How will you go faster?

General engine tech -- Drag Racing to Circle Track

Moderator: Team

User avatar
Stan Weiss
Vendor
Posts: 4801
Joined: Tue Feb 20, 2007 1:31 pm
Location: Philadelphia, PA
Contact:

Re: How will you go faster?

Post by Stan Weiss »

Ken_Parkman wrote: Wed Nov 23, 2022 12:15 pm
David Redszus wrote: Tue Nov 22, 2022 11:46 pm
David your own simulation data proves these statements incorrect.
It does nothing of the sort. What do you feel is incorrect? Could you please provide a concrete example with real numbers?
How is it possible your better calculated performance is with the lower of your chosen input parameters if there is not something more to be learned?
There is a great deal more to be learned; we have just begun. The changes in calculated performance were never intended to optimize performance; only that as various parameters change, so does performance.

The shape of the torque curve and the gear sets used can produce substantial differences in performance.
A short first gear will produce a very quick 60ft time and high G force.
But, if the remaining gears are incorrect, the total 1/4 mile elapsed time will be slow.

Given a specific torque curve, it is important to find the right combination of rear end ratio, gear ratios, and tire diameter.
The concrete example was several pages back where a simple corrected input gave a 17% improvement in acceleration. Pretty good bang for the buck. The second concrete example is your own simulation where the use of the higher power portion of the torque curve (even with gear inputs are not realistic) gave a much better quarter mile time.

The question what is wrong with the chart (if you are serious) is it cannot improve real performance as we cannot optimize the driveline with that data. As you used the torque input there is not a crossover for different gears and we have no idea how optimize even the shift point. You cannot begin to optimize performance when you don't even know when to shift. Replot it using a power input against rpm and then you can optimize shift points and then have data to work with for driveline optimization.
Ken,
It doesn't what units you use to plot the curves they will not cross.

Why is that? Because most dyno pulls stop just after peak HP. One would need data at a higher RPM's before the curves will cross.

Stan
Stan Weiss/World Wide Enterprises
Offering Performance Software Since 1987
http://www.magneticlynx.com/carfor/carfor.htm
David Vizard & Stan Weiss' IOP / Flow / Induction Optimization Software
http://www.magneticlynx.com/DV
Ken_Parkman
Expert
Expert
Posts: 660
Joined: Sun Jan 02, 2005 11:30 pm
Location: Ontario, Canada

Re: How will you go faster?

Post by Ken_Parkman »

This is actually the point, and how you optimize. That input shows a rpm drops of 21, 33, and 38% and the point is the curves do not cross. So it is not optimized and cannot be with this data.

You should run the engine high enough to understand how it drops off, and knowing the characteristics you can also then begin to optimize torque converter, ratios for rpm drop, and clutch if applicable. Then pick reasonable gears and shift points, factored for the rates of acceleration.

I always have an idea about the power characteristics above peak, and use that to optimize drivetrain, perhaps limited by a mechanical redline. The better the converter or the smaller the ratio drop the less you have to overrev the engine to go fast. This is why a CVT is perfect, but with a good converter you can get closer.

I have gone mid 10's in my bracket car with a taxicab junk parts engine; maybe 400 hp. The torque converter was worth more than the engine and allowed the engine operated in an 600 rpm band. Very much the curves crossed so you also have to include or at least understand the converter in your simulation. The actual engine torque peak rpm should never be seen in performance use unless you have drivetrain limitations, and if you can it is much preferred to fix the drivetrain.
Last edited by Ken_Parkman on Wed Nov 23, 2022 2:00 pm, edited 1 time in total.
skinny z
HotPass
HotPass
Posts: 2642
Joined: Tue Mar 11, 2008 8:42 am
Location: AB. CA.

Re: How will you go faster?

Post by skinny z »

Ken_Parkman wrote: Wed Nov 23, 2022 1:26 pm. The actual engine torque peak rpm should never be seen in performance use unless you have drivetrain limitations, and if you can it is much preferred to fix the drivetrain.
Do you mean to say that if my engine makes peak torque at 4000 RPM, and my converter flash stalls to 4000 RPM (as observed through data logs) that this not a good thing?
I was under the impression that you want the converter to "hit" at peak torque.
Thanks and my apologies if I've derailed the subject at hand but that comment caught my attention.
Kevin
Ken_Parkman
Expert
Expert
Posts: 660
Joined: Sun Jan 02, 2005 11:30 pm
Location: Ontario, Canada

Re: How will you go faster?

