Twin plug question
 

At what point do you need to consider twin plugging. And why not just split the spark from a standard distributor instead of a custom or very expensive dual distributor.

I'll be building a engine for my SWB at some point, I have 3 long blocks to chose the original Aluminum case, a 73.5 7R, and a 2.7 race prepped case.

Electricity takes the path of least resistance... so splitting the spark doesn't ensure both plugs fire with the same energy, if at all. Additionally, providing multiple paths changes the resistance, which in turn changes the voltage across the gaps.

If the point of twin-plugging is to get better control of combustion propagation in the chamber and prevent secondary ingnition (knocking), then unpredicatable primary ignition seems to be the last thing you would want.

Put another way, if it were that easy then there wouldn't be so many other options for twin plugging.

I knew there would be a good answer to this, I just didn't know what it was, Thanks.;)

With the original 911R/ 906 spec motor they used twin plugs because the piston had a high dome and the head design not optimal. but what if you are using later heads, like from a 71T, that flow better, and compression that is 10-1 or less. Cost no object , great, but it looks like twin plugging adds at least $3,500-4,000 to the build. I'm trying to be a bit more economical these days while still being extravagant.:D

Short answer - under 10:1 and <98mm bore you should be OK. A mild compression 2.7 should be fine. High compression not so much, same for a big-bore motor.

Of course, if you have the money you could always spring for the machine work for the twin-plug and just put old plugs in until you can splurge for your choice of spark options. If you are in the upper range of safe single plug compression (running conservative timing) you will benefit from twin plugging (and more aggressive timing).

Explanation follows:

Best info I can give you is out of Wayne's book + some general physics. The hemi-style combustion chamber (which Porsche's have, so do Harleys, it ain't just Dodge) with its high dome and off-center spark plug isn't great for flame front propagation, at least when compared to the more modern 4-valve center-plug design.

For a low compression engine, this is less of a problem. In order to increase compression, the piston dome is raised... to the point where it can actually prevent the flame front from crossing to the far side of the cylinder if the spark-advance isn't high enough. This of course raises the risk of spark knock. So, if you aren't running high compression (under 10 is usually the "magic number" but its actually dynamic compression which is also tied to cams, and the ability to keep the chamber temps in check) then the flame should easily cross the chamber.

With the large chambers (>98mm), the key problem is the distance between the spark plug and the far side of the chamber. Again, the options are either a controlled second flame front or more advance to get the flame to cross the chamber.

The later heads (2.2-2.4-2.7) are not as octane sensitive as the 2.0 heads so you can get away with a bit more compression before requiring twin-ignition on pump gas. Bore size & camshaft profile plays a big role here, as well. :)

People having access to 93 octane gas have more flexibility than ones with only 92, too.

More information on this subject can be found here http://www.rennsportsystems.com/2a.html

If you have your heads off just have them drilled for double plugs, then you have options for later

Aren't there some increases in HP due to a reduction in negative work done during the cylinder pressure cycle (because there is less timing advance required)?

Given the above, it is harder to get that last little bit of compression with a second plug drilled into the combustion chamber, could this negate the above?

If the advantage only goes to decreased octane requirement, then if one is building a track engine for use on 110 or greater octane fuel why run twin plug at all - especially when the compression obtained could be higher with a given piston, not to mention doing the job with a cheaper, simpler, lighter ignition system?

Thoughts on these points...

I am not sure how a spark plug would take away from compression or C/R.
With Hemi style chambers the big problem is getting the flame front over the top of the piston crown.
With twin plugs this problem is circumvented and as a plus the timing can be back offed with out performance problems

That said, twin plugging 9.5:1 and above will see great benefits in your ability to time the ignition ideally. With twin plugs, you will need much less advance than a single plug application and get snappier throttle response and some power increase.

This isn't my thinking, this is from Steve Weiner @ Rennsport Systems, so you can take it as gospel. :)

Higher octane gasoline doesn't burn significnatly faster than the lower octane stuff... just more predicatably and w/ less chance of unintentional detonation. As a result, even with race gas you still have to get the flame front across the cylinder to the far side. With high domed pistons this requires a lot of advance - hence the negative work, or for high enough RPMs just can't be done - resulting in unburned fuel and lost power.

