larchmont

(Hard to believe, ain't it? But of course I didn't know what anything meant. I memorized the formulas and knew how to use them and I got the right answers, even with complicated stuff -- but most of the time I had no idea what I was doing.)

fdl

rzee...see this is what i mean when I say that you guys still dont understand me. I am bad at explaining things for sure.

when I said ...

I mean that how can you say torque is tied to accelreation while at teh same time saying horsepower isnt, when horsepower is a function of torque (hp=torque*rpm/5252).

Now do you understand?

All I am saying is that HP is the better indicator, torque isnt enough.

rb1

Ah, that is an entirely different statement. Who said anything about shifting at peak torque? All I said was that peak acceleration occurs at peak torque.

In most well designed cars, you should shift near the redline for the fastest overall performance.

Someone actually has to run a simulation using the gearing and torque curve to determine the "perfect" shift points for a given car. What determines who wins is who has the greatest area under the torque curve and you don't shift at the torque peak (or the hp peak for that matter) to achieve this.

Again, this is simply not true for a given gear if we are assuming a flat torque curve. Your TSX will accelerate better in 1st gear at 6000 RPM than it would say in 2nd gear at 3000 RPM (I don't have the exact ratios in front of me), but this is because we are discussing wheel torque and you are using your 1st gear to deliver greater torque to the wheels.

I would put money on your acceleration in 1st gear being fairly close to (or even greater!) at say 3000 RPM (let's be sure we are in the meat of the torque curve) than it is at 6000 RPM. It will actually even be greater at 3000 RPM since the torque at 3000 RPM is more than it is at 6000 RPM.

rb1

OK. Here's the proof (if you'll accept if for a different car, although one with a flat torque curve over a large range much like the TSX).

There are two curves, one for a chipped car (not a flat torque curve) and one for a stock (fairly flat). Note the g-force produced by the acceleration over the RPM range:

The specified peak hp for the stock car is 150 @ 5700 RPM, but observe that peak acceleration occurs at about 2800 RPM.

fdl

rb1...I'm trying to get things back on topic...regardless of where the highest g forces of accleration are...you want to stay at peak hp for the best acceleration..and that is the main point I am trying to make. Hence why I think a shorter 2nd gear or taller 1st gear would be better. (yes i know a taller gear reduces torque to the wheels but I think there is room to do this in the TSX).

I just think the gearing in the TSX could be better and produce a better 0-60 , and 1/4 mile. That is my point right from the begining...that having your rpms drop to 4000 rpms just because you have more torque there is not a good idea.

spc

Please go ahead and do the race or experiment and let us know the exact time different. Remember to control the environment as equal as possible between the experiments.

Can't wait to see the result.

KingV

I think the key here is that high torque at low rpms will make your car "feel" sportier because you have more power off the line in every day driving. A low torque car might make 150 HP at 6000 RPMS while a car with a higher torque might make 150 HP at 3000 RMPS. Well the car with higher torque will feel faster even if both cars have a max HP of 200 HP because the bulk of the power is available at lower RPMS, and few people Rev up to 6000 RPMS every time they take off from a stop light.

Proo

This topic has gotten pretty far off, but just one comment to help those who still disagree with fdl's explanation... I've seen several references to F=ma. F does NOT equal torque. F is torque divided by the radial distance (which is proportional to the gear ratio), which is why you get more "F" with the same torque in a lower gear (eg. higher RPM = higher hp), and consequently higher "a".

Back to the topic, I do think that 1st gear should be taller. I personally don't care much about getting the absolute best 0-60 time. I find it hard to smoothly shift 1-2 under normal driving without taking it above 3k and waiting for a second or two for the needle to drop back down towards ~1.5k... I can't even make a left turn without shifting in the middle of it. I guess I bought this car for different reasons than some of you, and I think you should have all bought RSX-S's.

I also think 6th should be taller, because I think it's just stupid that I spend that much more on gas than an auto.