Post by Ken_Parkman »

BobbyB wrote: Wed Nov 23, 2022 12:38 pm
it looks to me like you should shift at 55 mph & 92mph to maximize acceleration ... Am I missing something?
If you shift at 55 mph the acceleration takes a sudden drop from .76 g to .58, and the 3-4 shift drops .41-.33. If we had a complete power curve we should be able to do a lot better than that.
Ken_Parkman
Expert
Expert
Posts: 660
Joined: Sun Jan 02, 2005 11:30 pm
Location: Ontario, Canada

Re: How will you go faster?

Post by Ken_Parkman »

skinny z wrote: Wed Nov 23, 2022 1:51 pm
Ken_Parkman wrote: Wed Nov 23, 2022 1:26 pm. The actual engine torque peak rpm should never be seen in performance use unless you have drivetrain limitations, and if you can it is much preferred to fix the drivetrain.
Do you mean to say that if my engine makes peak torque at 4000 RPM, and my converter flash stalls to 4000 RPM (as observed through data logs) that this not a good thing?
I was under the impression that you want the converter to "hit" at peak torque.
Thanks and my apologies if I've derailed the subject at hand but that comment caught my attention.
Correct. You are throwing away some performance - depending. But of course an 8" converter may not be the most drivable thing if it is a street strip application.

The above bracket car example was a mild hydraulic cam engine with a 4400 rpm torque peak, and it nosed over badly at 6000 rpm. The stall was 5400 and I shifted at 6000, which means it actually operated on a 600 rpm band.
skinny z
HotPass
HotPass
Posts: 2642
Joined: Tue Mar 11, 2008 8:42 am
Location: AB. CA.

Re: How will you go faster?

Post by skinny z »

Ken_Parkman wrote: Wed Nov 23, 2022 2:00 pm
skinny z wrote: Wed Nov 23, 2022 1:51 pm
Ken_Parkman wrote: Wed Nov 23, 2022 1:26 pm. The actual engine torque peak rpm should never be seen in performance use unless you have drivetrain limitations, and if you can it is much preferred to fix the drivetrain.
Do you mean to say that if my engine makes peak torque at 4000 RPM, and my converter flash stalls to 4000 RPM (as observed through data logs) that this not a good thing?
I was under the impression that you want the converter to "hit" at peak torque.
Thanks and my apologies if I've derailed the subject at hand but that comment caught my attention.
Correct. You are throwing away some performance - depending. But of course an 8" converter may not be the most drivable thing if it is a street strip application.

The above bracket car example was a mild hydraulic cam engine with a 4400 rpm torque peak, and it nosed over badly at 6000 rpm. The stall was 5400 and I shifted at 6000, which means it actually operated on a 600 rpm band.
Thanks for that.
Kevin
BobbyB
Pro
Pro
Posts: 489
Joined: Mon Apr 15, 2013 1:35 pm
Location:

Re: How will you go faster?

Post by BobbyB »

Ken_Parkman wrote: Wed Nov 23, 2022 1:55 pm
BobbyB wrote: Wed Nov 23, 2022 12:38 pm
it looks to me like you should shift at 55 mph & 92mph to maximize acceleration ... Am I missing something?
If you shift at 55 mph the acceleration takes a sudden drop from .76 g to .58, and the 3-4 shift drops .41-.33. If we had a complete power curve we should be able to do a lot better than that.
If I understand the legend correctly, 1st gear is 2.5 to 1, axle is 3.7 to 1, tire is 81.68 inches per tire revolution...

These number give 6577.4 engine RPM at 55 mph.

Ken, are you saying with more data, you could show that revving higher could give better acceleration performance?
David Redszus
Guru
Guru
Posts: 9633
Joined: Tue Nov 27, 2007 9:27 am
Location: Chicago
Contact:

Re: How will you go faster?

Post by David Redszus »

The question what is wrong with the chart (if you are serious) is it cannot improve real performance as we cannot optimize the driveline with that data. As you used the torque input there is not a crossover for different gears and we have no idea how optimize even the shift point.
The torque curve was furnished by Stan and used real world gear ratios, rear end ratio, and tire size. They are not my choice and are used as a real starting point for analysis.

You are right, there is no crossover of the acceleration curves. That is the problem. The shift points from this data are easily seen: run first and second to redline (6700) then shift and drop to 4000 and 4500 respectively.
The acceleration will drop from 0.762G to 0.545G going from 1st to 2nd, and will drop from 0.40G to 0.33G going from 2nd to 3rd. If the curves could intersect, we would not suffer an acceleration loss.
You cannot begin to optimize performance when you don't even know when to shift.
Knowing when to shift is pretty obvious using this data. If we are locked into this torque curve, we might consider alternative gearing. In fact we should consider alternative gearing.
Replot it using a power input against rpm and then you can optimize shift points and then have data to work with for driveline optimization.
There is no need to do so. The acceleration curve always tracks to the shape of the torque curve, always.

