what is rpm band of best performance?

[ketan]

i wanted to know that wen we change gears wat shud be engine's rpm at a particular gear and how to go about in selecting proper gear ratios for the gearbox.this wud be our first time in this competition so any help wud be of great use.thanxz

[ketan]

i wanted to know that wen we change gears wat shud be engine's rpm at a particular gear and how to go about in selecting proper gear ratios for the gearbox.this wud be our first time in this competition so any help wud be of great use.thanxz

[Chris Allbee]

We had a judge in detroit tell us that a real driver can make use of any size powerband and that aiming for a lot of power (or torque) below 10500RPM was a waste of time. :P

[John Valerio]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Chris Allbee:
aiming for a lot of power (or torque) below 10500RPM was a waste of time. :P </div></BLOCKQUOTE>

i would seriously beg to differ considering last years FSAE-East endurance course. With corners 4-7 (or 3-6, i can't remember, but if you drove enduro you know which ones i'm talking about) and the last corner before the pit straight being as slow as they were, if you had no juice below 10500 RPM you would have been royally screwed.

[Homemade WRX]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by John Valerio:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Chris Allbee:
aiming for a lot of power (or torque) below 10500RPM was a waste of time. :P </div></BLOCKQUOTE>

i would seriously beg to differ considering last years FSAE-East endurance course. With corners 4-7 (or 3-6, i can't remember, but if you drove enduro you know which ones i'm talking about) and the last corner before the pit straight being as slow as they were, if you had no juice below 10500 RPM you would have been royally screwed. </div></BLOCKQUOTE>
so were you beneath 10.5k in first gear for these corners?...I think that is point the judge is getting at...although a broader powerband makes it easier to drive and less demanding on the driver.

[John Valerio]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">
so were you beneath 10.5k in first gear for these corners? </div></BLOCKQUOTE>

uhh....yeah. the minimum OD for hairpins is 9 meters and i'm pretty sure at least the last turn on the enduro as tight as that.

[B Hise]

I know im digging many long and much discussed threads saying this, but those courses are junk.

With that said, I would work on a wide easy to drive powerband. Yes, you can design your car with a short very high powerband, choose your rear end ratio to maximize that, have a shifter kart champ drive for you and do well dynamically. But the fact of the matter is most people in FSAE arent very good drivers, and you should take that into consideration when you design your powertrain. Practicality, especially for a first year team, should have alot of weight in your decisions.

-Bryan

[darienphoenix]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by B Hise:
I know im digging many long and much discussed threads saying this, but those courses are junk.

With that said, I would work on a wide easy to drive powerband. Yes, you can design your car with a short very high powerband, choose your rear end ratio to maximize that, have a shifter kart champ drive for you and do well dynamically. But the fact of the matter is most people in FSAE arent very good drivers, and you should take that into consideration when you design your powertrain. Practicality, especially for a first year team, should have alot of weight in your decisions.

-Bryan </div></BLOCKQUOTE>

Also remember that you're marketing this thing to a 'weekend autocross racer'. Drivability is a good weapon to have on your table of reasons why people would want to buy your car.

[Homemade WRX]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by darienphoenix:
Also remember that you're marketing this thing to a 'weekend autocross racer'. Drivability is a good weapon to have on your table of reasons why people would want to buy your car. </div></BLOCKQUOTE>

I can't agree more and that would have been my counter to his reasoning...none the less the judges point is true. A good driver will slip a clutch to exit a corner when at the bottom of first gear...

[VFR750R]

Everybody knows Horsepower is what sells http://fsae.com/groupee_common/emoticons/icon_wink.gif

Wide flat powerband...yes...but how wide and flat is the question.

[B Hise]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Wide flat powerband...yes...but how wide and flat is the question. </div></BLOCKQUOTE>


yeah but answering that would require some sort of research/design/testing mess.

who has time for that when there are forums to post on? http://fsae.com/groupee_common/emoticons/icon_smile.gif

- Bryan

[Erich Ohlde]

here is a good answer, perfectly flat from idle to redline at 50 lb ft of torque for us.