But complaining about this is not going to help me, unless they change it in the next redesign and I go buy another TSX.

calbear2k1

This topic has been a mildly interesting read for me, and since I am studying up for a Physics exam this very minute, I figure I should chime in. While I hold a degree in physics, it seems like a mechanical engineer would be able to expalin some of this better since most advanced physics has nothing to do with mechanics.... so going back to my freshman class knowledge:

Torque is what makes your car accelerate. While it is true that you must factor in gearing, wheel radius, etc to get the final force to the wheels, those factors do not change when you are in the same gear. Horsepower is a non-intuitive number that provides an estimate of how quickly your car will accelerate, but as people have discussed over and over, some cars seem to out perform their HP and other seem to underperform their rating. The key is that the area under the torque curve IS the correct number to look at when figuring out acceleration. Peak torque is not that useful of a number for acceleration.

If horsepower were the deciding factor in acceleration, one would expect that a car would accelerate significantly faster at a higher RPM. From a TSX dyno , the car produces 170 HP at 6.5k rpms and around 85 HP at 3.3k rpms. Nominally, that would suggest that the car would accelerate 2x as fast at 6.5 k than at 3.3k, if one were to believe that HP is the deciding value -which it IS NOT. Anyone think the car pushes 2x as hard between 6 and 7 than between 3 and 4?


So what exaclty does hp measure?! For any given peak torque, the best acceleration will come from a perfectly flat torque curve. Since horsepower is torque x RPM x a constant factor, and area under a FLAT curve is just the area of a rectangle ( torque x RPM), horsepower is directly proportional to area under the curve. So if torque bands were perfectly flat, then horsepower would be a great judge of acceleration.

Now, this gets a little abstract. if a car produces peak torque at a low RPM, it is still likely that the peak horsepower will still be at a high RPM. since the torque falls off, the peak horsepower could understate the acceleration. Essentially, the HP would be finding the area under a crurve for a rectangle, while the torque curve falling would resemble a triangle. The opposite could happen as well - a car with a high RPM peak torque could overestimate acceleration. (damn if I could draw a picture here it would make more sense....). Of course that DOES NOT mean that cars with a low RPM peak torque will always have their acceleration understated and high RPM torque cars overstated because you need to see the torque curve.

Anyway, real life is a lot more complicated, so there are tons of other factors out there. But horsepower is a somewhat accurate estimate of acceleration, but for what I think is a rather non- intuitive reason. I'll have to think to see if I can come up with anything else HP measures....

rb1

While basically true, what happens is that you:

1. Multiply the crank torque times to overall gear ratio to get wheel axle torque.

2. Divide the wheel axle torque by the radius (as you noted) to get the actual linear FORCE applied to move the car.

3. Acceleration is determined exactly by this force, minus other forces (friction, air resistance) and the mass of the car.

[As a practical matter, car tires are very close to 1 feet in radius (generally within an inch or two), so wheel torque and linear wheel force are pretty close in passenger cars. ]

While you are technically correct that torque != force, they are exactly proportional to each other so that torque is substitutable for force in this dicussion. Instead of F=ma -> a = F/m, you simply introduce a constant 'r' to represent the constant multiplier -- for a given gear -- to convert torque to force (in cars it would always be greater for the ratio of linear force to torque):

a = (F * r)/m

Proo

I'm not sure if you're just joking with everyone and trying to get them to waste their time, but I suppose it is a good exercise. I'm not a mechanical engineer, but all engineers should have a certain aptitude for this. I am open to being proven wrong, but so far it all makes sense to me.

You said it yourself. Acceleration is a constant multiple of torque, depending on the wheel size. So we can disregard this, unless you are comparing by changing your wheel size. Where you went a little fuzzy is to say that to convert torque to force. You must essentially just multiply by the gear ratio to do so, which is the whole point.

Torque peak will give you the best acceleration IF you stay within a given gear. But downshifting will give you whatever torque you get at that speed (probably somewhat lower due to torque curve falloff at a higher rpm) in the lower gear multiplied by that gear ratio (probably significantly higher).

As for the relation between acceleration and power... you're right, I think. Intuitively to me, saying 170hp@6600 and 85hp@3300 means you have the exact same acceleration at those speeds, assuming you stay in the same gear. I believe that means you have the same torque at those two speeds.

