A short study on ignition timing and combustion

Deets55 said:

Slick,

If you start the fire in the middle of the pasture would it spread outwards on two front as quickly as it would spread inward from opposite ends?
I could be way off base on this but it is something to think about.

Pete

Click to expand...

Depends on which way the wind is blowing.

Seriously, if the combustion event follows the macro model, it would seem to be the same. But as I said in the previous post, I can visualize sonic waves colliding, reflecting and enhancing the ignition event, then starting the fire at opposite ends would be advantageous.

Slick

Slick,
I'll have to take your word on that, way out of my realm of expertise. Since this is all taking place at supersonic speed, in a 3" bore ( give or take a little) I would assume the reaction to be instantaneous. I was thinking along the lines of a true hemi, where the plug is centrally located in a domed head, between the valves. Ignition starts at a central location and radiates outward equally in all directions, no worry about colliding flame front and their affects on each other. I would be curious to know what an old hemi with a modern ignition would do. Not talking about the new hemis.
I'm just happy to have fuel, air and spark meet to produce forward motion without any adverse side effects.

Pete

I suggest that the only time the reaction proceeds at sonic speeds is when you get detonation. At every other time there is more advantage in extending the time taken for the complete combustion event to take place. It is about pressure rise verses crank rotation. Forgetting the chemistry for a moment, the physics are probably more important. If you have fuel with a very high antiknock rating such as methanol, you can raise the comp. ratio much higher without getting detonation and also increase the ignition advance, however the fuel/air mixture has to be richer to get the balance right. That usually places jetting beyond the range at which a normal petrol carburettor can cope. You find that the carburettor needles for methanol in a motor running 12 to 1 comp. and standard petrol ignition advance have a concave taper and a very fine point, so that the motor can run on the main jets at full throttle. Generally speaking, the more fuel you use while the mixture does not cause detonation, the more power you derive. Normal petrol needles have a convex taper, and with lower compression ratios a petrol needle gives better acceleration at mid-throttle openings when using methanol. With petrol the situation is much more critical, it is much easier to get detonation. In both cases the optimum mixture is just slightly on the rich side of too lean (detonation), however that is a bit dependent on load. If you raise the overall gearing of the bike, you might have to change the jetting slightly.
Cam timing comes into the equation a bit. I have advanced the standard 850 cam by 12 degrees to compensate for the two into one exhaust system. If max pressure occurs at 10 degrees after TDC, and the exhaust valve opens at 88 degrees after TDC, the push on the piston occurs for 78 degrees max. compared with ? for a standard motor. The early opening operates the standing wave in the exhaust system, and probably with a stronger pulse - so is it a matter of 'what you lose on the swings, you pick up on the merry-go-round' ?

There was a device being sold many years ago called a 'fuel miser'. It consisted of a spinning vane which was supposed to sit in the port where the manifold gasket resides. It was intended to give better mixing so the combustion would be more event would be more controlled. I always thought it was 'snake oil'. It would simply stuff the gas velocity in the port.

Acotrel,

Makes me wonder, is it possible to achieve a perfect environment for combustion, with such imperfect external conditions. Somewhere there is a limit to what can be done with these motors. I follow what you guys do and find it amazing. The extent that you, Comnoz, and others go to push the limits is admirable. From a laymans standpoint, I think that properly balanced, well flowing engine is the basis for a strong motor. I think the limiting factors on these Norton engines are set valve timing, inability to change ignition timing easily and preset air/fuel mixtures (carbs). I realize those of you who race have your hands tied on what you can do. Comnoz has addressed most of these problems rather successfully. I like reading the theory and the engineering explanations on combustion, but the bottom line, in my opinion, is getting the right mixture to fire at the right time. Thus fuel injection and variable timing.
I enjoy following this forum because there is always something interesting and informative to read. Thank guys.
Pete

texasSlick said:

I see the combustion as fitting the micro model. I do not see a flame wave spreading as that in the macro pasture analogy. I would like to see the event further slowed. Does anyone else see it this way, or am I just mole gas prejudiced?

Slick

Click to expand...

