- Joined
- May 28, 2001
What does a 3D map of your fuel table look like?
Is it anything similar to the efficiency curve of the compressor?
Is it anything similar to the efficiency curve of the compressor?
Advancement of fuel delivery?
- Thread starter Alky V6
- Start date
Hey Reggie.
I'll work on getting the latest 3D fuel map posted. It looks really cool. Nothing you'd ever expect.
That's a good idea. I'll overlay the engine airflow curve on the compressor map and post it for everyone. That should be interesting.
You coming down Saturday? Fuddruckers?
I used two different methods to determine the airflow requirement of the engine. I created a range with each method.
The compressor map is of an Airwerks S510. It uses a compressor wheel with an inducer diameter of 95mm. The compressor inducer diameter of the wheel I'm using is 91mm. So the efficiency islands may be skewed a little to the right on this 95mm compressor map, putting the data points closer to the surge line.
Lines a and b (blue trace lines) use the following formula:
[(1/2 cid x max rpm) / 1728] x VE x DR = cfm.
cfm / 14.5 = lbs/min of air.
The blue crosses from left to right are for VE variables of .80, .87, 1.0, 1.1.
DR is density ratio. The variable used there was 2.84.
RPM values used are 7800 rpm and 4000 rpm.
CID used was 224.
Lines c and d (red trace lines) use the following formula:
BHP / a = lbs/min of air.
Variables for a are 10.00 and 10.86.
Variables for BHP are 1130 and 289.
The compressor map is of an Airwerks S510. It uses a compressor wheel with an inducer diameter of 95mm. The compressor inducer diameter of the wheel I'm using is 91mm. So the efficiency islands may be skewed a little to the right on this 95mm compressor map, putting the data points closer to the surge line.
Lines a and b (blue trace lines) use the following formula:
[(1/2 cid x max rpm) / 1728] x VE x DR = cfm.
cfm / 14.5 = lbs/min of air.
The blue crosses from left to right are for VE variables of .80, .87, 1.0, 1.1.
DR is density ratio. The variable used there was 2.84.
RPM values used are 7800 rpm and 4000 rpm.
CID used was 224.
Lines c and d (red trace lines) use the following formula:
BHP / a = lbs/min of air.
Variables for a are 10.00 and 10.86.
Variables for BHP are 1130 and 289.
Attachments
I'm not an expert at estimating engine airflow and plotting it onto turbocharger compressor maps, so if anyone can see an error in my math, please chime in.
If there isn't any appreciable error in my math, then I really don't see where this 91mm is too big. In the sense of spool up time,... maybe. But, wouldn't that be more of a problem associated with the turbine side? The exhaust side being too free breathing?
As I work the boost pressure up in the future, the trace lines will continue upward through a very efficient range for this turbo. Particularly, the red trace lines and the blue line representing the 1.10 VE variable. Should make for some very efficient high boost conditions. :biggrin:
If there isn't any appreciable error in my math, then I really don't see where this 91mm is too big. In the sense of spool up time,... maybe. But, wouldn't that be more of a problem associated with the turbine side? The exhaust side being too free breathing?
As I work the boost pressure up in the future, the trace lines will continue upward through a very efficient range for this turbo. Particularly, the red trace lines and the blue line representing the 1.10 VE variable. Should make for some very efficient high boost conditions. :biggrin:
Here is the latest fuel table revision. Alky 3.2 25i.
This view shows how the wall has lessened and been extended by a large amount compared to past tables. The wall really hasn't come down as much as the fueling before the wall has come up. The extension of the wall was due to the boost rise ramp up blasting across the wall in less time as tuning changes advanced. A sure sign that boost is now building faster. So the changes are working. Good news.
In fact, the length of the wall, a real good indicator of boost rise time, is very close to what the length of the wall was when the T76 was on the car.
You'll notice how the fueling in relation to rpm has taken on a consistent upward ramping. Fueling appears to have peaked at around 5900 rpm. That would be the point of peak torque for the engine. Most tuning books will tell you that fueling requirements will tend to flatten off after peak torque.
At 2500 to 3000 rpm, there is a massive bump of fueling requirement. This has to do with the characteristics of the camshaft and pulse tuning of the manifolding.
At 3400 to 3800 the fueling dips down and from there the fueling becomes very linear in relation to rpm and manifold boost.
A smaller camshaft would not tend to have such large fueling waves in the fuel table and would be much easier for a beginning tuner to dial in.