BUT say your 3rd gear is 1:1 and 2nd is 2:1. If you switch from 3rd at 3300rpm to 2nd at 6600rpm, you will get twice the acceleration!

hjd73

From the article noted by fdl, to me the following sentence is the essence:

"But, you ask, isn't your acceleration greatest at the torque peak? Yes, it is! But only within that gear. The next gear down will give you even greater acceleration at the same speed, unless the vehicle speed is too high for that gear."

There is an other article on horsepower v torque at:

http://vettenet.org/torquehp.html

from which I think two excerpts are telling:

"The Integra GS-R, for instance, is faster than the garden variety Integra, not because it pulls particularly harder (it doesn't), but because it pulls *longer*. It doesn't feel particularly faster, but it is."

"It is better to make torque at high rpm than at low rpm, because you can take advantage of *gearing*."

larchmont

Deserves being highlighted, big-time. Seems to me it addresses and ties together a lot of the dispute on here.

I think we all sort of know this and instinctively take it into account, whether with MT or AT. And, hope this doesn't sound too nihilistic, or "sour-grapes-like" because of my not understanding a lot of the technical stuff (which I don't), but..... If HJD's excerpt does basically answer it, and if it's what we all know instinctively, it would be another example of how too much knowledge can bollix things up, if you get a little lost in the trees. But don't get me wrong -- I think this is a great and impressive discussion on here, not to mention over my head.

rzee

OK, Mr. "star physics student", that's just stating some known facts.
meter per second square that is.

larchmont

So how come you didn't say it?

(Remember, used to be, used to be.)

fdl

I think one point of confusion is whether we are talking about torque to the GROUND, or torque at the ENGINE. They are two different things ... with torque to the ground being more a function of gearing than anything else.

The bottom line is still that ENGINE torque, as a number, alone does not really tell us how a car will accelerate. It doesnt take into considerearion gearing and that is why horsepower is still the best indicator for acceleration, and the most important curve on a dyno if you want to optimize your overall acceleration. You can argue that torque...when plotted as a curve on a dyno chart will give us a good indication of how a car will accelerate...but in that case the reason it becomes usefull is because we are looking at it with reference to rpms ... and that is exactly what HP, as a measurement, does.

So to put into practice...if you are in gear x, at peak torque for your ENGINE.... even though you are at peak torque you would be better off one gear lower with a higher ratio...taking you out of your peak ENGINE torque but actually putting more torque to the GROUND. And the measurement that tells us what is the largest amount of torque we can put to the ground at the given speed, is HORSEPOWER ... which is why horsepower is the best measurement of acceleration. Again...i've said it many times...horsepower and torque are not independant of each other. Horsepower takes a raw number like torque and turns it into a more informative measurement, one that IMO defines how well a car will accelerate overall. We want to keep our engine at peak horsepower at all times, irregardless of ENGINE torque because peak horsepower is the place that the car is putting the most power to the ground for the given speed, even though its not making much torque at the ENGINE.

... and to bring things back on topic I dont feel that the gearing in the TSX best keeps the engine at peak horsepower.

larchmont

First off, I'll acknowledge that amateurs (like me) probably don't belong in this. So, stop me if I'm wrong, but it seems like this post confuses the discussion, because, as near as I can tell, FDL is messing with the existing concept of "torque" as it applies to a car. "Torque" means different things. With regard to cars, it's a jargonistic term, known and understood in a specific way, with specs listed for every car. (I imagine it's ENGINE torque?) "Torque" is also a general physics term, meaning twisting force; this can be applied in many ways, including other ways in a car. Sounds to me like FDL is talking about a different kind of "car torque," a different one besides what's usually measured and listed, and he's saying this other one would be more significant for what we're talking about here. But if that's the case, isn't it besides the point? -- because what we're debating here is the relative significance of "the usual kind of car torque." If you start talking about a DIFFERENT kind of "car torque," you're not talking about what's at issue here. Sounds to me like in effect you're arguing for an additional spec.

