Spark Timing
03-17-12, 10:20 PM
#1
Racing Rotary Since 1983
Thread Starter
iTrader: (6)
Spark Timing
here's another article from the Innovate site i found interesting... feedback invited
"Spark Timing Myths Debunked
A widely-held myth is that maximum advance always means maximum power. Here’s what’s wrong with this thinking:
The spark plug ignites the mixture and the fire starts burning. The speed of this flame front depends on the mixture, this means how many air and fuel molecules are packed together in the combustion chamber. The closer they are packed together in the same volume, the easier it is for the fire to jump from one set of molecules to the other. The burning speed is also dependent on the air-fuel-ratio. At about 12.5 to 13 air-fuel-ratio the mixture burns fastest. A leaner mixture than that burns slower. A richer mixture also burns slower. That's why the maximum power mixture is at the fastest burn speed. It takes some time for this flame front to consume all the fuel in the combustion chamber. As it burns, the pressure and temperature in the cylinder increases. This pressure peaks at some point after TDC. Many experiments have shown that the optimum position for this pressure peak is about 15 to 20 degrees after TDC. The exact location of the optimum pressure peak is actually independent of engine load or RPM, but dependent on engine geometry.
Typically all the mixture is burned before about 70 deg ATDC. But because the mixture density and AFR in the engine change all the time, the fire has to be ignited just at the right time to get the peak pressure at the optimal point. As the engine speed increases, you need to ignite the mixture in the combustion chamber earlier because there is less time between spark and optimum peak pressure angle. If the mixture density is changed due to for example boost or higher compression ratio, the spark has to be ignited later to hit the same optimal point.
If the mixture is ignited to early, the piston is still moving up towards TDC as the pressure from the burning mixture builds. This has several effects:
The pressure buildup before TDC tries to turn the engine backward, costing power.
The point where the pressure in the cylinder peaks is much closer to TDC, with the result of less mechanical leverage on the crankshaft (less power) and also causes MUCH higher pressure peaks and temperatures, leading to knock.
Many people with aftermarket turbos don't change the spark advance very much, believing that earlier spark creates more power. To combat knock they make the mixture richer. All that happens really then is that the mixture burns slower and therefore hits the peak pressure closer to the right point. This of course reaffirms the belief that the richer mixture creates more power. In reality the flame front speed was adjusted to get the right peak pressure point. The same result (with more power, less emissions and less fuel consumption) could be achieved by leaving the mixture at the leaner optimum and retarding the ignition more instead.
Turbo charging or increasing the compression ratio changes the mixture density (more air and fuel molecules are packed together). This increases the peak pressure and temperature. The pressure and temperature can get so high that the remaining unburned mixture ignites by itself at the hottest part in the combustion chamber. This self-ignition happens explosively and is called 'knock'. All engines knock somewhat. If there is very little unburned mixture remaining when it self-ignites, the explosion of that small amount does not cause any problems because it can't create a large, sharp pressure peak. Igniting the mixture later (retarding) causes the peak pressure to be much lower and cures the knock.
The advances in power of modern engines, despite the lower quality of gasoline today, comes partially from improvements in combustion chamber and spark plug location. Modern engines are optimized so that the flame front has the least distance to travel and consumes the mixture as fast as possible. An already burned mixture can no longer explode and therefore higher compression ratios are possible with lower octane fuel. Some race or high performance engines actually have 2 or three spark plugs to ignite the mixture from multiple points. This is done so that the actual burn time is faster with multiple flame fronts. Again, this is to consume the mixture faster without giving it a chance to self-ignite.
Higher octane fuel is more resistant to self-ignition. It takes a higher temperature and pressure to cause it to burn by itself. That's why race fuels are used for engines with high compression or boost. Lead additives have been used, and are still used to raise the self-ignition threshhold of gasoline, but lead is toxic and therefore no longer used for pump-gas. Of course a blown engine is toxic to your wallet."
"Spark Timing Myths Debunked
A widely-held myth is that maximum advance always means maximum power. Here’s what’s wrong with this thinking:
The spark plug ignites the mixture and the fire starts burning. The speed of this flame front depends on the mixture, this means how many air and fuel molecules are packed together in the combustion chamber. The closer they are packed together in the same volume, the easier it is for the fire to jump from one set of molecules to the other. The burning speed is also dependent on the air-fuel-ratio. At about 12.5 to 13 air-fuel-ratio the mixture burns fastest. A leaner mixture than that burns slower. A richer mixture also burns slower. That's why the maximum power mixture is at the fastest burn speed. It takes some time for this flame front to consume all the fuel in the combustion chamber. As it burns, the pressure and temperature in the cylinder increases. This pressure peaks at some point after TDC. Many experiments have shown that the optimum position for this pressure peak is about 15 to 20 degrees after TDC. The exact location of the optimum pressure peak is actually independent of engine load or RPM, but dependent on engine geometry.