The graph that plots G forces against MPH is called a "Cascade plot" and has been the standard used by the automotive industry for over 70 years. It never considers power as an input since the torque curve data does that already.

I can create a plot of G forces against rpm, if you like. Send my your email and I'll send it to you. Which you can post if you like.
Ken_Parkman
Expert
Expert
Posts: 660
Joined: Sun Jan 02, 2005 11:30 pm
Location: Ontario, Canada

Re: How will you go faster?

Post by Ken_Parkman »

BobbyB wrote: Wed Nov 23, 2022 2:17 pm
Ken_Parkman wrote: Wed Nov 23, 2022 1:55 pm
BobbyB wrote: Wed Nov 23, 2022 12:38 pm
it looks to me like you should shift at 55 mph & 92mph to maximize acceleration ... Am I missing something?
If you shift at 55 mph the acceleration takes a sudden drop from .76 g to .58, and the 3-4 shift drops .41-.33. If we had a complete power curve we should be able to do a lot better than that.
If I understand the legend correctly, 1st gear is 2.5 to 1, axle is 3.7 to 1, tire is 81.68 inches per tire revolution...

These number give 6577.4 engine RPM at 55 mph.

Ken, are you saying with more data, you could show that revving higher could give better acceleration performance?
You may have me there, I assumed the gear ratio's were as posted several pages ago and did not read the legend. Sorry about that.

But it still does not change the conclusion, but may change the optimization. This sounds like a mechanical rpm limit situation. Then this is a 3 speed automatic, and the optimization becomes in the converter. Substitute in a 8" converter with a 5400 stall, but it will be challenging to get a torque ratio curve for the input/output rpm.
Ken_Parkman
Expert
Expert
Posts: 660
Joined: Sun Jan 02, 2005 11:30 pm
Location: Ontario, Canada

Re: How will you go faster?

Post by Ken_Parkman »

David Redszus wrote: Wed Nov 23, 2022 2:27 pm
The question what is wrong with the chart (if you are serious) is it cannot improve real performance as we cannot optimize the driveline with that data. As you used the torque input there is not a crossover for different gears and we have no idea how optimize even the shift point.
The torque curve was furnished by Stan and used real world gear ratios, rear end ratio, and tire size. They are not my choice and are used as a real starting point for analysis.

You are right, there is no crossover of the acceleration curves. That is the problem. The shift points from this data are easily seen: run first and second to redline (6700) then shift and drop to 4000 and 4500 respectively.
The acceleration will drop from 0.762G to 0.545G going from 1st to 2nd, and will drop from 0.40G to 0.33G going from 2nd to 3rd. If the curves could intersect, we would not suffer an acceleration loss.
You cannot begin to optimize performance when you don't even know when to shift.
Knowing when to shift is pretty obvious using this data. If we are locked into this torque curve, we might consider alternative gearing. In fact we should consider alternative gearing.
Replot it using a power input against rpm and then you can optimize shift points and then have data to work with for driveline optimization.
There is no need to do so. The acceleration curve always tracks to the shape of the torque curve, always.

The graph that plots G forces against MPH is called a "Cascade plot" and has been the standard used by the automotive industry for over 70 years. It never considers power as an input since the torque curve data does that already.

I can create a plot of G forces against rpm, if you like. Send my your email and I'll send it to you. Which you can post if you like.
Ok David, I made some assumptions on your data, and that was wrong. Yes I am sorry. But your data also proves the maximum acceleration of the car does not track with the engine torque, as does your earlier simulation. Substiute in the above mentioned converter and replot and see how it changes the performance.
digger
Guru
Guru
Posts: 2722
Joined: Sun Jan 27, 2008 6:39 am
Location:

Re: How will you go faster?

Post by digger »

Rick! wrote: Wed Nov 23, 2022 7:04 am
digger wrote: Tue Nov 22, 2022 4:12 pm
I know you don’t like thinking outside the box with equations but

Tractive effort force = wheel power / velocity

So 400 hp (298,000 watts) at a speed of 60 mph (28.7m/s) will generate a tractive effort force = 10,383N (22,800lb)
Last time I checked, 1lbf = 4.448N, so your calculation should be closer to 2334lbf.
And now you have others haphazardly throwing around lbf and lbm. :) I so much prefer SI units over the British Engineering system...
I also prefer to calculate atmospheric parameters from the inputs of temp, humidity and absolute pressure (barometer.)
Carry on, I'm entertained by the snowmobile example since as it was a passion and vocation of mine for a almost two decades. When you get to the part about the track driver/track interface and track hysteresis as knobs to turn to go faster, I might come out to play.
Yes I converted the N as if it were kg…
digger
Guru
Guru
Posts: 2722
Joined: Sun Jan 27, 2008 6:39 am
Location:

Re: How will you go faster?