There are reams of pressure vs crank angle plots in the literature and in general consistently show an effective burn angle of 40 crankshaft deg with 20 deg of burn before TDC and 20 deg of burn after TDC, with pressure maximizing ~ 20 deg ATDC. The 40 degrees does not include the slow start-up (lag angle), nor the slow final burn (after burn). So whatever range of flame speed/propagation rate/burn rate (or whatever we want to call the moving front) actually occurs, our model needs to be consistent with this range of time intervals, i.e., the time required for 40 crank degrees to pass at whatever the subject engine speed is. The time interval associated with 40 crank deg occurring is relatively long, so it is difficult for me to see how anything sonic is going on in the chamber (with the exception of detonation).

Any literature I've seen where flame fronts moving in combustion chambers have been photographed through quartz windows or cylinder heads, support the notion of a brush fire moving from the ignition source outward. Similarly, where heads have been fit with pressure transducer arrays, they show the pressure building progressively and systematically across the combustion chamber, i.e., starting at the ignition source and moving outward. In the case of dual-plug heads 2 pressure waves are created, one at each ignition source. In either case, photograph or pressure transducer array, the event takes on the order of 40 crank degrees to complete, so again, our mental model must take into consideration the slow nature of this event, the burn across the chamber is not occurring at a sonic speed.

Following is a picture from "The Internal Combustion Engine in Theory and Practice", found on Google Books, that shows combustion in a flathead engine viewed through a quartz glass cylinder head over a series of crankshaft angles (picture series starts with ignition occurring ~ 29 deg BTDC and ends ~ 40 ATDC - unfortunately the intermediate pictures are not labeled with crankshaft angle - the end of the unlabeled rows would be approximately -2.5, 12, 26, 41 ATDC). You can see the spark in the extreme left of the picture and the flame propagate from there. The flame front appears ragged and uneven due to the turbulence in the chamber. Here again the pictures look an awful lot like hobot's pasture burning from the camp fire started on one end of the field. There is no indication of a 1000 points of light and instantaneous combustion, and as noted above, the effective burn angle (from 10-90% combustion) again progresses slowly, taking somewhere on the order of 40 crank degrees to complete.

I apologize for the redundancy above, but am simply trying to point out how slow the combustion event is relative to sonic events, and that our mental model must accommodate this low range of burning speeds that occur in real world combustion chambers.

My math may be skewed so please correct me if I am wrong. These numbers seems way too high, but I can't see where I may have made a mistake.

5,000 RPM translates to 83.3 Rev. per second.
Firing every other stroke would be 41.6 ignitions per sec.
40 degrees of rotation for burn time, translates in 5.2 times a second for flame propagation?

duh dudes its been proven in the end its all just one type of magic principle on top of and mingled with others, <ugh> having said pasture fires like WZ-Kurt describes I can assure you that when mower caught on fire in middle of chest high grass on 1/2 acre area the fire spread out from it on both side spreading in widening rim with common center at speed I could tamp out in time, while 2 bush fires started at far ends just before a rain, front winds suddenly spread combustion from two centers of expansion and combined with each other to send sparks/embers all over the place to spread way too fast to control, so called volunteer fire department but between my franics exhaustion inhaling smoke and rain hitting it was out before they got here. May not sound like real magic to the lessor trained-observant- experienced, but even in packed dryish grass sending embers far out from 2 burn centers, the burnt areas were not well formed circles but branched/fractual with some places that started up but went out and other places didn't burn at all plus there was an over all bias to the main flame -away from gravity. My hollaring spread faster than the flame fronts but no one to hear it but some scattering small animals.

When I burn up close to trees they can start to smoke and brust into flames w/o any sparks hitting just lots of IR photons. KIlls ones I want to season a while before heat harvesting.

Combustion is quantum hit and miss on micro scale combining and chaotic on macro scale mixing, when flame speed reaches sonic its detonating beyond deflageration burn. Some men can't control fire so carry extinguishers or water it down some. Our engine combustion depends a whole lot on humidity and the water released in chemical reactions. There is not enough excess combustion gas molecules pressure to push much on pistons, so pressure is mostly the heated Nitrogen expanding. Some fuel if combusted in a closed cylinder leave a strong vacuum.

Deets55 said:

My math may be skewed so please correct me if I am wrong. These numbers seems way too high, but I can't see where I may have made a mistake.

5,000 RPM translates to 83.3 Rev. per second at 5000 rpm.
Firing every other stroke would be 41.6 ignitions per sec.
40 degrees of rotation for burn time, translates in 5.2 times a second for flame propagation?

Click to expand...

You are correct that there are 83 rev/sec at 5000 rpm.