Typically all the mixture is burned before about 70 deg ATDC. But because the mixture density and AFR in the engine change all the time, the fire has to be ignited just at the right time to get the peak pressure at the optimal point. As the engine speed increases, you need to ignite the mixture in the combustion chamber earlier because there is less time between spark and optimum peak pressure angle. If the mixture density is changed due to for example boost or higher compression ratio, the spark has to be ignited later to hit the same optimal point.
If the mixture is ignited to early, the piston is still moving up towards TDC as the pressure from the burning mixture builds. This has several effects:
The pressure buildup before TDC tries to turn the engine backward, costing power.
The point where the pressure in the cylinder peaks is much closer to TDC, with the result of less mechanical leverage on the crankshaft (less power) and also causes MUCH higher pressure peaks and temperatures, leading to knock.
Many people with aftermarket turbos don't change the spark advance very much, believing that earlier spark creates more power. To combat knock they make the mixture richer. All that happens really then is that the mixture burns slower and therefore hits the peak pressure closer to the right point. This of course reaffirms the belief that the richer mixture creates more power. In reality the flame front speed was adjusted to get the right peak pressure point. The same result (with more power, less emissions and less fuel consumption) could be achieved by leaving the mixture at the leaner optimum and retarding the ignition more instead.
Turbo charging or increasing the compression ratio changes the mixture density (more air and fuel molecules are packed together). This increases the peak pressure and temperature. The pressure and temperature can get so high that the remaining unburned mixture ignites by itself at the hottest part in the combustion chamber. This self-ignition happens explosively and is called 'knock'. All engines knock somewhat. If there is very little unburned mixture remaining when it self-ignites, the explosion of that small amount does not cause any problems because it can't create a large, sharp pressure peak. Igniting the mixture later (retarding) causes the peak pressure to be much lower and cures the knock.
The advances in power of modern engines, despite the lower quality of gasoline today, comes partially from improvements in combustion chamber and spark plug location. Modern engines are optimized so that the flame front has the least distance to travel and consumes the mixture as fast as possible. An already burned mixture can no longer explode and therefore higher compression ratios are possible with lower octane fuel. Some race or high performance engines actually have 2 or three spark plugs to ignite the mixture from multiple points. This is done so that the actual burn time is faster with multiple flame fronts. Again, this is to consume the mixture faster without giving it a chance to self-ignite.
Higher octane fuel is more resistant to self-ignition. It takes a higher temperature and pressure to cause it to burn by itself. That's why race fuels are used for engines with high compression or boost. Lead additives have been used, and are still used to raise the self-ignition threshhold of gasoline, but lead is toxic and therefore no longer used for pump-gas. Of course a blown engine is toxic to your wallet."
03-18-12, 01:14 AM
#2
F-IT
iTrader: (5)
Join Date: Jun 2006
Location: ocala,fl
Posts: 995
Likes: 0
Received 0 Likes
on
0 Posts
This is why being too "consecutive" with timing maps can be just as bad as being too aggressive. We see cars all the time come through the shop with really conservative timing tables and the owners complaining of lack of power and detonation.
Also with our engine primarily having split timing it's really key to have that spot on.
Also with our engine primarily having split timing it's really key to have that spot on.
03-18-12, 06:35 AM
#4
Senior Member
Join Date: Nov 2008
Location: Czech republic
Posts: 357
Likes: 0
Received 0 Likes
on
0 Posts
If the engine can't be tuned for maximum brake torque timing on given fuel at given AFR mixture without detonation, then it means that you have exceeded fuel capabilities. But knock resistance is AFR dependent - stoichiometric AFR permits lowest levels of power without abnormal combustion, lean mixture - read lean with excess air, permits higher power but engine behavior is no good for automotive application. Rich mixture permits highest levels of power on given fuel. Knock limited power increases all the way and beyond 50% of excess fuel and this has been well known for at least 70 years....
This article didn't brought anything new and these basic things about air/fuel ratio and timing should be known to anyone who tunes anything
03-20-12, 07:45 AM
#5
"Elusive, not deceptive!”
