Science Buddies: "Ask an Expert"

Does this project idea have too many uncontrolled variables (like the condition of the engine, road conditions, amt. of stepping on the gas pedal, etc.)?

The question that I will be (hopefully) testing is:
Will using a lower-than -recommended grade of gas decrease fuel economy in a high performance engine?

The way that I was thinking about testing this is fully filling up the car with a high-performance engine with regular-grade fuel, then writing down the odometer mileage. After this, my parent would drive the car 25 (or 50) miles round-trip back to the pump, then refill the engine completely again, and write down both the # of gallons that it took to refill the tank, and the odometer mileage.

Then, I would calculate MPG from both the approx. 25 or 50 miles that we were supposed to drive, and the actual mileage that we drove.

I was originally going to be testing whether using higher-grade gasoline increased fuel economy (no matter what type of engine), but after doing research, I realized that cars had either high or low performance engines. Thus, I had to change my Question.

I want to know whether or not this project would have too many uncontrolled variables to have reliable data in the eyes of the judges, and if so, what kind of things that I might be concerned about.

My self-imposed deadline for coming up with a new project idea is today, (Jan 3), so it would be nice to have an answer, but I do know that I should have asked this sooner; I can push the deadline back a bit.

I want to finish in 1st place, and be nominated for Broadcom MASTERS (by finishing in the top 10% of the fair). I am willing to do what it takes to achieve these goals.

Hello and welcome to the forums:

In a word, yes, there are too many variables. The obvious one is that you probably have a mixture of octane grades in the gas tank. To get only a single grade in the tank would require completely emptying the tank and filling with the chosen grade of fuel. You would need to run this all the way to empty and then start with a different grade. You would want to do this several times to get enough trails to get around the other variables.

I've seen this experiment before done with small engines, such as those on lawnmowers. You can add a small amount of fuel to the tank and let it run dry and record the time taken for this to happen. This way you can run multiple trials, too. Also, if you can run this on multiple engines you will get better data.

This might be a better experiment if you try different additives instead of just octane grades. You could de-bunk the claims of fuel economy booster additives? Also, I'd be curious to know if the "preservatives" they sell the keep gasoline from going "bad" actually impact the fuel economy?

Let us know what you decide!
Keith

Hi Keith, thanks for the quick response.

That's a good idea with letting the tank run dry, but how would I know when to consider the tank "dry"? Would that be when the tank hit "E"?

kgudger wrote: You would want to do this several times to get enough trials to get around the other variables.

What kind of other variables were you thinking of? I forgot to mention about the mixing of the fuels (one of the reasons why I wasn't sure if I should do this), which I would make up for by doing as you suggested.

I like the additives idea - in between additives, I would need to drain all fuel with additives, then refuel and drain, then refuel and put in the additives, correct?

Thanks so much for your help! If you have any other ideas for a project regarding octane or fuel economy, please let me know; that way, I can keep my current background research.

Hi:

If you use a small engine for your tests, they have a very small fuel tank. You might put in, say, 50 mL of fuel, start the engine and a stop watch, and let it run until it quits. (That's what I mean by running dry.) This procedure means you don't have to drain anything - all of the fuel gets burned up. Each trial, the engine runs under the same conditions (idling until out of fuel.) You would want to run at least one trial before you measured trials so that the engine is completely warmed up (eliminating that variable.)

Keith

If i did the additives idea, then could I just use the car as we would use it in real life (different routes), or would I need to use the same route?

My mom (aka teacher), says that it doesn't matter, because you would use the car in real life. I think that you still need to have as controlled an environment as possible.

Also, would it be even reasonable to try to still test this with a normal car engine, given all the variables that using a real, used car engine would have?