Rx-8 gearing ratios?
 

The gearing ratios for the rx-8 are one unknown that we haven’t really discussed in much detail. There has been some talk and much speculation about the rx-8’s top speed, mileage, 0-60 times, ¼ mile times etc but they are *all* dependent on gearing.

Since the rx-8 was designed to seat 4 people and the gearing/diff ratios should take that into account and be much higher than the ratios on the last rx-7. What would everybody like to see the gearing ratios end up at?

I would propose the following aggressive and closely spaced gearing to give the best 0-60, ¼ mile, mileage and utility. Gears 1 thru 5 are aggressively geared and closely spaced. 6th gear is only overdrive gear and is spaced a bit apart from 5th . The .84 6th gear is mainly for fuel economy at highway/freeway speeds but will also lead to a very high top speed. Because of the renesis' wide power band and flat torque curve, the rx-8 will be very driveable (until we hear the dang buzzer). :D

Rear wheels = 225 45 18
Diff - 4.1
1st – 4.02
2nd – 2.46
3rd – 1.77
4th – 1.33
5th – 1.08
6th - 0.84

1st gear at 3500 rpm would be approx 16 mph.
2nd gear will hit 60 mph near the hp peak.
55 mph would be approx 2750 rpm in 6th gear.

Would there be any reason why the above gearing would not be practical? Below is also a chart which gives the speeds per rpm for each gear.

RPM 1 2 3 4 5 6
1000 4.6 7.5 10.4 13.8 17.0 21.9
1500 6.8 11.2 15.6 20.7 25.5 32.8
2000 9.1 14.9 20.7 27.6 34.0 43.7
2500 11.4 18.7 25.9 34.5 42.5 54.6
3000 13.7 22.4 31.1 41.4 51.0 65.6
3500 16.0 26.1 36.3 48.3 59.5 76.5
4000 18.3 29.8 41.5 55.2 68.0 87.4
4500 20.5 33.6 46.7 62.1 76.5 98.3
5000 22.8 37.3 51.9 69.0 85.0 109.3
5500 25.1 41.0 57.0 75.9 93.5 120.2
6000 27.4 44.8 62.2 82.8 102.0 131.1
6500 29.7 48.5 67.4 89.7 110.5 142.1
7000 32.0 52.2 72.6 96.6 119.0 153.0
7500 34.2 56.0 77.8 103.5 127.5 163.9
8000 36.5 59.7 83.0 110.4 136.0 ----
8500 38.8 63.4 88.2 117.3 144.5 ----
9000 41.1 67.2 93.3 124.2 153.0 ----

If the rx-8 were geared as such, the 0-60 times will *easily* be well less than 6 seconds and the 1/4 mile might be less than 14?
:eek:

Brian

Oops, the tabs in the chart disappeared. I think the below chart will be easier to read:

RPM````````1`````````2`````````3 `````````4`````````5`````````6
1000``````4.6``````` 7.5``````` 10.4`````` 13.8`````` 17.0`````` 21.9
1500``````6.8``````` 11.2`````` 15.6`````` 20.7`````` 25.5`````` 32.8
2000``````9.1``````` 14.9`````` 20.7`````` 27.6`````` 34.0`````` 43.7
2500``````11.4`````` 18.7`````` 25.9`````` 34.5`````` 42.5`````` 54.6
3000``````13.7`````` 22.4`````` 31.1`````` 41.4`````` 51.0`````` 65.6
3500``````16.0`````` 26.1`````` 36.3`````` 48.3`````` 59.5`````` 76.5
4000``````18.3`````` 29.8`````` 41.5`````` 55.2`````` 68.0`````` 87.4
4500``````20.5`````` 33.6`````` 46.7`````` 62.1`````` 76.5`````` 98.3
5000``````22.8`````` 37.3`````` 51.9`````` 69.0`````` 85.0`````` 109.3
5500``````25.1`````` 41.0`````` 57.0`````` 75.9`````` 93.5`````` 120.2
6000``````27.4`````` 44.8`````` 62.2`````` 82.8`````` 102.0`````131.1
6500``````29.7`````` 48.5`````` 67.4`````` 89.7`````` 110.5`````142.1
7000``````32.0`````` 52.2`````` 72.6`````` 96.6`````` 119.0`````153.0
7500``````34.2`````` 56.0`````` 77.8`````` 103.5`````127.5`````163.9
8000``````36.5`````` 59.7`````` 83.0`````` 110.4`````136.0`````----
8500``````38.8`````` 63.4`````` 88.2`````` 117.3`````144.5`````----
9000``````41.1`````` 67.2`````` 93.3`````` 124.2`````153.0`````----


Brian

I'd rather see it a little shorter than that, but maybe that's just me. Perhaps 5-10 MPH lower in each gear at redline. I don't do a lot of high-speed cruising, but I do value quick acceleration.

