How to make good test documentation from of A-B-A test

[Vekke]

Sad news.

We were not able to perform the test yesterday. We broke two tools when trying to remove the first of 6 stainless steel hubcaps which were installed to the truck.

22.5 inch Achter wieldop RVS 1 stuks [12011] - €85.00 : FSB-SHOP!, Truck@Vans interieurs en onderdelen

They were this type that there is bracket under bolts where you screw that hubcap with two different size flanged bolts. They were just so rusted that with our tools no chance. Put some rust off threre and the guy will try to remove those during next week.

[instarx]

Make sure the car's running gear and engine are at equilibrium before running the first test. This means fully warmed up. A 5 km drive will not be enough.

Instead of evening runs I suggest early morning runs. The atmosphere is at its most stable before sunrise and there is frequently no wind. Traffic is at its lowest in early morning too.

Make everything about the car the same so you can get comparable results across several days. For example, full fuel tank, no extra weight, tire pressures.

if possible, perform a calibration run or two before and after each test series where the car is set to zero state (i.e. no mods). This lets you know if things have changed during the series, and if each series is comparable to runs done at other times. These calibration runs also let you correct for factors beyond your control, such as seasons and temperatures change.

[Vekke]

You can see estimation when there will be least traffic. Usually in Finnish roads it start to get "silent" at 22.00. From 24.00 to 05.00 is best time and the lowest point is near 03.00:

Liikennetiedot Pääkaupunkiseutu - Foreca.fi

Go over those arrows to see estimations and real numbers.

[bondo]

Interesting thread. A-B-A tests have been reported here and elsewhere as a great way to calculate the fuel efficiencies of aerodynamic modifications to vehicles.

In this thread it is discussed things like atmospheric conditions (winds, temperature, etc.), even the aerodynamic effect of other vehicles upon the test vehicle, actively engaged in a test, may have an effect on the results of the test.

I have often wondered if the very cold temperatures and brisk cross wind may have effected this A-B-A test, which was also determined in just one day. From what I have read in this thread, my thinking is that Fubeca's test may be very flawed indeed, which he has actually agreed to. Not to belittle Fubeca for the hard work and fine job he did on his aerocap, he just needed better atmospheric conditions on the day he ran the A-B-A test below.

http://ecomodder.com/forum/showthrea...ead-11611.html


Bondo

[Piwoslaw]

Originally Posted by RobertSmalls

If you spend too much time out there, e.g. with 20 runs, other parameters like ambient temperature, driver attention, and the weight of fuel in your tank will start to drift, potentially skewing your data.

This is why A-B-A-B testing is better than A-B-A. If there is a trend, you will notice it in both trials.
Extrapolating this: It would be best to do one run with, one run without mods. Then again and again and again. But this of course takes up too much time, so A-B-A-B is the best compromise. If you have time left over, then continue to do another A-B as soon as possible.

[Vekke]

Hi,

Performed first tests and learned thing trought the hard way. Plan was to do A-B-A measurements in 3km long and level straight. We used only 14 FleXcaps because front capsels were so rosted that the bolt heads were ruined and we decided to leave them on.

Longer distance was about 11 long kilometers in one direction. Total length of one north to south run is about 28 kilometers.

Mittaustulokset 5-6.2.2011.jpeg - Windows Live

1. We drived the first run to north and got a result 38.7L/100km. Mistake number one was that we drived the car about 20 km before the first run.
2. Same leg to south 30,5 l that number (27.1) is for longer distance.
3. North still capsels off 39.5l. Those + plusses mean that car went by, one plus one car...
4. South 30,3l and (27.0)

When we first drived two times the straight and got north results varied 2% we decided to measure effect on longer distance, because we got better accuracy for that to south. First results for south were. 27.1 and 27.0 so accuracy was amazingly good between runs.

5. North same long distance as to south. 43,1 and 42,5. Measured results to two points on the road.
6. 29.9l (26,7l) As you can see from these figures they get smaller. That is caused the fact that car warm up during the runs bearing, oils etc. Now we had already driven the car before the test for 50 km before the first run.

Now it took about 20 minutes to install the capsels beacause we took some video material also. Car was runnin this time but temperature was -2 celsius so the car calmed little.

Now the first difference is that we did not warm the car for 20 km before first run capsels on so the car is "cold"
7. result was 44.2 liters which is 3,8% more than the one run (42,5l) we had for that long run to north capsels off.
8. 30,6 and to that longer distance 26,8 liters. That short leg car is still guite cold and gives the highest result so far but to the longer measurement it is already best 26,8.
9. Now here we had a human error and driver forget to put down one of the axels on the truck. We decided to measure same leg as in first two north runs. 38,6 liters which is the lowest by small margin.
10. Here police patrol stopped us checked our papers. They were wondering what the hell we drive aroung motorway in the middle of the night. It took 20 minuts and again car became "cold". Still we were able to get 31,4 and 27,3 liters. Now the warm up effect is 4.1 liters or 13% difference. This police incident pretty much ruined our last measurements because we measurements are not really comparable after this.
11. 43,1l
12. 27,1 liters. I have also wind data and try to find out if this can be caused by the wind. During measurements wind speed rose little.
13. We took 20 minutes coffee breack and too Capsels again off. 45,6 liters. Now this runs only difference measuremenet 7 is wind speed so this is good comparison. Difference is 3.1 %.
14. 32,7l and 28,1. Again best comparison to mearument 8 and difference is 4,6%. Difference between these two runs which are tests best comparisons is total 3,85% less fuel.
15. 44.0
16. 31,1l and 28,0 liters
17. 43,8 liters
18. 28,5 liters.

Average of all A off north 43,975 (4 runs)
average of all B on north 43,65 ( only 2 runs)
Difference 0.74 % and remenber those figures are not good comparison to each other...

A1 off south 26,933l (3 runs)
B on south 27,07 l (3 runs)
A2 off south 28,2l (3runs)
Better comparison is B to A2 and difference is 4 %. Difference between A and B south all runs is 1.8%

Difference between all A and B runs is 1.3% which is still good result. We have to do second test on that same place. Place worked out well but our testing werent perfect. Biggest factors were, mistake on the first measurement, and the police. Also we should have changed the capsels between every run to avoid that wind and also warming up effect. Next time we will install capsels in 5 minuts and take them out in 5 minuts so all runs are comparable to each other in better manner.

More data will come as soon as I am able to process it first. Also video footage.

[Piwoslaw]

Wow! That's a lot of work, Vekke. It's too bad that you had problems during the test runs, hopefully the next try will be better.

BTW Do have any idea how much fuel was used during the test? Doing that many 20km runs @ 30-40 l/100km must have used up quite a bit.

[Vekke]

Average consumption with emty trailer is circa 35 l/100 km. We drove about 50 km to test place and test driving for 250 km and back that 50 kilometers so total 350 km that means about 122 liters which is in money 156€. Whole test took over 6 hours to finish it.

I could drive with my seat with that money for over 2400 km easily at 5l/100 km. When you are driving empty truck you see bigger gains than with full load, but you can make rough estimation what is the effect of those FleXcaps with full load. Now we just need new test.

Now imagine that some of the trucks do run empty on our roads and try to imagine why? One of the best ways to save fuel is to get your gargo space full all the time and minimize empty driving.