The old HP vs TQ debate.

I'm involved in a discussion over on another forum. I wanted some sanity. I might be wrong, but I don't think I am. Take a peek at this photo:

This is a screenshot of a dyno sim for an engine I'm building. My question on the other forum was where should I shift for max acceleration.

My assumption was to shift around 6500-6700 so that the RPM drop puts me back just before the HP peak. In that way, I would spend the most amount of time in an RPM range that uses the most area under the HP curve. Well, then someone got into the "HP means nothing, its all torque" debate and it spiraled from there. I have one guy (who I respect for his knowledge over the years) who swears that HP doesn't matter, its all about shifting to maximize torque, which suggests that he wants me to shift at 5000 rpms or so. He won't answer me on why though.

Anyway... Here is my question for you guys to verify what I'm thinking. In that graph above, TQ is at 500 lb-ft at two RPM points; 3500 and 5000 rpms. (indicated by the yellow and blue lines) Let's say I'm in 4th gear (manual transmission) at 2000 rpms and mash the pedal all the way until 7000 rpms. At which of those two RPM points will the car be accelerating faster; the 3500 or the 5000, or neither? At what RPM point (2000-7000) will the car be accelerating at the fastest rate?

The only way to end this debate is to go to a drag strip and race. Try each method and then 'Speak with Data".

Cars are different, only way to know is to try several different ways. For example, people think you should shift the SHO's before 7000rpm due to graphs like you posted(peak power is 6700ish stock), but shifting at the rev limiter(7400) results in half second quicker ET's and usually 4-5mph higher.

HP is torque with an RPM multiplier. Higher HP with the same torque means you carried your torque farther up the RPM range. The benefit is you can stay in a lower gear longer and use that torque multiplication to help acceleration. For your question you would need to calculate what shift points get you the most torque to the ground the longest.

My interpretation, For the same gear acceleration:
3500=5000. ~4350 Is maximum. Calculating shift points requires knowing ratios.

Disclaimer: I am pretty sure I am correct but if someone wants to prove me wrong I am willing to learn.

MrJoshua wrote: My interpretation, For the same gear acceleration: 3500=5000. ~4350 Is maximum. Calculating shift points requires knowing ratios. Disclaimer: I am pretty sure I am correct but if someone wants to prove me wrong I am willing to learn.

For the sake of this argument, RPM drops are irrelevant. In the actual car, yes, but I'm speaking simply mathematically. But, RPM drops from 6500 in the 1-2, 2-3, and 3-4 shifts are about 1700 RPMs.

While I don't disagree with your assessment of this, why does anyone build high HP engines and rev them to 6700 rpms? By that rationale, I should be shifting at about 5000, and I know that is not correct.

iadr wrote: but you need to multiply the "curved" torque line, by the gear ratio multiplier, and do so for each gear. Doing so would take considerable patience with a pencil and graph paper, "raising" each line the correct amount...then hold one over the other, up to the light, but yes, it's the correct application of the physics. And it only has to be done once. I suppose you could print the curve multiple times and move it on the page, up by a factor, or something similar with software. FWIW, that's an unrealistically wide torque curve the software produced.

Well, here is this chart based on 1-2 and 2-3 shift ratio changes of this dyno sim.

what I'm saying is... the multiplied torques for each gear will still peak at the same RPMs, which suggests again that 5000 or 5500.

iadr wrote: In reply to curtis73: so 7100-ish in first and 6750-ish in 2nd

How do you figure? that is way above where the multiplied torque peaks.

I really am not understanding this whole thing. Everyone is saying to maximize torque, but then suggesting shift points that are way above maximizing torque.

I'm kind of following your diagonals thing, but not really.

So you're suggesting that I shift up when the RPM drop will put me to peak multiplied torque in the next gear?

How does that maximize torque if torque is dropping off immediately? Do you see what I'm asking? Everyone is saying maximize torque. Maximizing torque means keeping it between 3500-5000. That puts the most area under the curve.

iadr wrote: In reply to curtis73: You have to account for the gear ratio improving driveshaft torque.

That is correct, but gear ratio improving torque shifts the curve straight up, not to the right. Doesn't matter what gear you are in, the most torque coming out of the transmission will be happening between 3500-5000.

Mosey over to neons.org and look in the FAQ. There is an excellent essay on shift points by Chrysler engineer and road racer Erich Heuschele.

perhaps try thinking of it in terms of wheel speed. For every mile per hour you want to be in the gear that puts the most torque to the wheels. Create a table with speeds 0 to 100 miles per hour and indicate wheel torque at a given mph in each gear in columns by gear. For every speed you are simply selecting the gear with the highest wheel torque.

TQ man is wrong.

Shift points are dependent on POWER to the ground. Torque is simply twisting force, power is force applied over time. That makes the difference. A car with 300 ft-lbs and 300 HP will be slower than a car with 300 ft-lbs and 400 HP, all else being equal. To find shift points, you need power to the ground in each gear. In absence of measuring directly, you can approximate. Multiply the power curve by each gear ratio, this will "stretch" the power curve upwards for anything shorter than 1:1 and "flatten" it for overdrive. Then place them on a graph with road speed as the X-axis, and shift them so that the beginning rpm and end rpm of each in-gear power curve matches what the road speed in that gear at that rpm would be. You'll end up with something like the following graph (divide curves by 15% or so to fudge factor powertrain losses if you like) :

Theoretical ideal shift point occurs at the road speed where the two gear graphs cross (or redline if they don't cross), and knowing the gear and road speed can get you the rpm. Power graph + gear chart + math = WIN.