[VFR750R]

That's a terrible answer

[Homemade WRX]

I agree...simply not enough torque... http://fsae.com/groupee_common/emoticons/icon_razz.gif

[Erich Ohlde]

well you know, i had to pick a number.

[Alexandre D.]

The post was about gearing and shift point.

I don't think that standard gearing methods apply to FSAE (some of my teammates will disagree). The whole thing is not exactly clear especially if you add driver (none) experience. I think that the most significant (but not only) factor is corner exit speed. It depends of your engine. If you have a peaky torque curve, aiming for max torque may not be a good idea for stability. That's why you want a flat torque curve.
Your shift point should be around the RPM where for the same wheel speed, you have equivalent torque with the next gear. That is, having a higher ratio (lower gear) does not provide any advantage over the next gear. This is far from a complete answer, only a starting point.

How wide is another story. It really depends of your available torque and maximum traction. How wide may be the factor that determine the number of gears used. Again, not a complete answer.

There's a lot of thing that are specific to the engine you use. We have very high torque at very low RPM. But we don't use it a lot because delay is higher than at higher RPM. At lower speed the limit is the tire not the engine. How do you gear for such a scenario? Decreasing the ratio is to some extent a good idea but not an absolute solution.

The ultimate solution is to run some simulation with something like simulink and optimize it. Then try it on track. It's a tough job even if you only use 1 DOF. Some software may be able to do that if you have some data to input.


For a starting team, assuming you use a F4i, I THINK you should aim for somewhere near 60mph in third gear at high RPM. Sprocket are very cheap, you can buy more than one. Do some good old paper calculations with different scenario.

[formula_wally]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by VFR750R:
That's a terrible answer </div></BLOCKQUOTE>
Your a terrible answer, lol. JK

I move we scarp TQ and HP figures, and just integrate the area under the TQ curve, and adopt it as one number called "GRUNT"

[Big Bird]

This is where I bore everyone witless again about my engine performance theories. They usually go down like a lead balloon.

Personally I don't like "area under the curve"� measures as they ignore some driving practicalities in the pursuit of numbers. And I don't believe an engine that has its torque down low is necessarily any easier to drive than one with its torque higher up either. The critical factor is the "ramp rate"�, for want of a better term.

Back in the late 90's when they were still racing two strokes in MotoGP, word was that Honda could build an engine capable of well over 200hp. They didn't. They gave up on the "max area under the torque curve"� or "torque down low"� argument too. Instead their design goal was to give up some forward acceleration performance with the goal of designing a LINEAR response – e.g. ¼ throttle = ¼ acceleration, ½ throttle = ½ acceleration etc. Their most effective race engines were lower output (back around the 180hp mark), but were the ones that were "friendliest"� to the rider.

Think about it. Our goal is to design the car so we can hover around on the outside of the friction ellipse at all times. At maximum lateral acceleration (constant speed, mid corner), we have no capability to add any longitudinal acceleration or we spin out. As we feed in the power to make the transition out of the corner, we need points on the power curve corresponding to each of the "allowed"� longitudinal grip states – which means designing a proportion of the rev range where the engine ain't doing much at all.

Imagine the magical 50 lb ft of torque from idle mentioned above. (Sorry folks, I have no idea what a lb ft is, I'm just quoting Erich).

Torque (lb ft)
:
:
:******************************** 50
:
:
:
:
:________________________________ rpm


When transitioning from mid-corner to exit, the driver sitting on the lateral grip limit can:
1: Try to feed in the throttle, get 50 lb ft torque right away and break traction – not good
2: Drive slower mid-corner, reducing our lateral load to a point somewhere in the middle of the friction ellipse where we can apply the full 50 lb ft – not good either
3. Not apply any throttle until corner exit and the car is pointing straight ahead – not good either again.

Once again we are butting our heads up against that trade off between what is fastest in a straight line, and what is fastest around corners. We can't have both.

To me, the perfect engine curve has maybe a range of a few thousand revs or so where the engine output roughly matches cornering friction drag (for mid corner), a range of say 2000 revs where the torque is linearly increasing, and whatever top end you want for the straight line stuff.

