alcohol and evaporative emissions
 

The other issue is the evaporative emissions. High alcohol fuel blends have about .5-1 psi higher Reid Vapor pressure than normal gasoline depending on the alcohol concentration. Highest RVP is actually at about 10% concentration and goes down some with higher percentage blends.

That puts high alcohol fuel blends outside the current specifications for RFG. The major problem in California is that not all fuel sold in the region has ethanol blended in. If you mix a fuel that was blended with MTBE to have the proper Reid Vapor pressure with a fuel that is blended to have proper Reid Vapor pressure with ethanol as the oxygenate, you get a fuel blend with too high of an evaporation rate. That causes problems with photochemical smog from evaporative emissions, which is a big concern in Calif.

When the ban MTBE out there, if they mandate that all fuel must have some ethanol in it, than as I understand it the problem of in tank mixing of different fuels will no longer be a problem.

http://www.arb.ca.gov/fuels/gasoline/oxy/updatedwvr.pdf
http://www.nrel.gov/docs/fy03osti/32206.pdf
http://www.afdc.doe.gov/pdfs/6968.pdf



Larry

Back to WI Tuning ...
 

To get back on topic I thought I'd made a post that represents my oppinions on WI tuning ...

General Tuning, and Making the Switch to Water Injection

Mixture for a Fixed Injection Rate:

If stuck wtih a fixed relative injection rate, I would go with a 70/30 methanol/water mixture at a rate of 50% of your fuel flow on an intercooled engine. I'd have it setup to switch on 2 psi lower than you first start to see knock at. This would put the overal water injection rate at 15% of your fuel flow, and methanol at 35% of your fuel flow. Other people may say much less is advisable. This is just my oppinion.

I like that number because it keeps the water level down relatively low, yet the alcohol level high. Water and alcohol will cool the charge, and alcohol will raise the blended octane of your fuel right when you need it most.

On 91 AKI octane, the blended octane with 30% methanol would be 98 AKI octane (RON + MON)/2. That means you have the benefits of near race fuel + water injection when your WI is active.

If your engine doesn't knock at all at any PSI on this mixture, try reducing to a 40% injection rate, then 30% and so on so that you only use just barely more than you need.

Ideal Mixture and Setup:

A truly ideal WI setup would have a variable Meth/Water ratio. However, given the unnecessary complication of such a setup, I say that a 70meth/30water setup is a good place to start. The reason I chose this mixture is primarily that it allows brief trips into the severely lean section without knock.

My ideal controller would use zero WI whenever possible, and upon detection of light knocking begin adding WI while holding ignition timing constant if possible. (It'd need to pull timing for one or two rotations as WI cannot always be added quickly enough.) Then as the knock persisted continue to add more WI until a certain limit, at which point it would begin to lower boost pressure and slowly pull ignition timing until the knock abated.

That is essentially how my stock Saab's ECU works in the first place, with the exception that it first pulls timing, then adds fuel (interernal coolant), then slowly lowers boost as necessary. I think given enough programming skill, most ECU's could be altered to do the same. Instead of adding to the fuel curve, they just add WI from supplimentary injectors on the intake runners.

My Oppinion on Certain Encountered Problems:

Again I'm far from an expert, but these are just my thoughts to add to the discussion ...

Hesitation:

I think that when WI seems to be causing a loss in power this may be due to the over-quenching of the spark. Most cars use quench pads to create a faster burn. With WI this may be "blowing out" the flame Kernel.

I think one possible solution is to advance the timing several degrees. Remember that the quench pads only work near TDC, so the further you are from TDC the easier it should be to ignite the mixture. If timed properly the flame kernel should develop before quench takes place, then as the flame kernel is quenched the flame wave begins to propogate quickly. If you're experiencing hesitation, you're probably nowhere near the knock threshold, so advancing the timing a little to compensate for quench and a very low Flame Development Angle should be ok.

Another, slightly more obvious, solution is to lean the mixture. If you are running just straight water see how close you can get it to 14.7:1. As shown by the previous graphs, when running lots of water (IE enough to make th engine hesitate) 15:1 should be just as resistant to knock (or nearly) as 12:1.

If none of that works, either change the mixture to a mostly alcohol mixture with just a little water, or reduce the quantity of the mixture until there is no hesitation. (I think that should be a last resort, as clearly even a 50% mixture can work if setup right ... just perhaps not on all cars.)

Some Thermodyamics to Consider:

Since most engines are NOT like the engine in those graphs I posted, I thought I'd make list of things that can make your engine different. (Aside from the name of your engine's manufacturer.)