Since I'm an admitted amateur on this, I know I could be way off (and I know I'll be told if I am). But maybe this'll help clarify things.

fdl

Torque to the ground and torque at the engine are both twisting force, but torque to the ground is created by multiplying torque at the engine by gearing. What I have been arguing is that torque at the ENGINE is not a measurement we can use to define overall acceleration. HP which through its formula will consider gearing is the better measurement.

The reason torque at the engine and torque to the ground need to be distinguished in this thread is because at peak horsepower, you will be producing the maximum torque/power to the ground that is possible (through gearing), while at the same time your torque at the engine may be much lower that its max. And that is why Horsepower is king

calbear2k1

Horsepower is a decent indication of how fast a car will accelerate. This is because it does an INEXACT job of estimating area under the curve. Note the words "AREA UNDER THE CURVE" - if you do not have those in your argument for acceleration, well, your argument stading on weak legs. Basically you need to know more information than just peak HP or peak torque, you need to know these number for the entire range of RPM's.

IN THE SAME GEAR your car does not accelerate faster at maximum horsepower. The reason you accelerate upto the readline is that after you shift you are obviously in a different gear, and therefore are not getting the same multiplyer on the engine torque.

fdl

As "inexact" as it is...its a much better indication than ENGINE torque and is still the curve you want to use you when you are setting up your gears.

As long as you can stay at or near peak horsepower (via gearing) ...HP is all that matters because it provides the best indication of torque to the GROUND. Area under the curve wont matter. In otherwords..the closer to peak horsepower you can stay ...the more torque/power to the ground you are producing for the speed you are traveling.

P.S. Arguing that when staying in the same gear engine torque will define acceleration is a useless fact becauses its far from the real world, where we have gears, and where HP is the better indicator.

calbear2k1

So this whole argument is getting a bit silly. FDL, you are correct for the most part, and I was never trying to say that horsepower was not important. I just though that some of these justifications of why horespower creates acceleration were not quite physically sound..... but in fact, did I not say that horsepower is a good estimate of accleration because it estimates area under the torque curve. I could elaborate on that more, but I think it is pretty clear.

As for gearing, it seems like magazine 0-60 times dictate the car companies' chose their 1-2 gears so as to get the car to 60 before the shift to third gear. That means there is not that much choice in gearing. After a shift under normal acceleration, it would be good to have the car remain at a flat point on the torque curve, otherwise the car will not feel smooth at all. Though that would only be a problem if 1st gear were too tall. I can't imagine why honda wouldn't have made 1st gear taller if that would really have helped acceleration. They could have had an engineer work out a computer program in less than a day, so why wouldn't it be optimized?

So how about this for an agreement: torque causes acceleration, but horsepower better estimates it.

fdl

sure cal The whole "what causes (g-force) acceleration" was really a small side note and a digression from my main topic anyways... (and something I think that was an area of confusion because of the fact that some people equate peak ENGINE torque to max torque to the ground...which is simply not correct).

My origional point was that when going into 2nd gear..the rpms drop too low. It was argued that this would actually be better since the rpms are dropping to where there is the most engine torque...and as you know this is simply not true. You want the rpms to stay near peak HP, not peak engine torque ... as some have stated (because although there is less ENGINE torque..there is actually more torque/power being put to the ground ). And on this I'm sure we can now all agree.


EDIT: You edited your post and added somethign I have to somewhat disagree on. You said that you want shifting to put you in a place where the torque curve is more flat for smoother shifts? Again ...I think you are getting caught up in engine torque here..when its torque to the ground that going to matter. When those rpms drop too much...that means that we are actually getting a large drop in torque to the ground. By keeping the gear ratios closer we can have less change in torque to the ground when shifting and it would actually be smoother. Hard shifts to 2nd right now cause a bog because of the difference in ratios between 1st and 2nd.

calbear2k1

Yeah so I edited my post and it now makes less sense than it did before. Great job me.

Anyway, what I meant to say "Though that would only be a problem if 1st gear were too SHORT" - which is eaxctly what you are describing. I somehow always get short and tall backwards (and incidently I always get confused in an auto sport shift car by thinking + means more power and - means less power, thank god for rev limiters, but that's a whole other matter).