Ignition timing is complicated... almost a black art, but the science is fairly straight forward. The goal is to fire the plug early enough that peak pressure occurs just after TDC. This isn't, the only consideration though. Others include
- complete combustion - unburnt fuel/air is wasted power
- knock protection - from the spark side not allowing pressure/temps to exceed fuel autoignition conditions (higher octane, higher autoignition characteristics)
- Emissions - the higher the combustion temp, the more NOx produced.

To complicate matters, a lean fuel mix burns faster, a rich one slower. A rich mixture is also initially cooler, as the fuel absorbs heat during vaporization. This is why going lean is so dangerous - your chances of detonation go way up as the mix is at a higher initial temperature and generally more volatile. Its also why vacuum retard was added - it is partially tied to leaning out the idle/part throttle while the fuel system leans out the mixture for emissions & fuel economy reasons.

Additionally, the higher the RPM, the increased chamber turbulence better mixes the fuel/air. This is why spark advance is "all-in" early... the increased mixing increases combustion speed enough that more advance isn't needed.

Of course in a more modern spark tables take into account both RPM and either MAP or MAF inputs to correct the timing for RPM and load. These are tied to the fuel tables to run closer to the limits.

So, twin-plugging effectivly cuts the chamber in half which reduces the amount of advance needed. This in turn causes
- more complete combustion, especially at high RPMs - more power
- increased dP/dt - less pressure rise before TDC - less power lost
- more stable combustion - less mix for each plug to ignite

I'm sure its possible for you to run a "small" bore high compression in single plug configuration if you keep the timing conservative, but at that point you've created an unbalanced engine, with your ignition being the limiting factor.

In my opinion, this is no better than building a 400hp motor and trying to mate it to an unmodified 915 gearbox... you won't be able to use it to its potential.

Or what he said. ;)

The second plug drilling looks to me to add about 1 cc to the combustion chamber volume. See below photo of a plug recess.

One extra cc in the last engine I built would have lowered the CR from 10.2:1 down to 10:1

http://forums.pelicanparts.com/uploa...1245428270.jpg

True, I do not see that as a detriment. ( with the plug installed and all,, are you sure it would come dwn that much ? Just flip the head over with the plug in the hole and see how much you can get in there B4 running over )

Response to ^^^

I don't have a head in hand to check right now. There is not too much plug projecting into the recess though, just the electrodes and a bit of the porcelain insulator.

One CC makes a 0.2 point difference at the 10.2:1 C.R. level of the example engine I gave, the sensitivity to volume increases as the C.R. goes up though, adding one CC would have dropped the C.R. by 0.5 points if my engine were at 11.5:1.

The twin plug engines I have built do have great throttle response and seem to run smoother, but that is a subjective thing.

For street engines, the sensitivity to knock is decreased with twin ignition and one can use lower octane fuel for a given compression ratio. This is a huge advantage I think for the hot rodder.

The lower NOx of of the twin plug is good for smog emissions- something hot rodders do not worry about.

So twin plug has definite advantages.

However, if an engine were built that was to run on 110 octane fuel only, where knock was not an issue, and the choice was 11.5:1 C.R. and single plug or 11:1 CR and twin plug, I am not sure which would make the greater power.

We always use the 12mm plugs for the second and use a very small relief, one cc or less may be about right, then if you take off 3 to 5 thou off the sealing surface it may be about a wash.

http://i30.photobucket.com/albums/c3...r91/27head.jpg

To complicate matters, a lean fuel mix burns faster, a rich one slower.


Where did this idea come from ? This is incorrect, in fact the opposite is true. Lean mixtures burn slower. One demonstration of this is the lean intake backfire, which many CIS 911 mechanics are familiar with. The cold running mixture is too lean, for a variety of reasons, and the mixture is still burning when the intake valve opens and the fuel in the intake ignites. The fact that lean mixtures burn more slowly is the reason vacuum advance, not retard, is used. The less dense mixture at part throttle, low load cruise needs more time to burn completely and the engine can tolerate the extra advance because the cylinder pressure is low. This results in higher efficiency. Vacuum retard is strickly an idle H/C emission test device, it works mainly because of temperature, and has nothing to do with mixture formation. Ignition timing does not effect mixture preparation, it is adjusted in response to it. Conditions in the cylinder vary considerably under different loads and many absolute statement are false because of it. What happens at rest under idle conditions is largely irrevelent to what happens under load. Unfortunately, many misconception are created by playing with engines at idle.