1. Higher compression. The NACA engine was 7:1 compression. More compression means more heat, which means more cooling is needed. It also means it's harder to ignite the mixture, especially if you're injecting lots of water. The further away from TDC the charge is ignited the easy it will be to ignite, but too far advanced and cyllinder pressure's skyrocket.

2. Large Bore X Short Stroke. The NACA engine had a long stroke, and thus a relatively small bore. This means it's easier for the piston to dissipate heat as there is less distance for that heat to go to make it to the coolant through the cyllinder walls. Large bore engines may need more internal cooling.

3. Single Spark Plug. The NACA engine was a twin plug engine. But, because the plugs were both on one side of the combustion chamber, the burn rate was still somewhat sluggish. A DOHC or Hemi chamber should burn just as quickly, but with a single plug igniting a heavily water laden mixture is harder.

4. Computerized Spark Control. Some cars, like my newer Saab, have a pre-set spark voltage which is designed to be JUST enough to ignite the mixture at various boost levels. A high water content can bog down the car as it's not expecting to require that much ignition power at that boost level.

5. Quench Pads. As discussed earlier, these can blow out your spark if you have enough water. Quenching is greatest near TDC, so sometimes a very advanced ignition can compensate to some degree. High compression engines usually have a great deal more quenching, so high compression turbocharged engines may need a disproportionately large spark energy, or high ignition advance ... or a little less water. ::lol: (Easiest solution last!)

Anyway ... those are my thoughts for now. Any correlation to the real world is purely coincidental. :oops:

Adrian~

An interesting article has just appeared in Modified Mag on tuning with WI. It's a good article, and good for WI in general, but as tends to happen too short on detail.

It is reassuring that the conclusions reached on these boards are the one's espoused in the article.

November Issue Modified Mag
http://www.modified.com

Currently no link to the article on the site, maybe later after it's off news stands.

b_boy's summary is great - just which someone has car with that set up that will reflect the result of the theory - on a modern engine of course.

Hi Richard, hi Larry Hi Adrian....

sorry to disturb you but it might be possible that an italian army of forum subscribers is coming up to the gasoline forced induction forum to recive some informations... (hope to be 2... may be 2.000 :shock: ) :D
I'm deeply sorry but you know Richard I'm completely unable to manage GASOLINE engines... just DIESEL... :lol:

Regards

Daniele

Hello Prometeus,

999 Italians (average of 2...2000)!!! I 'd better strart learning italian.

Diesel is very popular in Europe, I am a bit behind with the techanology. I am learn fast.

When is the invasion? :D

I would like to discuss the effect of flame speed against power - ideally we want a very high knock resistant set up that can also burn very fast as well.

Almost all race fuels burn quite slowly except those specials that F1 car uses. If anyone could chime in on this topic, may be some information on various kind of fuel or additives that can increase up frame speed. Low Knock resistance fuel doesn't always mean fast-burning.

I have heard that if peroxide is disassociate into free radicals (the OH molecules that has temporary lost their links) will initial and speed up the chain reaction of hydocarbon with air (oxygen).

Don't forget Nitromethane.

Anyone?

Alcohol fuels burn faster than straight gasoline but the burn speed depends on the mixture as well.

Gasoline has its highest burn speed near 11.3:1 AFR

Max burn speed for ethanol and methanol fuels are at even richer mixtures. The higher burn speed of ethanol is one of the reasons E85 is more effecient than gasoline (aside from octane and charge air cooling).
That is also one of the reasons 50:50 water methanol WI needs less ignition advance than plain water.

http://www.epa.gov/otaq/presentation...mjb-051303.pdf

Check out the chart on page 9


Larry

Advanced topic this one :D

I've got somewhere a model of this I made a couple of years ago. The graph on my site has the incorrect shape, by the way, it is exaggerated bell shape to make a point.

Max speed for gasoline is at AFR 12:1 or thereabouts. Go leaner or richer and it slows down (effectively retarding the ignition)

It is a 3D graph, because mixture DENSITY also affects burn speed. More density (boost, nitrous) more speed. It is the *main* reason you need to retard the ignition under boost or nitrous, despite what most people think.