So back to the 1st second shift. As I said before, 2nd gear is dictated by the fact the car needs to get to over 60mph in second. Now honda can choose 1st gear to reach a certain top speed, we are going to redline it in that gear no matter what, and then after the shift we fall to an lower RPM - the taller the gear, the faster the car is going at the shift point and the higher the RPMs the engine is at after the shift. Sorry, I had to spell that all out for my own benefit as well....

So getting back to my point about the accerelation is basically constant for a flat torque curved car when IN THE SAME GEAR. So whether you drop in with the new RPM's in 2nd at 3k , 5k or anything above 2.5k (under which there is a drastic falloff of torque) you will have the same acceleration. Because in any given gear, the accel is directly proportional to the torque with a constant factor that changes with gearing. So my point is that if you don't fall below 2.5k, you are not bogging the engine. And since the car is designed not to fall below 2.5 k for shifts under normal acceleration, then it definately will not bog under redline acceleration.

fdl

ya...maybe bog isnt the best word. But because of the big jump in gear ratio between 1st and 2nd you kinda feel the car lug a bit when you hit 2nd gear,... (there is a noticable drop in acceleration that comes from switching to this taller (lower ratio) gear).

I get the whole 0-60 thing....they wanted 2 shifts...and a shorter 2nd gear would mean 3 shifts. ...

But .... here is what I am thinking. .... (bear with me)..

You want a high ratio in 1st for sure ..... but of course there has to be such a thing as TOO high. To me too high means where the ratio is so high you put so much torque to the ground that you get nothing but wheel spin at WOT. I dont think 1st in the TSX is quite that bad but there is an abundance of power in 1st..so much so that in stock trim stepping on the gas spins the tires quite a bit. So I am thinking there is some power to spare there...power that is being wasted .....some rooom to make 1st gear a bit taller. What this will mean is that yes...the slightly lower ratio will mean slightly less power...BUT there will be enough to spin the tires and accelerate just as fast as before (my hypothesis being that the power that was taken away would have just meant more tire spin anyways). Plus..because your gear is slightly taller..you can stay in it longer..meaning that although you have slightly less power than before ...you get that power for a longer period of time. And finally...you get to keep second where it is...stay at or near max hp..have a smooth 1-2 shift...and get to 60 in 2 shifts. Anyways....thats just what I am thinking...and I probably did a bad job of explaining it but in my head it makes sense

rzee

I did! I did! I even quoted myself there!

C'mon old man! I hope the Alzheimer's hasn't taken you completely.

larchmont

Did not! I hope the myopia hasn't taken you completely.


Splitting hairs, but IMO he put it in a different way and context and emphasis that made the difference.

Yeah, splitting hairs -- if we were putting this to a court or an arbitrator or something like that, you'd win.
But if we were putting it to a crowd of morons (like me), you'd lose.

rb1

More along the lines of fdl's earlier claims, consider the following (I may not have the numbers exact, but bear with me).

Suppose the TSX currently ends up at 4000 RPM after 1-2 shift at the redline. Now you change nothing else except the gearing so that after the 1-2 shift, you are at 5000 RPM.

The crank torque is the same, but now the "new" 2nd gear pulls (5/4) = 25% harder for the same crank input, resulting in 25% higher acceleration. The tradeoff is that now you have to shift earlier.

If you changed the gearing so that after 1-2 you are at 2500 RPM, then the engine pulls 2500/4000 = only 62.5% of what it did with the original gears, which some might perceive as "bogging".

calbear2k1

fdl... your first post was actually pretty clear, but the whole thread got sidetracked. In my experience, it seems like the 1-2 shift has been slightly problematic in both my 96 integra and 89 accord LXI. I think it just has to has do to with the fact that 1 to 2 has the biggest difference in torque to the wheels of any gear shift. The TSX is no doubt smoother than those two, but yes it seems easier to spin the wheels as well. A taller 1st would seemingly make the shift smoother, but I don't know about faster. Someone would have to write or use a computer simulation to find out. And I am assuming Honda would have done that already.

