Lean burn with negative split timing
04-10-2012, 10:33 PM
#201
Sorry, I'm conflating things a bit.
Lean combustion lowers peak combustion temperatures and EGTs, but it increases mean cycle temps.
The latter is the reason that lean mixtures can oxidize exhaust valves in piston engines. However, the exhaust gas ought to be cooler by the time it hits the exhaust port, and that might help the side seal springs.
On the other hand, it seems to me that higher mean cycle temps might increase issues seen on the face of the rotor, around the spark plugs for example.
Lean combustion lowers peak combustion temperatures and EGTs, but it increases mean cycle temps.
The latter is the reason that lean mixtures can oxidize exhaust valves in piston engines. However, the exhaust gas ought to be cooler by the time it hits the exhaust port, and that might help the side seal springs.
On the other hand, it seems to me that higher mean cycle temps might increase issues seen on the face of the rotor, around the spark plugs for example.
04-11-2012, 10:31 AM
#203
Regular 50/50 engine coolant only picks up approx 10F as it pass'es the exhaust side of the engine. Only 10 degrees F. So if the coolant is boiling around the exhaust ports then it is very close to that on the hot side of the engine?
I have measured the delta between the combustion side and the exhaust side in many scenarios and it remains constant. Approx 10 degrees F.
I would LOVE to see a single pass coolant system in the rotary engine. Not one that makes a turn in the rear iron and comes back to the front. But, one that would enter the front on both the hot and cold side and both exit in the rear. The side exhaust port changed everything.
Cooling the exhaust port more, the sparkplug areas and some stratigic engine coatings should allow for more timing and leaner a/f's without any risk to longetivity. Talking about over 15:1 for steady cruise with no side affects.
.
I have measured the delta between the combustion side and the exhaust side in many scenarios and it remains constant. Approx 10 degrees F.
I would LOVE to see a single pass coolant system in the rotary engine. Not one that makes a turn in the rear iron and comes back to the front. But, one that would enter the front on both the hot and cold side and both exit in the rear. The side exhaust port changed everything.
Cooling the exhaust port more, the sparkplug areas and some stratigic engine coatings should allow for more timing and leaner a/f's without any risk to longetivity. Talking about over 15:1 for steady cruise with no side affects.
.
04-11-2012, 11:03 AM
#204
Some clarification. Water free coolant does run hotter (ignoring the control from the thermostat.) Because it has a lower specific heat capacity than water. Water can boil even when the majority of the water is colder than the boiling point. This starts as nucleate boiling and ends up as departure from nucleat boiling. Nucleate boiling uses steam bubbles to transfer heat to the majority of the coolant and is more effective at heat transfer than no nucleate boiling. Departure from nucleate boiling happens when steam insulates the heat transfer surface, and gives a MUCH lower heat transfer. So wherever the engine is transfering the most heat to the coolant per area you run the risk of DNB. This is annother reason why water free coolant runs hotter, it doesn't get insulated by steam.
The end result is hotter coolant, but cooler internal temps (after DNB would have happens). Also a much higher temp for boilover. The question isn't if it works, but rather how much DNB is going on at given load/RPM, and how much the engineers had to account for. (Or what we can get away with water free coolant.)
Also, we are derailing this thread badly. I'll start a new thread when I have results on 50/50.
The end result is hotter coolant, but cooler internal temps (after DNB would have happens). Also a much higher temp for boilover. The question isn't if it works, but rather how much DNB is going on at given load/RPM, and how much the engineers had to account for. (Or what we can get away with water free coolant.)
Also, we are derailing this thread badly. I'll start a new thread when I have results on 50/50.
04-13-2012, 09:18 PM
#205
Posted a thread on waterless coolant in the Tech Garage. Unfortunately the results were not far enough off to account for what I'm seeing. So I dusted of the spare engine and started the strip down for rebuild. Now all I have to do is remove the dreaded Mazda 2 piece side seal.