One example of misguided 'experts': Running lots of NOS and running pig-rich as well. They think that the extra fuel cools down the situation, when in reality it simply slows down the burn speed so that they don't need to retard the ignition. False logic ofcourse, because if they were to lean down and retard the ignition properly they'd be better off. The extra fuel just messes up their overwhelmed spark plugs. :wink:

Different sources give different numbers for best burn speed, all seem to be in the mid 11:1 - low 12:1 range for gasoline --- far too many variables to give an absolute number I guess.

Ignition advance ( indirectly an indicator of burn speed ) also depends on combustion chamber design and squish etc. as we all know. Interesting chart on the change in required ignition advance in one of the NACA reports also.

Check out Figure 5 In NACA report E5E18 for a good example of both the change with afr and engine rpm. On the Aircraft engines they were working with they got minimum required advance at a fuel air ratio of .088 or approx 11.36:1.

In the book How to Tune and Modify Engine Managment Systems he plots a curve on page 127 that shows best power at fuel air ratio of about 0.083 (12.05:1) and fastest burn speed at 0.09 (11.1:1)


John you raise a very important point that people need to keep in mind --- every time they change fuel air mix or in our case WI rate or water/methanol mix you are in effect changing engine ignition timing by changing when peak cylinder pressure occurs in the cylinder.

At a given engine rpm and manifold pressure and AFR, there is only 1 ignition advance that will optimize cylinder peak pressure with the mechanical best crank angle for max engine effeciency (somewhere near 14 deg ATDC)

It is much better to be a bit late on ignition timing in a high power engine than it is to be a bit early. If your on the ragged edge and you reduce your WI spray rate or add alcohol to the mix, you are for all practical purposes advancing ignition timing.

The window of acceptable ignition advance for best power/effecincy is only a few degrees wide so you don't have a lot of room to play around with if your pushing the engine hard.


In figure 6 in the NACA report E5E18 on the next page following figure 5 you can see that at max power output (863 lb/hr air flow) the curves get a sharper peak and the max power nose of the curve is only about 8 deg wide, where at a lower boost pressure (437 lb/hr air flow) the same zone was almost 14 deg wide.

Larry

Thank you guys for chiming in.

I often wonder how difficult is to to tune to towards MBT with different burn rate, bore/stroke and combustion chamber design. For day to day and non-crucial tune (based on a square bore/stroke ratio), I think I would work on pump fuel plus 50:50 methanol/water and try to get the best possible outcome, because all the ingredients are easily available. This is what we have been trying to do all along.

I am hoping to find a way improve the flame speed on 50:50 mix and gain MBT without over-advancing the engine, do you think a small percentage of nitromethane will alter the flame speed in the right direction?

The best way to safely tune for MBT (minimum best torque timing) is to find max power and then back off timing until you see a 1% power drop. That will ensure your on the safe side of the "Hump". (or an aproximate equivalent is find knock then back off 2 - 3 degrees)

There is very little change in power for the timing that brackets max power and a timing spread of maybe 8 degrees from too advanced to too late timing. Very hard for even the big guys with the expensive toys to find the absolute peak.

As far as the nitromethane I'm not sure I understand it burns quite slowly but can't say for sure. I know propylene oxide will speed it up --- but unfortuantely is sometimes speeds up combustion when the car is not even running ---- makes things go boom and people drive over their crankshafts.

The ideal timing will only change slowly with boost, so you could safely find best timing at a moderate boost level and then gradually pull back timing as you add boost.

Larry

Just a quick note about NACA and SAE papers:

They are usually based on engines with combustion chambers inferior to those we use, so their AFRs might have to be richer just to catch up because of poor atomisation (usually they have no squish, swirl or tumble, hell not even fuel injection often)

Then we have the fuels themselves, they can vary a lot. Even recent SAE papers are based on test engines that would mimic model "T" quite well :lol: and fuels used could be worse than those found in the backstreets of Delhi in plastic containers. That is how you get the range of 80 to 100RON, while we'd be more interested in 95 to 115RON :wink:
The range is still 20RON but I'd bet that the flamefront would move differently.

I am trying to clarify between knock threshold and MBT.

Tuning these days, knock threshold arrives much early before reaching MBT, it appears that no one are too concerned about MBT. I wonder if it matters or not whether if MBT is the main aim anymore?

The general road car tuning strategy seemed to be accepted as follows:

Run as much boost until the flow linit of the turbo is reached, dumping as much fuel as possible until egt is below 900C. Wind on as much ignition retard as possible until knock disappears.

I was wondering if this common method can be improved? with or without water injection.

I agree I think trying to run ideal ignition timing is not as popular as it was in the past. I still think you should try to stay as close to ideal timing as possible.