Looks good,

I agree with Grimace to some extent, perhaps a little shorter... not as much as 10Mph though...

I do think that 6th should be a real long legged cruiser as suggested... I often cruise near 90-100Mph and when in germany I had the chance to cruise at 135+ for about 200 miles.... don't want to be ragging the engine at high speed cruising...

One thing though.... If you want to reach 60 in second, for 0-60 times... it must be 62 not 60 before the rev limit... because european tests are to 100Kph (62Mph).

Hi BlueAdept,

With the gearing that I suggested, 2nd gear would actually be going at 67.2 mph at the rev limit and 63.4 at the hp peak. I am hoping that the rx-8 will come out as aggressively geared from 1st thru 5th as my example (or even a little shorter).

The renesis is really different than anything out there though. Does anyone know any other production engine that comes close to (or beats?) 6000 rpm of greater than 90% torque?!?!?!

Brian

If the rx-8 6th gear was .84 as in the example, 135 mph would still be approx 1200 *below* the torque peak. The car would be *begging* you to at least take it 1200 more rpms to it's torque peak at 163 mph (or whatever the electronically limited top speed would be). :D

You wouldn't be ragging the engine and probably wouldn't even notice the revs were that high until you heard the buzzer. :D

Brian

Oh, absolutely,, I wasn't complaining about your numbers... but Grimace was suggesting lowering the ratios... I don't mind that, but 6th must be long legged...

As for the 0-62 thing... I don't think we are exactly sure where the redline will be...

Unfortunately, we will probably have to wait until at least January before we get more concrete info on many of the particulars. :(

Brian

Gear ratios OK to me. Perhaps a little taller, if anything.

I can't see taller myself... we're talking about a 6 speed box that can get to 60 in second.... even that's probably a sacrafice to get good 0-60 times.

Taller would be a mistake. You wouldn't get into the power band above 3500 RPM in first gear for 18-odd MPH. I'd favor shorter so that people get into the 4000 RPM powerband quickly. That way it feels more powerful. SHorter gearing takes advantage of the high-RPM, low-torque nature of the engine, while tall gearing would actually make the car feel sluggish.

Exactly... I agree they could all be a little shorter, except 6th which is just about right in my eyes,

I was playing around with CarTest (a car performance calculator) to see how different gearing might affect the acceleration of the RX-8. On my own, I came up with these values for gearing:

Final - 4.10
1st - 3.70
2nd - 2.65
3rd - 1.90
4th - 1.45
5th - 1.05
6th - 0.75

My ratios give the following results:

0-60: 5.35
1/4: 14.0 @ 102.0
Top Speed: 161 MPH in 121.6 seconds

The values suggested by Buger - His 1st and 2nd are shorter, while his 3rd, 4th, and 5th are taller:

Final - 4.10
1st - 4.02
2nd - 2.46
3rd - 1.77
4th - 1.33
5th - 1.08
6th - 0.84

Here are CarTest's results for Buger's ratios:

0-60: 5.40
1/4: 14.0 @ 101.4
Top Speed: 160 in 111.96 seconds.

This is only a computer simulation, so take it for what it is worth, but it appears that my slightly taller first two gears will actually help 0-60 and 1/4 acceleration, albeit ever so slightly. However, Buger's taller 3rd and 4th seem to greatly improve time to top speed. It seems that taller is better from this starting point. I'll play around with the numbers some more and see if I can improve on things.

Note that 6th gear has no effect on any of these tests, as top speed occurs at redline in 5th for both of our gear sets.

Other note: The simulations were based on a 2800 weight.