Edit: Make that a 3 piece side seal. I'm giving up for the night.
Edit: Make that a 3 piece side seal. I'm giving up for the night.
Last edited by Harlan; 04-13-2012 at 11:01 PM.
05-20-2014, 08:37 PM
#207
Wow I've learned a lot since then. Bringing this thread back because I'm trying it again with a bit more knowledge.
We all tried lean burn, but without any idea of what it should look like or where we were going. After some time, and some research I have some information to share. Most of this came from NACA reports (Wartime report 264 was the most helpful http://www.naca.central.cranfield.ac...a-wr-e-264.pdf) and studying what little information is available about the Shell OPEL P-1 (376 MPG | The Story of the 1959 Opel P-1).
Why lean burn?
Lean burn is NOT about throttling losses. Yes relieving throttling losses gains some MPG. Yes, lean burn lowers throttling losses, but that’s not all it does.
The Shell OPEL P-1 is very telling. It’s a big block V-8 with an intake heater and the entire cooling system insulated. This is not by accident a smaller engine would have had worse fuel economy!
Ideal thermal cycle efficiency is based on the difference between the hot side and the cold side. The smaller the difference the lower the efficiency. In the real world however there are other losses to contend with, so the higher the combustion temperature the more energy is lost to coolant and oil. On a rotary this is even worse than a piston engine because of the combustion chamber shape.
The peak combustion temperature is based on several things, the amount of energy created by combustion of the fuel/air, the specific heat capacity of the air (fuel specific heat capacity is mostly insignificant), and the amount of heat lost to the chamber.
The hottest combustion happens at stoich. Go richer and the temperature falls because of the cooling caused by vaporizing extra fuel. Go leaner and the temperature falls because of the specific heat capacity added by the extra air. Simply put you are trying to use the same amount of energy to heat something bigger on either side. You can only burn as much air as there is fuel and you can only burn as much fuel as there is air.
As you go leaner the combustion temperature drops. Twice the air (with the same fuel) means half the temperature rise. Because peak combustion temperature is lower, EGT is lower, losses to coolant and oil are lower. You can continue to go leaner until cycle efficiency becomes more limiting, or you run out of air, or the fuel can no longer completely combust in the time available. More displacement means more air available for a given HP output.
Most cars have a lean limit of around 17:1. This is when even with properly advanced timing combustion can no longer be maintained and misfires begin stealing power. In the Shell P-1 this was overcome by heating the fuel/air. The hotter the mix the better it burns and the leaner it can be before hitting the limit.
Mazda had a lean burn RX4. While I haven’t found much information on it, I do know why the rotary engine is uniquely suited for lean burn. The gasoline that does not vaporize pools at the rotor trailing edge. This is because of inertia, vaporized gasoline follows the air and quickly follows the chamber, liquid gasoline is flung at the rotor and then is swept around by the apex seal. The MSP design was built around the idea that if you prevent the liquid gasoline from escaping it can come back around for combustion on the next pass for better FE.
Current lean burn piston engines have a complex head design, or a specialized fuel injector that causes the spark plug to have a rich mixture surrounding it when combustion starts. This allows the first combustion (of the relatively rich fuel) to increase the dynamic compression of the lean fuel before it ignites, or in some cases causes the combustion of rich fuel to the heat air without fuel. The whole purpose of these systems is to make the lean fuel air burn fast like a much richer mixture so that the power can be used at lower temperatures.
Rotary engines already have a lean burn spark plug, it’s the trailing plug. As compression happens the richer fuel/air mix is at the back of the rotor. As it’s compressed the lean mix is in the bathtub and the rich mix is in the trailing plug recess. Fire the trailing plug first and the compression is much higher when the leading plug lights off the lean mix. Voila, lean burn.
As you go leaner, more advance must be applied because of the slower burn, but you are actually moving away from the AFR region where knock is the worst.