Food for thought, the aircraft folks probably have done more testing and development on high performance piston engine in high load environments than all the automotive folks combined.

Many of them have timing fixed near ideal timing for max power. The way they handle max power for take off is they richen the mixture until the engine runs rough and then crank up the boost to the maximum recommended manifold pressure ( just short of det).

I think we should look more at managing boost pressure curves and less with playing games with the ignition timing.

I think the issue is it is much easier from a control system point of view to solve problems by pushing the timing values all over the map. It is much more difficlult to get fast acting stable boost control with enough head room so you can reach knock limited performance at high rpm with boost rather than ignition timing.

Most street turbos simply can't deliver knock limited boost at high rpm.

So they fake it by jacking in a lot of ignition advance, and create an electronic variable compression ratio by lighting the fire a bit early. It works to a point, but in theory it should not produce as much power as maintaining ideal timing and running the boost necessary to reach the knock limit.

Larry

The main subject in these threads is Detonation. There is the assumption that It can be detected.

How do you all guys go about detecting detonation????

In real time -- the modern ECU's monitor for knock at least through the moderate RPM ranges. The Subaru ECU quits listening for knock at about 5700 rpm or so, as do many others becasue it is too hard the distinguish det from normal engine noise at high rpm, and engines are much less likely to have serious detonation at high engine rpms.

For after market solutions the "knock Link" is pretty popular. It is a add on knock detector with adjustable sensitivity. If set up right is appears to be quite effective.

In some cases -- especially on NA cars you can hear the detonation due to the characteristic knocking sound or sharp pinging sound.

After the fact, reading your spark plugs is still the final word on if you are experiencing detonation. Detonation causes some characteristic changes in the sparkplug appearence. In mild cases you get what is often called "salt and pepper " on the plugs. Small dark specks on the insulators (the pepper) are carbon blown off the inside of the combustion chamber. This appears first, then you get very small balls of aluminum that are so small you need a 10x magnifier to reliably see them. This is the "salt". It looks like white "dust" all over the spark plug electrodes to a casual observer. On close examination under magnification they are nearly perfect, brilliant silver colored balls of aluminum stuck to the spark plug electrode.

When you see these its time to back off, because they are bits of aluminum blown off your piston crown and combustion chamber.

The next steps are the plug it self begins to show the beating it is taking. The insulators can crack, the electrodes begin to get an "abraded" appearence -- all the sharp corners look like they have been chewed off by a very small animal.

Next sign is usually lots of smoke out the tail pipe as you hole a piston or break a ring.

Larry

Propylene oxide:
C3H6O -soluble in water and methanol

Nitromethane:
CH3NO2 -soluble in methanol

Will look into it.

With the Knock Link how do you set the sensitivity. Two little sensitivity will make the unit appear not to detect knock...Too much will make it seem as if there is knock when there is none...

VERY BIG CAUTION with Propylene oxide: it was banned in NHRA due to some serious accidents as I recall.

It is a suspected carcinogen, it is not compatible with copper and there are cautions against mixing with more than 2% water.

Propylene oxide reacts with water to produce propylene
glycol, dipropylene glycol, tripropylene glycol and higher molecular weight polyglycols.

Hazardous Polymerization: Will occur. May polymerize violently, especially
in the presence of aqueous sodium hydroxide, chlorine, ammonia, strong
oxidants, and acids. If polymerization takes place in container, there may be heat and a violent rupture of container. Hazardous polymerization can occur when in contact with highly active catalytic surfaces such as anhydrous chlorides of iron, tin, and aluminum; alkali metal hydroxides; and peroxides of iron and aluminum.

Violently reacts with acetylide-forming metals such as copper or copper alloys.
Conditions to avoid: Ignition sources, temperatures above 50?C or 122?F,
confined spaces

In bulk form it may explosively polymerize, and has a very low boiling point ( 93 deg F. ) which can lead to high pressures in containers, and unitended exposure to vapors.

http://www.bndrc.com/pdf/Pro-Oxide.pdf
http://www.scottecatalog.com/msds.ns...5?OpenDocument


Don't even think of messing with this stuff, it kills people!!


Larry

Sound just like the right stuff for the boys.
I was thinking mixing 95%-98% of water to PO just to offset the slower burn rate of WI or race fuel.

At longer polymer chains, would it solidify? another thing to find out.

Richard

It's got an adjustment screw.
Normally you have it fully counter-clockwise (full sensitivity) and only turn it a bit if it gives too many false 'reds'

It's a balancing act, it also depends on where the sensor has been fitted, or how it is coupled to the engine.