RedRotaryRocket

Good posts from you and the others. Why not use 2970 for the weight instead of 2800 like you've done. until we hear differently, we are better off using the last weight given by Mazda. if they better it by production, all the better but for now, let's see what the simulations come out with at the weight of 2970.
Just a suggestion.:rolleyes:

Boowana,

Ask and you shall receive. :D Actually, the numbers using 2800 lbs are probably too optimistic, since the software assumes that the weight includes the driver.

I re-ran the simulation using a weight of 3170 (2970 + 200 lb driver). Here are the results:

My gear set:
0-60: 5.9
1/4: 14.6 @ 98.2
Top Speed: 160 in 140.03 seconds

Buger's gear set:
0-60: 6.0
1/4: 14.5 @ 98.3
Top Speed: 159 in 124.62 seconds

Howzzzat?

Interesting software....

I don't think it could be that hard to write somthing along those lines that did an "Auto optimisation"...

And the nissan guys just realized what power to weight ratio meant. I hope those new calculations are too conservative because I would love to easily be in the 13's with a few mods. That really shows how weight effects performance.

Have you plugged in stats from known cars and seen how they matched up?

I also just realized that none are 1:1 which one of them will have to be. In most sets it is fourth but your fifth is close @1.08 so try the scenario with it like that please?

The software actually comes with stats from many vehicles...generally it's within 2 or 3 tenths as far as 0-60 or 1/4 mile times are concerned. Plus or minus 2 or 3 tenths is still a pretty big spread, but it's not too bad for getting "in the ballpark"

Hi spdfrk,

I was aware that 4th was usually the 1:1 gear but before coming up with my original gear ratios, I found that the s2000 had the below gearing:

Final - 4.10
1st - 3.133
2nd - 2.045
3rd - 1.481
4th - 1.161
5th - 0.971
6th - 0.811

Oddly enough, the s2000 has no 1:1 gear. Honda geared the s2000 so that the gears are nicely spaced and it actually has 2 overdrive gears. Too bad the s2000 doesn't have a renesis torque curve. ;)

Nice numbers RRR! I like that your 2nd thru 5th gears are a little shorter than mine. I think the only thing hurting your 0-60 time is the taller 1st gear.

I also think that your 6th gear is probably more in line with what Mazda has done in the past as rx-7s usually had their top gear rpms at around 2100 - 2200 at 55 mph.

That cartest software is pretty cool too. I downloaded the demo version and it puts out tons of graphs and info, even a mpg graph per gear! The demo version only lets you analyze the cars already in its database though. How about some revised gearing numbers:

Final - 4.10
1st - 4.10
2nd - 2.61
3rd - 1.84
4th - 1.39
5th - 1.06
6th - 0.75

I like your very short second gear but I made it just a *little* longer. I also made 1st gear shorter to maintain even spacing and because a sports car with seating for 4 adults should have ample torque in first gear. I slightly shortened gears 3 thru 5 and made 6th gear even longer as one of the goals of Mazda was that the car have much improved mileage. Could you try the below gearing numbers in Cartest?

Brian

Brian,

I ran your new numbers, but I made a change to your 5th gear...I made it 1.03 as that maximizes top speed.
Here's the results using your revised gearing (with 1.03 5th):

0-60: 5.80
1/4: 14.5 @ 98.0 MPH
Top Speed: 160.3 in 145.13 seconds

That's based on a test weight of 3170 lbs, so compare these numbers to the later set I generated...you've picked up 0.2 seconds to 60...pretty good!

Eric

It's DEFINATELY going to be 2940 lbs or LESS, so re-do those figures and lemme see those times again :)

yeah herc but ya gotta add the weight for the human ( or rotorhead) driving:p

Hrmm good point :)

How do they test cars in magazines and such?

