SCCAForums.com

This is a little rant, and hopefully some good discussion will follow.

When we're debating the capability of certain cars, we often cite peak HP numbers. It seems to me that most of the time in autocross the real differentiator is torque. And not just peak torque, but usable torque across the powerband. (HP is really a measure of how many revs an engine can maintain high torque levels over)

It seems to me that especially in daily driving situations, a more helpful number would be total area under the torque curve. So, if you have cars (like hondas) that make great power, but don't do it until 6,000 rpm, that's often not real useful on the street. I'm guessing there's lots of math involved (that's calculus, right?) to figure this out, but I'm also guessing that dyno computers could do this pretty easily, and a rough number wouldn't be that difficult to calculate manually based on a measurement every 100 rpm or so. (or more fine-grained with a computer)

What do you guys think, does that make sense? How do we bring this up to the SAE?

---

DS #313 | the rolling couch of doom | La-Z-Boy Racing

murph1379:

This is a little rant, and hopefully some good discussion will follow.

When we're debating the capability of certain cars, we often cite peak HP numbers. It seems to me that most of the time in autocross the real differentiator is torque. And not just peak torque, but usable torque across the powerband. (HP is really a measure of how many revs an engine can maintain high torque levels over)

It seems to me that especially in daily driving situations, a more helpful number would be total area under the torque curve. So, if you have cars (like hondas) that make great power, but don't do it until 6,000 rpm, that's often not real useful on the street. I'm guessing there's lots of math involved (that's calculus, right?) to figure this out, but I'm also guessing that dyno computers could do this pretty easily, and a rough number wouldn't be that difficult to calculate manually based on a measurement every 100 rpm or so. (or more fine-grained with a computer)

What do you guys think, does that make sense? How do we bring this up to the SAE?

 

http://www.sae.org/standardsdev/ - link to how standards are developed. Need to find someone in industry to champion the cause.

---

Al Chan

Area under the HP curve is the important factor.

 

 If you are just comparing one car to another, you can use torque or HP, but what makes a car quicker or not is area under the HP curve of interest.   This correction is important because, if you have two hypothetical cars, and one makes an average of 200 ft-lb of torque from 2000 rpm to 4000 rpm and other makes 200 ft-lb of torque from 4000 rpm to 6000 rpm, and both are geared to use their respective rpm ranges, the one that makes the torque up higher makes a lot more power, and thus, all else equal, will be much much quicker.

---

WTB: CSP Spyder parts. What you got?

HP is all about shifting the power curve up. Torque is good for daily driving, merging, and autox, but ultimately HP wins for straight line runs like dragging because you can keep the car in the high RPM range. I am in agreeance with you on the toqure thing. Finding the area under the curve is very important in events where midrange power is important. I'm barely even a novice when it comes to autox so I'm not going off of experience in this arena but....Take for instance two cars were everything is equal, If one make 200 Peak HP between 6500-7000 RPM but makes 75-125 in the 4000-6000 range and pit it against a car making 150 HP in the 4000-6000 range, the area under the curve may make a difference. There are about one million articles on the internet about this so take it with a grain of salt.

So a prime example of this might be the AS S2000 vs the C4 Corvette?

I'll bet you can find a dyno curve for these cars somewhere on the internet.

 

---

Marcus
ESP 89
www.margravemotorsports.com

Combining what several posters have said, what you really want to see is area under the curve for "torque to the wheels" from about 25 MPH to 60 MPH and divide it by weight of the vehicle.  That is pure acceleration, and includes gearing, tire diameter and mass.

I've seen a couple of spreadsheets that Boxboy did when comparing various stock class cars and they are very informative.

--Andy 

as in for example (using '04 Cooper S, 2nd gear)

the Cooper S makes > 1,100 lb/ft of torque anywhere from 3,200 to 5,800 rpm, peaking at 1,191 lb/ft around 4k

running 15" wheels with /50 rubber gives a diameter of almost 24" exactly, meaning the radius of 12" translates lb/ft directly to the pavement

that means the car sees a peak thrust of ~1,200 lbs at the FW contact patches, against a mass of 2,540 lbs (w/ driver)

that suggests I should see GTech max acceleration numbers in the order of .47g

guess what!  that is correct (at least on the warm days where tire slip is not a factor)

I wish I could post the theoretical graph and the actual GTech data, as the pictures are more compelling than the numbers

Of course with the FWD car, weight transfer limits the available traction at the front axle so we poor folks do not often see longitudinal acceleration much in excess of .6g (unless running wrinkle wall slicks...)

am I following here?

---

Charlie Thompson
'04 JCW Cooper [STX]
NER Cannon Fodder

I agree, except that you have to make the final conversion from wheel torque to force at the road (edit: someone beat me to it).  Then it's F = mA.  Simple physics.

Power = Torque * RPM / 5252

^ this relationship always applies, you just have to be careful to use it at the right points (i.e. if you're measuring torque at the wheels, you have to use the wheel RPM, not the engine RPM).  What you find is that you can change the gearing, which will change the torque, but not the power.

