Follow my progress with Wagonstein!

[91CavGT]

I'll be posting in this thread any updates I have for Wagonstein. For those that didn't see my other post, Wagonstein is a 1991 Chevrolet Cavalier station wagon. It has a built 3.1L V6 that's turbocharged which is not good for gas mileage, but the 5 speed transmission that was swapped into it does help. What really helps is the car has a full standalone Haltech computer!! I can program it to run however I want!!!


The 1st tank of gas was a baseline. Just to show how rough of a tune the car had on it.

14.6 mpg! 168 miles and 11.5 gallons of 93 octane gas


I just filled up tonight at the same gas station and the same gas pump. However, I have done some tuning to the car since the last tank of gas. I DID NOT touch the air/fuel mixture controls at all. I did however get the car to idle at 800 rpm instead of hunting for a 1200 rpm idle, I turned on closed loop operation (I can't believe it was turned off!!!), I lowered the coolant temp where the computer switches over from open loop to closed loop operation (it's at 140 degrees now, it was at 160 degrees), and I decreased the accelerator pump output at 0 rpm, 1500 rpm, and 3000 rpm.

20.2 mpg! 227 miles and 11.2 gallons of 93 octane gas!!


This is a huge improvement for not even touching the air/fuel maps!! Today I started looking at the ignition timing maps. While not under boost, the ignition timing is VERY conservative (no more than 30 degrees of advance while under light throttle and cruising) even though this motor has a compression ratio of 8.8:1 and I'm running 93 octane fuel!! Just in the last few hours I've advanced the timing as much as 10 degrees in some points of the power band. I'm going to keep advancing the timing till it either misses or I hear audible knock, at which point I'll back the timing off 3-5 degrees. Now, I'm not adjusting anything at all while the car is on boost, I'm only adjusting the settings that apply to the motor while it is under vacuum.


My goal is 25 mpg in town and 35 mpg on the highway, the ability to run a 13 second 1/4 mile with just changing tires, and be able to haul my wife and 2 kids somewhat comfortably along with cargo where ever we want to go.


I'll post results from my next tank of gas here.

[101mpg]

Good goals trying for a mixture of FE and usefulness. Keep us posted!

[91CavGT]

I filled up with gas tonight. Here are the results.........


250 miles driven
11.767 gallons of fuel used
21.25 mpg!


The mileage is still going in the right direction! This was once again with a few full throttle blasts for one reason or another. The only change I made from the last tank of gas to this one is advance the ignition timing under cruise and adjust the idle a bit better. Right now the car is running 37-42 degrees of advance at low load cruising speeds and 13-18 degrees of timing at idle.


I want to get a weaker wastegate actuator from a friend of mine so I can tune a "fuel economy" program. This will limit boost to 2-3 psi with a very slow buildup of boost. It will also be tuned to run on 87 octane. This combined should give me a bit more power off boost while still giving more power than stock while under boost. I'll keep the current program I'm working on as the "fun 93 octane" program. I need to wire up the electronic boost controller though to do this effectively. I also need to get a pair of 185/70/14 tires to put on the stock rims. Right now the car has 225/50/15 tires up front and 205/55/15 tires on the back. This also changes my gear ratio so it's running a couple hundred rpm higher on the highway.

So I'm thinking that the tire swap and the 87 octane tune should give me a 2-3 mpg increase over what I'm seeing now. Time will tell if this in fact is true.

[91CavGT]

Yet another update!


On this tank of gas the only changes to the tune were minor since it is starting to get close. After seeing at idle that the O2 correction was pegged at 20% decrease in fuel I leaned out idle air/fuel mixtures just a bit. There's still room for improvement though since the O2 correction now goes from 10% to 14% reduction in fuel. The biggest help I think though is that I lowered the engine off during deceleration from 2,000 rpm down to 1400 rpm. I can lower it more, but when the motor is cold, it tends to REALLY make the idle hunt since it revs up, but with zero throttle the computer kills the motor till rpms drop below 1400.


So, without further ado..........


275 miles on this tank of gas
12.138 gallons used
22.65 mpg!!


I'm now less then 2.5 mpg away from my goal of 25 mpg in town, yet the car still has not had a tune on the air/fuel maps!! Not only that, but the ar has wide 225/50/15 front tires and 205/55/15 rear tires. The stock tires are 185/70/14s. I'll be putting these back on the car later down the road. Traction will suffer, but overall gas mileage will increase and rpms on the highway will also decrease.

[theholycow]

Can't you have separate programs for cold vs. warm? In the cold program, DFCO could be 2000 rpm, and in the warm program it could be 1000.