If you study the two traces for pressure (top) and location (bottom) you will notice that our optimum rotary pressure location appears to be about 45 degrees ATDC.
You will also notice how much deviation we have between ignition events.
Keeping the AFRs, temps and humidity constant (very difficult) helps verify data.
Barry
You will also notice how much deviation we have between ignition events.
Keeping the AFRs, temps and humidity constant (very difficult) helps verify data.
Barry
03-20-12, 08:22 PM
#6
Racing Rotary Since 1983
Thread Starter
iTrader: (6)
might some of the variability come from timing of the data? how many CPS Hz are you logging?
one of the (many) reasons i am switching ECUs was i want alot more datapoints per second.
howard
one of the (many) reasons i am switching ECUs was i want alot more datapoints per second.
howard
03-20-12, 08:46 PM
#7
^ Howard brings up a good point. What is the sample speed and resolution of the crank trigger wheel?
MBT spark timing can be calculated in simplified form with a certain differential equation... certain stock ECU's use an on-board MBT calculation and then set a baseline off of that, with knock learning applied at the end
MBT spark timing can be calculated in simplified form with a certain differential equation... certain stock ECU's use an on-board MBT calculation and then set a baseline off of that, with knock learning applied at the end
Trending Topics
03-21-12, 07:46 AM
#8
"Elusive, not deceptive!”
^ Howard brings up a good point. What is the sample speed and resolution of the crank trigger wheel?
MBT spark timing can be calculated in simplified form with a certain differential equation... certain stock ECU's use an on-board MBT calculation and then set a baseline off of that, with knock learning applied at the end
MBT spark timing can be calculated in simplified form with a certain differential equation... certain stock ECU's use an on-board MBT calculation and then set a baseline off of that, with knock learning applied at the end
One Thought... when we tune at say 10.8 AFR and the mixture leans to 12.5 on a hard run we are advancing the spark in essence because of the faster burn rate.
If we could tune to 12.5 AFR any variation would retard the ignition.
Notice the chart below the greater variation at about 6000 rpm. It could be the water/meth cutting in... or a leaner/richer cell... or a cell that needs better timing.
Many variables that need to be balanced.
03-21-12, 10:11 AM
#10
"Elusive, not deceptive!”
I forgot to add that the HKS Twin Power might be causing that change on the log.
Does anyone know when it switches from CDI to Transistor ignition?
The variation on the initial burn is striking! An analogy would be like lighting a candle in a hurricane.
If the wick can be heated enough by the initial flare of the match we would get the best result. Unfortunately that only happens about one in five.
Most of the time it will take longer to get it going.
And idle is much-much worst. It clears up appreciably as soon as a slight load is added.
This initial light-off where the kernel starts to grow is the basis of the great variation that we see in the rest of the burn.
My guess is that exhaust reversion is the biggest addressable problem.
This could be analyzed with sensors in the intake and exhaust.
Unfortunately fast sensors are about $1K each.
03-21-12, 10:54 AM
#11
"Elusive, not deceptive!”
^ Howard brings up a good point. What is the sample speed and resolution of the crank trigger wheel?
MBT spark timing can be calculated in simplified form with a certain differential equation... certain stock ECU's use an on-board MBT calculation and then set a baseline off of that, with knock learning applied at the end
MBT spark timing can be calculated in simplified form with a certain differential equation... certain stock ECU's use an on-board MBT calculation and then set a baseline off of that, with knock learning applied at the end
ECU's using on-board MBT calculations sounds simple and the way to go. How can we adapt that system to our Rotaries.
I used an InstaCal program to read the HET sensor. The program had to be adjusted slightly for the switching between seeing the tooth vs seeing the space.
An interesting side note. In calibrating the system a double check is done by doing a high RPM ignition cut and then opening the throttle wide open.
You are checking the centering of compression on the TDC mark.
It is like doing a dynamic compression check at 3000 RPM? Guess what the pressure was.....
about 180 psi.
Barry
03-21-12, 07:44 PM
#13
"Elusive, not deceptive!”
It shows where the pressure rise is at its max rate of increase but not the rate itself.
My latest tuning experiment has the IMEP at close to 325 psi and the Max pressure at almost 1200psi.
I believe that there are other rotary users but no one else has posted.
Barry
03-21-12, 09:50 PM
#14
325psi is about 22 bar, that's along the lines I was thinking for IMEP. From what I've seen that's getting into or beyond the range of the most modern turbo direct injected gas and diesels engines. I'm curious what boost and timing (and on what turbo) that corresponds to.
