by "PardonMyFlemish16" (TheCoolKid)
Published 10/08/2017 at 10:12
by "PardonMyFlemish16" (TheCoolKid)
Published 10/08/2017 at 10:12
No Tags
STARS: 6
I have been playing with this program called CarTest for probably over a decade now. It is a decent simulator of car performance, mainly in a straight line (though it does have a lap time simulator lmao).
To me its best feature is that it allows you to modify cars as you wish. Want to double the horsepower? Go for it. Change/delete/add gears? No problem. Change tire size? Sure! All the parameters it shows are open to change.
Over time one thing the creator added was more transmission options. Back when it was a DOS program they allowed for 6 gears and either manual or conventional automatic. With each iteration more gears were allowed, and different transmissions (CVTs and dual clutches) were also enabled.
As a former Maxima owner one thing I lamented about the 4th gens were the abysmal transmission option. After driving various MT Nissans and Hondas I’m convinced Nissan cannot engineer a good shifter/clutch setup (which is part of why I went auto with my G37). But the 4AT sucked all the damn life out of that excellent engine, probably adding another 1-1.5s to the 0-60 time. Now one option would be to just engineer a better manual, but that wouldn’t affect performance much. What if we fit an old 2000 Maxima with the 222HP VQ30 with a modern 8AT?
Well again this software allows that easily, so I went for it. I took the original 4AT gear ratios and did the following:
Kept original final drive and gear ratios
Squeezed a ratio between each of the original gears
Added a short AF 1st gear for that off the line rip
This program has a crap load of great plots which is awesome for a data geek like me, with the ability to export values as well. Anyway, here’s what the outcome was:
The regular SE is the 5MT version which was also 100lb lighter. The 8AT version was nearly a second faster to 60 and through the quarter, and over 2 seconds faster to 100 (vs the manual version). Much of those gains come from that super short first gear which prompts a shift before 30MPH and accelerates the car at close to 1g. That’s like an AWD clutch dump, but without the drivetrain damage.
Is it possible this software is a bit optimistic? Maybe. The 4AT Maxima was good to 60 in the low 7s with this engine, so perhaps the 8AT would be good for the low 6s rather than the high 5s. But even still, just looking at the road torque plots is telling.
Again, that 8AT first gear delivers nearly twice the road torque. And then from there, you’re squeezing in more area under the curve, which translates to more acceleration (and better fuel economy).
Now obviously if you enjoy rowing your own there is no substitute for a 3 pedal setup. Though I will say not all manuals are created equal. But with automatics like this, only the best manuals deserve to survive.
Side note, I ran the test with 8 manual gears... it did improve performance from the manual but it was nowhere near as dramatic. It split the difference between the 5MT and 8AT due largely to the longer shift times. 6-7 gears spaced at about 30-35MPH is still ideal for manuals. I didn’t bother with CVTs because in real life they are limited in their ratio spread, though Nissan’s CVTs are excellent.... the CVT in the latest Maxima is the best auto transmission I’ve ever experienced.
"Maxima Speed" (maximaspeed)
10/08/2017 at 10:28, STARS: 0
Former 4th gen owner here. You done us proud son.
"Eric @ opposite-lock.com" (theyrerolling)
10/08/2017 at 11:50, STARS: 0
Nissan did a good job with the 5MT in the 1992 NX2000 my parents own. It’s possibly the best 5MT I’ve driven.
"If only EssExTee could be so grossly incandescent" (essextee)
10/08/2017 at 12:12, STARS: 0
My car was only rated .7 seconds slower to 60 for the auto, which isn’t bad when you consider that it was a close-ratio 5MT vs a 4spd auto.
"DC3 LS, will be perpetually replacing cars until the end of time" (dc3ls-)
10/08/2017 at 12:45, STARS: 0
The problem is that I don’t really thing there’s much room for manuals to improve. Sure the Corvette has seven gears, but I doubt that’ll catch on with other cars. Plus with more gears the more time you loose with all the extra shifts.
Although who knows. People in the 60s probably said the same thing about 4 speeds.
"PardonMyFlemish16" (TheCoolKid)
10/08/2017 at 13:33, STARS: 0
This is exactly what the problem is. Manuals are pretty much at their peak, especially with auto rev matching.
My G37 is my first auto car, but it probably won’t be my last.
"Aaron M - MasoFiST" (amarks563)
01/03/2018 at 20:14, STARS: 0
What’s the underlying mathematical formula that drives this program? The modeling assumptions are important to consider if these numbers are any better than pulling something out of your ass.
Tire friction is also important. Your example first gear ratio, with stock 2000 Maxima tires, would dramatically *increase* your 0-60 time because you’d spin your tires all the way into third gear, if not fourth with an 8AT.
"LongbowMkII" (longbowmkii)
01/03/2018 at 20:24, STARS: 0
the ubiquitous 8AT (ZF 8HP) has a (potential) 5.00 in 1st and 1.00 in 6th just as the baseline.
"PardonMyFlemish16" (TheCoolKid)
01/03/2018 at 21:24, STARS: 0
The underlying model just seems to be good old calculus (integrals). Take the power curve and apply it against all the resistive forces over time. The tire model is just a simple static/dynamic friction model. No accounting for weight transfer beyond the wheelbase I think. It’s not the most sophisticated program but it’s been accurate enough over a ton of cars for it to have decent credibility with me.
There are plenty of cars with similar total 1st gear gearing these days that put more power down through the front wheels. A 2000 Maxima will be on 2018 tires today so I don’t see that being an issue.
"Aaron M - MasoFiST" (amarks563)
01/03/2018 at 22:28, STARS: 0
While that’s fine and good, it does mean the 0-60 times are essentially made up unless the base test was redone on modern tires.
"PardonMyFlemish16" (TheCoolKid)
01/04/2018 at 02:17, STARS: 0
Simulators can only account for so much. If it can reliably produce results close to real life, it’s good enough. Not really something worth fretting over.