Try it with 3115 lbs :)

That's so it accounts for ME in the car :P (175 lbs hehehe)

I just bought cartest2000 and have some good news. According to the help menu in the program:

"CAR CURB WEIGHT: Enter the curb weight of the car excluding the driver, passengers, or fuel. "

Furthermore, if you click "create car specific model parameters", you will see where you can specify the "weight of driver". :D

Unfortunately, I just got done with a long day at work and I'm going to sleep now. I will try entering the rx-8 hp curve values into the program tomorrow morning and see what the program will come up with for a curb weight of 2900 lbs. :)

Brian

Cartest normally uses a hp/torque curve profile for cars which is modified to match the peak hp/torque numbers that are entered. Since the renesis hp/torque curves are very different than a normal piston engine, I modified the hp/torque curves to match the renesis. (see attachment)

Below are some of the parameters that were used:

hp:........... 250 @ 8500
torque:.... 159 @ 7500
redline:.... 9000 rpm
wheels: ... 225/45R18

Diff - 4.1
1st - 4.10
2nd - 2.61
3rd - 1.84
4th - 1.39
5th - 1.06
6th - 0.75

curb weight:........ 2900 lbs
driver weight:..... 160 lbs
drag coeff:.......... .29 (estimated using value from Mazda Millenia)
grnd clearance:... 4.9" (estimated using value from Mazda Millenia)
launch rpm: ........ 2000 (optimum est by cartest using all data)

Below are the some of the results:

0-60: ........... 5.74 seconds
1/4: ............. 14.36 @ 99.08 MPH
Top Speed: .. 160.69 in 131.37 seconds
Mileage: ....... 20.4/32.1 mpg (combined 24.4)

Below are some results with a 3000 lb curb weight:

0-60: ........... 5.89 seconds
1/4: ............. 14.49 @ 98.19 MPH
Top Speed: .. 160.39 in 134.70 seconds

Below are some results with a 2800 lb curb weight:

0-60: ........... 5.6 seconds
1/4: ............. 14.23 @ 99.97 MPH
Top Speed: .. 160.97 in 127.44 seconds

It will be interesting to see what the final weight and gearing specs end up at. This should give us all some idea of what may be possible.

Brian

Thanks Buger for the great info! Do you know it the 0-62 (0-100km/h) will be very different?

holy effin' eff... 0-60 mph in 5.6s?? again, holy...

i know that this approximation (using approximated and guesstimated numbers) is only an approximation (using approximated and guesstimated numbers), but still man, that's pretty damn fast for a sports sedan...

Thanks.:D

Much better.:p

I can see I will have to go on a major diet:( Oh well, it'll be worth it. Now I have a real reason to drop a few.

Buger,

Nice numbers. It sounds like your version of cartest is much more sophistocated than mine...mine must be some old demo version or something...it's dos based and I downloaded it for free almost two years ago! Needless to say, mine doesn't have a help menu and only accepts one weight, entitled "test weight". Since you've got the fancy-schmancy version, we'll let you generate the numbers from now on :D

Does CarTest2000 account for drivetrain configuration and losses?

That's a big 10-4 good buddy.

Or....

Yes.

:)

Hi Sheylen,

It shouldn't make much of a difference. Add about .02 seconds at that speed to the 0 - 60 time.

I got the new version from www.cartest2000.com. I think the same guy has been tweaking it a bit. I can run some other numbers if anybody wants to but it would be good to buy a version to reward the guy for all of his work.

Yes, it appears to estimate drivetrain losses pretty closely but you can modify the losses if you have more specific information. See the attached screenshot of some of the parameters.

Brian

Thanks again:cool:

Here is the attachment with some of the drivetrain loss parameters (it was too big earlier):

The performance of the 1999 Mustang gt coupe and the estimated performance of the rx-8 are pretty close. It would be interesting to look at how two closely matched performance cars with vastly different approaches compare. The Mustang has far greater torque but the Rx-8 will have a much higher redline.

Attached are a couple of curves comparing the 1999 Mustang gt coupe and the rx-8 acceleration in first gear. The graph on the left illustrates how the renesis can pull over a much wider rpm range (almost 60% longer) than the Mustang v8.

In the graph on the right, you can see how the torque curve of the example rx-8 was shortened to approx the same speed range of the Mustang by shorter gearing. Although the Mustang engine has a huge advantage in torque(302 - 159), it is also approx 300 pounds heavier in this example (3205 vs 2900). The Mustang still pulls harder (by approx .1g) at its torque peak but we can see that the sample rx-8 pulls harder at the low and high ends of its curve.