So the crank sensor--does it use the stock trigger wheel? How does it work? I can say from personal experience that expensive lab-grade is ~720 teeth for a resolution of 1/2 a crank angle degree. Most of the better production crank triggers use ~60 teeth for a resolution of every 6 crank degrees.
Honda's onboard cylinder pressure-based knock control for the aborted V10 NSX was designed to operate accurately for every crank degree. It seems like that would mean a trigger wheel with ~360 teeth. The system was also implemented as part of the misfire detection to protect the catalytic converters. It used two dedicated 100mhz processors, one for each bank, with a sample rate of 50 kHZ. As Howard has run into, a Power FC samples at about 20 hz hopefully.
So the crank sensor--does it use the stock trigger wheel? How does it work? I can say from personal experience that expensive lab-grade is ~720 teeth for a resolution of 1/2 a crank angle degree. Most of the better production crank triggers use ~60 teeth for a resolution of every 6 crank degrees.
Honda's onboard cylinder pressure-based knock control for the aborted V10 NSX was designed to operate accurately for every crank degree. It seems like that would mean a trigger wheel with ~360 teeth. The system was also implemented as part of the misfire detection to protect the catalytic converters. It used two dedicated 100mhz processors, one for each bank, with a sample rate of 50 kHZ. As Howard has run into, a Power FC samples at about 20 hz hopefully.
Last edited by arghx; 03-21-12 at 10:02 PM. Reason: NSX
03-22-12, 07:24 AM
#15
Senior Member
Join Date: Nov 2008
Location: Czech republic
Posts: 357
Likes: 0
Received 0 Likes
on
0 Posts
Max, it is a TFX unit and uses an Optrand sensor.
It shows where the pressure rise is at its max rate of increase but not the rate itself.
My latest tuning experiment has the IMEP at close to 325 psi and the Max pressure at almost 1200psi.
I believe that there are other rotary users but no one else has posted.
Barry
It shows where the pressure rise is at its max rate of increase but not the rate itself.
My latest tuning experiment has the IMEP at close to 325 psi and the Max pressure at almost 1200psi.
I believe that there are other rotary users but no one else has posted.
Barry
Otherwise such IMEP would indicate about 345 lbf·ft indicated torque and assuming it happens around 7500 rpms, about 490 IHP. From NASA enablement program it can be noted that two rotor wankel of similar geometry and displacement around this power level losses at least 70 HP as friction horsepower. So brake power should be around 420 BHP.
They also show heat release rate and pressure curves with peak at 20°ATDC (Direct injection with pilot injection). Maybe you should exploit very low boost setting and optimize mixture and timing for higher peak pressures in earlier crank degrees. Maybe even shut down water injection as uneven flow rate could be large part of the combustion scatter. But thats just my opinion
03-23-12, 05:40 PM
#16
"Elusive, not deceptive!”
325psi is about 22 bar, that's along the lines I was thinking for IMEP. From what I've seen that's getting into or beyond the range of the most modern turbo direct injected gas and diesels engines. I'm curious what boost and timing (and on what turbo) that corresponds to.
I am running a TO4s on some early logs and a TO4Z on later ones.
All runs are done on a HKS 10# spring but boost depends on air temp and turbo. About 12-14 psi.
Clint from TFX says maybe we have enough advance and the next test would be to see how much we lose by backing it off some.
So the crank sensor--does it use the stock trigger wheel? How does it work? I can say from personal experience that expensive lab-grade is ~720 teeth for a resolution of 1/2 a crank angle degree. Most of the better production crank triggers use ~60 teeth for a resolution of every 6 crank degrees.
My thought on this is that ignition firing accuracy is critical... but timing where the pressure wave is not so much so. That being said it is probably within a half degree.
Honda's onboard cylinder pressure-based knock control for the aborted V10 NSX was designed to operate accurately for every crank degree. It seems like that would mean a trigger wheel with ~360 teeth. The system was also implemented as part of the misfire detection to protect the catalytic converters. It used two dedicated 100mhz processors, one for each bank, with a sample rate of 50 kHZ. As Howard has run into, a Power FC samples at about 20 hz hopefully.
Honda's system can readjust individual cylinders for best power on the run.
We can only adjust to the weakest rotor in its weakest range.
I am running a TO4s on some early logs and a TO4Z on later ones.
All runs are done on a HKS 10# spring but boost depends on air temp and turbo. About 12-14 psi.
Clint from TFX says maybe we have enough advance and the next test would be to see how much we lose by backing it off some.
