Surface ignition (pre-ignition) vs. Knock.. why more damage with one vs. other?
03-03-2005, 08:04 AM
#1
Honda-Tech Member
Thread Starter
Surface ignition (pre-ignition) vs. Knock.. why more damage with one vs. other?
I've had alot of time to kill these days, thank god I still have accessible internet at work 
Anyways, I've been doing alot of thinking and reading up latley, particularly on cylinder pressures and I noticed something that made curious.
My understanding (and I know this is a 'grey' area with so many people placing different definitions on different things, bare with me) with regards to knock is that it is the uncontrolled and spontaneous ignition of end gases which the controlled flame front has not yet reached. Pressure and tempature increases from the flame generation and travel cause these end gases to lose stability, increase is volitility and self-ignite creating large pressure oscillations in the combustion chamber.
Now if what I just stated is correct and agreeable, move on. If not, i'm open for further clarification
On the other hand is surface ignition or 'pre'-ignition, basically an ignition source other than the spark plug i.e. overheated valves, carbon deposits etc.. Now here i'm a little shady, but it seems to me based on my reading that such auto-ignitions are not nearly as damaging, at least on their own accord, as the kind of pressure spikes seen via knock. I did however read (and this one really supprised me, never thought if it..) that the increase in pressure/tempature of the unburned end gases, affected both by the controlled flamefront generated by the spark plug AND by the rogue flame(s) generated by the pre-ignition event, can actually lead to a knock event at the SAME TIME.
Basically, pre-ignition leading to knock.. whoa.
Now my question is, why is the igntion of the end gases in a knock event more damaging than the ignition of end gases during a pre-ignition event? What makes the massive pressure spike generated by early spark so much greater than one that wouldn've been set off by a random hot spot?
Sorry for the length of the quesiton but I just had to get this out there
Anyways, I've been doing alot of thinking and reading up latley, particularly on cylinder pressures and I noticed something that made curious.My understanding (and I know this is a 'grey' area with so many people placing different definitions on different things, bare with me) with regards to knock is that it is the uncontrolled and spontaneous ignition of end gases which the controlled flame front has not yet reached. Pressure and tempature increases from the flame generation and travel cause these end gases to lose stability, increase is volitility and self-ignite creating large pressure oscillations in the combustion chamber.
Now if what I just stated is correct and agreeable, move on. If not, i'm open for further clarification
On the other hand is surface ignition or 'pre'-ignition, basically an ignition source other than the spark plug i.e. overheated valves, carbon deposits etc.. Now here i'm a little shady, but it seems to me based on my reading that such auto-ignitions are not nearly as damaging, at least on their own accord, as the kind of pressure spikes seen via knock. I did however read (and this one really supprised me, never thought if it..) that the increase in pressure/tempature of the unburned end gases, affected both by the controlled flamefront generated by the spark plug AND by the rogue flame(s) generated by the pre-ignition event, can actually lead to a knock event at the SAME TIME.
Basically, pre-ignition leading to knock.. whoa.Now my question is, why is the igntion of the end gases in a knock event more damaging than the ignition of end gases during a pre-ignition event? What makes the massive pressure spike generated by early spark so much greater than one that wouldn've been set off by a random hot spot?
Sorry for the length of the quesiton but I just had to get this out there
03-03-2005, 08:20 AM
#2
Honda-Tech Member
Join Date: Jan 2004
Location: Vancouver, B.C., Canada
Posts: 1,323
Likes: 0
Received 0 Likes
on
0 Posts
my understanding.
pre ignition = the A/F mix lighting up before the piston reaches TDC and tries to fire the piston in reverse, while damaging, its not as bad as detonation because the piston and rod still have room to travel an only serves to slow down the rotation of the motor
Detonation on the other hand is when the A/F mix fires exactly when the piston/rod is at TDC. this forcces the piston and rod straight down on the crank and squishes your berings and lifts the head causing massive damage and WAY higher cylinder pressures. sometimes this cracks the cylinders and pistons and/or bends rods.
pre ignition = the A/F mix lighting up before the piston reaches TDC and tries to fire the piston in reverse, while damaging, its not as bad as detonation because the piston and rod still have room to travel an only serves to slow down the rotation of the motor
Detonation on the other hand is when the A/F mix fires exactly when the piston/rod is at TDC. this forcces the piston and rod straight down on the crank and squishes your berings and lifts the head causing massive damage and WAY higher cylinder pressures. sometimes this cracks the cylinders and pistons and/or bends rods.
