What you need to know about Brake Specific Fuel Consumption

[serialk11r]

Sorry, I meant at idle. I typed it, and then erased it for whatever reason. Variable duration eliminates throttling loss.

[mort]

Originally Posted by drmiller100

Right now, I don't think ANYONE has a good way to actually calculate the losses due to throttling.

Once you can calculate it, then you should be able to explain it.

Hi drmiller100,
Interesting. I use this fact: the amount of power to pump past a restriction is the mass flow times the force. To make an estimate, I use the intake manifold pressure minus 1 atm, or measure the manifold vacuum and convert to psi as the force. I figure the mass flow by taking the fuel flow and multiply by the A/F. This is my back of the envelope formula, you can just about do it in your head. I've been accused of missing compressibility. I'll discuss that at the end.

You can compute the exact power to pump across a restriction, given the properties and operation of the engine in question.

HP = mass flow (corrected for compressibility) times force (corrected for compressibility) times conversion factors.

This can be computed from the displacement of the engine in cubic inches, the engine speed, and the absolute manifold pressure in psi, you get this:
HP = D * RPM * Pm * (k / k - 1) / (12 * 2) * ((Pa / Pm) ^ (k - 1 / k) - 1) * (1 / 33000)


Pa is 1 atm, 14.7 psi.
k is 1.41 for air (adiabatic index or ratio of specific heats)

That is (cu in) * rpm * (lb /sq in) * (compressibility correction) times (1 / inches per foot * revs per displacement) * pressure ratio ^ (compressibility correction) * units conversion
(ft lb / min) * (hp / (ft lb / min))

I have a Volvo 240 with the b23 engine. Disp = 140 cu in. At 750 rpm idle the manifold vacuum is 20 in hg = 9.8 psi

HP = (140 * 750 * 4.9 * 3.44 / 24) * ((14.7 / 4.9) ^ .29) - 1)) / 33000

Throttling HP = 0.838

Now you can calculate pumping power across the throttle.

You'll also notice this is almost the same formula given by the air compressor page except that that page is for compressing 1 atm to higher pressure, and my formula sucks 1 atm in to a lower pressure.

In my back of the envelope formula I estimate mass from the A/F and fuel flow which automatically compensates for compressibility because it is the actual mass of air. Measured psi is the actual force. Notice in the exact formula that correcting the mass term for compressibility involves multiplying by (k / (k - 1)) and correcting the pressure ratio you raise the ratio to the power of (k - 1) / k)) and subtract 1. When the pressure ratio is 2.406 these factors cancel. That happens at manifold vacuum pressure of 6.1 psi. That's not far from the range of manifold vacuum for part throttle operation.

Note "Figure C-3 ...For flow rates less than about 60 percent of the choked flow, the effects of compressibility on the mass flow rate are less than 5 percent." Internal Combustion Engine Fundamentals - John Heywood


Heywood also has this graph, for his test engine: pumping is less than friction.


In that graph the 200 kPa bmep is about 11 HP, the pumping power is about 3 hp and the rubbing friction is over 5 HP


-mort

[mort]

Originally Posted by Old Mechanic

Bottom line for me is the Mercedes SB that claimed 7%.
Mech

Hi Old Mechanic,
I'm sure they would have bragged a bigger number if they could have. 7% seems pretty good improvement for removing a throttle from a diesel. 21:1 compression ratio!

-mort

[Olympiadis]

Originally Posted by E4ODnut

Having said that, and emissions concerns aside, I see no reason why conventional gasoline engines can't get improved BSFC at lower engine loads by leaning the mix, adjusting spark timing to suit this leaner mix, and opening the throttle somewhat to maintain the required power. The larger throttle opening will reduce pumping losses or what ever you want to call it, as well as improve the thermal loss. The exact cause of the improvement doesn't really matter, what matters is that the end result should be better low load BSFC.

Correct. Being a tuner, I have done this many times over the years.
You must use a wideband O2 as a reference. The narrow band O2 is unsuitable for anything but very near 14.6:1.
The key to any tune is the degree of control the tuner has over the ECM, - calibration and algorithm.
I have yet to use a M.S. system, but I would guess that you start out with a good advantage in the control department.
If you are able to re-define the load-cell matrix for fuel trims, then you can increase the tuning resolution in your sweet-spot area of your main fueling table. This will help greatly with running lean but quickly adding fuel to cover additions to load while driving normally.
You will find that when cruising near the lean-limit, that even a small amount of added load can cause surging and/or spark-knock (not the same as detonation).
Most cruising occurs in a fairly narrow band in your fueling table. If you can increase the resolution of the table here (increase the number of cells), then it will make your tuning work easier.

Many of the wideband O2 sensors, like ones from Innovate, can simulate a narrow-band output signal in a fashion that is skewed to the tuners desire. This allows your ECM to perform the adaptive fueling algorithm without major issues.
Since I don't work with M.S. I am unaware what the system can handle in that area.

Those tuners who insist on adding the maximum amount of spark advance while running lean both make their job of tuning for smooth operation harder, and end up negating much of the fuel mileage gained from running at the leaner AFR.

If you watch your wideband closely and have the ability to change spark advance values in real-time, then you will see that advancing the spark has the effect of leaning the AFR as read on your gauge. Most people assume a leaner burn in the chamber when in reality the fuel delivery has not changed.
The idea is to work towards the lowest injector pulse-width that you can get away with while maintaining a given speed.
Don't go into it trying to work towards the most spark advance or the leanest reading as seen on your wideband gauge.