ChassisEar:

http://tools.batauto.com/images/products/ST06600.jpg

ChassisEar for fine tuning, KnockLink for safety use.

Gert

...KnockLink for safety use.

While I haven't tried the KnockLink per se, I have experimented with a similar device made by another manufacturer. If one's car is fitted with an OEM knock sensor fine, but if not, it gets more complicated. Knock sensors come in two basic types, resonant and non-resonant, also known as flat response type sensors. The resonant sensors tend to be more application specific. Each motor has its own resonant frequency at which knock occurs, being dependent upon cylinder bore among upon things.

Using a resonant type knock sensor in a motor other than the application it was designed for can result in false warnings. In my personal experience testing 2 different GM resonant knock sensors in a 4 cyl turbo motor, the sensor would regularly confuse valve train mechanical noise with knock.

While it is true that the display sensistivity may be turned adjusted, without actually being able to hear detonation one runs the risk of "desensitizing" the unit excessively. Inaudible knock could be occuring without one being aware of it.

Below is an interesting primer on knock sensors and the signals they produce:

http://deviantmethods.com/bigmoose/pages/knock.htm

Here is another site with more information on flat response type sensors:

http://www.delphi.com/pdf/ppd/sensors/et_flat_knock.pdf

For a universal application I'd be inclined to go with a flat response type sensor, which is also what Bosch Motorsport recommends with their stand alone racing ECU's. However, I have yet to find an aftermaket logger that will read their low voltage output without an amplifier.

espritGT3

This Delphi 'flat repsonse' sensor looks like the Bosch one that comes with KnockLink

If Knock link comes with a flat response sensor, that's certainly a step in the right direction. Now if I could only hear the knock (mid engine car) to calibrate it.

espritGT3

Mine came with part nr.# 0 261 231 006

Gert

I decided to opt for a J&S Safeguard, configured coincidentally, for the exact same Bosch sensor. J&S allows one to choose the type of knock sensor they prefer, resonant or flat response.

http://www.jandssafeguard.com/images...0261231006.jpg

I considered a KnckBlock but wasn't able to obtain much technical information on the unit from KnockLink's US reseller. Do you have any experience with it? Does it retard ignition sequentially?

espritGT3

KnockLink is just a display thing (and a pretty ugly one at that!)
It doesn't retard or anything, just passively light up LEDs

John,

I'm aware that KnockLink is just a display, but apparently they make a kncok retard as well, called "KnockBlock". While the local dealer could give me a price on it, I've never seen one and additional information on its operation is scarce. Being familiar with the KnockLink I thought you might know somehting about their knock retard unit as well.

Thanks,

Mike

Yes, it is ugly. I solved it as follows:

http://www.celica.dds.nl/plaatjes/knock/knk05k.jpg

As soon as the red light comes on, I step off the throttle. Happends almost never, only during mapping.

Gert

I've seen a similar setup drilling holes in a A-pillar pod.
Looks neat the way you've done it.

It's good to resolder the LEDs anyway, some come with faulty connections, and the top ones never light up (would you believe, eh!?)

Mike, I didn't know about KnockBlock, cheers :smile:

Thanks John.

Nice installation Gert. Very clean.

I have done lots of testing this year. I have some untuned figures to share.

http://www.geocities.com/sdminus/ind...?1151874216342

Scott

sdminus,
Nice wright up! If you happen to extend your test futher pleace keep us posted! I would love to see second round of test's with AFR and timing been ajusted to sute.

Thanks. I was unsure where to or not but decided to have a crack.

I have tuned for meth but not the other 2 yet. I may well try the 50:50 mix as it looks quite good on paper. I have transmission probs at the mo and also a strange occurance of the methanol jets getting clogged. This confused me slightly becasue methanol usually breaks stuff down. Any way back on topic.

I am happy to dispaly any info if it makes life easier. It is nice to be appreciated. Thanks

Scott :D

Scott,

Thanks for the write-up! Nice results.

What percentage of fuel are you injecting the water/methanol at?

-Craig

Peak petrol delivery at 100 IDC is 44GPH or 3400cc/1 they r=peak at about 85 ish idc

The water/meth and 50:50 runs was 100% @ 4 Gph or 252cc/1 at peak.

I have since retuned with meth @ 5.5GPH or 570cc/1 ish and achieved 340 WHP @305 RWTQ

I am going to re tune again but i am unsure what to do. I may try 50:50 with a higher percent mix of 515cc/1 or 8 GPH

or try and tune the one i didnt include in the graphs. nitro !