03-03-2005, 08:31 AM
#3
Re: Surface ignition (pre-ignition) vs. Knock.. why more damage with one vs. other? (VTC_ohgodnotaga
I dont agree with all the specifics of that. I believe the differences have to do with what caused the uncontrolled combustion, rather than its relation to TDC.
uncontrolled combustion is bad mmm kay
uncontrolled combustion is bad mmm kay
03-03-2005, 08:54 AM
#5
Honda-Tech Member
Join Date: Jun 2003
Location: DFW, TX, USA
Posts: 197
Likes: 0
Received 0 Likes
on
0 Posts
03-03-2005, 09:17 AM
#6
Honda-Tech Member
Thread Starter
Re: (Turbo E)
<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by Turbo E »</TD></TR><TR><TD CLASS="quote">my understanding.
pre ignition = the A/F mix lighting up before the piston reaches TDC and tries to fire the piston in reverse, while damaging, its not as bad as detonation because the piston and rod still have room to travel an only serves to slow down the rotation of the motor
</TD></TR></TABLE>
I think I see what you're saying, but afaik the mixture is ALWAYS fired before TDC regardless, it's the when the end gases ignite on their own that apparently causes the massive spike in cyl. pressure.. anybody?
pre ignition = the A/F mix lighting up before the piston reaches TDC and tries to fire the piston in reverse, while damaging, its not as bad as detonation because the piston and rod still have room to travel an only serves to slow down the rotation of the motor
</TD></TR></TABLE>
I think I see what you're saying, but afaik the mixture is ALWAYS fired before TDC regardless, it's the when the end gases ignite on their own that apparently causes the massive spike in cyl. pressure.. anybody?
Trending Topics
03-03-2005, 10:30 AM
#8
Member
Re: Surface ignition (pre-ignition) vs. Knock.. why more damage with one vs. other? (VTC_ohgodnotaga
Pre ignition = normal burn that takes place before optimal time. Sometimes caused by hot spots, etc.
Detonation = uncontrolled explosive ignition of the air/fuel mixture. Usually caused by incorrect air/fuel ratio, etc.
Detonation = uncontrolled explosive ignition of the air/fuel mixture. Usually caused by incorrect air/fuel ratio, etc.
03-03-2005, 11:11 AM
#9
Honda-Tech Member
Thread Starter
Re: (mopar junkie)
<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by mopar junkie »</TD></TR><TR><TD CLASS="quote">Read that link I posted. 8 pages of info written by a GM engineer. </TD></TR></TABLE>
Excellent article, i'm reading it now, but I noticed:
"..therefore LPP is always 14 degrees ATDC for any engine."
Now I was under the impression that every engine design has unique characteristics that would vary LPP?
</TD></TR></TABLE>Excellent article, i'm reading it now, but I noticed:
"..therefore LPP is always 14 degrees ATDC for any engine."
Now I was under the impression that every engine design has unique characteristics that would vary LPP?
03-03-2005, 11:49 AM
#10
Honda-Tech Member
Thread Starter
Re: (VTC_ohgodnotagain_CiViC)
Wow... there's alot of info in that article that blew my mind! I never gave burn rate too much thought, but the idea of combating detination by ADVANCING the spark, heh, like I said wow! Concepts I never thought of, thanks again for the link
03-03-2005, 12:26 PM
#11
Honda-Tech Member
You're right that optimal LPP varies.
R/S ratio is one of the biggest factors influencing it.
Also the detonation frequency he speaks of varies too. The prime frequency is determined primarily by the bore of the motor, and then harmonics occur too.
R/S ratio is one of the biggest factors influencing it.
Also the detonation frequency he speaks of varies too. The prime frequency is determined primarily by the bore of the motor, and then harmonics occur too.
03-03-2005, 12:28 PM
#12
Honda-Tech Member
Thread Starter
Re: (blundar)
<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by blundar »</TD></TR><TR><TD CLASS="quote">You're right that optimal LPP varies.
R/S ratio is one of the biggest factors influencing it. </TD></TR></TABLE>
Kinda funny, the last statement of the article is something like "beware of blanket statements".. go figure.
R/S ratio is one of the biggest factors influencing it. </TD></TR></TABLE>
Kinda funny, the last statement of the article is something like "beware of blanket statements".. go figure.