I believe that with proper tuning very impressive fuel economy will be possible on the RX8. There is no scary ledge as you go lean, no need to go lean fast to protect the engine, just transition there and make adjustments. And no mystery heat built up in the engine, nor extra heat on the side seals quite the opposite in fact. We cruise at the worst possible region for heat buildup and detonation.
To this end I’ve disconnected my rear o2 sensor and set the OL exit to 500 RPM. This has given me complete control over fueling, and now the fun can begin. I’m doing this in baby steps, the first being to find the lean limit on near stock timing. Today I ran at ~15.3 AFR, EGTs are near normal, possibly lower. Since the only way to be sure about fuel economy is odometer vs pump I’m not going to post tune by tune FE updates, instead I’m going to post tank by tank. Timing is stock leading with a -5 trailing split until 4k rpm or 60% load, then transition to stock.
Also I do have det cans set up, so I am continuously monitoring for detonation, but the science says it's unlikely!
We all tried lean burn, but without any idea of what it should look like or where we were going. After some time, and some research I have some information to share. Most of this came from NACA reports (Wartime report 264 was the most helpful http://www.naca.central.cranfield.ac...a-wr-e-264.pdf) and studying what little information is available about the Shell OPEL P-1 (376 MPG | The Story of the 1959 Opel P-1).
Why lean burn?
Lean burn is NOT about throttling losses. Yes relieving throttling losses gains some MPG. Yes, lean burn lowers throttling losses, but that’s not all it does.
The Shell OPEL P-1 is very telling. It’s a big block V-8 with an intake heater and the entire cooling system insulated. This is not by accident a smaller engine would have had worse fuel economy!
Ideal thermal cycle efficiency is based on the difference between the hot side and the cold side. The smaller the difference the lower the efficiency. In the real world however there are other losses to contend with, so the higher the combustion temperature the more energy is lost to coolant and oil. On a rotary this is even worse than a piston engine because of the combustion chamber shape.
The peak combustion temperature is based on several things, the amount of energy created by combustion of the fuel/air, the specific heat capacity of the air (fuel specific heat capacity is mostly insignificant), and the amount of heat lost to the chamber.
The hottest combustion happens at stoich. Go richer and the temperature falls because of the cooling caused by vaporizing extra fuel. Go leaner and the temperature falls because of the specific heat capacity added by the extra air. Simply put you are trying to use the same amount of energy to heat something bigger on either side. You can only burn as much air as there is fuel and you can only burn as much fuel as there is air.
As you go leaner the combustion temperature drops. Twice the air (with the same fuel) means half the temperature rise. Because peak combustion temperature is lower, EGT is lower, losses to coolant and oil are lower. You can continue to go leaner until cycle efficiency becomes more limiting, or you run out of air, or the fuel can no longer completely combust in the time available. More displacement means more air available for a given HP output.
Most cars have a lean limit of around 17:1. This is when even with properly advanced timing combustion can no longer be maintained and misfires begin stealing power. In the Shell P-1 this was overcome by heating the fuel/air. The hotter the mix the better it burns and the leaner it can be before hitting the limit.
Mazda had a lean burn RX4. While I haven’t found much information on it, I do know why the rotary engine is uniquely suited for lean burn. The gasoline that does not vaporize pools at the rotor trailing edge. This is because of inertia, vaporized gasoline follows the air and quickly follows the chamber, liquid gasoline is flung at the rotor and then is swept around by the apex seal. The MSP design was built around the idea that if you prevent the liquid gasoline from escaping it can come back around for combustion on the next pass for better FE.
Current lean burn piston engines have a complex head design, or a specialized fuel injector that causes the spark plug to have a rich mixture surrounding it when combustion starts. This allows the first combustion (of the relatively rich fuel) to increase the dynamic compression of the lean fuel before it ignites, or in some cases causes the combustion of rich fuel to the heat air without fuel. The whole purpose of these systems is to make the lean fuel air burn fast like a much richer mixture so that the power can be used at lower temperatures.