[drmiller100]

question:

>I have a Volvo 240 with the b23 engine. Disp = 140 cu in. At 750 rpm idle the manifold >vacuum is 20 in hg = 9.8 psi


when you say 20 in hg, do you mean "20 inches of vacuum in reference to ambient?"

If so, wouldn't the final pressure be more like 5 psi absolute, therebye doubling the horsepower number?????

1 horsepower is something like .1 gallons per hour, which is something like 10 percent fuel economy. I think.

[drmiller100]

you can run hondas in leanburn to where the wideband 02 sensors won't read the mixture - not wide enough.

With MS you can map anything you want. and if that is not enough, you have access to the source code for the chips. And if that is not enough, you have complete schematics, chip sets, and lots and lots of help rebuilding the hardware to become whatever you desire.

MS2 has even more options, and MS3 even more.

[serialk11r]

Originally Posted by drmiller100

question:

>I have a Volvo 240 with the b23 engine. Disp = 140 cu in. At 750 rpm idle the manifold >vacuum is 20 in hg = 9.8 psi


when you say 20 in hg, do you mean "20 inches of vacuum in reference to ambient?"

If so, wouldn't the final pressure be more like 5 psi absolute, therebye doubling the horsepower number?????

1 horsepower is something like .1 gallons per hour, which is something like 10 percent fuel economy. I think.

Yea I think 20 inch vacuum is typical, that's what I found when I tried to look it up. And engines of that size usually idle at .3gph if I'm not mistaken.

[E4ODnut]

Olympiadis,
Now we are talking.

I have some experience tuning, primarily Megasquirt but I did use the Haltech E6K when I first started this hobby, some would say an obsession. No matter, all programmable EFI systems are similar in basics. I use speed density, batch fired. This is because I am limited in my tuning resources. I don't have access to a dyno, so all my tuning is on the road, or on the water as the case may be. I use WB EGO feed back and EGT for tuning reference only. I only use EGT as a "sanity check". If I was serious about it then I would need thermocouples for each cylinder, which is not practical for me. I'm also of the belief that EGT tuning has it's place, but because it is much more application specific, I prefer to reference EGO. I have feature that can can calculate fuel burn in litres per hour so I can monitor fuel consumption at any given load.

I tune mainly for economy and reliability. Power output is important, but definitely not a priority. I also run open loop only. Some people may prefer to use closed loop but I agree with the camp that says you give up some control with closed loop, and it is often used as an excuse for sloppy tuning. I don't use EGR.

In the days before EFI and the feed back it can provide, a lot of dedicated tuners made relatively primitive technology perform very well across the complete power band. If I can improve even a small amount on this with the late '80s/early '90s technology I'm using I'll be quite satisfied.

Tuning for best power is relatively easy, and I stress relatively. Best economy is a lot harder. I agree with your theory that the best way should be to aim for minimum pulse width (leanest mix) then adjust spark to suit. This isn't as easy as it would appear. It's especially difficult to do this on the road. I'm finding a bit easier to do it on the boat, almost like a dyno because it's easier to hold a given load. In both cases, using a conservative spark map, I try to hold a given cruise load and lean the mix until I get a "lean surge", then fatten it up a bit until the surge is gone. I note the EGO feed back. Then I try to hold the same load and advance or retard the timing to see if I gain or lose power and note the EGO feed back. I'm finding that I can vary the timing a lot before performance or EGO changes much. I don't reference throttle position at all but prefer to use Manifold Absolute Pressure as an indication of engine load. I tune at or near sea level only but I do have barometric pressure compensation if necessary.

So far I've found that at loads under ~ 75 KPA MAP, EGTs are not an issue and it is almost impossible to induce detonation with any sane ignition timing. I tune for target AF ratios of between 16:1 and 17:1 below ~ 80 KPA. I reason that with pressures above that I am in need of some serious power and ramp up to ~ 13:1 at ~ 90 KPA and ~ 12:1 at 100 KPA. Some of my engines like it a bit richer, some a bit leaner, but not a whole bunch.

As for WOT spark tuning I just revert to the old stand by of "advance until it pings, then back off a couple of degrees". I may be leaving some power, or economy on the table. but without I dyno I don't know. At least I'm comfortable that I won't hurt the engine this way.

I have knock detectors on the engines, but these have to be calibrated by ear, so they just serve as one more open loop feed back. All of my sensors can be data logged, which is the most valuable tool in my arsenal.

The drill is, come up with some realistic base settings, go for a drive and data log. Go home analyze the logs, make changes, go for another drive. Repeat. It never ends, but that's the fun part.

[mort]

Originally Posted by drmiller100

question:

>I have a Volvo 240 with the b23 engine. Disp = 140 cu in. At 750 rpm idle the manifold >vacuum is 20 in hg = 9.8 psi


when you say 20 in hg, do you mean "20 inches of vacuum in reference to ambient?"

If so, wouldn't the final pressure be more like 5 psi absolute, therebye doubling the horsepower number?????

1 horsepower is something like .1 gallons per hour, which is something like 10 percent fuel economy. I think.

Hi drmiller100,
Yes, well spotted.
.84 HP

-mort

[drmiller100]

Originally Posted by serialk11r

Yea I think 20 inch vacuum is typical, that's what I found when I tried to look it up. And engines of that size usually idle at .3gph if I'm not mistaken.

so.........

if we could figure out how to run a gasoline engine without a throttle plate, our CITY mileage would improve 20 to 30 percent.

probably not as much gain on the highway.