I twin plugged a 2.7 head with the big plug recess volume like the one above once and ended up with a 69 cc combustion chamber volume. It's hard to make a high C.R. starting with that.

This is contrary to what I have known for years. Without getting into a long and complicated discussion, can you provide proof of your claim. I think you misunderstand the entire concept, but I have also been wrong before. I am eagerly awaiting further explanation of your theory.

I am quite surprised that someone has not responded to your claims already, unless you are correct. On the CIS engines, I thought the mixture burned too fast and backfired into the intake before the valve closed. ????

On the CIS engines, I thought the mixture burned too fast and backfired into the intake before the valve closed. ????


Hello Ed,

On a 4 cycle SI engine, the spark plug does not fire until after the intake valve closes at the beginning of the compression stroke. How does that fit with your suggestion ?

If you want to clear up your misunderstanding, I suggest a basic book on combustion science or BOSCH's "Gasoline Engine Management". Like I said, most absolute statements about SI engine are false, because conditions vary so much the answer is usually not a single point or linear relationship, but a peak with contradicting curves on either side. For example, the generalization that "leaner mixtures are hotter" by itself is false. A lean mixture by definition is above an AFR of 14.7:1. EGT actually peaks around stoich and falls off as the mixture gets leaner. One reason pilots are trained to use "rich of peak" and "lean of peak" fueling. "Leaner is hotter" is only true rich of stoich.

If you want to understand fuel burn rates, there is a good chart on page 127 of Jeff Hartman's EMS book. Almost any basic text on fuel and ignition timing will state that a rich mixture burns faster than a lean mixture. Flame speed can vary from 20 to more 100 ft/sec. Burn rate is highest at a rich mixture near 11.1:1 AFR and drops as the mixture leans. It also drops as the mixture gets richer than 11:1, but in a N/A SI engine, you are generally above this outside a subzero cold start. Understanding an ignition timing map can also help. The spark plug is fired to create peak cylinder pressure around 14 degrees ATDC to maximize work done. The burn rate varies with mixture, but is relatively constant for a fixed mixture. Obviously, as rpm increases, you need to fire the plug earlier, because the window of time to accomodate the burn speed gets shorter. Lets say you determine that 30 BTDC give best torque under WOT. Now, at part throttle, low load cruise, the same ignition timing map will give another 10-15 degrees of timing advance. Why ? Because the leaner, less dense mixture supplied under light load takes longer to burn and if the ignition point was not advanced, more of the energy in the fuel would be wasted in the exhaust. If a lean mixture burned quicker than the rich mixture supplied at WOT, the ignition timing would not need to be advanced.

Paul - thanks for the correction. I appreciate having my mistakes/misunderstandings pointed out so I can learn.

Paul - after reading your follow-up, I think I found part of my confusion...

For a given load condition, there's lean (below stoich), stoich (14.7 to 1), and rich. That said, you don't really want to run stoich at higher loads, so even though its technically still rich, its refered to the engine as "running lean" as in "its leaner than I'd prefer." I know going lean causes all kinds of issues with turbo engines (much moreso than NAs because of intake temperatures).

Also, doesn't the cooling effect of fuel come into play - the heat absorbed by fuel vaporization helping to reduce the chance of detonation? Perhaps that effect is adding to the confusion - all things being equal lean burns slower, but lean and hot vs. rich and cool one effect offsets the other, at least with respect to detonation? Just a though.

Very interesting. Looks like I need to go back to the books. Thanks for the info Paul.

edit:
Got it. Lean mixtures burn slower which lengthen the combustion time which causes more heat throughout the combustion process thereby raising EGT (exhaust gas temps). My thinking has been flawed on this for years.

Paul - after reading your follow-up, I think I found part of my confusion...

For a given load condition, there's lean (below stoich), stoich (14.7 to 1), and rich. That said, you don't really want to run stoich at higher loads, so even though its technically still rich, its refered to the engine as "running lean" as in "its leaner than I'd prefer." I know going lean causes all kinds of issues with turbo engines (much moreso than NAs because of intake temperatures).