So the crank sensor--does it use the stock trigger wheel? How does it work? I can say from personal experience that expensive lab-grade is ~720 teeth for a resolution of 1/2 a crank angle degree. Most of the better production crank triggers use ~60 teeth for a resolution of every 6 crank degrees.
My thought on this is that ignition firing accuracy is critical... but timing where the pressure wave is not so much so. That being said it is probably within a half degree.
Honda's onboard cylinder pressure-based knock control for the aborted V10 NSX was designed to operate accurately for every crank degree. It seems like that would mean a trigger wheel with ~360 teeth. The system was also implemented as part of the misfire detection to protect the catalytic converters. It used two dedicated 100mhz processors, one for each bank, with a sample rate of 50 kHZ. As Howard has run into, a Power FC samples at about 20 hz hopefully.
Honda's system can readjust individual cylinders for best power on the run.
We can only adjust to the weakest rotor in its weakest range.
Barry
03-23-12, 06:19 PM
#17
"Elusive, not deceptive!”
How exactly is IMEP calculated? Actual pressure trace can be measured, but mean effective pressure is calculated from engine torque and displacement. Or does the software somehow calculate extracted energy from given cycle based on pressure or temperature drop?
Libor, really I don't know.
I had to provide an Excel sheet showing displacement for each degree of rotation.
The pressure is read for each degree so the combustion pressure curve can be drawn. All the rest are calculations I guess.
A side note. This is where the actual displacement of our rotaries comes in... we know the pressure so if you double the actual displacement,
as some want to do... bingo... 1000HP at 12lbs boost!!!!
Otherwise such IMEP would indicate about 345 lbf·ft indicated torque and assuming it happens around 7500 rpms, about 490 IHP. From NASA enablement program it can be noted that two rotor wankel of similar geometry and displacement around this power level losses at least 70 HP as friction horsepower. So brake power should be around 420 BHP.
This is crank HP and not RWHP also. So drop another 15-18%
They also show heat release rate and pressure curves with peak at 20°ATDC (Direct injection with pilot injection). Maybe you should exploit very low boost setting and optimize mixture and timing for higher peak pressures in earlier crank degrees. Maybe even shut down water injection as uneven flow rate could be large part of the combustion scatter. But thats just my opinion
Libor, really I don't know.
I had to provide an Excel sheet showing displacement for each degree of rotation.
The pressure is read for each degree so the combustion pressure curve can be drawn. All the rest are calculations I guess.
A side note. This is where the actual displacement of our rotaries comes in... we know the pressure so if you double the actual displacement,
as some want to do... bingo... 1000HP at 12lbs boost!!!!
Otherwise such IMEP would indicate about 345 lbf·ft indicated torque and assuming it happens around 7500 rpms, about 490 IHP. From NASA enablement program it can be noted that two rotor wankel of similar geometry and displacement around this power level losses at least 70 HP as friction horsepower. So brake power should be around 420 BHP.
This is crank HP and not RWHP also. So drop another 15-18%
They also show heat release rate and pressure curves with peak at 20°ATDC (Direct injection with pilot injection). Maybe you should exploit very low boost setting and optimize mixture and timing for higher peak pressures in earlier crank degrees. Maybe even shut down water injection as uneven flow rate could be large part of the combustion scatter. But thats just my opinion
I have not seen any combustion pressure charts for any engines that did not have scatter. Clint says that some diesels don't have much scatter.
I am chicken to shut the water off but that test needs to be done.
Happy for your input.
Barry
03-23-12, 08:25 PM
#18
Measuring cylinder pressure traces has a lot to do with spark timing, but yes we are perhaps going on a tangent here about Barry's equipment.
Barry, you were asking how an OEM would calculate MBT spark timing in a way that makes sense for onboard use in a stock ECU. I don't know the exact specifics with a rotary. On a piston engine you can start at the 60% mass fraction burnt time in crank degrees and then calculate backwards from there.
"60% mass fraction burnt" comes from generally agreed ballpark figures. Normally peak combustion chamber pressure on a good cycle occurs between 12 and 15 degrees ATDC (-12 to -15 BTDC), and at this time about 60% of the mass inside the cylinder has been burnt.
We use a first-order differential equation to describe the basic relationship:
instantaneous range of change in combustion gas mass = the gas density * the flame area * the combustion rate
then set up a series of separate calculations that take this mathematical relationship and apply it to the time period from 0% fraction burnt through 60% burnt. You are working backwards to figure out how long the whole process should take to have peak pressure at the point in the cycle. A lot of it requires lookup tables or single values (constants) in the ECU to make the model work.
