:Outlet-coolanttemperature, % . . . . . . . , . . . . . . . . 250Inlet-oil temperature,% . . . . . . . . . . . . . . . . . . 185Inlet-valve ttming, degrees B.T.C.Opening . . . . . . . . . . . . . . . . . . . . . . . .48Closing . . . . . . . . . .
39、. . . . . . . . . . . . . . 118Exhaust-valvetiming, deaees A.T.C.Opening . . . . . . . . . . . . . . . . . . . . . . . .104Closing . . . . . . . . . . . . . . . . . . . . . . . . . . 26Sperk timhg, degrees B.T.C.Inlet . . . . . . . . . . . . . . . . . . . . . . . . . . . 28Exhaust . . . . . . . . .
40、. . . . . . . . . . . . . . . . . 34Compression ratio . . . . . . . . . . . . . . . . . . . . . .6.65Friction data were obtained bymotorhg the engine after apreignitiontest with the fuel shut off.Several sources of preignition (exhaustspark plu and corrodedexhaust valves) were rated by operatingthe
41、engine at the basic setof conditions and increasingthe manifold pressure until preigni-tion was encountered. The power level was noted at this point andtaken as the rating of the preignition source. In order to obtainpreignition at a given power level, it was then necessary only tochoose the appropr
42、iatespark plug or exhaust valve. Several mnsoould usually be made before an appreciable chenge in rating madeit necessary to replace the preignition source.When a run was to be made, the engine was operated at thedesired power level but with the fuel-alr ratio considerablyricherthen 0.070. When oper
43、ating conditionswere steady, the fuel flowwaa slowly decreased until the engne was running at the basicfuel-air ratio (0.070)where preignitionwas encountered. Theengine was permitted to run under preignition con the curves were so adjusted alongthe scale that the points where preignitionbecame rapid
44、ly acceler-ated (the of ignition about 60 B.T.C. coincided. Comparisonsbetween the curves are made easier in this manner by eliminatingfrcm considerationthe period of poor reproducibility. The loca-tion of the zero point on the time scale therefore has no signifi-cance. Figure 4, then, shows the rep
45、roducibility of the preigni-tion runs after preignitionhad begun to advance rapidly.In most cases data were taken from the frsmes of the motion-picture film that showed operating conditions at l-second intervals;data showing rapidly changing conditionswere taken at shorterintervals. The percentage i
46、ncrease in the peak pressure at snyinstant over that for normal combustionwas obtained from therelative height of the pressure-timediagras aa measured frau thephotographs of the oscilloscopescreen. The accwacy of tlieheight measurement is estimated to be about 5 percent. The meas-ured values of indi
47、catedmean effective pressureare not correctwhere the rate of change is rapid because of inertia of the manom-eter fluid and because of the speed-regulationcharacteristics ofthe dynamometer used.The curves shown in figure 4 have been faired to represent emaverage of the individualruns and are conside
48、red to be typical Mthe behavior of the engine during preignition at the given set ofoperating conditions. All the curves presented have been obtainedin this manner from several runs at each operating condition. Forthe sake of simplicitythe original data ere shown only in figure 4.Backfiring. - Most
49、of the runs for which data are presentedwere terminated by bacldiring into the induction system. The pointat which backfiring occurred is indicatedby arrows on the curves.In order for preignition to cause bacldiring, the time of ignitionmust be considerably earlier than that at which the intake valvesclose (llE!OB.T.C.). Enough pressure rise must take place