The example rx-8 has the following gearing:

Diff - 4.1
1st - 4.10 * 4.1 = 16.81
2nd - 2.61 * 4.1 = 10.701
3rd - 1.84 * 4.1 = 7.544
4th - 1.39 * 4.1 = 5.699
5th - 1.06 * 4.1 = 4.346
6th - 0.75 * 4.1 = 3.075

The 99 Mustang GT has the following gearing:

Diff - 4.1
1st - 3.37 * 3.27 = 11.02
2nd - 1.99 * 3.27 = 6.507
3rd - 1.33 * 3.27 = 4.349
4th - 1.00 * 3.27 = 3.27
5th - 0.67 * 3.27 = 2.191

With the above gearing, the renesis will turn almost 17 times before the rear wheels turn once in first gear. The mustang engine will turn a little over 11 times before its rear wheels turn once in first gear. The difference in gearing and weight mitigate the peak torque advantage that the mustang has. With the above gearing, simulated world (not real world) testing shows that the rx-8 actually beats the mustang in 0-60, 1/4 mile and top speed.

Two more graphs compare the drive power of the 2 cars thru the gears. You can see how the higher drag coefficient affects the top speed of the mustang.

If you would could you try adding 50 hp to the top end. Just to show us what kind of times the mazdaspeed version of the rx8 would produce. I would think the 50 hp would come in late in the powerband.

Thanks
Ryan Fritts

Top Speed
 

What kind of swept area are you using for the top speed calculation? If the RX-7 had about 1.9 square meter frontal area, how does the RX-8 compare.

What was the coefficient of drag of the RX-7? 0.35?

Also since the Miata 6 speed has a 1:1 5th gear, so should the RX-8. That benefits top speed because the power doesn't flow through a gear mesh in the transmission.

Re: Top Speed
 

0.35 is very high these days... I'd be surprised if a modern design was above 0.30... infact I'd be surprised if the RX-7 was that high!

A quick web search seems to indicate that the FD RX-7 had a drag coefficient of 0.30.....

Hi Ryan,

I am at work now but will try using the single turbo torque profile with the extra 50 hp and the added weight tomorrow morning at home.

Hi Mike,

As there has been no real info on the frontal area of the rx-8, I let cartest estimate it from the given parameters. It is an optional field and the help menu says that "CarTest will estimate it if it is not known more precisely". I could probably try some values to geta ballpark figure of what cartest is using though.

The drag coefficient of the 3rd gen rx-7 was .29. I estimated .29 for the rx-8 as well since the Mazda Millenia also had the same coefficient. Surprisingly, even the 2003 Toyota Corolla has a low drag coefficient of .296. It may be more conservative to estimate the rx-8 drag as .30 but no more than that. A convertible version of the rx-8 (if it is ever made) might have a drag of something close to .35 but the rx-8 coming out next year should have .29 or .30.

It is a good idea to try a 1:1 ratio in 5th gear as it is much more common on 6 speeds than the gearing #s I used. There are several cars out there without a 1:1 ratio gear but you make a good point that having a 1:1 5th gear could marginally reduce frictional losses for that gear.

I think the best ratio for top speed was somewhere around 1.04 or 1.05 to 1 because the drag losses curve meets the 5th gear torque curve close to the highest peak. A 1:1 5th would be more fuel efficient though and would probably be more in line of what Mazda is planning. I will try a 1:1 5th tomorrow morning and move 3rd and 4th to space them evenly with the new 5th gear.

Brian

A google search on the web comes up with conflicting info for the Cd of the 3rd gen. Most places generally quote .29 or .30.

I remember reading that the base had a .29 Cd while the r1 had a .31 because of the added body items.

I also remember reading that the 2nd gen rx-7 had a similar Cd of .29 to .31 depending on the model.

I fully expect the rx-8 to come in at a similar range of .29 - .30 for the base model. The Cd doesn't make too much of an impact at lower speeds but is the major factor in limiting top speeds of cars.

The top speed of the rx-8 is definitely not a factor for me as I would not like risking a ticket at anything over 100 anyway.

Brian

Most Japanese/German (soon to be american) companies like to use Cw instead of Cd. Well Honda doesn't.

The Insight is 0.25 Cd, Toyota claims that the Lexus LS430 is also 0.25 (my ass!!!) I would believe 0.25 Cw if the car was lowered to its bump stop (air suspension) and the wheels weren't rolling in a wind tunnel.

Cw differs from Cd in that Cw is what the drag from blowing air at a car. Cd is what results from hitting stationary air (Real Driving)


I would hope that the RX-8 has low front/rear lift rather than a remarkably low Cd

How do you test Cd then if stationary in a wind tunnel doesn't cut it?