The time for 1 revolution is 1/83 of a second = 0.012 sec

Firing occurs on half the revolutions so 41 fires/revolution

The time for 40 deg to pass during a single revolution is 40/360 * 0.012 sec = 0.0013 sec or 1.3 milliseconds allocated to the effective burn angle.

You can see all these numbers in the data table on page 13 of this thread.

Unfortunately, from there I don't know what you had in mind? If you could clarify the "40 degrees of rotation for burn time, translates in 5.2 times a second for flame propagation?" perhaps we could help. I struggle with what you mean by "5.2 times a second for flame propagation?", when the engine is firing 41 times/sec.

Got jammed up on how to come up with actual burn time in seconds.
Was trying to get a real world grasp on how fast that whole process took place. Looked back on 13. Thanks WZ507

Guess you can tell I read the instructions after I finish the project.

Pete

A very interesting subject and lots of good explanations to give me good visuals of combustion processes. After reading it all, I can only say I'm amazed and envious of what comnoz has accomplished with his Norton in regards to ignition and FI and if I could convince him to market only one mod I think it would be the high energy ignition combined with a knock sensor or even a more or less std. EI with a knock sensor with the ability to control the spark curve. His ride report on his engine running 50* cooler with the ignition seamlessy rolling back the ignition timing to keep him out of detonation was one of those 'ah ha' moments to me, especially since it was happening way before any audible warning, now that would be priceless. An ignition with those capabilities on an old brit bike would save a lot of grief for some of us.
To add a bit more to the dual plug/ignition on reciprocating aircraft engines, I'd like to say that it's not that the dual plugs give the engine more power but make up for the very conservative timing that costs the engine that power to start with, you cannot hear detonation in an aircraft engine so they are limited to the minimal advance to stay clear of detonation, and of course, dual plug dual ignition is for redundancy and safety. It's very easy to lead foul that bottom spark plug with improper leaning, any recip. pilot knows this. Some modern piston engine aircraft (esp. homebuilt/experimental) run one std. magneto and one electronic mag and at altitude where less than full power is available, they can run up to 38* or more advance (std. mag usually 20-25*) on the elect. mag and see no difference when shutting off the standard magneto, if I remember correctly, this advance allows them to lean more aggressively also for increased fuel economy which is very important to aircraft.
On the single offset plug vs. dual plug vs. center plug discussion I think each engine design probably would benefit differently then another. I know the Triumph splayed port head aims the incoming charge to the opposite side of the chamber from the spark plug, whereas the TR6/7 and T140 parallel port heads aim the charge at the spark plug and the designers said that because of the more efficient T140 chambers they were requiring a hotter spark plug, however, this chamber is just a 650 chamber with a small chamfer to match the 750's bore to the 650's chamber so what gives? Coincidentally the T140 reputedly has more detonation concerns even with 7.9-1 cr then the old 650s with 9-1 cr. so it makes me wonder if the hotter plug may be a band aid fix for intake & spark plug placement issues only to cause hot running detonation issues. I'm not sure this is relevant to the post but it got me thinking about the Triumph, so there you go...Mark

Good comments Mark. I had an early T140 for many years and have looked inside a number of Brit twins. One thing I like to do is to examine the carbon build up in the chambers and on the piston crowns. Whilst I never made a study or records of these, I noted some heads had a tendency to have more crust on one side or the other and sometimes darker black than other areas. OK, that in itself if not science and darkness of the carbon more a result of poor mixture or it could be simply poor combustion and the cooler deposit an agglomeration of unburnt fuel particles.
In comparison, I found the Commando heads generally had a more uniform distribution of carbon. A shallow dome chamber and flat top pistons seems to be a more efficient environment for flame propagation, hence the less advance the Norton has compared to the Triumph.
A friend built a very modified 850 and he centre plugged the heads. The heads also had bigger valves and I'm not sure if that was beneficial with the close proximity of the spark plug nose given the overlap. But it ran strong however, without a before and after dyno recording there is no evidence that centre plugging made any difference.

@WZ507 October , 3:40 PM

I see from your pics, the macro model rules. In Hobot's video, the micro model seems to be dominant. Damfino what the real case is. Such dichotomies in scientific data can often be resolved by analysis of the details of the experiments. We are hampered by not having more info on such details. The frame rate of the pics would provide more info to determine the flame propagation speed. It is a simple calculation: flame propagation (ft/sec) = change in distance (ft) across cylinder / frame x frame rate ( frames / sec). Does your reference give this frame rate?