Rotary engines already have a lean burn spark plug, it’s the trailing plug. As compression happens the richer fuel/air mix is at the back of the rotor. As it’s compressed the lean mix is in the bathtub and the rich mix is in the trailing plug recess. Fire the trailing plug first and the compression is much higher when the leading plug lights off the lean mix. Voila, lean burn.
As you go leaner, more advance must be applied because of the slower burn, but you are actually moving away from the AFR region where knock is the worst.
I believe that with proper tuning very impressive fuel economy will be possible on the RX8. There is no scary ledge as you go lean, no need to go lean fast to protect the engine, just transition there and make adjustments. And no mystery heat built up in the engine, nor extra heat on the side seals quite the opposite in fact. We cruise at the worst possible region for heat buildup and detonation.
To this end I’ve disconnected my rear o2 sensor and set the OL exit to 500 RPM. This has given me complete control over fueling, and now the fun can begin. I’m doing this in baby steps, the first being to find the lean limit on near stock timing. Today I ran at ~15.3 AFR, EGTs are near normal, possibly lower. Since the only way to be sure about fuel economy is odometer vs pump I’m not going to post tune by tune FE updates, instead I’m going to post tank by tank. Timing is stock leading with a -5 trailing split until 4k rpm or 60% load, then transition to stock.
Also I do have det cans set up, so I am continuously monitoring for detonation, but the science says it's unlikely!
05-21-2014, 11:28 AM
#208
I would like to see your results!
my timing map has some negitive split in the cruise regions, but nothing dramatic. I haven't yet wrapped my head around what timing maps really mean in corrilation to the rotor
I know the old school guys used the crank shaft to figure out timing, and thats 1/3 differents
so I was pondering this morning, say the crank is 21 degrees btdc does that really mean the rotor is 7 degree BTDC?
my timing map has some negitive split in the cruise regions, but nothing dramatic. I haven't yet wrapped my head around what timing maps really mean in corrilation to the rotor
I know the old school guys used the crank shaft to figure out timing, and thats 1/3 differents
so I was pondering this morning, say the crank is 21 degrees btdc does that really mean the rotor is 7 degree BTDC?
05-21-2014, 12:00 PM
#209
Driving my unreliable rx8
Farza, I am not positive, but I'm pretty sure it's not based on crank. It's based on the piston, or in our case the rotor face being at top dead center. So if it is 21. Then that one face is 21 degrees before top dead center.
05-21-2014, 02:34 PM
#211
21 deg BTDC is what it is, the rotor doesn't move up and down like a piston so I can understand the confusion, but ignition timing is still based on crank angle. Negative split is required to overcome lean burn limit, but just setting negative split in your ignition map is not going to cause lean burn. Also be careful about running negative split under high load or high rpm the trailing insulator can get hot enough to cause preignition.
Team, while I'm impressed by your basic math skills, I don't really understand what you are referencing, and just because documents are ancient does not make them wrong.
Team, while I'm impressed by your basic math skills, I don't really understand what you are referencing, and just because documents are ancient does not make them wrong.
05-21-2014, 06:53 PM
#214
Back when both military effectiveness and airline profits depended on piston engines, there was a comparatively huge amount of research done on their performance at the limits. For example, the Wright R-3350 Turbo-compound engine not only managed to make up to 3700 hp, but deliver 34% thermal efficiency at cruise power.
F = ma is 400 years old, but its usefulness remains ...
F = ma is 400 years old, but its usefulness remains ...
05-22-2014, 03:26 AM
#215
Here is the current 'do':
Ran today at ~16AFR. No notable change in EGT, however my exhaust is a bit quieter. Next flash will have more split and go a little leaner. First tank of gas is 20.2MPG, since any number of things can effect indications and I don't have a way to do back to back runs, I'm not counting any gain until I'm beyond 22MPG or drop at least 100deg EGT. Anything less could just be a day to day difference. At least with this setup there is nothing tricking the sensors so indications are real.
