WRC and Group N Torque

[cougarrange]

This discussion will supplement OldF's very interesting threads of similar title, but with the focus on torque characteristics of the restricted WRC and Group N engines, using the Subarus to illustrate.

Both WRC and Group N engines develop very high peak torque outputs:
- Impreza WRC approx 700 nm (516 lb-ft) at 3000 rpm
- Impreza N12b approx 560 nm (412 lb-ft) at 3250 rpm

Astronomical numbers, these, yet the cars are rarely driven at peak torque.

The WRC car is typically in the 4500-6000 rpm range, where torque ranges from 550 nm (405 lb-ft) at 4500 to 370 nm (270 lb-ft) at 6000, following a roughly linear decline with increasing rpm due to the 34 mmm restrictor. So, a WRC car puts out an average torque of 460 nm (339 lb-ft) in its rev band. Not so very different from a well tuned street WRX or STi, as we shall see below.

The Group N car is typically driven in the 4000-5500 rpm range (check out an excellent video on Youtube of Makinen testing an N12 and watch his rev counter.) In this range, the GRN torque is 475 nm (350 lb-ft) at 4000 declining to 312 nm (230 lb-ft) at 5500. So, the GRN car is averaging 394 nm (290 lb-ft) in its rev band.

By comparison, my 2.0 liter VF34 street car puts out around 490 nm (360 lb-ft) at 4750 and is still making 325 nm (240 lb-ft) at 7000 rpm. In the band 4500-6000 (the optimum with RA gears and 4.44) the average torque is 440 nm (325 lb-ft), demonstrating the advantage of not running a restrictor. Of course, without antilag, the torque is not always immediately available like it is in the WRC and GRN cars.

High boost gives the WRC and GRN cars huge torque at low rpm, but the restrictor progressively takes it away with increasing rpm.

Why don't they drive in a lower rev band, say 3000-4500?
Because as the same speed they would be in a higher gear and the torque at the wheels would be lower. It is faster to stay with the declining torque curve in a lower gear.

How often do the WRC and GRN cars truly benefit from the massive low rpm torque?
I'll defer to those who have driven these cars on rally stages, but my guess is that stage starts and slow hairpins are about the only times when they are driven at peak torque.

In conclusion, the WRC and GRN heroics we see on the stages have more to do with driving talent from "another planet" than from peak torque or horsepower. I am even more impressed.

Superg

[OldF]

You are quite right cougarrange, the best torque on wheels are on the declining part of the torque curve. Below are few links to graphs that shows that’s true (here I’m again with my pics).

Mitsubishi Evo 9 (566 Nm @ 3300 rpm):

http://i267.photobucket.com/albums/ii284/OkdF/Mitsubishi_wheel_torque.jpg

As you can see the gearshift should be made at such revs that with the next higher gear the revs should drop to revs where the peak torque is. For the Mitsubishi in question it’s 3300 rpm. In practise the revs should be few hundred higher because it takes some time to do the gearshift but that depends how fast gear shifter you are.

For the Mitsubishi the gearbox ratios are:

1st gear: 3,000
2nd gear: 2,000
3rd gear: 1,470
4th gear: 1,111
5th gear: 0,857
and the final drive ratio is: 3,307

For example the revs when shifting from second to third gear can be calculated by dividing the 2nd gear ratio with the 3rd gear ratio and multiplying the result with the revs where the peak torque is.
2,000 / 1,170 * 3300 = 4490 rpm.

I’ve also seen in some forums some discussion which one is more important, the power or the torque. If you look at the graph (red vertical line) you can see that the torque at the wheels (in the graph the wheel torque is the torque / wheel) is higher on 4th gear than on 5th gear although the engine torque is higher on the 5th gear than the 4th gear. The power again is just opposite. As you can see the gearshifts should be made where the power curves are crossing so as the power curve is always as high as possible. By this the power is more important but remember there’s no power without torque. So, which on one is more important? I would say both are important.