RB1, I am not saying the 1-2 shift should drop the car from the redline to 2.5k. That would mean 1st gear is really, really short, and when the car is not driven hard - say shift at 4.5k instead of 7k - then you really would be bogging.

illmeltxwithyou

Try punching the clutch in really fast keep your foot on the pedal about half way , make sure you shift rather quickly or your burn your clutch. It takes practice with this car. I do agree they should of geared it a bit different but what can you do? We're just gonna have to get used to it

drchaos

You can do an an energy balance.

drchaos

You can do an an energy balance.

E = 1/2 M v^2 + 1/2 I w^2

where M is mass of the car, v linear velocity,
I the total moment of inertia of wheels and
driveshaft and crankshaft, and 'w' (think omega)
as the rotational velocity of the wheels.

w is related to v by the tire radius if you dont' slip.

Then dE/dt = applied power P(t)

dE/dt = C v dv/dt for some constant C

so E = 1/2 C v^2

so v^2 is proportional to integral of applied power over time.

That means for all out drag racing you want to maximize the average power output, i.e. around peak horsepower.

Still, if you are driving a car normally, good engine torque (as opposed to at the wheel torque, which is proportional to acceleration) helps you go now without shifting, so there is an advantage for useful drivability.

as they say, "people buy horsepower, but they drive torque".

larchmont

Looks like Drchaos shut up everybody pretty good.

TinkySD

it's a bit of an overgenerlization without having any considerations for gearing of the vehicle in question. High rpm horsepower is only good if you have the gearing to take advantage of it.

I'm not disagreeing with drchaos in anyway just saying i dont' believe energy balance tells the whole story. Actually i find a good way to think about this whole arguement is a CVT transmission; which work buy infinitely varying the gear ratio to keep the engine at peak horsepower. In the case of the tsx a CVT equipped auto would be quicker than a 6mt in a straight line.

RogerPodacter

Good stuff here...:P

oh the memories from undergrad engineering...

TinkySD

I went the engineering path as well though I am about 5 years removed from my physics series

calbear2k1

DrChoas- That is some good stuff. If only I had used some equations, my posts would look less wishy washy. But nothing I said really contradicts you, I was just saying you need to know more than HP.

Tinky, that point about a CVT is pretty cool. It had never occured to me that a CVT would maximize acceleration. Maybe that was because I associate a CVT with the Subaru Justy, and that cannot possibly be technologically advanced can it?

Anyway, I probably should not have said justifying using horsepower to find acceleration was "not physically sound", but rather relying on horsepower only is not a complete way to think of it. That is because with gears, one is not always at the peak HP. The torque curve is especially important in 1st gear since you cannot shift down. So here area under curve is important.

I also remember a thread a while ago where the 325's 200hp was compared to the TSX's 200 HP. Someone (Buff-daddy?) suggested that Germans compared to Japanese tend to get more out of the same HP. Is is not the case that automakers don't know how to measure HP (except Mazda). If the cars are kept around peak, they will perform the same, which is what everyone believes about the TSX: Off the line the TSX is slower, but at highway speeds, there should not be any difference.

So I might not have been too precise in what I was saying before, but I think I might know what the hell I am talking about now. Good thread huh?

EDIT: 325 only has 184 hp....

fdl

Thank-you


And yes no doubt having lots of engine torque , especially down low makes great day to day driveability..but thats not really what we are discussing.

rb1

Of course not. I was just trying to provide a practical example with some real numbers as to how the wheel torque after a shift at the redline could be improved by closing up the gear spacing (to take advantage of hp per fdl), along with an example of how increasing the spacing makes it worse. No biggie.

dnb

As TinkySD says, this is an overgeneralization. You could just as well have gone down the path

E = 1/2 M v^2 + 1/2 I w^2

since M v^2 >> l w^2,

E ~ 1/2 M v^2

dE/dt = Mv dv/dt
dE/dt = F v

d^2E/dt^2 = F dv/dt

which would indicate that v^2 is proportional to the double integral of F (torque).

disclaimer: I wrote the above not as proof of anything but as an example of better living through handwaving.

DEVO

calbear2k1... thanks for the info.