Cylinder Design: Bore and Stroke vs. Flame-front propagation

[Turbo20V]

So I have been thinking a lot about the mathematical optimization of cylinder geometry lately.

I seems as though the intersection point of flame-front propagation and the dynamic stresses for a standard gasoline engine yield a cylinder volume of around .5 liters.

Obviously low speed engines can have larger cylinder volumes because there is more burn time, and conversely high speed engine operation can only sustain small cylinders.

Now since most modern cylinder designs are 4v the spark plug is located at the center of the combustion chamber, and the flame-front propagates radially from that point.

Is there a formula that predicts the combustion front velocity?

What about the heat? how does the heat develop as a function of combustion and time?

The pressure in the cylinder I assume is a function of the base pressure at TDC, and the heat of combustion as it causes the air to expand.

Now I do understand that this is multivariate, and non linear. There is certainly a lot of things going on. However I think it all boils down to this:

You have a radius of combustion. The spark happens and the flame-front rushes to the sides. Combustion is exothermic, heat is generated. The heat energy expands the volume of the gas. Dow during this process the piston is moving down, so the volume is changing in a nonlinear fashion.

Does the complete combustion happen over a relatively small crank angle? I imagine at low speeds it does, but what at high speeds? 6000-8000+ RPM? How much crank angle change happens during that combustion time?

I would imagine that engine developers have modeled this fully, however the raw results are often obscured by race regulations.

Anyone have any insight on this? Actual combustion design experience?

 

[dicer]

Due to turbulent conditions in the combustion chamber, the direction of the flame front is not very predictable.
If the gas was steady state then it may propagate radially from the ignition source given that all the area is homogeneous.

 

[mattsooty]

"Obviously low speed engines can have larger cylinder volumes because there is more burn time, and conversely high speed engine operation can only sustain small cylinders."

Whilst this is patially true, a higher engine speed means greater gas speed, turbulence, scavenging & squish velocity and less internal EGR & quench - all of which increase the rate of combustion.

The faster the linear combustion speed the less knock sensitive the combustion chamber design - which can be an issue with larger combustion chambers.

In terms of your quest for formulae: Im pretty sure our friend Heywood touchs on this but not 100% sure since I havent looked in a good while!

At part load, low engine speed the combustion event can last for roughly 90deg but at higher speeds for the same load this can be easily reduced by half. Saying that its all application specific.

MS

 

[Turbo20V]

Who is heywood?

My linear combustion speed are you referring to the propagation of the flame of the piston speed?

 

[SomptingGuy]

http://lmgtfy.com/?q=heywood+combustion

The engine bible. Covers virtually everything.

- Steve

 

[mattsooty]

Turbo20V - Steve said it! The only problem is that the book is aimed at quite a technical audience, not that I am second guessing your back ground - its just something to bear in mind if you decide to purchase it.

By "linear combustion" I mean: -

Normal combustion that moves away from the ignition source in a linear fashion (although technically it should be called turbulent combustion deflagration) and at subsonic speeds

As opposed to Detonation: -

In which the combustion is explosive in nature and occurs at trans sonic or supersonic speeds (hence the noise associated with this type of combustion).

MS

 

[j2bprometheus]

Turbo20V-

Basically, the duration in crank angle degrees is relatively constant over the "normal" range of engine speeds.

This is because turbulence increases with engine speed.

Turbulence wrinkles the flame front. The more wrinkled it is, the shorter the duration.

Laminar flame speed is mostly a function of temperature. NASA published some papers on this. The flame actually propagates, on a local level, at the laminar flame speed. But, since the flame front is wrinkled, the turbulent flame speed is much higher than the laminar flame speed.

It is possible to expand a burning charge so fast that the flame goes out, especially if running very lean or diluting with lots of EGR. Here's a reference on that:

Smith, O.I., Westbrook, C.K. and Sawyer, R.F. (1979). "Lean limit combustion in an expanding chamber." Seventeenth Symposium (International) on Combustion, 1305-1313, The Combustion Institute, Pittsburgh, Pennsylvania.

This is the paper where they were able to extinguish the combustion by rapidly expanding the charge.

Many spark ignition engines have used more than one spark plug per cylinder, often for reliability reasons --- like on private aircraft engines. Two spark plugs per cylinder results in shorter combustion duration. Each spark starts its own flame front, so it takes less time for all of the fuel-air mixture to burn.

j2bprometheus

 

[SomptingGuy]

Turbo20V,

Sorry for the lmgtfy post. I love that site. Hope you didn't take offence.


- Steve

 

[Turbo20V]

Not only did I not take offense, I appreciate the link. I have added the book to my cart.

My background is in mechanical engineering technology. I have a concentration in fluid and thermal sciences as well as stress analysis.

In my personal life I am completely fascinated by automobiles of all kinds, as well as jet engines, trains, and all sorts of turbo-machinery.

I have 5+ years of experience at a turbo-machinery engineering business. I spent most of my time instrumenting test rigs, and developing data acquisition software, and then running the rig and collecting data.

Although that was interesting, I really wanted to design entire machine concepts. Although I had done that a few times, it wasn't what I was doing most of the time like I had wanted. So one day I just decided to leave my job and finish off my BS in mechanical engineering. On Tuesday I start my last semester, and in may I will have graduated!

Now of course my dream job would involve engine development for some sort of auto racing application. Or oven an OE manufacturer. Frankly I just love engines and cars. There is no doubt about what field I really want to be in. I personally engage in every from of motor-sports I can including: Autocross, Drag racing, Track events, Pro solo etc.