Subaru Spec C (570 Nm @ 3500 rpm):

For the Subaru the gearbox ratios are:

1st gear: 3,333
2nd gear: 2,385
3rd gear: 1,750
4th gear: 1,333
5th gear: 1,040
and the final drive ratio is: 3,545

http://i267.photobucket.com/albums/ii284/OkdF/Subaru_wheel_torque.jpg

Mitsubishi and Subaru in the same graph:

http://i267.photobucket.com/albums/ii284/OkdF/Mitsu_Sub_wheel_torque.jpg

As you can see the Subaru has higher torque at the wheels but less top speed.


It would also been nice to have a comparison with a S2000 car but I don’t have any gear and final drive ratios for a S2000 car. The only thing I could do was a comparison on the highest gear. At last year NORF Anton Alen told that Abarth had tree different ratios for top speeds of 165 km/h, 175 km/h and 185 km/h. He chooses to drive with the ratio for 165 km/h top speeds. If the top speed is reached at 8400 rpm the torque is approximately 230 Nm (peak torque 250 Nm @ 7000 rpm. Source: S2000 database). At top speed with 6th gear the torque / wheel would be about 359 Nm. For the Mitsubishi the torque / wheel at 165 km/h with 5th gear would be 256 Nm and for the Subaru 357 Nm. The torque / wheel at the peak torque and highest gear would be for the different cars:

S2000: 6th gear, 250 Nm @ 7000 -> torque / wheel = 390 Nm
Mitsubishi: 5th gear, 566 Nm @ 3300 rpm -> torque / wheel = 402 Nm
Subaru: 5th gear, 570 Nm @ 3500 rpm -> torque / wheel = 526 Nm

[cougarrange]

Nice charts, OldF. What software do you use?

It is interesting that the theoretically optimum Impreza Group N shift point for peak torque in the next gear is 4500-4800, depending on the gear, yet the video of Tommi testing an N12 shows he is shifting at about 5500. I wouldn't have thought that shift lag would require 700-1000 rpm beyond the optimum shift point, especially with antilag. Can you explain why he is staying in gear to higher rpm than "optimal"?

By the same reasoning, I am suprised that the Impreza WRC shift point is 6000 rpm. I can't find any WRC gear ratios to confirm this, but I suspect 6000 rpm is well above the theoretical optimum shift point for max torque in the next gear at 3000 rpm. Any thoughts?

Have you ever found any WRC gear ratios?

thanks
Superg

[OldF]

I’m using Excel spreadsheet.

Sorry, I don’t have any explanations why the gearshift is made about 1000 rpm above theoretical. In fact I tried today with my own car and when shifting quite slowly the revs dropped by about 1000 rpm but with a N-grp car with a dog box when you don’t have to use the clutch the gearshift should be able to do much faster. It would be nice to have some comments from some driver.

About the WRC cars I think it’s much about the attributes of the engine. For example Juuso Pykälistö told in an interview that with the Citroen the gearshift was made far beyond 6000 rpm compared to the Peugeot where the gearshift was made little above 5000 rpm. Also he said that the Citroen has broad rev band but with less torque on low revs compared to the Peugeot. I think he knows what he’s talking about because he’s been driving both cars.

No, I don’t have any WRC gear ratios. I think this is one of the secrets among others that the WRC teams don’t publish.

[OldF]

cougarrange, do you have the link to the Tommi Mäkinen test video? I’m not sure if I found the right one.

typicall mitsubishi evo 9 (N4) has 280 - 290 HP and 500 - 550 NM

[cougarrange]

Makinen testing the N12. Partway throguh the video, the lighting imptoves so you can see the tachometer. I reckon he is usually between 4000 and 5500 rpm.


http://www.youtube.com/watch?v=gsSrfK4AGzA

[HaCo]

Acording to mag.gpweek.com (free online magazine), Gr N's will have 32mm restrictor and more freedom in rear suspension design in near future. This to get them on the same level as S2000!

[Mirek]

But gr. N cars already have 32 mm restrictor :D

[cougarrange]

Two more on board videos with telemetry showing rpm, gear, and speed.

Loeb Rally Italy 2008
http://www.youtube.com/watch?v=KD3yTIXikG4
Rev band mostly 3000-6000 rpm, with some brief moments above and below.

