Speed/mpg testing

[Formula413]

I did some speed/mpg testing on my car recently. '97 Ford Escort, 2 liter 4 cylinder, 4 speed auto. Just a few minor aero mods. The speeds I tested were 40, 45, 55, 60 and 65 mph. I couldn't find a good road to do 50 mph so I left it out. I didn't bother to do any slower speeds because the torque converter won't stay locked below about 37 mph. I recorded the mileage with the SGII. To get the data I did three runs back and forth, getting the car up to speed, setting the cruise, and re-setting the trip meter. Roads were as level as possible, but the two way runs should eliminate any error from grade. Engine was at operating temp before making any runs, and electrical load was minimized (no lights/radio/hvac). Ambient temperature was 17*F. I'm sure these numbers would be significantly higher in warmer weather but that should not affect the slope much.

40 mpg - 40 mpg
45 mph - 38.2 mpg
55 mph - 36 mpg
60 mph - 33.4 mpg
65 mph - 31 mpg

Anyone have some good graphing software?

The data from 55 to 65 is pretty much in line with what I have seen on graphs for other vehicles, but it still drives home the point of the huge FE hit taken at speeds above 55 mph. I was a little surprised to see the significant decline from 40 to 55 though. Am I missing anything? Questions/comments?

[groar]

Originally Posted by Formula413

Anyone have some good graphing software?

Just did this with OpenOffice. Not really pretty as I don't know how to set scales :



This is inline with the straight line the mpg vs. mph data are drawing.

Denis.

[cmittle]

Out of curiosity was it in 4th gear for all of the tests?

[Frank Lee]

It's very linear and I would have expected it to be less so- more fe drop as speeds go up.

[Formula413]

Thanks Groar. Yes this was all in fourth gear, my car will only lock the torque converter in fourth gear. As far the slope of the graph, it does start to decline more sharply after 55, but I was expecting it to be more level from 40 to 55. This is a generic graph from the fueleconomy.gov website:

[TestDrive]

Originally Posted by Frank Lee

It's very linear and I would have expected it to be less so- more fe drop as speeds go up.

Frank, Based on his statement in the first post - "I couldn't find a good road to do 50 mph so I left it out." - I'd guess the remaining tests were conducted on at least two different roads, perhaps at different times/conditions. eg a few mph difference in wind speed?

Originally Posted by Formula413

This is a generic graph from the fueleconomy.gov website:

The generic "graph" is quite likely an artist's illustration intended to emphasize a point as opposed to an actual graphing of data gathered from testing.

[Formula413]

Originally Posted by TestDrive

Frank, Based on his statement in the first post - "I couldn't find a good road to do 50 mph so I left it out." - I'd guess the remaining tests were conducted on at least two different roads, perhaps at different times/conditions. eg a few mph difference in wind speed?

55-65 was done on the highway, 40-45 was on a rural road. It was not a particularly windy day, but as I said I made runs in both directions, so wind should have been pretty much eliminated as a factor.

[TestDrive]

Originally Posted by Formula413

55-65 was done on the highway, 40-45 was on a rural road. It was not a particularly windy day, but as I said I made runs in both directions, so wind should have been pretty much eliminated as a factor.

It would minimize the factor but not eliminate it.
The wind gains won't exactly match the wind losses.
The same is presumably true for the variation in slope of the roads.

If you were able to do 50 mph runs on both roads, you could correlate the two data sets, but as is an addition margin of error has been introduced. Perhaps +-0.1 mpg???

[Formula413]

I would guess that cars passing me on the highway probably introduced more inaccuracy than anything. It's not easy to collect good data in the real world, not to mention I just can't seem to find a nice quiet stretch of road around here. Maybe I just need to do testing at 3 am. LOL

[TestDrive]

Wasn't trying to find fault with you data gathering, just suggesting a revised interpretation of the data gathered thus far.

One of the problems with 3:00 AM testing is the deer and other wild life that treat the highway as their playground about then.