Post by digger »

David Redszus wrote: Wed Nov 23, 2022 11:27 am
So at any speed maximising the force at that speed comes from gearing it to running at maximum power.
Yes, but it also runs at constant torque as well as constant power.
Therefore with cvt maximum force across speed range comes from running at maximum power continuously and infinitely variably change gearing. This provides maximum torque to the wheels or linear force to sled tracks
Since engine power and torque are constant, the remaining variable is the clutch.
It changes clutch ratio which multiplies wheel torque, but the power does not change. The trick with a CVT is to tune the shape of the clutch shift curve for acceleration and/or top end.
A load cell dyno'd 800 hp engine in a small lightweight car with skinny tyres never produces 800hp in the car.
Excellent observation. No matter how much torque or horsepower is seen on a dyno, the actual horsepower delivered to the track will depend on grip and traction. Wheelspin indicates the upper limit has been reached; more power does no good whatsoever.
I also prefer to calculate atmospheric parameters from the inputs of temp, humidity and absolute pressure (barometer.)
Using those atmospheric parameters we can calculate absolute air density at any altitude or temperature; in fact the data logger can compute and display absolute air density.
We also predict the percent power loss or gain, the fuel enrichment change needed and air drag.
Carry on, I'm entertained by the snowmobile example since as it was a passion and vocation of mine for a almost two decades. When you get to the part about the track driver/track interface and track hysteresis as knobs to turn to go faster, I might come out to play.
Snowmobile racing now consists of drag, oval track, snocross and cross country racing using both two stroke and four stroke engines. The sleds are computer tuned, fuel injected, and some are turbocharged.

Drag sleds now produce almost 1000 horsepower which cannot be put down to the ice. We use on-board computer programs to control boost so as to maintain traction. Belt slippage and track to ground slippage is measured as is acceleration and speed for every foot of track. But there is always more to learn.
It’s running at constant engine torque but not at its maximum engine torque that’s why using power is a simpler method as it just needs to run at the maximum power and the rest comes out in the wash
User avatar
Stan Weiss
Vendor
Posts: 4801
Joined: Tue Feb 20, 2007 1:31 pm
Location: Philadelphia, PA
Contact:

Re: How will you go faster?

Post by Stan Weiss »

Using data from one of Kevin / Skinny Z simulations. One two shifted @ 7500 RPM and two three shifted @ 7300 RPM.

Stan
G_Force_vs_Time.gif
G_Force_vs_Time_HP_Torq.gif
You do not have the required permissions to view the files attached to this post.
Stan Weiss/World Wide Enterprises
Offering Performance Software Since 1987
http://www.magneticlynx.com/carfor/carfor.htm
David Vizard & Stan Weiss' IOP / Flow / Induction Optimization Software
http://www.magneticlynx.com/DV
skinny z
HotPass
HotPass
Posts: 2642
Joined: Tue Mar 11, 2008 8:42 am
Location: AB. CA.

Re: How will you go faster?

Post by skinny z »

So many sims Stan I've lost track.
Was that one the 470HP@6000 as an answer in another thread?
Kevin
User avatar
Stan Weiss
Vendor
Posts: 4801
Joined: Tue Feb 20, 2007 1:31 pm
Location: Philadelphia, PA
Contact:

Re: How will you go faster?

Post by Stan Weiss »

skinny z wrote: Wed Nov 23, 2022 4:14 pm So many sims Stan I've lost track.
Was that one the 470HP@6000 as an answer in another thread?
Kevin,
I believe that is the one I used see below. Note these graphs are for illustration purposes only as there are factor which can / will change these curve somewhat. One would need information about the car to get closer, but it still comes down to what runs best at the track.

Stan
G_Force_vs_Time_Input.gif
You do not have the required permissions to view the files attached to this post.
Stan Weiss/World Wide Enterprises
Offering Performance Software Since 1987
http://www.magneticlynx.com/carfor/carfor.htm
David Vizard & Stan Weiss' IOP / Flow / Induction Optimization Software
http://www.magneticlynx.com/DV
Post Reply