Describing combustion

[Fahlin Racing]

Well, hopefully we could get somewhere here in putting some kind of description on the combustion process. Here and there I read flame fronts traveling through one another as flame propagation occurs. Among other things that occur when things are off. How would any of you guys put this into words? Sure the diferent types of cylinder head designs will produce different motion within the cylinder, but, its sort of difficult to begin besides mentioning the electrical spark crosses the plug-gap and provides the heat to ignite the air/fuel within the cylinder....

[David Redszus]

The internal engine combustion process has been studied for over 100 years. We have pretty good idea of what happens but it is more difficult to quantify and visualize the process. Following is a very simplified description; jump in anytime.

When the spark plug fires, an electrical arc jumps across the gap at a temperature of approx 60,000F which will ionize the air space in the gap. Fuel molecules will be decomposed into their elements (carbon, hydrogen, oxygen) and diatomic gases (oxygen, nitrogen) will also be decomposed into singular elements of N and O. When and if this occurs, the elements begin to change partners and form a series of new compounds consisting of various hydrocarbons, free hydrogen, OH radicals and nitrogen compounds. This process is referred to as oxidation which we call combustion.

The process does not occur smoothly, either from the perspective of time nor space. As the compounds in the spark plug gap combust, the temperature is raised to approx 4700F which causes the gas to expand and move outward. Some temperature and outward motion is lost due to heat being absorbed by the spark plug electrodes. This resultant delay is often referred to as ignition delay and every engine has it. The initial flame kernal diameter is the size of the gap.

Once past ignition delay, the expanding gas pushes the flame front outward. The combustion occurs along the surface area of the flame kernal. As it grows larger, the surface area is increased and the combustion process is accelerated. This growing fireball does not grow uniformly in time nor space. Some unburned mixture pentrates the flame front and is combusted after the flame front has passed.

The rate at which combustion occurs is a function of temperature, pressure and motion. Increased temperature will accelerate flame speed. Pressure is increase the rate at which heat from the burned gas is transferred to the unburned gas. This flame speed is called laminar flame speed. Mixture motion produced by piston to head position is called squish velocity and is added to laminar flame speed to produce a total or turbulent flame speed.

The time that it takes for combustion to occur can be expressed in units of time or crank angle degrees. Various terms can be applied such as burn angle, burn time, flame speed, etc. Laminar flame speed is a function of heat and pressure and is fairly constant while turbulent flame speed will increase with increasing piston speed.

Since a flame front is actually a thin (approx 2mm thick) moving surface, with unburned gas in front and burned gas behind (or inside), the flame front will cease when it runs out of combustible materials or loses heat. The combustion process will stop if the local mixture is too rich or too lean and when the flame front hits any cooler surface such as a valve, piston or combustion chamber. An aborted combustion process will produce partially combusted hydrocarbon deposits known as carbon deposits. Flame fronts cannot pass through each other since that would require one front to burn an already burned mixture. But rapidly expanding pressure waves certainly can pass through each other.

The usual source of a combustion pressure wave is due to detonation which is the rapid combustion of a volume of unburned mixture. An unburned volume of charge that combusts spontaneously will produce a pressure wave that travels at local sonic velocity and rides on any available particle motion that might exist. These rapidly moving pressure waves will bounce off of all chamber surfaces and will make the characteristic "rattle" sound of detonation. But actually, detonation or knock, will produce three knock frequencies, one in each direction within the chamber.

Normally detonation will occur after top center. If, however, the charge is ignited too early, combustion pressure can increase substantially before top center. When this occurs, the combustion pressure is trying to push the piston down while the piston is still on the way up. We call this pre-ignition which can have a spark source or a surface temperature source.

A fuel will never ignite as a function of pressure alone; heat must be present. But a combination of heat and pressure will change the point of ignition. This makes setting the correct igntion timing for a turbocharged system very interesting. Even if the mass of fuel being delivered is constant, the mass of air ingested will vary with engine speed and induction resonant tuning, so the air/fuel ratio will constantly change. Now the fuel mass curve must be modified to accomodate the changing air mass curve. And often it is not fully adjusted. But spark plug and piston surface deposits reflect the average over time, it does not reveal the true instantaneous combustion air/fuel mixture. Even Lambda sensors which measure the presence of oxygen (not air/fuel ratio) often present a time averaged value which does not represent the true process.

The above is merely an elementary overview of the combustion process. There is still much be be learned regarding combustion pressure curves, misfires, cyclic variations, fuel properties, etc. And power production.

[dwilliams]

Nice description, David!

For people who want even more depth, I usually recommend "Combustion" by Glassman. It's expensive, but available via Inter-Library Loan.

Though it deals with heavy-duty chemistry, it was written for practical engineers who wanted to know how to burn things up. (don't we all?) There's lots of useful information in there for engine guys, along with some nifty exotic stuff like aluminum/Diesel slurries and how some noble gases like argon can act as combustion accelerants.

[Fahlin Racing]

Thank you david. I will agree on the visualization part. Sorry I didn't get back to this sooner, I have been busy working and currently going through the paper work for a house. I don't have much time to read the whole thing today, I will get back to this soon.

Hope everyone's doing well!

[Fahlin Racing]

Since sharp edges are heat sinks, what do you David or anyone believe the flame does once it passes this edge? If you sit and watch a camp fire, the flames travel around the wood similar to how air flows around a corner in a head or intake, I think at least. I think from time to time, if the edge is sharp, and when the flame from combustion passes this, there could be a minute vortex adding heat to the piston top in addition to the normal travel of the flame where ever this sharp edge may exist. Once we, soften these edges or edge, the heat sink is reduced and allows a more efficient process and we make more power. In short, the flame, I believe flows similar to how airflow does.