Mitsubishi Evo 9 in the Brazilian Rally Championship 2007
http://www.youtube.com/watch?v=4bMjMbuzQB0
Rev band 3000-5000 rpm, averaging 4000 rpm.

If anyone knows of other examples of on board videos of WRC or Group N with telemetry, please post.

Superg

[cougarrange]

Two more onboard videos showing rpm, etc.

Evo9 Group N (Brazil)
http://www.youtube.com/watch?v=b_a08RGjTfM

Richard Burns Impreza WRC 2001 Argentina
http://www.youtube.com/watch?v=wq7K7uKmVPY

[Viking]

I’m using Excel spreadsheet.

Sorry, I don’t have any explanations why the gearshift is made about 1000 rpm above theoretical. In fact I tried today with my own car and when shifting quite slowly the revs dropped by about 1000 rpm but with a N-grp car with a dog box when you don’t have to use the clutch the gearshift should be able to do much faster. It would be nice to have some comments from some driver.

About the WRC cars I think it’s much about the attributes of the engine. For example Juuso Pykälistö told in an interview that with the Citroen the gearshift was made far beyond 6000 rpm compared to the Peugeot where the gearshift was made little above 5000 rpm. Also he said that the Citroen has broad rev band but with less torque on low revs compared to the Peugeot. I think he knows what he’s talking about because he’s been driving both cars.

No, I don’t have any WRC gear ratios. I think this is one of the secrets among others that the WRC teams don’t publish.

Loeb, Sardinia
http://www.youtube.com/watch?v=KD3yTIXikG4
Youtube quality, but if the graphics is right you can see him shift at 6000+rpm and dropping some 1000rpm/gear . He seems to use it between 4000-6000+rpm

[OldF]

http://www.youtube.com/watch?v=kD6par7ez_M

Sebastian Loeb, Jordan rally, end of stage one (at about 4:35). The speed is 204 km/h and revs are approx. 7200 rpm. It seems that he try to keep the revs between 5000-6000 rpm.

It’s more difficult to keep the revs between optimum in a rally compared to track.

[HaCo]

]But gr. N cars already have 32 mm restrictor :D

I meant the other way around :( :D

[janvanvurpa]

]But gr. N cars already have 32 mm restrictor :D
Yep, the GpN has been at 32mm and the GpA or A8 at 34mm since 1 Jan 1993

[cougarrange]

I used an instantaneous acceleration calculator to compare the performance of an Impreza WRC with my well-tuned modified 2.0 liter WRX (VF34).

Link to the calculator:
http://vlsicad.ucsd.edu/~sharma/Potpourri/perf_est.html

Assumptions:
Impreza WRC:
Flywheel torque 515 lbft declining to 270 lbft (source: 2004 Race Engine Technology magazine article on the Subaru WRC engine)
Wheel torque 369 lbft at 3000 rpm declining to 193 lbft at 6000 rpm
Ratio of wheel to flywheel torque consistent with the Dyno Dynamics dyno I used to test the VF34 WRX
Weight 2856 lb (WRC regulation minimum weight plus a full tank of fuel)
Since WRC homologated gear ratios are secret, I used ratios from a published test of the Prodrive 1997 Impreza WRC 6 speed manual dog box (1- 2.92, 2- 2.13, 3-1.74, 4-1.45, 5-1.23, 6-1.04, diff ratio: 3.89)
Shift time 50 msec (assumed paddle shift)

Modified VF34 WRX:
Max wheel torque 259 lbft at 4750 rpm (Dyno Dynamics verified)
Torque in excess of 200 lbft 4100-6200 rpm
Weight 3080 lbs
RA gearset (1-3.083, 2-2.062, 3-1.545, 4-1.151, 5-0.825, diff ratio: 4.44)
Shift time 300 msec (manual)

Both cars:
CD 0.37; frontal area 23.7 ft2
Driver weight 200 lb
Tire size: 205/60-15 gravel tires

Results:
WRC VF34 WRX
Peak Power (HP at wheels) 242@4500 251@5400
Peak Torque (lbft at wheels) 369@3000 259@4800
Top speed (MPH) 117 140
30-60 MPH 2.6 sec 3.15 sec
30-90 MPH 7.35 sec 8.3 sec
60-90 MPH 4.75 sec 5.1 sec
Shift points 4978-5118 rpm 6570-6582 rpm

The predicted 60-90 MPH time compares closely with my Vf34 WRX road test times of 4.8-5.2 sec. Note: my road test was at 6000 feet elevation and the predicted performance is at sea level.