Regarding a sonic front: please note I said a sonic wave originates from the ignition point; creating the environment for atomic collisions to result in combustion.....this does not mean the burn velocity outward from the ignition point is sonic. Mole gas theory assumes the combustion reaction rate can keep pace with the sonic wave, it does not assume nor predict all the fuel combusts at the sonic rate, only that a small probability of atomic collisions occurs at one instant in time which result in a combustion reaction. Mole gas theory is three dimensional; a flame propagation rate is two dimensional. In mole gas theory, fuel combustion rate (burn rate pounds fuel/ sec) is more appropriate than flame speed (ft/sec)

The micro model uses the probability of molecular collision that results in a combustion nuclei. This probability is not fixed, but determined by molecular density, mixture temperature, fuel dispersion, fuel reaction kinetics. In short there is no "fixed" flame speed.

Slick

Theory and Practice of Infernal Combusting engines, the source of Kurt's flat head view and maybe if a piston can out run flame pressure rise with light detonation near TDC. Start @ pg 11 and hope can stop before finishing tonight
http://books.google.com/books?id=mX1-OJ ... e&q&f=true



TheKnockBox lets you hear detonation before consuming engine.
https://www.youtube.com/watch?v=GGJZ6kKzLus

texasSlick said:

@WZ507 October , 3:40 PM

I see from your pics, the macro model rules. In Hobot's video, the micro model seems to be dominant. Damfino what the real case is. Such dichotomies in scientific data can often be resolved by analysis of the details of the experiments. We are hampered by not having more info on such details. The frame rate of the pics would provide more info to determine the flame propagation speed. It is a simple calculation: flame propagation (ft/sec) = change in distance (ft) across cylinder / frame x frame rate ( frames / sec). Does your reference give this frame rate?

Regarding a sonic front: please note I said a sonic wave originates from the ignition point; creating the environment for atomic collisions to result in combustion.....this does not mean the burn velocity outward from the ignition point is sonic. Mole gas theory assumes the combustion reaction rate can keep pace with the sonic wave, it does not assume nor predict all the fuel combusts at the sonic rate, only that a small probability of atomic collisions occurs at one instant in time which result in a combustion reaction. Mole gas theory is three dimensional; a flame propagation rate is two dimensional. In mole gas theory, fuel combustion rate (burn rate pounds fuel/ sec) is more appropriate than flame speed (ft/sec)

The micro model uses the probability of molecular collision that results in a combustion nuclei. This probability is not fixed, but determined by molecular density, mixture temperature, fuel dispersion, fuel reaction kinetics. In short there is no "fixed" flame speed.

Slick

Click to expand...

With respect to hobot's youtube combustion video and determining whether it suggests a micro or macro combustion process, that is a tough one for me to call. First I think we'd need a top view, not a side view, to watch the flame front move away from the point of ignition if we are trying to assign micro vs macro to the process. Second, it appears (to me) that the burn is occurring really late, predominantly when the piston is descending rapidly, which doesn't fit with the real world information I've seen. In a real (not model) engine, effective burning would typically occur evenly across TDC, where we'd expect to see lots of flame action before, at, and just after TDC, with the after burn continuing during piston descent. However this isn't the case in the youtube example where it appears to me that most of the burn occurs with the piston descending. Again, perhaps a top view would clarify much for me. Possibly their chamber needs 35+ deg of lead to accomplish this, as the lag angle may be of significant length and take up most of the time prior to TDC. At any rate it seems to me to be a suspiciously late combustion process. In spite of the video title containing the phrase "Spark Advanced 20 Degrees", maybe they mean it was advanced 20 degrees from some, unknown to us, starting point. Who knows?

To your question regarding the frame rate of the pictorial series, there was none given. However, I think we already have all the information needed to make any determination desired from the pictorial series, because we have the crankshaft angle of each picture, which can also be expressed in time, if we know the engine speed. We don't know the engine speed, but we do know that flame speed remains nearly proportional to piston speed, so we can assign any engine speed we want and know that if the engine speed doubles the flame speed will double, or conversely, if the engine speed is halved, the flame speed is halved. So pick any engine speeds you desire and you can calculate flame speeds therefrom. The flame speeds you come up with should be of the same order of magnitude as those given in the table on page 13 of this thread, since those all assumed a 40 deg effective burn angle.

