Peugeot 205 5th gear swap (8.2% improvement at constant 50 mph)

[HypermilerAX]

My father introduced me to hypermiling 3 years ago and then I discovered ecomodding myself. And now, my father asks me about ecomodding when he saw what I did on my car. He’s interested in having a higher 5th gear on his 205 turbo diesel.
So I did some research about the different gearings on the XU engines. The car has already the longest available differential (17x59) but the 5th gear has room for improvement. Stock gearing is 37x28 and 38x25 is mounted on some other cars with shorter differential. We already have a spare engine which has the gearbox with the parts we are looking for so we have no reason not to do it. The mod is easy to do as it doesn’t require to demount the gearbox. It’s like on mine, just a cover to remove to have access to the 5th gear.

Stock gearing is 40,3 km/h at 1000 rpm and we would have 46,4 km/h at 1000 rpm, that’s 15,1% improvement. The mod should be even more effective considering the fact that the stock 5th gear is not that high. Max speed is reached at max RPM (4300) which means that it’s designed for performance. On my AX, top speed in 5th was reached (before mod) at 4300 rpm whereas max RPM is at 5000 rpm. So the gear was designed for economy and I still had nice improvements so imagine on the 205. His best with no mods at all (not even tire pressure) is 68.2 mpg by driving no faster than 45 mph.

Pictures of mod and results will follow.

EDIT:

Recap :

Before :
Temperature : 25,7°C
Pressure : 1021,1 hPa
Humidity : 43%
Air density : 1,197
Wind speed : 13 km/h NNW
Sunny
Engine at operating temp
Tire pressure raised to 2,8 bar

Fuel consumed : 840 g = 1015,7 mL (827 g/L at 25°C)
Distance driven without EOC : 28,56 km
Consumption : 3,556 l/100 km
Recorded average speed : 80,26 km/h

Corrected consumption (80 km/h and air density 1,2) : 3,547 l/100 km

After :
Temperature : 20°C
Pressure : 1020 hPa
Humidity : 77%
Air density : 1,220
Wind speed : 12 km/h NNW
Clouds but dry road
Engine at operating temp
Tire pressure raised to 2,8 bar

Fuel consumed : 748 g = 901,2 mL (830 g/L at 20°C)
Distance driven without EOC : 27,30 km
Consumption : 3,301 l/100 km
Recorded average speed : 80,23 km/h

Corrected consumption (80 km/h and air density 1,2) : 3,255 l/100 km

Before : 66.16 mpg
After : 72.28 mpg

This is exactly what I expected. It's an 8.2% improvement for a 15.1% higher ratio (MPG gain = 55% of ratio change).

[HypermilerAX]

We did the before test this afternoon. We have a nice stretch of a road not far away which is good for testing. It's a straight 14,5-km road with a 0,048% grade (according to local topographical map). It's a 2-lane road but with few traffic. We were never in a vehicle's wake, just some passing cars on the left lane.
First, we filled up to the top at a station and 2 bottles too. We went to the starting point, filled up again the tank to the top and stopped at a mark we could recognize. I throttled at 100% load and shifted at 2000 rpm to reach the cruise speed of 80 km/h on GPS. 300 m before turning to do the other direction, I shut down the engine and EOCed to the stop line to avoid any varying idling time due to passing cars. That distance was subtracted to calculate the consumption. Again, I throttled at 100% load and did the other direction in the same conditions. I shut down the engine again at the end and EOCed to the starting point, distance was subtracted too. The course was recorded with a GPS to know the exact speed. We fillup again to the mark, we put one full bottle (1,00 L) and 28,0 g of the second one. So that's 1033,9 mL (827 g/L at 25°C).