Regardless of how the fuel table looks, it has created a very smooth running engine.
This view shows how the wall has lessened and been extended by a large amount compared to past tables. The wall really hasn't come down as much as the fueling before the wall has come up. The extension of the wall was due to the boost rise ramp up blasting across the wall in less time as tuning changes advanced. A sure sign that boost is now building faster. So the changes are working. Good news.
In fact, the length of the wall, a real good indicator of boost rise time, is very close to what the length of the wall was when the T76 was on the car.
You'll notice how the fueling in relation to rpm has taken on a consistent upward ramping. Fueling appears to have peaked at around 5900 rpm. That would be the point of peak torque for the engine. Most tuning books will tell you that fueling requirements will tend to flatten off after peak torque.
At 2500 to 3000 rpm, there is a massive bump of fueling requirement. This has to do with the characteristics of the camshaft and pulse tuning of the manifolding.
At 3400 to 3800 the fueling dips down and from there the fueling becomes very linear in relation to rpm and manifold boost.
A smaller camshaft would not tend to have such large fueling waves in the fuel table and would be much easier for a beginning tuner to dial in.
Regardless of how the fuel table looks, it has created a very smooth running engine.
Attachments
Working the view around.
You cannot make a mechanical fuel injection system to deliver a fuel curve like this. Although, if you're using a mechanical fuel injection system, you're only going to be concerned with 4000 rpm and above, and there it would work well due to the linear delivery characteristics of a mechanical fuel injection system.
At 2440 rpm to 2550 rpm you can see a very steep rise in fueling requirement. You would think that must look very ugly on the O2 readout. Actually, it delivers a smooth reading and the engine runs through it without the slightest hint of a stumble. Very smooth. You just can't tell when the engine has gone through that spot in the fuel map.
Now, if I were running the fuel table flat through that section and relied on O2 correction to take care of that, do you think I'd end up with a smooth transition through that spot? I have a hard time believing that any sensor would react fast enough with the proper authority in time to make that section seemlessly smooth.
You cannot make a mechanical fuel injection system to deliver a fuel curve like this. Although, if you're using a mechanical fuel injection system, you're only going to be concerned with 4000 rpm and above, and there it would work well due to the linear delivery characteristics of a mechanical fuel injection system.
At 2440 rpm to 2550 rpm you can see a very steep rise in fueling requirement. You would think that must look very ugly on the O2 readout. Actually, it delivers a smooth reading and the engine runs through it without the slightest hint of a stumble. Very smooth. You just can't tell when the engine has gone through that spot in the fuel map.
Now, if I were running the fuel table flat through that section and relied on O2 correction to take care of that, do you think I'd end up with a smooth transition through that spot? I have a hard time believing that any sensor would react fast enough with the proper authority in time to make that section seemlessly smooth.
Attachments
If anyone has any questions pertaining to any section of the fuel map, please feel free to ask. There are some interesting reasons behind some small sections in the fuel map.
In case some are wondering why.
I'm freely sharing my fuel table, first of all, because I'm really not worried about someone using it, as it is. Every engine is going to have it's own unique fuel map. So many variables are involved. But most importantly I'm hoping others can get some clues as to what direction to take their own fuel map. If you really want a sweet running engine, you have to dial the fuel map in. Especially if you're running a big cam.
In case some are wondering why.
I'm freely sharing my fuel table, first of all, because I'm really not worried about someone using it, as it is. Every engine is going to have it's own unique fuel map. So many variables are involved. But most importantly I'm hoping others can get some clues as to what direction to take their own fuel map. If you really want a sweet running engine, you have to dial the fuel map in. Especially if you're running a big cam.
Two very important things to realize when setting up your fuel map which could help you get it done quicker.
1) Fuel requirement will basically follow the torque curve of the engine. As torque increases, fuel requirement will increase. Both are closely tied to the VE of the engine. The VE of an engine being determined mainly by camshaft and head specifications.
In general, expect the fuel table to have a slight rise in relation to rpm up to the point of peak torque. After peak torque, it could flatten out or continue to rise at less of an upward slope.
1) Fuel requirement will basically follow the torque curve of the engine. As torque increases, fuel requirement will increase. Both are closely tied to the VE of the engine. The VE of an engine being determined mainly by camshaft and head specifications.
In general, expect the fuel table to have a slight rise in relation to rpm up to the point of peak torque. After peak torque, it could flatten out or continue to rise at less of an upward slope.
