mousecatcher

Yeah I have my own worksheet I've developed over the years for this. How do you do the shift point markers, is that just a manual addition to the graph or do you have a formula automatically calculate it?

I've always just used rpm on the Y axis, and only just now realized that isn't the correct variable. It should ideally be acceleration, or some proxy for it. Since I'm trying to evaluate gear and final drive ratios, I'm thinking I can ignore aero effects. Wheel diameter (radius) needs to be taken into account but only for speed calculation.

Thinking of acceleration, doesn't the dyno take gear ratio and final drive ratio into account when calculating torque/hp curves? So we can just multiply dyno torque by gear ratio and final drive ratio for FT = final torque. Or, just the same, we can use force F = FT/r, but I don't think we need that since r is constant.

Well I'm wondering why you used HP in your first chart, and how you generated the Force data for your second chart. I guess I'll just generate the graphs using variables as above and see how it looks compared to yours. But also I wanted to have the discussion and also maybe learn some of your graph-fu.

mousecatcher

I've convinced myself from https://www.physicsforums.com/thread.../#post-1231053 that HP is what we care about.

Mech33

I just manually added those after the fact for clarity.

Right, you can ignore aero effects because it has no impact on relative power available in each gear at a given vehicle speed. You only need the wheel radius, final drive, and gear ratios for the speed calculations as you say.

Dyno's in general just measure (or infer through other measurements) the force on the dyno wheel surface. This force can be translated to power via the relation Power = Force * velocity. And then the power is translated to engine torque via the measured engine RPM (Power = Torque * angular velocity).

I just find overlapping the power vs. speed the easiest way to show the impacts of different gearing choices. Sometimes folks don't believe that the results are the same if you instead use wheel force, so sometimes I generate that instead for comparison. In the end the force curves are just the power curves divided by the velocity (and corrected for proper units, i.e. 1 hp / mph = 375 pounds force), so the curve intersections always occur at the same speed.

Warwick Morris

My take is that the most certain outcome is going for more power, i.e. X51 kit. Then you're certain to be faster at every track. I wonder how fully installed costs work out for changing, just the gears, just the FD and the X51+/-4.0l.

Mech33

BGB Motorsports

This has been my stance since Day 1. I was sort of disappointed by the gearing hysteria created by the journalists when the car was released. Don't get me wrong; we developed the gears for Guard in 2013 and i believe in their potential and given the reliability issues with the OE 3rd gear, i think it's a no brainer. I just don't think the gearing "sucks" like others do. I am a race track guy and i appreciate a lengthy 2nd gear for those slow hairpins that most tracks have. I never felt like the 3rd gear was too long because the power curve on my car was like a staircase. People bemoan the gearing in the GT4 because the power falls off immensely at 6900 RPM because the motor A.K.A. the air pump is starving for more air. Take a look at the power curve of my beloved baby vs. a bone stock GT4. Believe you me, never did i ever feel the need for gearing because my power curve was legit. I sold the car to one of my good customers in Rhode Island that did one of our original 981 swaps. He has a 420 hp Boxster with short 2.7L gearing because that's what the car started life as and he drives that on the street; for the track he has my GT4 that has race headers, Cobb Stage II tune, BMC filters, lightweight clutch and full X51 upgrade. He added the tried and tested Guard LSD, bigger monoballs, brake lines and all safety upgrades in the form of belts and a 4-point.

I asked him how his track day went at Thompson yesterday and his report back was "while there was 1 other Cayman, the rest were Z28s and Vettes but the X51 didn't let down against the big V8s albeit on a shortish track!"

I took my car to Daytona and in bone stock form i couldn't clear 162mph; when i returned with the full Stage II package i was 172 - 173 mph through the tri-oval every time. the car needs more power over 6K RPM and there's no two ways about it; if you're pulling car lengths all the way to redline you're not worried about gearing!



Stage II X51 vs. Stock GT4

mousecatcher

I have convinced myself that isn't true, because aero drag isn't a constant. Using the OEM (1st) graph in https://rennlist.com/forums/gt4/8822...l#post12683867 as my reference HP data, that dyno is somehow magically guessing the powertrain losses to extrapolate from WHP to crank HP. Cleary it's a heuristic but I'm sure it's within reason. You can see that the losses are not linear, and clearly we want to use WHP (which includes those losses) not crank HP in power/speed/gear charts.

Given that we take into account non-linear powertrain losses, shouldn't we also take into account the non-linear aero losses, esp. since they dominate at even moderate speeds. The drag (->net power) at 60mph is >> the drag at 50MPH and so we should take this into consideration when deciding a shift point. If the aero drag were constant, or even linear, we could ignore it. But it increases with the cube of the speed so I'm thinking this should be factored in.

Can you or someone else make a comment on that?

From charts shown here, it seems close to ideal for the track, but it's also easy to understand that for street driving, mostly under 60MPH, and highway, mostly under 100MPH, there's not a lot of rowing of the gears ... may as well have bought an automatic!

Mech33

The aero drag force isn't a constant at all speeds of course, but it *is* a constant across different gears at a given speed (since the aero drag is a function of speed, among other physical characteristics of the outside of the car). My point was that you don't need to take into account aero drag when determining cross-over points of available power between gears (for calculating ideal shift points, or optimizing gear ratios for acceleration, etc.).

To put it simply, at every instant in time, if there is more engine power available by shifting gears (because the engine makes more power at the RPM it would transition to in that next gear), then you shift. The amount of the aero drag on the car at that instant doesn't change that.

mousecatcher

ah right. the aero drag is vs vehicle speed, not engine speed. so it can be ignored for our purpose.

whereas the parasitic "internal" drag through the drivetrain, is sensitive to engine speed. so we should consider that. there's actually sensitivity to overall gear ratio but that would be hard to model and likely doesn't introduce significant error.