The basic things in the ECU that need to either be calculated through a model or looked up in tables:
1. Ignition delay--time between the plug getting the signal to fire and the beginning of the burn
2. Unburned gas density & combustion mass
3. Flame velocity--broken into a laminar (smooth-ish) component and turbulent component
4. EGR, AFR, and water temperature compensation
This is the kind of stuff you find in a modern stock ECU. It's far from perfect, but used in conjunction with other adjustment maps it's more accurate than having a different lookup table for every situation. It's also a lot cheaper than having a system that calculates MBT by directly measuring combustion chamber pressure (those do exist in experimental form but cost is way too high).
.
Barry, you were asking how an OEM would calculate MBT spark timing in a way that makes sense for onboard use in a stock ECU. I don't know the exact specifics with a rotary. On a piston engine you can start at the 60% mass fraction burnt time in crank degrees and then calculate backwards from there.
"60% mass fraction burnt" comes from generally agreed ballpark figures. Normally peak combustion chamber pressure on a good cycle occurs between 12 and 15 degrees ATDC (-12 to -15 BTDC), and at this time about 60% of the mass inside the cylinder has been burnt.
We use a first-order differential equation to describe the basic relationship:
instantaneous range of change in combustion gas mass = the gas density * the flame area * the combustion rate
then set up a series of separate calculations that take this mathematical relationship and apply it to the time period from 0% fraction burnt through 60% burnt. You are working backwards to figure out how long the whole process should take to have peak pressure at the point in the cycle. A lot of it requires lookup tables or single values (constants) in the ECU to make the model work.
The basic things in the ECU that need to either be calculated through a model or looked up in tables:
1. Ignition delay--time between the plug getting the signal to fire and the beginning of the burn
2. Unburned gas density & combustion mass
3. Flame velocity--broken into a laminar (smooth-ish) component and turbulent component
4. EGR, AFR, and water temperature compensation
This is the kind of stuff you find in a modern stock ECU. It's far from perfect, but used in conjunction with other adjustment maps it's more accurate than having a different lookup table for every situation. It's also a lot cheaper than having a system that calculates MBT by directly measuring combustion chamber pressure (those do exist in experimental form but cost is way too high).
.
03-24-12, 08:10 AM
#19
Racing Rotary Since 1983
Thread Starter
iTrader: (6)
i still am amazed at this series of data...
the top chart is a burn speed comparison between straight gasoline and E85. note the HUGE difference as boost (X line) is added!
gasoline around 11-12 degrees and E85 (alcohol) at over 30! note how linear alcohol behaves as boost increases and how very progressive gasoline changes, as well as how much it degrades burnspeed-wise under boost.
then importantly, there are EGTs... this is especially interesting as i assume these EGTs do result from proper timing. often people complain of too hot egts or too cold egts and the condition may be because of timing not being where it should be. (firing out the exhaust or too early).
here we have proper timing and look at the advantage of alcohol.
the first comparative plot boostwise shows 810 V 770 or 1490 V 1418.
the next similar boost plot is 870 V 800 or 1598 V 1472
finally the last (highest common boost plot 950 V 820 or 1742 V 1508!
the data supports my last dyno session:
my fuel mix was 93 octane pump gas as base fuel w 31% to total BTUs being provided my methanol (alcohol).
at 24 psi here are my EGTs:
N12
4800
1438-1452-1474-1502-1530-1517-1453-1487-1464
timing IGL w 11 split
11-----12---12------12-----13-----13-----14---14---14
going back to 2004, the start of my methanol, my egts would linearly climb to about 1780 at 8000.
the difference is the amount of alcohol being used. my new setup uses two EV14 1000 CC injectors overdriven to 123 psi rail pressure which adds 69% flow to the 1000. they were running at 74% duty.
i am thinking that i may be able to use more advance.
howard
Last edited by Howard Coleman; 03-24-12 at 08:13 AM.
03-24-12, 08:33 AM
#20
Howard,
I know myself or someone else posted that in another thread somewhere. It's slightly out of context so I don't want to have any confusion about it--I can email you the source document if you'd like to see the whole study. Technically, none of the graphs show the exact length of the combustion period (say, the time from 10% mass fraction burned to 90%). You're completely on track in your overall interpretation though.
1. The top chart is commanded ignition advance. It is essentially the commanded spark timing (when the plug has been fired) punched into the engine dyno system. If you ever watch an engine dyno at work, you can do step-by-step adjustments of rpm, load, timing etc in a steady state rather than changing maps in a laptop and going WOT on a chassis dyno.