Scott

Scott

I have finally transferred your great write-up:

Water injection Study Models
This is a brief but concise documentation of the following water injection study carried out in March 06.


First of all let me explain why I decided to do the experiment.

After reading various forums and trying water injection within a controlled environment (A regional drag strip), I decided the best way to cut through some of the rumours, myths and facts surrounding various mixtures, and possible gains and losses associated with those mixtures.

From information gained on the Internet the general feeling was that injecting water or other chemicals into the intake tract would/could un-harness possible power gains for very little effort.

I used my own car for these tests, Mazda RX-7 FD3S twin turbo. The car itself is highly modified so was a stable platform from which to launch the test.

The data was captured using FC Datalogit software and analysed using Data log lab.

The water injection was the new Coolingmist vari-cool controller operating a multi nozzle system, one nozzle being at the throttle body and the other being just after the intercooler.

The tests were performed using the same fuel map and ignition maps on the same stretch of track, minutes apart. The fueling was set at 11:1 at peak torque which was considered safe for this tune at a boost level of 14.7 psi with a safe level of ignition advance.

http://www.aquamist.co.uk/forum/gallery/scott/1.jpg

The above power graph is the control for the test. 11:1 AFR @ 14.7 psi at peak tq. Air temps were 34 deg C at stand still dropping to 30 deg at 7000 RPM, peak knock was 33 @ 6000 RPM.

http://www.aquamist.co.uk/forum/gallery/scott/2.jpg

The above power graph is for water injection. Max flow was at full boost of 14.7 psi and was 4 GPH of water. The AFR?s seemed to be a touch richer on this run dipping from 11:1 to 10.8:1 in most of the full boost cells of the fuel map log (this seemed very odd with water not being a fuel and also displacing air). Air temps were a bit cooler at stand still 25 ?C and had a reduction at 6000 rpm to 19? (a drop of 6 ?C in a matter of seconds). Max knock was 33 at 6800 RPM

http://www.aquamist.co.uk/forum/gallery/scott/3.jpg

This power graph is 50:50 Methanol: Water. The flow rate was the same as the previous run.The AFR?s were again a touch richer than the control dipping to 10.5:1 in a high RPM cell but were generally between 11:1 and 10.8:1. The AFR?s do seem to be a little unstable with this mixture as compared to other runs. Air temps were back up again for this run to 31 deg standing and dropped to 27 deg at 6700 RPM. The knock peaked to 32 @ 6000 RPM.

http://www.aquamist.co.uk/forum/gallery/scott/4.jpg

This power graph is for Methanol. The flow rate is the same as the previous runs. The AFR?s were very rich on this run, which makes sense. A very cool 10.3:1 at its richest and barely coming out of the 10:1 range at all.

http://www.aquamist.co.uk/forum/gallery/scott/5.jpg

Air temp @ standstill was 28 ?C and dropped to 26 ?C as soon as the system started to run but lowered no further than this. Knock peaked at 31 at 4000 RPM but as a whole was extremely low whilst on boost.

http://www.aquamist.co.uk/forum/gallery/scott/6.jpg



http://www.aquamist.co.uk/forum/gallery/scott/table.gif

The above table is a simple representation of the results from the runs. The results run in number order. Ie 1 being the best & 4 being the worst. I have highlighted the best as blue and the worst as red.

Methanol is the best overall mix in this test. But 50:50 mix seems to be superior.

N.B please note all of these results are on a untuned basis. There is still plenty of room for additional tuning and power gains over this tune. Once tuning has take place different mixtures may well take over as the top mixture. I predict there is around 50 hp more in most tunes. The hp figures were kept low intentionally to keep the tune safe due to other mixtures that were used in this test but not displayed in this document.

In brief different mixtures and chemicals will give different results. The data above is just the tip of the iceberg as far a tuning these chemicals goes. This data clearly shows that there is a very small margin for real performance gains without proper tuning, but also shows that a bit more safety can be achieved with a bit of thought into the mix.

I hope this info is useful and helps clear up some myths concerning this subject

Scott Bishop

Very nice Scott. :D
My own experience agrees with your results. That there are no significant power gains without leaning towards 12.5:1 and/or adjusting ignition and/or running more boost.

If everything stays on the 'safe-for-non-WI' side, then all you get is extra knock headroom. (which is not bad, but not exciting :wink: )