Although only the pictures in the top row of the series are labeled wrt crankshaft angle, each picture differs by 2.4 degrees from it's neighbor. Thus the correct labels for the pictures by row, from left to right would be the following.

Top row, -29.0, -26.6, -24.2, -21.8, -19.4, -17.0
2nd row, -14.6, -12.2, -9.8, -7.4, -5.0, -2.6
3rd row, -0.2, 2.2, 4.6, 7.0, 9.4, 11.8
4th row, 14.2, 16.6, 19.0, 21.4, 23.8, 26.2
5th row, 28.6, 31.0, 33.5, 35.9, 38.3, 40.7

Using the above crank angles, one can see in the pictorial series that by ~ 10 deg ATDC, the flame has traversed somewhere around 90% of the combustion chamber. The youtube video doesn't dovetail nicely with such an image, at least the image in my mind's eye, but perhaps perspective is everything and a top view of the youtube combustion process would set it all right for me.

@WZ507

I concur with most of your post above, and considered making many of your points myself, but opted for brevity.

But here is where you leave me....

"We don't know the engine speed, but we do know that flame speed remains nearly proportional to piston speed, so we can assign any engine speed we want and know that if the engine speed doubles the flame speed will double, or conversely, if the engine speed is halved, the flame speed is halved. So pick any engine speeds you desire and you can calculate flame speeds therefrom. The flame speeds you come up with should be of the same order of magnitude as those given in the table on page 13 of this thread, since those all assumed a 40 deg effective burn angle."

Why is flame speed proportional to piston speed? If we sparked a charge with a locked crankshaft, (zero piston speed) the flame still spreads at some rate determined by chemistry and physics (assuming the head does not blow off). Look at it another way, at TDC, piston speed is zero....do you really expect the flame to stop spreading momentarily?

It is a pity the author(s) of your material did not note the engine rpm; that, and the crank angle information of the photos, and the cylinder bore size could shed some light on the issue of flame speed propagation.

It is unscientific to draw firm conclusions from either Hobot's video, or your reference, unless we have the full text of the author's experiment available. Even then, authors often leave "holes" that critics can drive wedges into.

Slick

texasSlick said:

@WZ507

I concur with most of your post above, and considered making many of your points myself, but opted for brevity.

But here is where you leave me....

"We don't know the engine speed, but we do know that flame speed remains nearly proportional to piston speed, so we can assign any engine speed we want and know that if the engine speed doubles the flame speed will double, or conversely, if the engine speed is halved, the flame speed is halved. So pick any engine speeds you desire and you can calculate flame speeds therefrom. The flame speeds you come up with should be of the same order of magnitude as those given in the table on page 13 of this thread, since those all assumed a 40 deg effective burn angle."

Why is flame speed proportional to piston speed? If we sparked a charge with a locked crankshaft, (zero piston speed) the flame still spreads at some rate determined by chemistry and physics (assuming the head does not blow off). Look at it another way, at TDC, piston speed is zero....do you really expect the flame to stop spreading momentarily?

It is a pity the author(s) of your material did not note the engine rpm; that, and the crank angle information of the photos, and the cylinder bore size could shed some light on the issue of flame speed propagation.

It is unscientific to draw firm conclusions from either Hobot's video, or your reference, unless we have the full text of the author's experiment available. Even then, authors often leave "holes" that critics can drive wedges into.

Slick

Click to expand...

You've obviously taken my words literally regarding the relationship of piston speed to flame speed, and correlated instantaneous piston speed with flame speed. I apologize for my inability to articulate what I mean in a more understandable form. Perhaps if I had said engine speed rather than piston speed or had I said "average" piston speed it would have been clearer or more easily understood. The gist of the point I attempted to make was that as engine speed increases, flame speed across a combustion chamber increases proportionately. I referenced piston speed, rather than engine speed, because it is the piston that increases turbulence in the combustion process, which in turn leads to higher flame speeds. That said, please allow me to try again.

Flame speed is nearly proportional to "average" piston speed due to the increased turbulence that accompanies higher "average" piston speeds. This attribute of engine combustion is what allows the use of the same "best power" ignition timing at 3000 rpm as 6000 rpm. At 6000 rpm the dominant combustion event (effective burn angle) has only half the time to occur that it would at 3000 rpm, yet it still occurs in about 40 deg of crankshaft angle (half the time required at 3000 rpm, but same 40 deg of crank angle) due to increased turbulence, and in turn increased flame speed.