03-03-2005, 12:39 PM
#13
Honda-Tech Member
Join Date: Jun 2003
Location: DFW, TX, USA
Posts: 197
Likes: 0
Received 0 Likes
on
0 Posts
Re: (VTC_ohgodnotagain_CiViC)
<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by VTC_ohgodnotagain_CiViC »</TD></TR><TR><TD CLASS="quote">
Now I was under the impression that every engine design has unique characteristics that would vary LPP?</TD></TR></TABLE>
I agree with you. Other than than his PPP idea I believe the rest is pretty accurate.
Now I was under the impression that every engine design has unique characteristics that would vary LPP?</TD></TR></TABLE>
I agree with you. Other than than his PPP idea I believe the rest is pretty accurate.
03-03-2005, 12:47 PM
#14
Honda-Tech Member
Re: (VTC_ohgodnotagain_CiViC)
to make it simple, pre ignition happens before the spark event, detonation happens after.
from what i have seen if the afr and egt's are right there is pretty much nothing you can do with the tune to stop pre ignition. the plugs are too hot, the engine is overheating, etc.
detonation is initiated by the spark plug, some of the same things that cause pre ignition can cause detonation.. the engine too hot, crappy combustion chamber, etc. but generally detonation can be tuned out by adding fuel, pulling timing, or adding octane.
on interesting thing if you are a user of aem ems.. if you monitor knock volts ( which is what they say to monitor for knock) you will not see pre-ignition. knock volts only samples the knock sensor for a specified period of time starting at the ignition event. if you hear "detonation" but dont see it in the logs, compair knock volts to knock raw and knock tooth. knock tooth will tell you what cylinder the noise occured on, knock raw is a contenous sampling of the knock sensor. if there is pre-ignition only knock raw will show it.
from what i have seen if the afr and egt's are right there is pretty much nothing you can do with the tune to stop pre ignition. the plugs are too hot, the engine is overheating, etc.
detonation is initiated by the spark plug, some of the same things that cause pre ignition can cause detonation.. the engine too hot, crappy combustion chamber, etc. but generally detonation can be tuned out by adding fuel, pulling timing, or adding octane.
on interesting thing if you are a user of aem ems.. if you monitor knock volts ( which is what they say to monitor for knock) you will not see pre-ignition. knock volts only samples the knock sensor for a specified period of time starting at the ignition event. if you hear "detonation" but dont see it in the logs, compair knock volts to knock raw and knock tooth. knock tooth will tell you what cylinder the noise occured on, knock raw is a contenous sampling of the knock sensor. if there is pre-ignition only knock raw will show it.
03-03-2005, 12:50 PM
#15
Honda-Tech Member
The stuff on page 4/5 of that article re: tuning + combustion chamber design is so overlooked in 99% of the builds that people do. *cough* listen to Larry *cough*
The comments about plugs are a very useful thought.
You get an idea of how conservatively cars are tuned from the factory by comparing some of his guidelines on page 7 to "accepted" guidelines in the tuner community.
All in all a pretty good article.
The comments about plugs are a very useful thought.
You get an idea of how conservatively cars are tuned from the factory by comparing some of his guidelines on page 7 to "accepted" guidelines in the tuner community.
All in all a pretty good article.
03-03-2005, 12:58 PM
#16
Honda-Tech Member
Thread Starter
Re: (blundar)
Definatley, I'm gonna be going over this a few times I'm sure.. I'm applying this to what happened to M.A.R.C.'s motor and trying to identify the ultimate culprit..
03-03-2005, 01:01 PM
#17
Honda-Tech Member
Re: (VTC_ohgodnotagain_CiViC)
<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by VTC_ohgodnotagain_CiViC »</TD></TR><TR><TD CLASS="quote">Definatley, I'm gonna be going over this a few times I'm sure.. I'm applying this to what happened to M.A.R.C.'s motor and trying to identify the ultimate culprit..</TD></TR></TABLE>
what'd hte pistons look like
what'd hte pistons look like
03-03-2005, 01:02 PM
#18
Honda-Tech Member
Re: (JDogg)
<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by JDogg »</TD></TR><TR><TD CLASS="quote">
on interesting thing if you are a user of aem ems.. if you monitor knock volts ( which is what they say to monitor for knock) you will not see pre-ignition. knock volts only samples the knock sensor for a specified period of time starting at the ignition event.</TD></TR></TABLE>
Monitoring knock voltage is a placebo for a real knock monitoring solution. Signal/noise goes to **** as RPM increases - no way around this. AFAIK, the AEM EMS does not do any bandpass filtering or FFT transforms of knock sensor data. Bandpass filters improve knock resolution tenfold by focusing on the characteristic knock frequency of the motor and eliminating a lot of ambient noise. Most knock sensors themselves are designed to have some form of mechanical bandpass filtering going on (like a bandpass subwoofer box) but doing additional filtering of the knock signal electronically can really help. (You really could probably build a filter to put inline between your knock sensor and your EMS to do this, btw.) Doing a FFT (Fast Fourier Transform) further improves on your signal resolution. A FFT transforms a signal into a sum of component sine waves of characteristic frequency and amplitudes. If you know your motor's characterisitc knock frequency, a FFT transform will tell you exactly how much acoustic energy is present at that frequency, and primary harmonics.
