As some of you know I built my entire engine around fuel efficiency and performance. When I started this project over ten years ago I made a list of major factors that would aid in the engines efficiency.
Here's my list in the order of importance.
1) Low numerical "Throttled BSFC" numbers.
2) Lean Burn Technology.
3) Fast-Burn Head design.
4) Reduced Pumping Loss Engine design.
5) High Fuel Atomization.
6) Programmable EGR system.
7) Advanced PCV system.
My list of modifications to reach these goals.
1) Low numerical "Throttled BSFC" numbers.
My goal was to be able to reach a BSFC number of
.25 or better (.25 BSFC is based in lbs) at part throttle. At one point I had the idea to decrease the engine size to under a liter and use a smaller engine and run at a higher load with a better BSFC number to help with FE. But then after talking to Ben Strader of EFI University and a few others, I decided against it.
I understand that car manufacturers base their engines BSFC at high load or part load. This will give them a better BSFC number, but it doesn't tell the whole story as far as the consumer driving down the freeway at light load.
IMO very light load freeway driving is by far the most used by the consumer.
One other thing I would like to address on this subject is I see a lot on this forum and other forums that ones thoughts are you will achieve your optimum BSFC at heavy load. So we feel like we need to run the engine at a higher load to get a better BSFC number.
I agree to some extent , but keep in mind all EFI systems have accelerator pump values that are read off the TPS. The accelerator part of the EFI program does what carburetors did with their accelerator pumps in their day. It adds fuel based on the speed of the throttle plate being opened to prevent a lean back fire. Usually the EFI throttle pump system will have three RPM opening points with a decay setting on all three. It will also have a coolant base adjustment built into it. All three will add to fuel injector opening time base on %.
So back to my point if you are trying to to get to the ultimate BSFC number with high load your also dumping more fuel in through the throttle pump part of the program. Something that is not included in the posted BSFC charts of different engines.
Now lets look at the BSFC WOT and or high load part of an engine. All engines have to rely on the fuel for thermal management. Incoming fuel is one of the largest factors for cooling down the combustion temperature that takes place durning the intake cycle. So at high load or WOT the engine is wasting a lot of fuel just for cooling purposes only. This kills BSFC numbers. Plus at high load you burn more fuel due to a higher VE. So more fuel results in more heat, more heat means more fuel that's needed to cool combustion temps.
One other factor is at a high load the air velocity is much slower. Slower velocity means less atomization of the fuel. So now more fuel is needed to make up for poor atomization due to lean spots and overly rich spots in the combustion chamber.
So as you can see I was left with a problem when I wanted to introduce my next modification "Lean Burn" There was no way i could run 20:1 + A/F at high load and keep the engine from detonation. So I had no choice then to concentrate on light load great BSFC design.
So that brings us to...
2) Lean Burn Technology.
Lean Burn Technology goes hand in hand with great BSFC numbers. My thoughts were to use as little fuel possible to produce 20whp. To do this I would need a BSFC number around .25 so lean burn was my only option.
I came up with a prechamber design that would be able to ignite and start combustion with A/F ratios in the stock combustion area around 25:1 to 30:1.
At these A/F ratios the heat is reduced (good) and flame speed slows down drastically (bad). Plus NOx levels increase(bad). One thing I found out is at very lean conditions atomization improves a ton.
But back to heat. I noticed in my data logs that my water temps drop when it goes into lean burn. This was a relief because the higher A/F ratio requires a slight higher load. This is also where the turbo comes into play. With a higher load with an turbo engine makes more boost and reduces pumping losses.(good)
But now the bad, flame speed decrease and now I have to start the ignition timing sooner.
So that brings us to...
3) Fast-Burn Head design.
My Honda multi-valve pent-roof design has a fairly fast burn rate when it comes to head design. This helps with increasing flame speed so less ignition timing is required.
So that brings us to...
4) Reduced Pumping Loss Engine design.
My main contributor to helping with pumping loss is the turbo.
Plain and simple it creates a higher then ambient pressure so as the piston cycles downward its being aided by pressure acting on it durning the intake stroke. But what about the compression stroke? High compression engines will take more energy then lower compression engines when it comes to the compression stroke.
My engine is around 7.5:1. You can turn it over to where it feels like there isn't any spark plugs in it. So at very light load it would turn very easily.
One other thing I think that is happening is there are some left over exhaust gases that will aid in NOx from a very low compression engine.
So that brings us to...
5) High Fuel Atomization.
All I have done here is I'm running a higher then normal fuel rail pressure 60psi. to help aid fuel atomization.
6) Programmable EGR system.
7) Advanced PCV system.
These two items are what I'm working on now.
The programmable EGR system will be where I can control how much I want to recirculate.
The advance PCV system will have an electric evac system to reduce crankcase pressure down to almost a vacuum.
This will help with sealing the lose ring pack and reduce pressure in the crank case that will help when the pistons are cycling downward.
Well this is my own home brew engine design and so far it has been successful. I have gone against the grain on a few other ideas. But sometimes thinking outside the box is necessary no matter what the out come good or bad.
Today
Other popular topics in this forum...
So I choose a design that anyone could drive and get the same results(normal driving). It takes more of the driver out of the equation.