Also, doesn't the cooling effect of fuel come into play - the heat absorbed by fuel vaporization helping to reduce the chance of detonation? Perhaps that effect is adding to the confusion - all things being equal lean burns slower, but lean and hot vs. rich and cool one effect offsets the other, at least with respect to detonation? Just a though.



Hello Chris,

You make these statements with such confidence, you should really consider running for public office..... You are still confusing the basics and there is no offset of effects in regard to detonation. Detonation by definition is the spontaneous combustion of "unburnt fuel". It is precisely because your previous statement (burn rate) is false that a rich mixture is less likely to cause detonation than a lean mixture. The rich mixture burns much faster and therefore there is no "unburnt fuel" to spontaneously combust to cause detonation. The lean mixture is burning slowly, raising temperature and pressure, while the "unburnt fuel" hangs around waiting to be burned or explode.

When discussing fuel mixture, it can be expressed many ways, AFR, lambda, fuel air ratio, by weight and volume. When using AFR, like 14.7:1, lean is above, not...lean (below stoich). If you are not afraid of textbooks, Stone's IC Engines has a good section on combustion science and for lighter reading Hartman did most of his homework on this subject for his books.

Paul - "public office" - uh... not a chance in hell. I work in/for the government, military actually, and the politics here have conviced me that I don't ever want to be a politician. I'm only renting my soul to Uncle Sam, not selling it.

Sorry if I come off as overly confident, I don't intend to. For whatever reason text, without tone of voice & all the nonverbals, doesn't do a good conveying anything but the literal words (maybe I should use more :)s). Part of it is probably the military thing, they teach us to write/speak with confidence.

Back on topic...

I'm not afraid of textbooks... I just don't necessarily have access to them. I have more "free time" lately (hence the more frequent posting) because I'm getting ready to move and work isn't loading me down with long term work. If I can find a copy, I'll take a look.

So, to make sure I understand - because the rich mix burns faster, the flame progression "outruns" the rise in temperature and pressure and is able to consume all the mix. In a lean burn, the flame front travels slower, allowing the rising temperature and pressure to destabilize the remaining mix, which causes detonation.

This effect is amplified in high compression motors because the higher compression (that sounds redundant) causes increased temp/pressure even pre-spark. Additionally, the hemi-style combustion chambers Porsche uses (used?) because the offset spark plug leads to a long flame travel (single plug) and the high-dome piston isolates the mix on the far side.

Have I got this right now, or am I still somewhere off in left field?

Ed wrote,

Got it. Lean mixtures burn slower which lengthen the combustion time which causes more heat throughout the combustion process thereby raising EGT (exhaust gas temps). My thinking has been flawed on this for years

Hello Ed,

Almost there. The statement you made is only true within a certain range, in this case, rich of stoich (14.7:1 AFR). EGT vs mixture is a curve, it peaks around stoich and goes the other way (drops) as you lean. Leaner is cooler, lean of stoich. One reason absolute EGT alone is not a good reference and not a good indication of abnormal combustion or detonation.

Hello Chris,

Just kidding. Yes, rich burns faster. The slow combustion problem can be clarified somewhat by considering the diesel engine. The cylinder temperature is so high from the compression that the injected fuel auto ignities immediately and is burned very quickly, so there is no end gas waiting to be burned. Turbo diesels can be boosted up without regard to pressure spikes from abnormal combustion and will continue making power as long as more fuel is supplied. A chart that compares the actual in cylinder pressure under normal combustion and under detonation is eye openning. The Hemi head with a wide VIA has an interesting history. Pent roof 4 valve engine were used in racing as early as 1912, but FIAT introduce a GP 2 valve hemi with a supercharger in 1923. Racing is a sport of copying what works and the FIAT hemi was copied by everyone and the supercharger masked the superiority of the 4 valve pent roof until Honda did the science in the early 60's, then everyone copied them. The basic problem with the hemi is that you have to use a wide VIA to fit larger enough valves, this raises the combustion chamber and increases volume, which requires domed pistons to get a decent N/A compression ratio. The Italians used VIA's over 90 degrees, Jaguar brought it down to 70 in the 40's, the 911 is 59, Lotus down to 54 . The shrouding of the spark plug is less of an issue when you are running a low compression supercharged engine, especially on alcohol based fuels. The N/A hemis of the 50's ran twin plugs, the Maserati 250F ran as high as 15:1 compression and it's power output was largely a chemical experiment. Twin plugging does not "benefit from more aggresive timing", it needs less to get the job done and more timing is not desireable.