The second thing to pay attention to took me months to realize on my own when I first started tuning my system. I'll try to make it as clear as I can to help you all fast track your own fuel map development.
I would imagine this point is going to be less of a consideration on camshafts that are spec'd a little above stock where overlap or short circuiting is very low, if there is any at all. This could be why tuners like to see turbocharged engines with small cams. Much less work getting the fuel map dialed in.
Large camshafts that increase the power band width and engine efficiency in the upper rpm, where an appreciable amount overlap or short circuiting exists, will ultimately cause waves to set up in the fuel table. Especially in the area below the operating range that the camshaft was designed for. As rpm comes off idle and is run up through the rpm band, there will be points along the way where the a/f ratio reading changes drastically, even after the most basic overall fuel curve has been established.
I would imagine this point is going to be less of a consideration on camshafts that are spec'd a little above stock where overlap or short circuiting is very low, if there is any at all. This could be why tuners like to see turbocharged engines with small cams. Much less work getting the fuel map dialed in.
Large camshafts that increase the power band width and engine efficiency in the upper rpm, where an appreciable amount overlap or short circuiting exists, will ultimately cause waves to set up in the fuel table. Especially in the area below the operating range that the camshaft was designed for. As rpm comes off idle and is run up through the rpm band, there will be points along the way where the a/f ratio reading changes drastically, even after the most basic overall fuel curve has been established.
Looking at my fuel map, there is basically 2 deviations off of the basic fuel curve. The first is a dip at 2440 rpm. The second being a wide and high rise in fueling requirement that peaks at 3000 rpm. The fueling then falls to a point at 3700 where things calm down and the camshaft starts working the powerband it was designed to do.
Here is the important #2 tip.
Any dip or rise in fueling requirement will tend to continue that trend throughout the map range for that particular rpm.
The real challenge is working out the true shape of these hills and valleys in the fuel map. It requires precise locating of rpm and map breakpoints in your fuel table.
Any dip or rise in fueling requirement will tend to continue that trend throughout the map range for that particular rpm.
The real challenge is working out the true shape of these hills and valleys in the fuel map. It requires precise locating of rpm and map breakpoints in your fuel table.
Of course, fuel requirement will also trend upward with increased map, but that one is pretty obvious to anyone with enough tuning knowledge to tackle their own tuning, so I didn't think I should need to include it in my important tuning tips list. But,... the number two tuning tip should be carefully heeded. Of all the books I've read on EFI tuning, I have never found anything that gave up that important tuning tip.
Some interesting tid bits of information about methanol.
Heat promotes disassociation of methanol.
The process of methanol disassociation absorbs heat and cools the surrounding atmosphere.
Pressure inhibits methanol disassociation. The higher the pressure, the more methanol disassociation is inhibited. In an internal combustion engine, the temperature rise due to compression slightly outweighs the drawback of loss of disassociation due to the pressure rise from compression.
The main compound that is made available for combustion after the disassociation of methanol is hydrogen. So basically, the engine is a hydrogen burning engine.
Heat promotes disassociation of methanol.
The process of methanol disassociation absorbs heat and cools the surrounding atmosphere.
Pressure inhibits methanol disassociation. The higher the pressure, the more methanol disassociation is inhibited. In an internal combustion engine, the temperature rise due to compression slightly outweighs the drawback of loss of disassociation due to the pressure rise from compression.
The main compound that is made available for combustion after the disassociation of methanol is hydrogen. So basically, the engine is a hydrogen burning engine.
Back to the subject of EGR to help get the mixture up to a good temperature where more disassociation of the methanol is possible for a better burn.
The target temperature range for good vaporization and disassociation of methanol is 600 to 700 degrees F. This would be after heat from compression and just before the spark plug lights the mixture off. A little beyond 700 degrees F and autoignition can become a problem.
One form of EGR can be had by camshaft timing. High overlap cams tend to have spots in the rpm band where exhaust reversion into the chamber, and even up into the intake tract occurs. From the camshaft development work I did for my engine I was well aware that there were a few points of high and low exhaust reversion throughout the rpm band. Previous engine sim work brought this out.
Now that the fuel table is very close to being done, I decided to go back to the engine simulator to see if the high and low spots in the fuel table corresponded with anything in the sim. Amazingly, the amount of exhaust reversion corresponded almost exactly to the high and low spots in the fuel map. 100 rpm off here and there in some spots. Right on in most. It appears that the higher the percentage of exhaust gases that is pushed backed into the chamber and up into the inlet track during overlap reversion, the more fuel that was needed at that point in the fuel map. When the amount of reversion lessened, less fuel was needed.