The one place where power becomes important on an autocross course is when you have aero.  When you get into air drag, power is what determines how fast you can go.

BTW, anyone have any data on what kind of force an A-Mod applies at the wheels?  I'm doing a kind of feasability study to see if its possible to run a competitve electric car in A-Mod.  I think the motors are there, but the batteries needed may be a bit heavy.

---

Dave Heinig

07 GXP Z0K (Thanks Rob!)

A Mod 8th mile or 4th mile times might help.  Weight can be arbitrarily determined, right?  A good guess? 

The complexity is that don't A Mod cars use two or more gears on course?  Electric would be single curve, not multiple - smoother. 

But doesn't the principle still apply?  A literal calculation of the total area beneath the curve (whether simple or spiked due to gear changes) for the range of speed that is the target, would give a useful answer would it not?

Getting that number looks a bit tricky if gear changes are permitted, as the discussion at http://people.stfx.ca/bliengme/ExcelTips/AreaUnderCurve.htm suggests that if a good trendline can be established there are useful Excel tricks to assist the non-mathemeticians.

(Are you among this crew? http://www.evdl.org who are clearly going fast!)

Another approach would be posted data regarding forward acceleration from Max Q or other on board data acquisition.  That too is tricky because the data is probably limited by traction on the low side of the target range of speed, wouldn't you think?  In other words such cars can spin their wheels anytime the driver wants to up to a certain speed (when the aero kicks in, right?).  Here's a potentially useful thread I think: http://sccaforums.com/forums/thread/237679.aspx

I'll rummage around and see if I can find a dummies way to get total area under the curve in Excel given an array of (rpm, pounds thrust at contact patch) that doesn't care if the data is smooth or not. With that data I could plug in mass (car weight) and compute acceleration (g) arbitrarily.  Might be fun.

unless someone has done it already?

---

Charlie Thompson
'04 JCW Cooper [STX]
NER Cannon Fodder

most fast A-Mod cars run a CVT trans, so keep the 2 stroke in the power band and let the CVT change it's ratio, no shifting at all.

 For racing an electric, check out ProEV.com (from memory) he is racing an AWD electric Impreza.  Frying batteries at the Amps needed has been a big problem.  Same thing with the electric drag cars.  You should be able to solve the battery issue with a few hundred thousand dollars of batteries.  Unless the Ultra capacitor folks arn't just blowing smoke on their technology.

The numbers I came up with for the electric drive are 243 lb-ft of motor torque from 0-90 mph, falling off after that, with a max power of 268 hp @ 5780 RPM.  This would be geared down so the thrust would be 1077 lbs at the contact patches.  The max power of 268 hp would happen at 1360 wheel rpm, or 90 mph.  It requires 1600A at 144V, so 2 strings of 12 batteries.  The batteries are 12 lbs each, so the total battery weight is 288 lbs.  The motors are 28 lbs each, so 112 lbs of motors for a total drivetrain weight of 400 lbs.

---

Dave Heinig

07 GXP Z0K (Thanks Rob!)

A good way to comair power would be like this.

 SAE corr Wheel TQ per RPM across the effect range

ex.
3000 RPM 300RWTQ
4000 RPM 350RWTQ
5000 RPM 370RWTQ
6000 RPM 320RWTQ
net RWTQ 1340
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
div by 4 points
335 Avarage RW TQ

Some dynos like a hub driven dynapack will calc your avarage for you.

 

---

NotoriouSS Tad
96' Impala SS - Johhny 5 - 4300+lb - 12.8-1/4
NA 390RWTQ+ 431RWHP unlocked 91
suspension and motor build ar autocross oriented

You could go farther and multiply that by the gearing you would be in on course

then multiply that by your output gear ratio

then divide that by the weight of your ride

EX.
335 Avarage RW TQ
X - times
1.62 second gear
~~~~~~~~~~~~~~~~~~~~
542 TQ tranmission output to differential

then mulitply by the diffential or ouptut gearing

542 TQ tranmission output to differential
X - times
3.73
~~~~~~~~~~~~~~~~~~~~
2021 foot pounds to the wheel