Torque (lb ft)
:
:
:.......................... ********* 50
:.................... *
:............... *
:......... *
:******
:________________________________ rpm

Compared to the 50lbft from idle curve, there is less area but I'd almost guarantee the reduction in output early in the rev range would make the car quicker on track. That is not just fluffy "driver isn't good enough" stuff, a look at the tyre friction ellipse demonstrates a requirement for it. Our 04 car was pretty good at that, power delivery was nice and linear and controllable. To paraphrase our 04 engine guy Shiraz (on one of the other threads), we designed an engine curve that "just sort of looked nice"�.

Anyway, that is enough for one day. Cheers all

[markocosic]

If only we could drive-by-wire and link it to wheelspeed and steerign angle... http://fsae.com/groupee_common/emoticons/icon_wink.gif

Agree with Geoff here - we've put one person to look at nothing but the accelerator pedal &lt;-&gt; torque at the wheels/torque requirements at the wheels relationship.

One of the things I snapped at FSG were these plots:

http://www.cosic.org.uk/galleryv2/main.php?g2_itemId=9411
http://www.cosic.org.uk/galleryv2/main.php?g2_itemId=9414
http://www.cosic.org.uk/galleryv2/main.php?g2_itemId=9417

because the design judges were /very/ interested in them and the justifications for them...

We were tempted to ask the rules comiitee this:

"Hello!

Rule 3.7.3 (B) states:

-----
(B) Once the vehicle is approved to compete in the dynamic events, the ONLY
modifications permitted to the vehicle are:
*Adjustment of tire pressure
*Adjustment of brake bias
*Adjustment of the suspension where no part substitution is required, (except that springs, sway bars and shims may be changed)
*Adjustment of mirrors
*Adjustment of belts and chains
*Adjustment of wing angle
*Adjustment to engine operating parameters, e.g. fuel mixture and ignition timing
*Replenishment of fluids
*Replacement of worn tires or brake pads

The vehicle must maintain all required specifications, e.g. ride height, suspension
travel, braking capacity, sound level and wing location throughout the
competition.
-----

We would like to vary the throttle response between the ˜steady state' dynamic events (acceleration/skidpad) and the ˜dynamic' dynamic events (autocross/endurance) ad another vehicle setup/tuning parameter.

Eg - traditional s-shaped pedal position vs mass airflow for the ˜dynamic' events and a large ˜insensitive' region at low load for the skidpad and high load for the acceleration run.

Drive-by-wire is prohibited, but would it be legal to (for example) switch the ˜cable-cam' on the throttle butterfly shaft between events or have an adjustable ˜cable-cam' with multiple profiles that we can slide the cable between? Might this class as an adjustment to engine operating parameters, is it the intention of the rules to force testing and research into a compromise or is there another line of thought again?"

But haven't done as of yet (not until we've an idea how much it really counts for). Reckon it'd be allowed or not?

[Big Bird]

Hey Marko,

Thanks for the photo links mate, they are really cool. Its nice to see that someone recognizes that there is a human bolted into this whole conundrum!

Cheers mate,

[Big Bird]

By the way, I'd reckon you would get away with your adjustable cam idea at the event. That sounds like an engine operating parameter to me. Or if they kicked up a fuss you could chrome plate your cam and pass it off as a mirror adjustment http://fsae.com/groupee_common/emoticons/icon_smile.gif

Cheers,

[Nick McNaughton]

Geoff's on the right track. Drivers are picky buggers, but his torque curve is probably something they could live with.

There's a few other things in the mix: engine braking torque on overrun is one of them. There's more of it at higher revs, so if that's where you put powerband the average driver is going to get bitten in a bigger way if s/he lifts off too much at the wrong place. Poor tip-in response can mess this up too... but a lot of it is the diff's fault..

Traction control plays a part too. If you nail the traction control setup, a peaky crappy torque curve becomes a lot more drivable on corner exit... all the peaks disappear in a cloud of misfires no matter what the driver's doing. Of course, Geoff's boys avoid this by using the single cylinder method of traction control. http://fsae.com/groupee_common/emoticons/icon_smile.gif

[JagLite]

It depends entirely on your engine, how it is tuned and geared. As has been posted, go for a nice fat torque band and not peak HP. For your first year I recommend that you keep the engine very stock and simple. Many don't do well due to engine problems that they designed in.