If the Impreza WRC had a manual shift (300 msec shift times) the times would be even closer: 30-60 3.1 sec, 30-90 8.45 sec, 60-90 4.75 sec.
Of course, the WRC also has a nearly 200 lb weight advantage. And the WRC's dynamic engine characteristics benefit from antilag that my car does not have.

Note that the calculator does not attempt to model standing start times, as there are too many launch variables. With over 500 lbft at 2000 rpm, I assume the WRC would bludgeon my car in a standing start!

So, again the incredible performance of the WRC cars is largely due to driving skill "from another planet." A well tuned 2.0 liter street WRX is nearly as fast in an absolute sense, but no doubt way slower through a rally stage.


Superg

[M5]

So, again the incredible performance of the WRC cars is largely due to driving skill "from another planet." A well tuned 2.0 liter street WRX is nearly as fast in an absolute sense, but no doubt way slower through a rally stage.


If fast had meant only engine and driver, it would have been easy building a WRC car, but ask Suzuki - it is a bit more than that !

(Otherwise American cars would have had a chance ;) )

[cougarrange]

I don't mean to minimize the importance of suspension, differentials, body structure, weight distribution, etc. The entire package is obviously critical.

[cougarrange]

A couple documented times to restore the WRC mystique:

2003 SWRT newsletter, WRC2003

0-60 (gravel) 4.0 sec
0-60 (tarmac) 3.2 sec

Superg

[cougarrange]

And perhaps the best documented numbers, Autocar's 0-100-0 of the Ford Focus WRC in 2006: 0-30 1.25 0-60 3.34 0-100 8.91

[Mirek]

Škoda Octavia WRC Evo II 2001, Tour de Corse settings. 0-100 km/h (62 MPH) 3,98 sec, 0-200 km/h (124 MPH) 16,04 sec (close to the top speed). 1000 meters (0,61 mile) from zero speed in 22,87 sec. Acceleration 80-170 km/h (49,7-105,7 MPH) on fifth gear 7,75 sec. Acceleration 100-200 km/h (62-124 MPH) on sixth gear in 11,42 sec. It was dry and +7°C. Those are average numbers from measuring in both directions.

[OldF]

It is interesting that the theoretically optimum Impreza Group N shift point for peak torque in the next gear is 4500-4800, depending on the gear, yet the video of Tommi testing an N12 shows he is shifting at about 5500. I wouldn't have thought that shift lag would require 700-1000 rpm beyond the optimum shift point, especially with antilag. Can you explain why he is staying in gear to higher rpm than "optimal"?

I made a comparison how much less torque the Mitsubishi and Subaru have on higher revs compared to optimum.

Mitsubishi:

Optimum 4950 rpm 4490 rpm 4366rpm 4278 rpm
Revs Gear 1 Gear 2 Gear 3 Gear 4
5000 5,98 % 14,68 % 17,12 % 18,70 %
5500 18,91 % 26,45 % 28,46 % 29,93 %
6000 30,34 % 36,77 % 38,46 % 39,65 %

Taking the revs to 5000 rpm on 1st gear produces 5,98 % less torque on the wheels compared to 2nd gear. Taking the revs to 6000 rpm produces 18,91% less torque compared to 2nd gear.

Subaru:

Optimum 4891 rpm 4770 rpm 4595 rpm 4486 rpm
Revs Gear 1 Gear 2 Gear 3 Gear 4
5000 5,15 % 7,47 % 10,85 % 11,78 %
5500 18,59 % 20,59 % 23,48 % 24,32 %
6000 32,12 % 33,82 % 36,11 % 36,87 %

I don’t how much loosing 10% – 20 % of optimum torque would affect the performance but loosing 20% - 30% must have an affect on the performance.

the WRC and Group N Torque is good it seems, the speed rates are quite amazing