Conditions :
Road : 14,5 km in each direction - 0,048% uphill grade in first NW direction (7 meters on 14,56 km)
Temperature : 25,7°C
Pressure : 1021,1 hPa
Humidity : 43%
Wind speed : 13 km/h NNW (we faced it in the first direction, it's hard to have less wind here)
Sunny
Engine at operating temp
Tire pressure raised to 2,8 bar

Fuel consumed : 1033,9 mL
Distance without EOC : 28,56 km
Consumption : 3,620 l/100 km
Recorded average speed : 80,26 km/h

That is 64.99 mpg at 49.88 mph.

Next week, we will change the cogs and wait for a day with similar wind conditions and do another test.

[renault_megane_dci]

ABA-ing on short distances is questionnable even if your operating mode sounds bullet proof

[HypermilerAX]

Most of the inaccuracy comes from the fillup, be able to refill to the same level. I quickly calculated that a one centimeter level difference is about 0.05 l/100 km on the distance we drove. Level difference was no more than 5 mm.

But I think that 30 km done that way are more accurate than 1000 km. More accuracy on the volume fillup but much more variables that I don't control.

Ideally, I would like to pump the diesel directly out of a bottle I would weigh before and after.

[renault_megane_dci]

Originally Posted by HypermilerAX

But I think that 30 km done that way are more accurate than 1000 km. More accuracy on the volume fillup but much more variables that I don't control.

I beg to disagree.
What you're testing this way is the efficiency at a very specific RPM and in specific conditions.

Performing a lot of miles on the other hand is a way to check the mod in various conditions, all in all, real life.

There are ways I can think of to improve mileage in very specific conditions but in real worl driving they end up worst than the factory setup.
Balance is a difficult task.

[HypermilerAX]

Originally Posted by renault_megane_dci

I beg to disagree.
What you're testing this way is the efficiency at a very specific RPM and in specific conditions.

Yep, that's what I want to test. Gain at a steady 50 mph. Most of our driving is at steady speed at a specific RPM so that test will be interesting to have.
We will see the gains on a tank anyway by driving but we'll have all the other variables : traffic, weather, speed, type of roads...
I probably won't gain as much as in the test since nobody drives 100% in 5th gear. But I'm optimistic for the gains, I'm hoping for 3,30 l/100 km, 3,25 would be amazing.

[renault_megane_dci]

That will be interesting to be able to categorize the effect of a measured improvement inside everyday use.
Depending on your drive, it could affect the FE in good or bad and if it is good, maybe you will see a return of just a portion of the measured improvement.
I'm looking forward to it.

[HypermilerAX]

We did the swap yesterday, everything went fine. To test the gearing, I turned the brake disc and compared to the primary cog (=engine). I got 8 revs on engine = 7 revs on wheel but the other wheel was on the ground so it has to be divided by 2 due to the differential. 8 revs on engine = 3,5 revs on wheel = 0,4375. Theory says 38/25*17/59=0,438. Not bad at all.
We did some stationary tests with one side up. 5th gear works smoothly, brake disc spins really fast compared to 4th gear.
And finally, ultimate test on the road. It was like I expected : nicely high geared but not too much. The engine has a lot of low-end torque so it works great. We'll do the other test next week.
We also plan on mounting an intercooler (which exists stock for the engine) which increases torque from 16 to 18.6 mkg at 2100 rpm. This mod should also increase efficiency.

Pictures : old gears / new gears / before / during / after

[MetroMPG]

Fantastic that it was such a simple swap.

I wish I could do the same for my Firefly (Metro/Swift)... alas, all the gears were the same from different variants. They only used different final drive ratios from model to model.

Looking forward to the final results!

[HypermilerAX]

I took the car one week to do a MPG test. I drove at 56 mph average (GPS) on 70% of the commute, the remaining at 45/50 mph. I got an impressive 3,43 l/100 km (3,53 with GPS distance). At same speed and same roads, I made 3,89 l/100 km some time ago but it was in winter and I didn't use as much P&G as now. But still, the 5th gear is doing well. My father couldn't believe it!
Tomorrow, we'll do the after test.