Finding the Brumby’s Drag and Rolling Resistance

I’ve previously written on how to find the Rolling Resistance and Aerodynamic Drag for your vehicle.  This video shows me doing the tests on our electric Subaru Brumby, then filling in the spreadsheet.  You can find the instructions on how to do this yourself here.

Before I selected an electric motor to install in the Brumby, I put some numbers into a spreadsheet to calculate what power I’d need.  The spreadsheet I used seems to have disappeared from the internet now, but here it is for posterity: EVDrivePower.xls I don’t know where this came from, or how good it is, so use it with the usual grains of salt.

Two of the numbers required by that spreadsheet are the Rolling Resistance and the Aerodynamic Drag.  These were not readily available for the Brumby, but I found instructions on how to find these two values experimentally for your own vehicle here.  I’ve now updated the instructions and re-written them myself here.

Lessons Learned

Straight, flat, seldom-used roads are rare where I live!  There are plenty of roads with little traffic, but these are generally the older, smaller roads, which follow contours and farm boundaries.  Looking on Google Maps I was able to find a few likely candidates, I was looking for a stretch of about 1 km without bends.  Although these were straight, they were generally not flat.  The one I finally found was a newer section of an old road, where it had been re-done to install a bridge for a new highway to go over it at right angles.  I was able to view this on Google Street View, which gave me an idea that it was quite flat.  When I got there I found that there is a slight slope from one end to another, and a few dips along the way.
Another option would be to use one of the highways, but do the testing at night or early in the morning when there might not be much traffic around.  Slowing down from 80 to 0 km/h several times is not very good manners when there are other road-users about.

Results

I decided to test the Brumby in three states.  Tarp on, tarp off and towing a trailer.  When towing the trailer I had the tarp on.

With the tarp on, the vehicle has a more streamlined shape.  My assumption was that this would be the most aerodynamic shape and would have the least drag.  Not that it would be great, the vehicle has a roo bar on the front and is very square at the back.
Without the tarp, air can flow into the tray and will come up against the tailgate.  This seems likely to add quite a bit of drag.
The trailer adds a third set of wheels, and also looks to be a terrible shape aerodynamically.  I’d guess that both the drag and the rolling resistance would be increased with the trailer.  Note that there is a 300 kg weight increase with the trailer, but the equation should allow for this.

The coast-down speed confirmed that having the tarp on the vehicle makes a difference.  Starting at 80 km/h, the coast-down speed at 50 seconds was 28 km/h with the tarp on, 26 km/h with the tarp off, and 21 km/h with the trailer.  As expected, the trailer adds a big penalty for efficiency.

	Coast-down (km/h)	Rolling Resistance	Drag
Tarp On	28	0.017	0.651
Tarp Off	26	0.022	0.472
Trailer	21	0.023	0.751

Going from tarp off to the trailer makes sense, the rolling resistance and drag figures increased.  But going from the tarp on to tarp off resulted in an increased rolling resistance and decreased drag figure.  This doesn’t seem right, and I suspect that my experimental technique might be to blame.

When I performed the tests I originally decided that I should start at the same place each time.  This wasn’t as easy as I anticipated, as my road wasn’t long enough for a coast-down from 80 to 0 km/h.  After a few replicates of practice, I found that I had to start further back than I had originally thought.  Unfortunately those first few replicates are included in the data for the tarp on, meaning that some replicates were on a slightly different section of road than others.  This wouldn’t matter if the road was flat, or even if it was even, but it is not either of those things.
I also didn’t populate the spreadsheet out to 70 seconds on any of my attempts.  I think that a slower initial speed would have been a better choice.  Perhaps 70 instead of 80 km/h.