Dumb QOTD: Why does load induct pre-ignition / detonation?

[NewbVetteGuy]

I really feel like this is one of those things that I SHOULD understand and it's going to be obvious when someone says the answer, but I don't understand why pre-ignition / detonation happens under load...


In my understanding, better cylinder filling, and hotter air temperatures and coolant temperatures should result in higher combustion temperatures and therefore moving closer to pre-ignition / detonation. I'd totally expect detonation to happen at wide open throttle, at tuned induction and exhaust RPMs, and I get why it's more likely at lower RPMS, but I just don't understand how load contributes...

Can someone please just state what appears to be obvious to everyone else for me?



Adam

[David Redszus]

I really feel like this is one of those things that I SHOULD understand and it's going to be obvious when someone says the answer, but I don't understand why pre-ignition / detonation happens under load...

In my understanding, better cylinder filling, and hotter air temperatures and coolant temperatures should result in higher combustion temperatures and therefore moving closer to pre-ignition / detonation. I'd totally expect detonation to happen at wide open throttle, at tuned induction and exhaust RPMs, and I get why it's more likely at lower RPMS, but I just don't understand how load contributes...

Can someone please just state what appears to be obvious to everyone else for me?

Adam

Pre-ignition is a function of temperature.

An increase in load is an increase in BMEP.
An increase in BMEP of 3 BAR will raise piston temperature by 54F deg.
Similarly, an increase in TCR of 3 points will raise piston temperature by 54F deg.

If the temperature rise causes the chamber temperature to exceed the fuel auto-ignition
point, we have pre-ignition.

[David Redszus]

Pre-ignition occurs before the spark plug fires.

Detonation typically occurs after.

Thanks
Randy

While that is typically true, there are exceptions. Anything that moves the combustion pressure curve
leftward can be considered pre-ignition regardless of when the plug actually fires.
A reduction of the ignition delay period is an example.

[BLSTIC]

Are you confusing "Load" the tuning parameter (i.e. manifold pressure/TPS) and "Load" the resistance on the engine (i.e. vehicle weight resisting acceleration)?

"Load" the resistance affects detonation in that it introduces *time* at some particular rpm, which strengthens any particular waveform being generated (as the wave forms have an opportunity to repeat and get stronger) and heats up things like cylinder walls, spark tips, valves, etc because it's been under... "load" the tuning parameter... longer.

It's why some engine management parameters have reduced timing in high gears/speeds/after long periods at high power. Because in a top speed run all the resonances have been stacking for as long as they can and everything slowly creeps hotter and hotter

[Nut124]

Why is it that WOT at 2800rpm, 32deg ignition, uphill, knocks but 6000rpm, WOT, @ 36deg ign advance runs great?

Now, I cannot say for sure that the ignition advance is as stated, but it does seem that low-mid rpm is more likely to knock, ping. Why?

[Schurkey]

Why is it that WOT at 2800rpm, 32deg ignition, uphill, knocks but 6000rpm, WOT, @ 36deg ign advance runs great?

Now, I cannot say for sure that the ignition advance is as stated, but it does seem that low-mid rpm is more likely to knock, ping. Why?

1. Greater cylinder filling at lower RPM (Higher cylinder pressure)
2. More time for combustion chamber temperature to heat the air/fuel mixture at lower RPM. (Higher air/fuel mix temperature)
3. Less turbulence in the chamber at lower RPM. (Slower "controlled" burn)
4. Fuel/Air mixture differences in the RPM range (potentially leaner at some rpm)

[Big Al]

How often does pre-ignition really happens compared to knock after the spark?

[modok]

If the engine RPM is higher then....
either the combustion is happening faster, or,
it isn't.

if it isn't then the peak pressure will be lower, assuming the peak is still timed in the right general degrees

if it is happening faster then why? how? well, probably from improved mixing, and that is also reducing the possibility of detonation.

[David Redszus]

There are a number of engine design parameters that will affect the octane number
requirement to avoid knock. They are:
1. Fuel composition and sensitivity
2. Combustion chamber geometry
3. Charge motion
4. Spark advance curve
5. Temperatures: inlet air, manifold, water
6. Fuel ratio and mixture preparation
7. Ambient conditions: pressure, temperature, relative humidity
8. Load
9. Operating speed

How often does pre-ignition really happens compared to knock after the spark?

Good question. Knock can be detected by combustion pressure deltas, or by frequency.
Knock sensors typically use frequency as a basis for detection.

But the detection of pre-ignition is a bit more difficult; we have no field grade sensors for that job.
Of course, pressure measurement and ion sensing can always be used.

The problem is complicated by the fact that pre-ignition, often called the silent engine killer, can
lead to detonation as temperatures rises.

Pre-ignition will have a considerable impact on the ignition delay period which is the main
cause of combustion instability.

[MadBill]

Why is it that WOT at 2800rpm, 32deg ignition, uphill, knocks but 6000rpm, WOT, @ 36deg ign advance runs great?

Now, I cannot say for sure that the ignition advance is as stated, but it does seem that low-mid rpm is more likely to knock, ping. Why?