2. The middle chart is the degrees ATDC when 50% of the mass in the combustion chamber for this experimental engine was burned. A higher number is considered later combustion phasing, meaning that the engine is going to run less optimally overall and have higher EGT. If you look closely, it says on the chart that their target is 6 degrees ATDC for 50% mass fraction burnt. This isn't possible at higher loads on 98 octane RON fuel (roughly 93 octane pump gas in the USA).
3. The bottom is exhaust temperature measured before the turbo, which is straightforward enough.
As you were saying, the ethanol fuel allows more spark advance as load/boost increases (1st graph). Thus the combustion does not have to be delayed (2nd graph) and turbine inlet temperatures are significantly lower (3rd graph). I think you should consider switching to ethanol versus methanol for further improvement in knock resistance. I have a Honda study related to this--Honda engineers seemed to like Toluene a lot (despite its need to be heated) and methanol not so much. Let me know if you want to check it out.
I know myself or someone else posted that in another thread somewhere. It's slightly out of context so I don't want to have any confusion about it--I can email you the source document if you'd like to see the whole study. Technically, none of the graphs show the exact length of the combustion period (say, the time from 10% mass fraction burned to 90%). You're completely on track in your overall interpretation though.
1. The top chart is commanded ignition advance. It is essentially the commanded spark timing (when the plug has been fired) punched into the engine dyno system. If you ever watch an engine dyno at work, you can do step-by-step adjustments of rpm, load, timing etc in a steady state rather than changing maps in a laptop and going WOT on a chassis dyno.
2. The middle chart is the degrees ATDC when 50% of the mass in the combustion chamber for this experimental engine was burned. A higher number is considered later combustion phasing, meaning that the engine is going to run less optimally overall and have higher EGT. If you look closely, it says on the chart that their target is 6 degrees ATDC for 50% mass fraction burnt. This isn't possible at higher loads on 98 octane RON fuel (roughly 93 octane pump gas in the USA).
3. The bottom is exhaust temperature measured before the turbo, which is straightforward enough.
As you were saying, the ethanol fuel allows more spark advance as load/boost increases (1st graph). Thus the combustion does not have to be delayed (2nd graph) and turbine inlet temperatures are significantly lower (3rd graph). I think you should consider switching to ethanol versus methanol for further improvement in knock resistance. I have a Honda study related to this--Honda engineers seemed to like Toluene a lot (despite its need to be heated) and methanol not so much. Let me know if you want to check it out.
03-24-12, 10:33 AM
#21
Racing Rotary Since 1983
Thread Starter
iTrader: (6)
"switching to ethanol versus methanol for further improvement in knock resistance"
i think i would need some selling on that idea.
Latent heat per gallon (cooling capability)
Methanol 3136 BTUs
Ethanol 2398
gasoline 952.... just thought i would throw that in BTW that includes racegas
perhaps that isn't the whole story. pls enlighten me.
i do know that most full tilt drag racers run meth rather than ethanol. currently various race sanctioning bodies are running ethanol but the reason is for, uh, another forum... political type
bottom line for me is taking a step back and looking at the data, the difference is profound.
howard
i think i would need some selling on that idea.
Latent heat per gallon (cooling capability)
Methanol 3136 BTUs
Ethanol 2398
gasoline 952.... just thought i would throw that in BTW that includes racegas
perhaps that isn't the whole story. pls enlighten me.
i do know that most full tilt drag racers run meth rather than ethanol. currently various race sanctioning bodies are running ethanol but the reason is for, uh, another forum... political type
bottom line for me is taking a step back and looking at the data, the difference is profound.
howard
03-24-12, 12:05 PM
#22
Senior Member
Join Date: Nov 2008
Location: Czech republic
Posts: 357
Likes: 0
Received 0 Likes
on
0 Posts
Second point, methanol does preignite very violently. Flashpoints and evaporative cooling has nothing to do with it and when you hit the google search with right keywords, you will find many research papers which examines this phenomena.
Methanol due to its lower peak flame temperatures can be slightly more efficient than gasoline fuels on energy basis, but on the mass basis, its abortion. Many OEM manufactures are making studies of highly boosted small capacity spark ignited engines with direct injection of methanol or ethanol as mean of charge coolant (note:main fuel supply is usually port injected gasoline) which allows them to operate engine with lambda 1 and MBT timing ie extremely efficiently, but they are well aware of disadvantages of methanol and consequently, ethanol is number one choose.