Now is the time for me to be sending out resumes and i am hoping to find myself involved with engines in some way.

 

[mattsooty]

@j2bprometheus

The combustion duration (however you measure it 0-100%, 10-100% Mass Fraction Burnt, Heat Release or whatever) changes as a function of charge motion/turbulence. Which is heavily influenced by engine speed and also throttle opening - so it is inaccurate to say:-

"Basically, the duration in crank angle degrees is relatively constant over the "normal" range of engine speeds"

At idle, with the large amount of internal EGR that is present, the burn duration can be excessively long - a fact that is highlighted in the high cycle to cyle variance of IMEP.

In the same ilk at, lets say, 2000rpm 0Nm clutch torque the combustion speed will, as a percentage, be much lower than 2000rpm 100Nm clutch torque.

If this were not the case why would we ever need to change the spark advance to meet MBT of a positive ignition, internal combustion engine.......?


MS

 

[j2bprometheus]

mattsooty-

All else being equal, duration in crank angle degrees increases slowly with engine speed. See p. 394 of Heywood's book for a graph & text.

"Both intervals increase by a factor of 1.6 for a factor of four increase in engine speed."


If you dilute the charge more using either more air or more EGR or more humidity, the duration gets longer.


j2bprometheus

 

[Turbo20V]

Well I see the Heywood book getting referenced more and more. Clearly this is the Go-to reference around here.

Are there any other books that are of this caliber that I should know about?

 

[mattsooty]

@j2bprometheus

I agree with both of your points.

However, with optimisation of spark advance & fuelling for higher speeds/loads the effects become less easy to represent graphically and indeed model. Especially with the addition of things such as VVT and/or external EGR.

MS

 

[SMOKEY44211]

A few years ago I read an article quoting a BMW engineer who stated optimum displacement was .9 liter per cylinder, stroke being 80% of bore size, and somewhere between 8+10 cylinders. (to overcome accessary losses water pump, oil pump ect.) This provided the best BSFC at WOT. Regarding flame propagation this is a dynamic speed. Biggest factor is pre combustion pressure. More pressure = faster flame front. Best torque timing is always higher at part throttle vs WOT. I should add that all of this is for a normally aspirated engine. Throw in some inert gas via EGR or valve timing that would add exhaust gas contamination, flame speed goes down. Abnormal combustion that results in detonation is not neccessarily a bad thing but most gasoline engines are not robust enough to withstand the near instant rise in pressure and temp.. Finally the turbulance generated from squish at TDC also helps to reduce the amount of burn time needed to consume the fuel.------Phil

 

[Turbo20V]

Question for owners of the HARDCOVER Heywood engine book...

Are the illustrations in COLOR?

I was going to pick up the MUCH cheaper paperback version, but since it is in black and white, i didn't want to lose any graphical information it may contain.

Thank you.

 

[vagman2]

Hi,

I think this would be very interesting article for everyone:

http://www.theoldone.com/articles/The_Soft_Head_1999/

BTW the whole site is like a book.
Happy reading

 

[SomptingGuy]

I have the soft-cover Heywood and had never even thought thought the hard-cover it might have colour and be better for it.

Save your money. Spend it instead on the "Bosch Automotive Handbook".


- Steve

 

[j2bprometheus]

Smokey44211-

Pressure does affect flame speed. There is some data in Rose & Walker's book ('Technical Data on Fuel') in Table 5.13, p. 265 that shows flame speed slowly DROPPING as pressure goes up. The fuels shown are propane and methane.


But temperature is more important than pressure. There is an old NASA report that basically shows this result:

laminar_flame_speed is proportional to temperature to some power ( 1.5 to 2).


Here's the report (it is available on the web for free since it is by NASA):

Technical Report 1061 – Effect of initial mixture temperature on flame speed of Methane-Air, Propane-Air and Ethylene-Air mixtures. Gordon L. Dugger. 1950


So changing operating conditions in an engine to increase pressure result in temperature going up -- the temperature rise causes the flame speed to increase, thus shortening the burn duration. The pressure actually caused a small drop in flame speed but the temperature increase probably overwhelmed the pressure effect.

Also, changing engine operating conditions might change the turbulence, which would change burn duration.


j2bprometheus

 

[SMOKEY44211]

j2bprometheus Yes I have to agree with you and the data you've provided clearly speaks for itsef. Somewhere in the collection of technical literature I've read that touches on this subject makes reference to the OH radical presant in the fuels you mention and also found in pump grade gasoline. According to these sources that radical acts a detonation suppresant above temps. of 1400 deg. F( independant of pressure) and flame speed diminishes also. My equipment is not sophisticated enough to verify or despute this. My conclusions stated in previous post are based on observations testing spark ignition, liquid cooled, gasoline fueled engines. Best torque timing requirements change based on the parameters described. It would be worthy of investigation but I doubt the aluminum pistons commonly found in most engines would survive a steady diet of operation in that elevated temperature. I know steel pistons are starting to find their way into some diesel engines perhaps it may open possibilities in spark ignition engines also.-------Phil

 

[j2bprometheus]

Smokey44211-

By the way, I had the author's names wrong for the book ('Technical Data on Fuel') .

The correct authors are Rose, Spiers, & Cooper, not Rose & Walker. And they are really editors, not authors.


j2bprometheus