"When unburned fuel/air mixture beyond the boundary of the flame front is subjected to a combination of heat and pressure for a certain duration (beyond the delay period of the fuel used), detonation may occur. Detonation is characterized by an almost instantaneous, explosive ignition of at least one pocket of fuel/air mixture outside of the flame front. A local shockwave is created around each pocket, and the cylinder pressure will rise sharply – and possibly beyond its design limits – causing damage."

A major factor is the above-mentioned short but fixed delay period. At 2,800 RPM, the delay time is exceeded and so detonation occurs. A t 6,000, the normal flame front has time to sweep through the region before the countdown hits zero, thus no detonation.

[hoodeng]

Why is it that WOT at 2800rpm, 32deg ignition, uphill, knocks but 6000rpm, WOT, @ 36deg ign advance runs great?

Now, I cannot say for sure that the ignition advance is as stated, but it does seem that low-mid rpm is more likely to knock, ping. Why?

Spark knock or detonation can occur over the full RPM range. Spark knock occurs over a number of crankshaft degrees of rotation, these degrees of rotation take shorter time to occur as RPM increases.

Under 3/3.5K RPM spark knock is typically audible, after that you can not hear it even though it can still be occurring, your dog though will hear it as it has a higher hearing range of up to 60 k/htz, we are only good for about 20 k/htz.

This is why it is commonly stated by people describing the condition as saying it went away at 3/3.5K RPM.

Guys that run dyno's will tell you their knock sensors can fire at all sorts of RPM, they can also tell you certain engine/chassis noises can be picked up and interpreted by knock sensors as spark knock.

Cheers.

[4sfed]

For a particular fuel, auto ignition is caused by a combination of three variables ... time, temperature and pressure. Increasing throttle position increases cylinder pressure, low engine speed increases the time available for auto ignition, and charge temperature will be higher at low engine speed (heat soak).

Auto ignition can occur before the plug fires (preignition), or in a secondary location after the spark (detonation), as pressure and temperature build in the chamber.

[naukkis79]

der 3/3.5K RPM spark knock is typically audible, after that you can not hear it even though it can still be occurring, your dog though will hear it as it has a higher hearing range of up to 60 k/htz, we are only good for about 20 k/htz.

This is why it is commonly stated by people describing the condition as saying it went away at 3/3.5K RPM.

That's not how knock sensors are tuned. Knocking results block vibration, so knocking sensor is tuned to block vibration frequency. Knock listening happens at same frequency at whole rpm range.

Knocking, if happens at high rpm range is still very audible but knocking probability is much greater at lower rpm range. Engine can also be so loud in high rpm that it masks knocking, but using something like stethoscope to block helps hearing.

[BCjohnny]

All else being equal one major cause of detonation, at low rpm & large throttle openings, my understanding is .......

Flame speed 'outruns' piston descent causing a pressure spike of the residual unburnt charge, commonly called 'end gas', exceeding the knock rating (octane) of the fuel producing detonation ...... even with otherwise 'correct' ignition timing

The 'fuller' the inducted combustible charge, the greater the pressure rise relative to piston descent (crank angle)

This is why some chambers 'knock' more than others ..... 'dirty' quench chambers don't leave enough charge to cause a problem, 'cleaner' open chambers burn relatively more fully ...... it's just the inbetweeny ones where the end gas gets funnelled into a 'partial' taper, that seem to give the most problems ....... generally speaking anyway

In most cases simply reducing the throttle opening, and thus the mass of the combustible charge, lessens the pressure rise relative to piston descent such that the 'end gas' doesn't detonate

Pre-ignition is what it is, normally a chamber 'hot spot', leading to 'compression ignition' prior to the spark timed event, due to localised elevated temperature exceeding the knock rating of the fuel ..... which can then lead to triggering the above

If only it were so simple ...... and obviously all influenced externally by other factors

[David Redszus]

Detonation can be detected by the vibration that is produced in the head and block.

The frequency of the knock vibration will depend on chamber temperature and
chamber geometry and is not a function of rpm.

Both chamber shape and dimension will determine the knock frequency.

For a chamber sized 80mm x 50mm x 20mm, the knock frequencies will be
6.6KHz, 10.5KHz and 35KHz respectively. Knock frequencies are not rigidly fixed and
will vary by about 400KHz for the same cylinder. The knock frequency in the
vertical axis will vary with piston position which changes with crank angle.

Some knock sensors are tuned to a specific frequency range while others can measure
a broad frequency band. If a tuned knock sensor is used, it will produce a voltage at
the selected frequency which can be fed into a data logger or simple ignition retard system.

If a broad band knock sensor is used, it will pick up all frequencies, knock or not, and will
require an ECU to read and respond only to selected frequencies.

Typically, knock will occur at the torque peak rpm since the combustion process is nearly
optimized and temperature and pressure are highest. But, the propensity to knock also
depends on the octane sensitivity of the fuel. With identical octane ratings, some fuels
will knock at lower speeds while other only at higher speeds. A low rpm knock means
little in a race engine, but avoiding high speed knock becomes critical.

Normally, one or two cylinders will have a tendency to knock while the others do not.
Once identified, the problem can be corrected with timing or fuel adjustments.

Knock is often intermittent and will randomly walk from cylinder to cylinder.
This makes knock control much more difficult.