The percentage of exhaust reversion throughout the rpm band practically mimics the wave in the fuel map throughout the rpm band. Even the sharp rise in fuel requirement from 2440 rpm to 2550 rpm is mimicked by a similar sharp rise in the percentage of exhaust gas reversion at that same exact point. Very interesting stuff.
By 5200 rpm the amount of exhaust gas reversion has fallen (1.8%) and slightly ramps back up as rpm increases. 3.1% at 6750 rpm. Maybe that's one reason why this thing gets so crazy as the rpm increases beyond 5200 rpm.
The target temperature range for good vaporization and disassociation of methanol is 600 to 700 degrees F. This would be after heat from compression and just before the spark plug lights the mixture off. A little beyond 700 degrees F and autoignition can become a problem.
One form of EGR can be had by camshaft timing. High overlap cams tend to have spots in the rpm band where exhaust reversion into the chamber, and even up into the intake tract occurs. From the camshaft development work I did for my engine I was well aware that there were a few points of high and low exhaust reversion throughout the rpm band. Previous engine sim work brought this out.
Now that the fuel table is very close to being done, I decided to go back to the engine simulator to see if the high and low spots in the fuel table corresponded with anything in the sim. Amazingly, the amount of exhaust reversion corresponded almost exactly to the high and low spots in the fuel map. 100 rpm off here and there in some spots. Right on in most. It appears that the higher the percentage of exhaust gases that is pushed backed into the chamber and up into the inlet track during overlap reversion, the more fuel that was needed at that point in the fuel map. When the amount of reversion lessened, less fuel was needed.
The percentage of exhaust reversion throughout the rpm band practically mimics the wave in the fuel map throughout the rpm band. Even the sharp rise in fuel requirement from 2440 rpm to 2550 rpm is mimicked by a similar sharp rise in the percentage of exhaust gas reversion at that same exact point. Very interesting stuff.
By 5200 rpm the amount of exhaust gas reversion has fallen (1.8%) and slightly ramps back up as rpm increases. 3.1% at 6750 rpm. Maybe that's one reason why this thing gets so crazy as the rpm increases beyond 5200 rpm.
Of course a downside to introducing EGR is, it's inert and not dense, and it's very presence in the cylinder displaces and reduces the amount of air available to burn fuel. If the space is occupied by EGR, and thus taking up more space per mass than cooler air, then you will not be able to burn as much fuel. Because of less available air. This effect will compete with the one you are trying to combat.
Of course in your case, injecting N20 just for spooling, it's not as cut and dried a picture.
Seems to me the best way you have readily available here to add heat to your air charge is to do it directly with your inter"heater".
Maybe you could also let the N20 expand out a nozzle well upstream of the intake, ahead of the inter"heater" even. That should reduce the charge cooling effect it's having significantly, no? Could adjust the control scheme to account for the delay.
TurboTR
Of course in your case, injecting N20 just for spooling, it's not as cut and dried a picture.
Seems to me the best way you have readily available here to add heat to your air charge is to do it directly with your inter"heater".
Maybe you could also let the N20 expand out a nozzle well upstream of the intake, ahead of the inter"heater" even. That should reduce the charge cooling effect it's having significantly, no? Could adjust the control scheme to account for the delay.
TurboTR
Another spooling trick (which works) is, shoot compressed air directly at the compressor blades. Like blowing on a pin wheel. Small carefully drilled holes through the compressor housing, at the correct angle serve as the nozzles. Could store the compressed air in like an air shifter bottle 
You'd get like 1 shot per run
One problem with the scheme is though, the compressor blades are angled the wrong direction to be blown like this. The turbine blades are correctly angled. But carefully placed nozzle holes for the compressor side still work well.
Edit- hey there we go- instead of using compressed air, just also spray the nitrous through the pinwheeling holes. Eureka! Lol
You'd get like 1 shot per run One problem with the scheme is though, the compressor blades are angled the wrong direction to be blown like this. The turbine blades are correctly angled. But carefully placed nozzle holes for the compressor side still work well.
Edit- hey there we go- instead of using compressed air, just also spray the nitrous through the pinwheeling holes. Eureka! Lol
All good food for thought, Todd.