Ran today at ~16AFR. No notable change in EGT, however my exhaust is a bit quieter. Next flash will have more split and go a little leaner. First tank of gas is 20.2MPG, since any number of things can effect indications and I don't have a way to do back to back runs, I'm not counting any gain until I'm beyond 22MPG or drop at least 100deg EGT. Anything less could just be a day to day difference. At least with this setup there is nothing tricking the sensors so indications are real.
05-24-2014, 01:04 PM
#216
Ok, assuming 16.2:1 AFR
( 1 + 1.5/14.7) x 20 = 22.0
which only applies when it's actually in lean burn mode, based on how finicky a rotary is to the slightest load change in cruise mode like head wind, slight grade increase, etc. I can't see getting near that gain
( 1 + 1.5/14.7) x 20 = 22.0
which only applies when it's actually in lean burn mode, based on how finicky a rotary is to the slightest load change in cruise mode like head wind, slight grade increase, etc. I can't see getting near that gain
05-24-2014, 04:09 PM
#218
My goal is MPG from lean burn, although the lack of carbon buildup is a good thing too.
Been running at 17+AFR and discovering some interesting things about spark plug fouling. Unfortunately a lot of my data is now suspect/bad, so back to 16:1. When the trailing plug fouls it really makes negative split worthless, and I think I managed to lower combustion temperatures out of my plugs heat range I'll know for sure when I pull them.
I really wish I could live tune, it would make a much simpler process.
More after I wake up.
Been running at 17+AFR and discovering some interesting things about spark plug fouling. Unfortunately a lot of my data is now suspect/bad, so back to 16:1. When the trailing plug fouls it really makes negative split worthless, and I think I managed to lower combustion temperatures out of my plugs heat range I'll know for sure when I pull them.
I really wish I could live tune, it would make a much simpler process.
More after I wake up.
05-24-2014, 06:51 PM
#219
Did those other engines intentionally inject oil into the combustion area? There are other reports that discuss the issues encountered with exhaust residue buildup on conventional engines. Even F = ma is a generalization that doesn't cover all the potential detail complexities to be completely accurate ... the reality of our existence does not occur in a vacuum
05-26-2014, 04:02 PM
#223
Ok, so here's how I''m tuning: The goal is for minimum EGT while cruising.
Adding timing has done nothing good, but pulling leading has dropped EGTs about 40-50F. I've been fixing some secondary issues as well, including possible brake dragging so this isn't definitive, but it's the best improvement I've seen yet, and the lowest cruising EGTs I've ever seen.
I'm going continue pulling timing at 16:1 until EGTs begin rising again, then I'm gonna start leaning out, and then pushing trailing timing.
Picked up some very minor knock at very low loads. When I close the throttle for a moment then open it sometimes there is a quick chirp, it's probably due to the ECU catching up on the load change. Regardless it is VERY low level, much quieter than the knock that happens on start-up.
Hmmm... Jump/drop in MPG around 3000 rpm. That's a new one to me, but I don't doubt it indicates it. Secondaries kicking on and messing the the math maybe?
Adding timing has done nothing good, but pulling leading has dropped EGTs about 40-50F. I've been fixing some secondary issues as well, including possible brake dragging so this isn't definitive, but it's the best improvement I've seen yet, and the lowest cruising EGTs I've ever seen.
I'm going continue pulling timing at 16:1 until EGTs begin rising again, then I'm gonna start leaning out, and then pushing trailing timing.
Picked up some very minor knock at very low loads. When I close the throttle for a moment then open it sometimes there is a quick chirp, it's probably due to the ECU catching up on the load change. Regardless it is VERY low level, much quieter than the knock that happens on start-up.
Hmmm... Jump/drop in MPG around 3000 rpm. That's a new one to me, but I don't doubt it indicates it. Secondaries kicking on and messing the the math maybe?
Last edited by Harlan; 05-26-2014 at 04:04 PM.