Long term polyisobutylene test w/GTAT Additive (Viscon) about to start
 

I just ordered a bottle of GTAT fuel additive that I intend to put through the ringer over the next 4-6 months. The goal is to find out if it has any measureable impact on fuel economy. I want to do as much as I can to avoid confounding the results so I will be doing 3-4 cycles of 2-3 tanks on/2-3 tanks off the additive. That would give me as many as 24 data points to analyze at the end. It should also help mitigate the effects of other variables such as fuel makeup, driving habits, weather, etc.

One idea that I am playing with is to reset the ECU as soon as possible each time I switch on or off of the additive. There's a chance that the ECU will make adaptations if the fuel behaves differently with the additive so I figure this is the best way to start from the same place each time.

The vehicle is a 2002 Honda Accord EX Coupe. 2.3L I-4 & 5-speed manual. I have a fuel log at GH where I will post results but will also update this thread as I make progress.

Any input would be welcome if you guys have suggestions for improving the integrity of the experiment.

Good plan, and thanks for
 

Good plan, and thanks for making this thread and doing this experiment. I would record your distance and fuel used for each trip, as well as ambient temperature if possible.

If you feel so inclined, I'd recommend you to also post your gas usage in the Gaslog here. Oh, and make an entry for your car as well in the Garage

I'd be interested in your
 

I'd be interested in your results. I used to work for a company that uses Viscon and I did some very informal experiments with it and I never saw the benefits that were touted, but as I said, it was highly informal and unscientific.

Keep us posted!

good luck, brick. (and
 

good luck, brick. (and welcome to the site!)

difficulty of testing is just one reason i've avoided additives. it's challenging to do quick successive tests (controlling for ambient conditions), and even harder to do A-B-A testing once you've added something to the fuel (until you use it all up). it's tough to conclusively figure out if the item works or not.

we need something like a donor test car with its fuel system modified to be able to run small consecutive batches of precisely measured fuel (like the recent & inspired lawnmower fuel brand experiment thread), except with the benefits of testing in an actual computer-controlled car engine (preferably an OBD2-monitored one!).

I agree...
 

It's very true that additives are tough to test. The difficulty in going on-off-on is part of the reason that it's going to take me so long to collect the data. Doing just one tank on and one tank off can never guarantee that the you can dilute enough not to skew the results. 3 on and 3 off is still pushing it but that's pushing the envelope of how long I can go before weather conditions and other factors start to get too different from one trial to the next. Even if I do this right it would take a very impressive effect to wash out of the noise. (Continuous change I can handle, but discontinuity poses analytical issues.) I guess I'll let you know around, say, August. (ha)

I have been playing with the idea of building a very simple device to test things like additives, water injection, and whatever else you can come up with. The picture in my mind is basically a vertical steel cylinder with a heavy piston that is pushed upward by a single combustion event. If implemented properly this would allow an experimenter to measure work output simply by how high the piston was pushed upward against gravity. Metering air, fuel, and additives would be pretty simple, too. Can that be made safe? I think so but it would take some care. I think it would be worth my time to develop it because its simplicity takes out many of the thermal and mechanical variables present when running a small motor, even in a lab situation. I worked in engine labs as an undergrad and things can get complicated even with lots of funding and people much smarter than myself running things.

i say go for it. we can
 

i say go for it. we can always hope for a clear outcome!

for me, another specific obstacle to this kind of testing is how little i drive: unless i do a road trip, i'm lucky to fill my (40 L) tank once a month. it would take me until well past august to run your proposed test!

i've said to krousdb, sometimes i wish i had to drive more, if only to practice my ever-improving techniques behind the wheel. i don't get enough saddle time.

that's what makes the scangauge invaluable (as you obviously already know). instant data allows for more test control.

Got it! And a slight delay...
 

A quart of additive showed up on my doorstep yeserday! The only hitch is that I have decided to delay testing just a bit. I have just started to improve my FE driving technique and krousdb and xcel (Wayne) have been pretty concerned that this will throw too much noise in. I don't think it will be a problem as long as my improvement is gradual and progressive, because the on/off/on additive use would provide plenty of control data to sort out that kind of trend. But if improvement due to technique is discontinuous I will have problems. In the interest of integrity I am going to push my FE hard without the additive for my next few tanks to see what my progress looks like. If the line is relatively smooth I will throw in the additive, if there are spikes I will hold off until my numbers level out.

So were there ever any results on this test? I'm interested, and if the claims are correct, I'd like to figure out the mix of how much to put in per tank.

out of date...

If the stuff is any good you should start using it so you can start SAVING GAS! Again you should see how high an MPG tank full you can get even if it is under not so ideal conditions. And how much does it cost per gallon of gas mixed so we know what the savings is going to be when you start seeing results.