That is the best I can really put into words when I wonder about how the flame might move across the piston attributes pertaining to heat absorption.

[David Redszus]

A sharp edge, or a thin strip, both have similar characteristics regarding heat abosrption; high surface to volume ratios.

A sharp edge will indeed gain heat more rapidly but it will also lose heat more rapidly since the amount of heat (not just temperature) is not very great. While a surface edge can indeed become an ignition source it would usually be quenched quickly by the much lower temperature of the mixture.

But if the unburned mixture is raised in temperature close to its ignition point, then any additional heat from any source, becomes an ignition point.

Two aspects of combustion worth remembering.
The flame front will move towards a hotter region. The flame front does not expand uniformly in all directions, it seeks a heat source. Or rather, any source of heat will lower the ignition threshold and move the flame in the direction of the heat source. As an example, an overheated exhaust valve will pull the flame front toward the valve since that is the route of least thermal resistance.

Secondly, while local heat sources will alter the direction of the laminar flame path, the combustion process is dominated by turbulent motion which is a violent mixing action. The turbulent mixture motion produces eddies, swirls and penetrations of unburned mixture into and behind the flame front. While the flame front may reach all the way across the chamber, it does not mean that all of the mixture is fully combusted.

Combustion pressure traces do not reveal the true nature of the combustion process. Neither do photographic methods correctly depict flame activity since the process is three dimensional and we can only see the surface shape. The best tools to visualize the combustion process is to employ combustion modeling simulations. But it may not be absolutely necessary to visualize the combustion details as long as we have a reasonable understanding of the global combustion process. Which is difficult enough.

[Fahlin Racing]

Very interesting stuff. Have a couple more questions again David.

What speed is normally seen as far as 'common' flame speeds are concerned during combustion inside the engine? (does a common flame speed even exist, hmmm)

Are there differences between fuels when observing flame speeds?

[David Redszus]

What speed is normally seen as far as 'common' flame speeds are concerned during combustion inside the engine? (does a common flame speed even exist, hmmm)

I am not aware of a 'common' flame speed; each engine (and state of tune) has its own characteristic flame speed. The factors used to evaluate flame speed are: rpm, bore diameter, spark plug location and burn angle (number of crank degrees it takes for the flame to reach the farthest point).

Using an engine with a 4" bore (100mm), with a 55deg burn angle, running at 6000rpm we find the following:
time required for 55deg rotation at 6000 rpm = 1.5278ms (.0015278secs)
flame travel distance is approx one half the bore dia 0.10m/2 = .05m
flame speed is distance traveled divided by time, .05m/.0015278sec = 32.7m/s
(32.7m/s = 107.75 ft/s = 157.3 mph)

If the same engine were running at 9000rpm with the same burn angle the average flame speed would be 49m/s (235.7mph).

Note that these are average flame speeds for the entire combustion cycle. Instantaneous flame speeds will be considerably higher.

This method considers the time and distance traveled by the flame. Another perspective is to consider the time and mass fraction that is burned. Then the units would be grams per second which is a much more useful calculation of heat release.

Are there differences between fuels when observing flame speeds?

Yes, different fuels will have different laminar flame speeds. But the differences are very small. More significant factors are ignition delay, chamber temperature and chamber turbulence, not to mention chamber geometry.

You could perform a laminar flame speed test yourself. Simply pour a thin stream of fuel on the ground in a straight line about 25ft long. Light one end and use a stop watch to see how long it take the flame to reach the other end. That distance divided by the time is laminar flame speed, but it is obtained at ambient temperature and pressure which is not the same as that found in the combustion chamber.

[whitehendrix]

i wish i could somehow extract and absorb everything david knows.

[Fahlin Racing]

The combustion occurs along the surface area of the flame kernal.

Since the burn cycle begins on the surface area of the kernal as you have stated above, then would this mean on a misfire pertaining to a rich mixture, the fuel penetrated that surface and extinguished the chance of ignition David? (Rich misfire even with the proper heat range plug)

[David Redszus]

The combustion occurs along the surface area of the flame kernal.

Since the burn cycle begins on the surface area of the kernal as you have stated above, then would this mean on a misfire pertaining to a rich mixture, the fuel penetrated that surface and extinguished the chance of ignition David? (Rich misfire even with the proper heat range plug)

A misfire describes a lack of ignition. Combustion abberations that occur after ignition are deemed malfires and can take many forms.
I'm afraid that I do not fully understand your question.

[Fahlin Racing]

I believe you answered it. I was just thinking about wet-flow (or any extremely rich charge of air/fuel) could/would flow into the flame kernal and stop the burn in a particular cylinder, which you mentioned, malfire.

[David Redszus]

Wet fuel (droplets) that enter the combustion flame front without sufficient oxygen to support combustion will produce...
Carbon deposits.

[Fahlin Racing]

Wet fuel (droplets) that enter the combustion flame front without sufficient oxygen to support combustion will produce...
Carbon deposits.

I am thinking this situation brought up the development of nitrous?

[David Redszus]

Wet fuel (droplets) that enter the combustion flame front without sufficient oxygen to support combustion will produce...
Carbon deposits.

I am thinking this situation brought up the development of nitrous?

The impingement of unburned mixture into and behind the flame front will occur with or without the use of nitrous or any other oxygenate. Impingement is a function of mixture turbulence. Carbon formation is a function of incomplete burning, either as a result of insufficient oxidant or a lowering of temperature when in contact with a cooler surface.