There are some good articles on the web for pilots, who actually adjust mixture under load, rather than turn idle screws. Try 43, 18, 63 and 64 here:

http://www.avweb.com/cgi-bin/texis/scripts/avweb-search/search.html?query=detonation+myths&x=29&y=9

Part of Jeff Hartman's EMS book is here but some of the best stuff you have to pay for.....

http://books.google.com/books?id=CaAP6B4TspcC&pg=PA34&lpg=PA34&dq=hartman+ mixture+egine+management&source=bl&ots=DzzgjRyftW& sig=GHyDvWznOw1j6mQSLngZZrkzmh0&hl=en&ei=IWpCSoXML pWltgfO7fHRCQ&sa=X&oi=book_result&ct=result&resnum =1

Many thanks.

Off topic, but as I understand, in diesels the main problem is getting the fuel actually exposed to the air - the outside of the droplets combusts but the inside is deprived of air exposure. Thus, the limiting factor is usually fuel atomization or structural integrity, which is why the modern common-rail TDIs have such high injection pressure. Once you hit that fuel-atomization limit you get the black smoke, but even then the overall combustion is lean.

Back on subject - I figured alcohol (or other octane boosters) would reduce the need for twin plugging. I think I understand what you are saying. Optimum timing is when you get peak pressure right after TDC. Depending on conditions, you may need more or less advance, but you will always need more single plug compared to dual plug because the volumes involved. Dual plug provides a more rapid rise in pressure causing less negative work from rising pressures before TDC. Because this rise is faster, and there's less shrouding of the spark plug, it reduces the chance of unburnt pockets of mixture that can detonate.

More "agressive" timing then would only be desireable only if it can drive peak pressure closer to (but not before) TDC to extract as much energy as possible? The increase in cylinder pressure & temperature from the earlier spark can cause problems with lower quality fuel. I ask because in my other car (also NA), I can purchase a remapped chip that requires premium fuel, and would buy me a few extra horses by adding more timing advance and remapping the fuel. This of course decreased fuel efficiency. Of course, this could just be marketing BS.

Thanks for the "light" reading :) I'll add it to the list of stuff I'll have to start reading for my MS in Astro Engineering.

Chris wrote,

Optimum timing is when you get peak pressure right after TDC.... More "agressive" timing then would only be desireable only if it can drive peak pressure closer to (but not before) TDC to extract as much energy as possible?

Hello Chris,

No, the ideal time for peak pressure is around 14 ATDC for mechanical advantages, not before. Twin plugs start the burn from both sides, and the peak arrives earlier, so you need less advance to get the peak at 14 ATDC. The shape and size of the combustion chamber effects the time it takes to reach the peak and you adjust the timing advance to arrive at the same place. Peak output with less advance means a better combustion chamber.

One reason for 24,000 psi high pressure diesel injection, is that once you go over 4000 rpm, the window of time is very small to inject all the fuel. Air cooled engines love alcohol fuels, it is a very effective way to get heat out of the cylinder head. There are some interesting E85 turbos.

Ahh, the small turbo-diesels with their "high" redlines. My Dad has a 7.0L F350, its redline is mighty low, but its got torque :).

14 deg ATDC -- got it.

I was going to ask if anybody had built an E85 930 but figured it was off-topic. I do know the Koenigsegg has CCX variant running on E85 that generates more power. SAAB also has a turbo (concept I think) that increases the boost via electronic wastegate when the ECU detects E85. Both are interesting concepts as "they" keep increasing the ethanol content in fuel.

Once you get by the corn politics, ethanol becomes much more interesting. MIT has a paper that indicates that a small amount of ethanol injected directly into the cylinder allows much high boost pressures and advance with much lower detonation. Ethanol also allows a much higher EGR dilution than gasoline, which is effectively a reduction in displacement on the fly. You will soon see something like a 5 liter V-8 , with cylinder deactivation and high (+25%) EGR dilution that can have the efficiency of a 1.8 liter 4, and the power of +1 bar boosted 5 liter. All with a single plug........