[URL=http://s140.photobucket.com/user/95ti/media/Shiftpoints.png.html][/URL]

Green points show when torque from lower gear intersects with torque from next gear, this is torque measured at the wheel taken from the data you provided. In some cases you can see the lines intersect and the next gear will provide more torque. With your gearing this does not happen so you need to avoid shifting until redline to get maximum acceleration.

iadr wrote: In reply to curtis73: Heck no! The output torque is input torque exactly multiplied by the gear ratio. That's what gear trains do, why we have them...! When I say up and to the right, I am talking in reference to the typed columns of numbers, as shown on the page. Not anything to do with the curve, not anything to do with revving up.

We're on the same page... I knew your diagonals were for the chart, not the graph. I was simply saying this:

I made lines that represent the output torque of the transmission in second and third gear. The greatest amount of torque would be sent to the wheels at the same RPM regardless of the gear. If I am trying to spend the most amount of time delivering the greatest torque to the driveshaft, how is that accomplished by going way above the peak? Looking at the chart and the graph, the biggest "meat" of the torque is 3500-5000.

Don't get me wrong... I agree with your shift point suggestions, I'm just trying to put the two together. How do those shift points maximize the AVERAGE overall torque making it out the back of the transmission?

In reply to jsquared:

You beat me to it, my custom graph slowed me down!

jsquared wrote: TQ man is wrong. Shift points are dependent on POWER to the ground. Torque is simply twisting force, power is force applied *over time*.

See... this is what I argued.

In my original post, I asked if the car would accelerate faster at 500lbft at 3500 or 500lbft at 5000. I think (and I could be wrong) that acceleration is greater at 5000 because of the higher HP. More work can be done because HP is higher.

I was about to create an excel/graph to educate those in the wrong... but jsquared already posted one and is 100% correct.

I'm ashamed that this is still a debate that pops up on this forum :(

iadr wrote: Curtis, we're both both posting fast- I've done multiple edits on each of my posts. Not changing the meaning, but cleaning up wording and formatting errors. I don't have a photo host handy. But on your pic, you want to rev it until the rightmost tip of the blue line (2nd gear) drops below the hump of the yellow curve (3rd).

I understand your description. And I do have a firm grasp on all the individual topics involved with this discussion, I'm just having trouble putting it together.

On the one hand, your (correct) theory of where to shift doesn't fit into my brain because I am constantly revving above the torque peak where torque is falling fast. I don't understand how that maximizes the torque that makes it to the driveshaft. In order to maximize torque to the driveshaft, it seems like RPMs should be toggled equally before and after the torque peak. Am I making sense?

If acceleration is the product of getting the most torque to the driveshaft for the greatest amount of time, how does that equate to utilizing RPMs that are way above peak torque?

Like I said, I agree with everything you're saying... just can't make it fit into my brain. I'm going to take a break and re-read everything you've all said. Then I might make a graph to demonstrate what I'm trying to say.

ProDarwin wrote: I was about to create an excel/graph to educate those in the wrong... but jsquared already posted one and is 100% correct. I'm ashamed that this is still a debate that pops up on this forum :(

I'm not ashamed that I'm trying to understand a concept. I understand each individual thing, just having trouble putting it together.

curtis73 wrote:

ProDarwin wrote: I was about to create an excel/graph to educate those in the wrong... but jsquared already posted one and is 100% correct. I'm ashamed that this is still a debate that pops up on this forum :(

I'm not ashamed that I'm trying to understand a concept. I understand each individual thing, just having trouble putting it together.

Sorry, worded that poorly... The part that makes me ashamed is the debate aspect of this, and the number of people that post spreading misinformation.

There is nothing wrong with trying to learn more/understand the concept.

My IT Neon has a 7400 rpm chip. The torque starts to drop off fairly dramatically above 7000, but if I shift at 7000, people pass me, because 7400 in third is putting more torque to the axles than 5500 in 4th.

Also, I came here to say that horsepower is how fast you hit the wall, torque is how far you move it.

curtis73 wrote: In my original post, I asked if the car would accelerate faster at 500lbft at 3500 or 500lbft at 5000. I think (and I could be wrong) that acceleration is greater at 5000 because of the higher HP. More work can be done because HP is higher.

this....

following your power/gear multipliers... you should be shifting somewhere between 6000, and 6500, the rpm drop you stated lands squarely in the peak power of the following gear.

ProDarwin wrote:

curtis73 wrote:

ProDarwin wrote: I was about to create an excel/graph to educate those in the wrong... but jsquared already posted one and is 100% correct. I'm ashamed that this is still a debate that pops up on this forum :(

I'm not ashamed that I'm trying to understand a concept. I understand each individual thing, just having trouble putting it together.

Sorry, worded that poorly... The part that makes me ashamed is the debate aspect of this, and the number of people that post spreading misinformation. There is *nothing* wrong with trying to learn more/understand the concept.

Ah... understood.