then divide that into your weight

3400lb car
Div
2021 TQ
~~~~~~~~~~~~~~~~~~~~~~
1.68 pounds of car per pound of TQ in main course gearing!

---

NotoriouSS Tad
96' Impala SS - Johhny 5 - 4300+lb - 12.8-1/4
NA 390RWTQ+ 431RWHP unlocked 91
suspension and motor build ar autocross oriented

NotoriouSS:

1.68 pounds of car per pound of TQ in main course gearing!

You forgot the last step.  You need to measure force, not torque.  Assume you're on 275/35R15 Hoosiers, which have a diameter of 23".

 2021 ft-lbs / (11.5/12) feet = 2109 lbs of thrust

Then convert to SI and F=mA

9381 N = 1,542 kg * A

A = 6.08 m/s^2 = 0.62 g average acceleration.

---

Dave Heinig

07 GXP Z0K (Thanks Rob!)

DaveH:

NotoriouSS:

1.68 pounds of car per pound of TQ in main course gearing!

You forgot the last step.  You need to measure force, not torque.  Assume you're on 275/35R15 Hoosiers, which have a diameter of 23".

 2021 ft-lbs / (11.5/12) feet = 2109 lbs of thrust

Then convert to SI and F=mA

9381 N = 1,542 kg * A

A = 6.08 m/s^2 = 0.62 g average acceleration.

most educational... 

I am quite curious what your projections are for this 400 lb drivetrain, with the rest of the car wrapped around it! .

Your earlier post seems to confirm my suspicion that electric drive can produce flat torque across the target range, and that just blows my mind.  a car that would pull (what's the number?  )gs regardless of speed!  no powerband?  crikey!  am I paying attention here?

---

Charlie Thompson
'04 JCW Cooper [STX]
NER Cannon Fodder

cmt52663:

Your earlier post seems to confirm my suspicion that electric drive can produce flat torque across the target range, and that just blows my mind.  a car that would pull (what's the number?  )gs regardless of speed!  no powerband?  crikey!  am I paying attention here?

 As I understand it, that's correct, and it's one of the exciting aspects of electric racecars. Also exciting:

- lower CG

- better weight distribution

- better traction control (a motor at each wheel makes it real easy for a computer to distribute power very precisely)
 

---

DS #313 | the rolling couch of doom | La-Z-Boy Racing

DaveH:

NotoriouSS:

1.68 pounds of car per pound of TQ in main course gearing!

You forgot the last step.  You need to measure force, not torque.  Assume you're on 275/35R15 Hoosiers, which have a diameter of 23".

 2021 ft-lbs / (11.5/12) feet = 2109 lbs of thrust

Then convert to SI and F=mA

9381 N = 1,542 kg * A

A = 6.08 m/s^2 = 0.62 g average acceleration.

Dave could you plz break that down again al little simpler for my pee sized brain??

how did you go from tire size to  (11.5/12) feet?

Could you define A =s?

thanks

 

---

NotoriouSS Tad
96' Impala SS - Johhny 5 - 4300+lb - 12.8-1/4
NA 390RWTQ+ 431RWHP unlocked 91
suspension and motor build ar autocross oriented

Ok, you've calculated torque at the wheels.  Torque is a measure of moment, the unit is ft-lb.  To translate that to the thrust delivered, you divide by the length of the moment arm (the radius of the tire).  With a 23" diameter tire, the moment arm is 11.5", or 11.5/12 feet.

This gets you 2109 lbs (force) of thrust delivered to the pavement.  To make the numbers easy, you convert this to Newtons, the SI unit.  This gives you 9381 Newtons of force.

From physics, F = mA.  You're applying a force to a mass, it will result in acceleration.  You have to make sure everything is in SI units to make it work without a correction constant, so:

9381 N = 1542 kg * A.

This gives you an acceleration of 6.08 m/s^2.  1g = 9.8 m/s^2.  Convert the units back and you get an acceleration of 0.62 g.

 EDIT: I'm a retard.  F = mA works in lbs (force), lbs (mass), and g's.  2109 lbf = 3400 lbm * 0.62g

---

Dave Heinig

07 GXP Z0K (Thanks Rob!)

cmt52663:

I am quite curious what your projections are for this 400 lb drivetrain, with the rest of the car wrapped around it! .

This doesn't show enough batteries.  The reduction gears are also missing.  I just threw that together to see how to package everything.  It would have to get redone once the suspension geometry is calculated anyway.

cmt52663:

Your earlier post seems to confirm my suspicion that electric drive can produce flat torque across the target range, and that just blows my mind.  a car that would pull (what's the number?  )gs regardless of speed!  no powerband?  crikey!  am I paying attention here?

Actually, electric drive produces a linearly decreasing torque curve.  They generate maximum torque at zero RPM and zero torque at some maximum speed.  The motor I'm looking at will generate 5495 lb-ft of torque at zero RPM and 144V.  It will generate a maximum power of 1580 hp at 3000 RPM.  The problem is, that max torque figure is at 9000A.  The max power figure is 4500A.  If you ever push it that high the only thing you'll do is let all the smoke out.

In reality, the motor controller has to limit amperage to an acceptable value, 400A in this case.  Torque is directly related to amps in a DC motor.  The torque stays flat until you reach the motor's natural curve then it linearly decreases to zero.  At 144V, that point is at 5780 RPM.  This is also the power peak.  Power falls off rapidly after this.

Assuming a weight of 900 lbs (class minimum) the thrust delivered is 1075 lbs.

4781 N = 408 kg * A

A = 11.71 m/s^2 = 1.2 g

Now for the downside.  The motors will fry if you flow 400A through them for more than about 7 or 8 seconds.  The batteries have to supply 800A per string.  They've been shown to output 900A for 10 seconds.  You have to run regenerative braking to make it through a run.  You have to have a support truck with a generator to run the charger between runs.  You'll probably have to replace the brushes in the motors after every weekend.  And if you break stuff, it's really expensive.

---

Dave Heinig

07 GXP Z0K (Thanks Rob!)