I would figure there must be an optimum size hole and amount of holes for the pinwheel idea. Multiple holes evenly spaced around the compressor wheel at the optimum angle and at the right size to generate the highest possible force on the compressor wheel. Nitrous bottle pressure would certainly be high enough to work effectively.
I wonder if the turbo could be forced to spool up fast enough to go into surge?
The nitrous purging solenoid could be routed to supply the gas for the pinwheeling system.
The launch procedure would be a little tough. Starving the main nitrous system momentarily to supply the pinwheel system, then switching back to supplying the main system after a certain boost level. Or, not even having to switch back to the main system because enough boost has been built up. Hmmm. You got me thinking.
I would figure there must be an optimum size hole and amount of holes for the pinwheel idea. Multiple holes evenly spaced around the compressor wheel at the optimum angle and at the right size to generate the highest possible force on the compressor wheel. Nitrous bottle pressure would certainly be high enough to work effectively.
I wonder if the turbo could be forced to spool up fast enough to go into surge?
The nitrous purging solenoid could be routed to supply the gas for the pinwheeling system.
The launch procedure would be a little tough. Starving the main nitrous system momentarily to supply the pinwheel system, then switching back to supplying the main system after a certain boost level. Or, not even having to switch back to the main system because enough boost has been built up. Hmmm. You got me thinking.
There's the delicate balancing act.
Do you prevent egr and then have a mixture which is too cold to properly dissociate and lose power that way?
Or, do you allow a certain amount of egr that will promote dissociation of the mixture so that more of the mixture, even though some of it has been displaced by the egr, will burn resulting in more power than would have been accomplished without egr?
Keep in mind too that I'm running 9.27:1 static compression. That is certainly not the optimum CR to use with methanol. Higher static compression is relied on mainly to supply heat from compression to help with the dissociation of the fuel mix. Maybe I traded combustion chamber space (more piston dish volume) for a certain volume of heat source (egr) to replace some heat from compression? And, egr starts to work on the intake charge mixture well before compression starts to take place. More time is made available for dissociation to take place.
Even though my static compression is lower than optimum, the engine doesn't really do too bad. I've been at 11.4:1 with a previous build, and I have to say the difference is not noticeable off idle. Maybe the egr is offsetting the loss in compresssion heating? Lower static compression does promote a higher percentage of residual exhaust gas contamination of the following intake charge. Am I just trading heat source for heat source?
I'm also gaining some overall cylinder space volume with the lower compression to accomodate the egr volume without having to sacrifice as much intake mixture volume.
According to the engine sim, I'm seeing a peak of over 10% egr at around 3000 rpm. That's the peak of the big bump of fueling in the fuel map. Maybe that explains why the nitrous hits real hard at initial hit (help from egr) and then lays down and ramps back in as rpm and map increase. The hit starting at 2440 rpm and passing immediately through a region of high egr percentage.
It's pretty obvious that between 2550 and 3500 rpm, egr is providing an extraordinary amount of dissociation of the methanol to occur. That is a massive amount of extra fuel I had to throw at it to get O2 readings under control. Without the extra fueling, a massive lean spot occurs in that rpm range.
Some may be thinking that the necessary extra fueling is due to fuel that is short circuiting out of the combustion chamber during overlap. A high overlap cam does have points in the rpm band where short circuiting does occur. That is true.
RPM ranges of high short circuiting percentages are also areas of low egr percentages, and vice versa. So if you have an rpm where you have the highest percentage of egr, you will also have the lowest percentage of short circuiting, and vice versa.
Some may be thinking that the necessary extra fueling is due to fuel that is short circuiting out of the combustion chamber during overlap. A high overlap cam does have points in the rpm band where short circuiting does occur. That is true.
RPM ranges of high short circuiting percentages are also areas of low egr percentages, and vice versa. So if you have an rpm where you have the highest percentage of egr, you will also have the lowest percentage of short circuiting, and vice versa.
The fuel map I posted is indicative of the fueling requirement of the engine without the nitrous injection.
What if converter stall speed, off the nitrous, could be moved to 3,000 rpm? The point of highest egr heating help. Previously, the stall speed, off the nitrous, was at 2440 rpm. The point of lowest egr heating help.
What if converter stall speed, off the nitrous, could be moved to 3,000 rpm? The point of highest egr heating help. Previously, the stall speed, off the nitrous, was at 2440 rpm. The point of lowest egr heating help.
Similar threads