Re: Top Speed
 

Trying different values, it appears that cartest estimated the frontal area to be 23.821 sq ft. This seems a bit high to me as I found a website that had the Mazda 929 at 20.76 while being both wider and taller than the rx-8. Whatever the case, 20.44 sq ft (1.9 sq m) seems to be reasonable number so I will use that number for the rx-8 from now on. A quick check on cartest shows that the change in frontal area increased top speed by 5 mph and marginally decreased the 0-60 and 1/4 mile times.

While researching some of the variables of drag, I found some good info on Cd and frontal area. An excerpt from the site (http://www.thirdgen.org/newdesign/te...deffects.shtml) is below:

Coefficient of drag is the most widely used number in relating the Aerodynamic efficiency of a car, and is very misleading. It is just a measure of how much out of 100% (like a barn door) a vehicle lets air flow around it. Typically sports cars have coefficients of between .28 and .45 (Viper roadster uggh!). The better (lower) the number is, the easier it is for air to pass around a car. It can be misleading because I recently read an ad in MT,and it claimed that the '97 Infiniti Q45 has the best CD of any production car (I think it was .27). You don't realize that the car has 10" more height, and a few inches more width than a third generation f-body (length isn't a factor). This means that it has more sq feet of air to push out of the way.

A good example of frontal area is the Mazda Miata. The car has a greater CD than an f-body (camaro?), but because its frontal area is tiny (I think it is 16.5 sq ft), it is more aerodynamically efficient.I use an equation to compare cars with different CD and Frontal Area figure, the number generated is merely a factor to be compared 1:1.

CD x Frontal Area(sq ft) = factor #
It works well for direct comparisons, because doing it the long way would take 10 minutes to compute (true engineering method).

Trying different values in the optional field tire circumference, it appears that cartest estimated the value to be 82.44 inches while a 225/45R18 should have a tire circumference of 81.59 inches. Real world circumference will be a little smaller because of the weight of the car so I just used a value of 81 inches. Note that this has the effect of additionally gearing the car an additional 1.75%.

The changed parameters for the new test are below:

frontal area: ................ 20.44 sq ft (instead of cartest est 23.821)
wheel/tire circumference: 81 inches (instead of cartest est 82.44)

Diff - 4.1
1st - 4.10 * 4.1 = 16.81
2nd - 2.65 * 4.1 = 10.865
3rd - 1.83 * 4.1 = 7.503
4th - 1.33 * 4.1 = 5.453
5th - 1.00 * 4.1 = 4.1
6th - 0.75 * 4.1 = 3.075

curb weight:........ 2900 lbs
launch rpm: ........ 1800 (optimum est by cartest using new data)

Below are the some of the results:

0-60: ............ 5.64 seconds
1/4: ............. 14.32 @ 99.32 MPH
0-160: ........... 57.28 seconds
Top Speed: ....... 168.49 in 144.82 seconds

Brian

Brian,
Is that estimated 0-60 time for the 300 hp mazdaspeed version?

That was for the na version with revised gearing, front area and wheel parameters.

I just did a quick run with 100 pounds added weight and using the standard torque curve with the turbo profile and the below hp/torque figures:

hp: .......... 300@8500 rpm
torque: ... 220@6500 rpm

The rest of the parameters are the same as for my previous post. Below are the some of the results:

0-60: ............ 4.77 seconds
1/4: ............. 13.43 @ 107.49 MPH
Top Speed: ....... 178.47 in 171.67 seconds

I really wanted to do some research to estimate what numbers would be reasonable to input but I didn't have the time. The above numbers seem a bit unreal don't they?

Brian

Those top speed seem very unrealistic. I would hope that the RX-8 (regular) would do 160, and a 300 hp version should do 170 (hopefully).


A wind tunnel can be used for estimating the aero drag, and so can a computer model. Empirical testing determines how much high the real Cd is.

Anyone remember 'roll down' testing in Motor Trend many many years ago.

Not like it matters anyway, you'll never get up to that speed, even on a track, unless it's an oval :D

I have gotten up to 150 in an E36 M3 on a straightaway, but otherwise I stay around the 80-85 limit on the highways.