You can find a lot of good information on knock detection from the VEMS Group / MegasquirtAVR crew - http://www.vems-group.org/wiki...ction
on interesting thing if you are a user of aem ems.. if you monitor knock volts ( which is what they say to monitor for knock) you will not see pre-ignition. knock volts only samples the knock sensor for a specified period of time starting at the ignition event.</TD></TR></TABLE>
Monitoring knock voltage is a placebo for a real knock monitoring solution. Signal/noise goes to **** as RPM increases - no way around this. AFAIK, the AEM EMS does not do any bandpass filtering or FFT transforms of knock sensor data. Bandpass filters improve knock resolution tenfold by focusing on the characteristic knock frequency of the motor and eliminating a lot of ambient noise. Most knock sensors themselves are designed to have some form of mechanical bandpass filtering going on (like a bandpass subwoofer box) but doing additional filtering of the knock signal electronically can really help. (You really could probably build a filter to put inline between your knock sensor and your EMS to do this, btw.) Doing a FFT (Fast Fourier Transform) further improves on your signal resolution. A FFT transforms a signal into a sum of component sine waves of characteristic frequency and amplitudes. If you know your motor's characterisitc knock frequency, a FFT transform will tell you exactly how much acoustic energy is present at that frequency, and primary harmonics.
You can find a lot of good information on knock detection from the VEMS Group / MegasquirtAVR crew - http://www.vems-group.org/wiki...ction
03-03-2005, 01:06 PM
#19
Honda-Tech Member
And oh yeah, V4 of the VEMS product is going to use Ion Sensing for its primary means of detonation detection. Arguably, ion sensing is far superior - 15,000 RPM has just as much resolution as 1000 RPM unlike an acoustic sensor, as long as you can crunch the numbers fast enough to do the math required.
03-03-2005, 01:16 PM
#20
Honda-Tech Member
Join Date: Jun 2003
Location: DFW, TX, USA
Posts: 197
Likes: 0
Received 0 Likes
on
0 Posts
VEMS as in the megasquirt spin-off?
ION sensing is the future. Some of the megasquirt guys are working on an implementation as well.'
There's some good links in this thread. Lots of technical info that goes way over my head. :lol:
http://www.msefi.com/viewtopic.php?t=4733
ION sensing is the future. Some of the megasquirt guys are working on an implementation as well.'
There's some good links in this thread. Lots of technical info that goes way over my head. :lol:
http://www.msefi.com/viewtopic.php?t=4733
03-03-2005, 01:29 PM
#22
Honda-Tech Member
Re: (blundar)
is there going to be a standalone ion sensing setup?
yeah.. the ems's knock sensing is pretty archaic, no better than hooking a microphone up to the engine and listining through some headphones.... but its better than nothing. most of the time its pretty useless, especially in a loud motor with loose valves and alot of piston-wall clearance. its impossible to get a good base noise table because the noise is never consistent.
as for the band pass, how would you determine which frequencys to filter?
yeah.. the ems's knock sensing is pretty archaic, no better than hooking a microphone up to the engine and listining through some headphones.... but its better than nothing. most of the time its pretty useless, especially in a loud motor with loose valves and alot of piston-wall clearance. its impossible to get a good base noise table because the noise is never consistent.
as for the band pass, how would you determine which frequencys to filter?
03-03-2005, 02:02 PM
#24
Re: (blundar)
So uhh... anyone want to tell me how to hook one of these puppies up... Mitsubishi Ionic Current Detection module
03-03-2005, 02:43 PM
#25
Honda-Tech Member
Join Date: Jun 2003
Location: DFW, TX, USA
Posts: 197
Likes: 0
Received 0 Likes
on
0 Posts
I don't think hooking it up would be the difficult part. Extrapolating the data into something useful would be. The sensor doesn't kick out a knock signal. The ECU software has to derive the PPP. Unless there's something magical going on in that box...