For the tires, I don't share your prediction. According to http://www.miata.net/garage/tirecalc.html you're looking at these tiny differences:
.3 inch diameter
1 inch circumference
11 tire revolutions per mile
1.4% speedometer error, for .9mph difference at 60mph.

I think such a small gearing change won't affect FE much. The difference will be between aero improvement and RR consequence. Unless they are constructed significantly differently, the narrower tire at the same pressure should mean increased rolling resistance. I think at city speeds the RR will be more important than aero. If the narrower tire accepts much higher pressure then it will probably be better.

[91CavGT]

Originally Posted by theholycow

Can't you have separate programs for cold vs. warm? In the cold program, DFCO could be 2000 rpm, and in the warm program it could be 1000.

For the tires, I don't share your prediction. According to http://www.miata.net/garage/tirecalc.html you're looking at these tiny differences:
.3 inch diameter
1 inch circumference
11 tire revolutions per mile
1.4% speedometer error, for .9mph difference at 60mph.

I think such a small gearing change won't affect FE much. The difference will be between aero improvement and RR consequence. Unless they are constructed significantly differently, the narrower tire at the same pressure should mean increased rolling resistance. I think at city speeds the RR will be more important than aero. If the narrower tire accepts much higher pressure then it will probably be better.

Unfourtunately, there is one setting for DFCO. There is one primary ignition and one primary fuel map. From there, everything else is just a matter of another map to tell the computer how much ignition or fuel to give under differant circumstances. There is one setting for DFCO, one setting for accelerator pump at differant rpms(but not at differant temps), one coolant/fuel correction, one coolant/ignition correction, one IAT/fuel correction, and one IAT/ignition correction. There are a few other minor adjustments that can be made, but these are the major ones.


Yeah, I went to that exact same calculator after I made the last post. I thought it was a bigger differance but I guess not. More than likely, the tires on the car now are shorter than a typical 225/50/15 tire is and that is where the differance is coming from, but it will only be 100 - 200 rpm drop at 65 mph is all.


It was my understanding that a narrower tire would provide less RR than a wider tire because there is a smaller contat patch. I'm currently running 35 psi in the 225/50/15 tires. I'll have to look at the 185/70/14 tires that I have to see what their maximum psi rating is.

[theholycow]

Nope, I used to think that too, but having achy tired legs while bicycling made me open my mind and really read up on it. The contact patch is the same size no matter the width of the tire, as long as the pressure is the same. If you have 50 psi and 500 pounds on it, you'll have a 10 square inch contact patch whether the tire is an inch wide or ten inches wide (except, with such an extreme example, there will probably be measurable interference from sidewall stiffness too).

The contact patch changes shape, however; with a narrower tire it's longer. So, to have the longer contact patch, more of the sidewall has to deform. That's where rolling resistance comes from (there's also some RR from tread block deformation too).

Bicycling enthusiasts are very superstitious and very marketing-driven, to the point that misinformation infects even dependable data sources. As such, even the ones who aren't superstitious or marketing-driven believe the same stuff. Additionally, the ones who really dig in and learn the stuff still find that (in a bicycling context) the aerodynamic advantages outweigh the RR -- and also, on a bicycle the narrower tire is usually rated for higher pressure, so it ends up with equal or better RR anyway. On a car, that's not usually the case.

[91CavGT]

Wow, that is actually pretty logical when you really stop and think about it. Thanks!!

So, in short, if I can get some taller and narrower tires that will lower my highway rpms dramatically (it's at 2400 rpm at 65 mph now) AND I can run higher tire pressures then it would be beneficial to FE. Otherwise I won't see much if any increase.

[theholycow]

That's probably right. I don't know exactly how the RR drag plays off against gearing and aero drag issues, and I don't know exactly how much additional pressure would be required to make a narrower tire equal a wider tire's RR.

Since you already have all those tires mounted on wheels, I'd love if you could do some coastdown testing and put some numbers behind my understanding of the theories. With wheel fairings it would be possible to almost entirely eliminate any aero drag error from the RR measurement. However, that sounds like an awful lot of work and especially difficult to do scientifically and accurately, so umm.......yeah.

[theholycow]

Actually, just after posting that I realized it's probably easy to calculate the amount of pressure required for the same RR. All you have to do is figure the contact patch size and shape and have equal lengths.

Also, in your situation, you have different wheel sizes; I don't know how sidewall height affects RR, but I'd guess that lower sidewall height increases RR because the sidewall has to deform more sharply and closer to the rim. Argh! There's too many variables to do the experiment. We'd have to be working with the same brand/model tires on the same size rims at the same sidewall height with the same tread pattern and the same load and the same temperature and the same pressure and.......eek....there's no realistic way to get accurate results from this without a big budget and a lab.