Mech33

I don't think you need to worry about any of that. The engine-side drags are accounted for in the engine power curves, and the angular velocities of everything from the output main shaft of the transmission to the rear wheels are proportional to vehicle speed (so that's not changing with gear at a given speed). The transmission input shaft drags are all linked to engine speed so again that should be accounted for in dyno-derived engine power curves.

mousecatcher

yes, exactly. i didn't mean that any analysis had to do something to account for drivetrain losses. the dyno already does it. i guess you're right, that vehicle speed already takes into account the different drag for different gear ratios.

ajw45

Agree with everything and love the x51 upgrade! Between 5k and 8k rpm the tach sweeps to redline so much faster than stock and the engine just feels so responsive at the top end, it really does dramatically change the character of the engine even with the stock gearing. At the track I can see the top end come into play down the straights as the distance from cars in front of me closes faster with rpms, especially with turbo cars like M3/4's that start to get reeled in as both cars run through the gears.

That said, as a city dweller that only gets to the track a handful of times per year, I'm looking forward to shorter gearing. I used to short shift a bit after 7k rpm since there wasn't much left in it after that but post-X51 the car is still making good power to 8k rpm so it's like having an extra 1k rpm to play with. Basically, the x51 made the top end better and more... toppy... but it's still equally out of reach at street legal speeds. IMO the stock GT4 already has too much power for my driving environment so maybe this is all moot but I like the idea of the extra in-gear responsiveness that a shorter 2-4 could bring and the ability to enjoy the x51 top end more (legally) even if it's purely experiential and completely non-functional performance wise.

mousecatcher

At long last (I have a day job you know), here is my power/gear/diff analysis workbook. You can select 2 choices of each parameter set (custom tire diameter also if you have different size than oem, eg track setup) and compare them visually. Big thanks to Mech33 for providing valuable insights on what is important. I used a simpler version of this workbook (no power data, just gearing and diff) when I was doing lotus stuff intensely. There are many many combinations of lotus gear and diff choices, so it was quite hard to make a proper evaluation without a tool like this. Even for the GT4, there are enough choices that you need a tool like this, and you need the power analysis as well.

In the full power graph and the operating power graph, I don't draw a vertical line between shifts. It's incredibly difficult to convince excel to do this, especially for parameterized data. It's still quite readable without those indicators.

I'll echo earlier comments that the OEM gearing choice is actually quite good for track duty, considering ultimate top speed and the flatness of the powerband. If you play with the workbook, you'll see that more powerful engines that are also peaky, exhibit high power losses at the shift point, You are still above the OEM power at the shift point (for most combinations) at that speed range, but you can see how you get a totally different feeling in the car as it pulls again to peak power, vs steady (but lower) power like with OEM. This makes total sense since you shift when still at peak power, catching the next gear at a lower point, vs with OEM where the peak is well before redline and the power drops off until you meet the next gear. It's very neat to see it visually. Mech33 did capture that with his earlier charts -- but I hope I've done a cleaner job with my workbook.

I think it makes an argument to tune for an earlier power peak if you can get the power band flatter, within reason of course. You could create a theoretical flat torque curve and model that to test the theory ...

Please note the importance of the speed range chart. Before evaluating any other data, validate that the speed range vs gear choice is usable. For each track, there is going to be a critical long corner or two where you'll want to be in a single gear. If your power/gear/diff stack forces you to shift mid-corner or too soon before exit, you may end up regretting that choice regardless of any other factor. Do keep in mind that the chart reflects WOT power. You might have a corner where you are right at a gear transition according to the speed range chart, but in practice you are in the next higher gear anyway to keep the torque manageable, and feathering the throttle until you start to exit.

As this is the final drive thread, I've prepared the 2 money graphs for an OEM car vs OEM with AQ 4.62 diff. This data is pretty much the same type of data that Mech33 has provided recently in this thread, so if you're following along it doesn't take too much effort too understand. In the operating power chart, I also show the total gain for a given speed range. I selected 35-75 randomly (you can choose any range with the workbook). Here you can see that the AQ diff is a net loss for this speed range. Of course you can game the speed range to make it a net win or net loss, so this isn't any kind of statement on whether or not it's a good mod. You can also understand why it's a net loss, as explained by Mech33 earlier.

Hope this workbook is helpful to RLers!

stout

^ Wouldn't argue with any of this…for the track. Or, possibly, for a GT4 with significant engine mods. But I guess I'm the boogeyman, as I was one of those reviewers (sorry! ) who raised concerns with the gearing in the GT4 at the original Portimao launch.

Second and third seemed too tall, and I remember wondering ever so briefly if it was turning too many revs in sixth on the one freeway section back. Those concerns didn't fade when the first press car came through San Francisco—and they moved from "is this a problem?" to "yep, it's a bummer" when I drove my own GT4 as a daily for two years and 15,000 miles. Second good to 82~, third good for 112~, and then sixth too short for freeway cruising—turning 3500-3700 all the way down to LA while going with the flow...meaning one or even two stops for fuel, with the large tank. Again, I can see where the gearing in the GT4 is fine for track use and possibly Germany...but I felt the GT4's gearing let the car down on the street in the U.S. and still do. Maybe fixing the flat spot in the powerband would help, or adding hp as some here have stated. But after selling my GT4 and then logging nearly 10,000 miles in a 991.2, I find the latter has perfect gearing for the street and that I'd love a GT4 with 1-5 from the 991 and 7th from the 991 in 6th for freeway cruising. If I still had my GT4, I might get a shift **** made that says:

2-4-5
3-5-5

YMMV, of course!



pete

Mech33

Nice work! Hard to argue with the physics.