To your point about no piston motion at TDC and burn still occurring, of course burn continues in the absence of piston motion, because in a dynamic situation even though piston motion ceases momentarily at TDC, turbulence and flame propagation clearly does not cease.

If you really want to know further details of the pictorial series reference, here it is for you to pursue.

Withrow and Rassweiler, "Slow Motion Shows Knocking and Nonknocking Explosions Jour. SAE 39, 297 (1936)

Note how recent it is - 1936! The reference contained additional pictorial series showing the point at which detonation occurred - always in the aftergas and relatively late in the combustion process when the flame had traversed 2/3 to 3/4 of the combustion chamber.

In closing I thank you for the civil nature this challenging dialogue has taken.

WZ-Kurt is repeating what the early researchers found and published. The rising piston slaps a pressure density wave just ahead of it which helps force reactants together physically to heat them more and brings them closer so the photons and free electrons and radicals have more chance to reach activation energy - so we lucked out that combustion speed rises to about match time available. This is why we can't just plug in a set formula of spark timing to get best pressure just after TDC as its not just the spark advance that determines when most pressure hits. Reading the Master Engine contesters and old fart dragsters I believe that slight detonation, ie: more than one area exploding before spark induced flame fronts reaching/mixing fully - allows hi rpm combustion to keep up with piston speed after TDC. One can limit spark advance enough to avoid even slight detonation, ie; comnoz digital knock senor limiter, but that would not win the power-torque contests. The most powerful piston dragster engines burn/melt spark plugs away half way down track to become pure diesel compression ignition engine operating through hydraulic lock density at TDC.

Multiple spark plugs don't add but a few percentage points of extra power and mainly benefit chambers that have stuff in the way of flame spreading from one area - ie: Theory/Practice tried 3 plugs in flat head for nil ignition speed increase until they lit off 17 plugs everywhere at once, to approach detonation supersonic combustion speed.

I have some real experience with the HP diff between good spark compared to weak spark.

During one of my "battle of the twins" road races at Laguna Seca I was running with Ducs and Harleys and holding my own when I noticed the power dropping off. There was no missing but the motor just felt weak and whimpy. The tricked out Harley that I had left behind began to catch back up. Then suddenly the motor just stopped. Pulling into the pits I found no spark and searched until I found that the wires in the Boyer pickups had broken inside the points cover. They had failed gradually strand by strand until there was barely enough current to fire the plugs. This was when the motor began to progressively run weaker - until the last strand broke and then there was nothing.

^^^ A similar phenomenon was reported during the development of the California Hotrod Ducati way back when by Jennings and co for Cycle magazine.

I was under the impression that detonation near the end of combustion was caused by the end gas being highly compressed by the advancing flame front to the point of instability.
Advances in both fuels and combustion chamber design along with better understanding of tumble, swirl and squish pretty much pushed the problem into the back seat for most ic applications, its only when pushing archaic designs too far they reappear.


A guy in the local guzzi section was doing his PhD in combustion in ic engines. He built a research engine with quartz windows so he could film the process in slomo. It would be useful now to access some of his results. His professor/supervisor was doing research into grooving the piston crowns. He had some good results, apparently the flame speed increased and the event was generally more stable with less inclination to detonation. No names were ever attached to sponsors of the research, far too sensitive.

At about this time, early/mid eighties, there was a lot of money being thrown at stratified charge engines.
All wasted because the govt of the day mandated catalytic converters, despite the evidence that SC could reduce pollutants more efficiently and with better fuel economy.


re the discussions on twin plugs, I did the main guzzi many years ago now, when leaded fuel disappeared over here (later here I think than usa). The engine would run very hot and loose loads of power.
I fitted lower compression big bore pistons, almost flat topped, this was more to remove the big humpy pistons euro spec Le Mans had. To restore a little compression I milled the heads. This gave about a 8.5ish.1 cr. rather than the claimed 10.1.
The second plug ends up almost a mirror image, but you can only fit a 10mm plug.
Using a dyno, the ignition was retarded about 3 degrees. The twin plugging didn't boost top end noticeably, but strengthened the midrange, indeed it was better than before loosing the compression. I think some of this was down to tightening the squish down to 0.030". At this clearance, the squish area on both piston and head stayed clean, but with no witness marks.