03-24-12, 12:47 PM
#23
Racing Rotary Since 1983
Thread Starter
iTrader: (6)
autoignition temps
ethanol 793
methanol 878
gasoline 495
i understand that if you cool the manifold it is subtractive to cooling in the chamber. it is also indicative of the difference between gasoline and alcohol. clearly alcohol has alot of remaining cooling capacity as it (also) works in the chamber. i could never make over 500 rwhp SAE on 93 octane. i would have lots of knock and break things. most who do run 100% meth AI find the logged Power FC knock drops to under 15 at max torque and max power. meth is magic based on my dyno experience going back to 04.
that said, i am very interested and open to explore ethanol V meth as my AI injectant.
i am going to call a friend this weekend and ask him. he runs 100% meth on his 2 rotor 13 brew, makes over 1000 rwhp, and enjoys surprising longevity to his motors which he builds.
empiricism and theory together is a good formula.
howard
ethanol 793
methanol 878
gasoline 495
i understand that if you cool the manifold it is subtractive to cooling in the chamber. it is also indicative of the difference between gasoline and alcohol. clearly alcohol has alot of remaining cooling capacity as it (also) works in the chamber. i could never make over 500 rwhp SAE on 93 octane. i would have lots of knock and break things. most who do run 100% meth AI find the logged Power FC knock drops to under 15 at max torque and max power. meth is magic based on my dyno experience going back to 04.
that said, i am very interested and open to explore ethanol V meth as my AI injectant.
i am going to call a friend this weekend and ask him. he runs 100% meth on his 2 rotor 13 brew, makes over 1000 rwhp, and enjoys surprising longevity to his motors which he builds.
empiricism and theory together is a good formula.
howard
Last edited by Howard Coleman; 03-24-12 at 12:50 PM.
03-24-12, 01:30 PM
#24
Senior Member
Join Date: Nov 2008
Location: Czech republic
Posts: 357
Likes: 0
Received 0 Likes
on
0 Posts
You must realize that this test is done in somewhat atmospheric conditions and doesn't reflect what happens in the chamber of engine. From research papers covering this phenomena in engines, it can be noted that methanol through certain catalyst process and also due to its volatility falls apart and creates unstable combustion species which spontaneously ignite at much lower temperature and pressure.
Also, just think about temperature of spark plug electrode tip. Its much higher than these values. Based on this, every engine should preignite to death even with methanol...
You should ask him, and maybe ask yourself, why methanol fueled engines are run so rich on lambda scale.
When you consider difference in fuel mass between gasoline fueled engine with certain lambda, and methanol fueled one with same lambda and huge cooling properties even increasing effect, you must ask yourself, why so much is needed? And why race gasoline could do same even without all the cooling. Answer is preignition of this very tricky fuel.
Also, just think about temperature of spark plug electrode tip. Its much higher than these values. Based on this, every engine should preignite to death even with methanol...
You should ask him, and maybe ask yourself, why methanol fueled engines are run so rich on lambda scale.
When you consider difference in fuel mass between gasoline fueled engine with certain lambda, and methanol fueled one with same lambda and huge cooling properties even increasing effect, you must ask yourself, why so much is needed? And why race gasoline could do same even without all the cooling. Answer is preignition of this very tricky fuel.
03-24-12, 09:03 PM
#25
For what it's worth, I've heard a few people speculate that the burn rate of today's pump gasoline is significantly different than it was 10-20 years ago. Someone who has spent a while tuning Nissan engines claimed that common pump-gas SR20 setups today are making the same power with less ignition advance vs what he experienced 5-10 years ago. I'm not sure how well this translates over to rotary engines, but I thought it was interesting.
PS, E85 is great for piston engines, I've seen a turbo engine run 10-15 degrees more ignition advance after already having been tuned to the 'MBT' timing values (Minimum Best Torque timing, or whatever people like to call it) without a significant loss of power, or 'squiggle' in the dyno's power readout, or audible detonation. The tuner did it just to see if they could make the engine detonate on E85, apparently he was already planning to remove the engine and upgrade the internals soon.
PS, E85 is great for piston engines, I've seen a turbo engine run 10-15 degrees more ignition advance after already having been tuned to the 'MBT' timing values (Minimum Best Torque timing, or whatever people like to call it) without a significant loss of power, or 'squiggle' in the dyno's power readout, or audible detonation. The tuner did it just to see if they could make the engine detonate on E85, apparently he was already planning to remove the engine and upgrade the internals soon.