I also recommend that you gear the final drive ratio very low. Try for a top gear top speed of 70 mph (110-120 KPH) and make first gear so low it is not used. Slipping the clutch works but should not be required as it burns out the clutch. Most motorcycle engines have 6 speed tranny's but we shoot for gearing to just use 2nd and 3rd. Even 2-4 is better than slipping the clutch in first gear to power out of corners. It is very easy to break the back end loose when slipping the clutch.

Or use an engine with a CVT or build your own. (Constant Velocity Transmission that is, not the engine!) No shifting and you tune the CVT to keep your engine in whatever RPM it has the best power. Snowmobiles use CVT's and many of the racing machines are putting out double the HP of an average FSAE motor.

Then get lots and lots of practice driving the car. Attend any events that will let you enter. Hillclimbs, Autocross, Time Trials, or anywhere you can drive. Ask around or advertise for a race driving instructor to teach you how to drive the car you build.

Have fun. Don't kill the joy by being too complex. Simplify and add lightness.

James

[ketan]

thanx everyone for being so co-operative....

[GSpeedR]

Just use traction control (as mentioned above). Put all of the power right at 100% throttle, and then you can incorporate a clip onto the throttle pedal. Basically when the driver has the pedal floored and flips his toes down, the clip engages and the throttle is stuck at 100% even when the driver lifts. With TC switched on the driver no longer worries about throttle application and lets the computer sort it out. Disengage traction control for the burnout event (there is a burnout event, right?) and try to design some way for the computer to brake and steer the wheels also.

[Steve Yao]

I agree with Geoff's thoughts on linear torque delivered for best utilization of available grip. However, I think he actually meant linear torque delivery versus throttle position. Which was paid special attention to by the team in Marko's photos(though it does not really look like they got it).

I also agree that a flat torque 'curve' (it would acually be a 'table' wouldn't it?) all the way across the RPM range would be ideal at full open throttle (FOT). The linearity, Geoff speaks of, should be in the response of throttle position versus torque at every rpm. So, yes we want 50 ftlb, full RPM range, FOT. But we ALSO want 10 ftlb, full RPM range, 20% pedal position. AND 25 ftlb, full RPM range, 50% pedal position. etc.

After all, if you have a torque curve that ramps across the RPM range from almost nothing to max torque with FOT, it would be mighty difficult just getting the car rolling short of revving to 8k and burning the clutch.

I am not on my engine team, so please let me know if I am full of it on this.

[BeaverGuy]

I think we really need to realize what the theoretical Ideal would be and the actual Ideal is. A perfectly horizontal torque curve is great if you never want to shift and the engine can rev forever. However, unless you are using a CVT you can't do that so a curve that ramps up and down would seem to be desireable. The up ramp so that your throttle actually works and the down ramp so that you have both rpm and torque matching on the up shift.

a torque curve that looks like this

------
|
----

can net you a great driving car if stay above the dip. But you won't neccesarily do so and regardless of your throttle control mechanism when you cross that jump drivers will have problems and complain. In '05 we gave up some peak torque to eliminate a torque jump and make the curve look "nice" because we had had problems with torque jumps previously.

I do agree that the linear torque regulation should be in respect to throttle position, but that linearity should be in the middle of the throttle travel where the most control is needed. That is also why I would say two different throttle cams would be unneccesary. The fine control needed during steady state cornering in skidpad is also needed during throttle control in a dynamic corner in endurance/autocross and I would be suprised if the throttle range for skidpad is outside that for autocross/endurance. I would say something much more linear than what was in the pics would be desireable while still maintainging a soemewhat insensitive region at low throttle.

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">After all, if you have a torque curve that ramps across the RPM range from almost nothing to max torque with FOT, it would be mighty difficult just getting the car rolling short of revving to 8k and burning the clutch. </div></BLOCKQUOTE> That is how every car short of an electric has to work. It may not be 8K RPM but there is a speed disparity of the engine and the rest of the powertrain when you start and the clutch or torque converter will "slip" to accomadate that speed difference.

[Steve Yao]

I am still not clear on how a torque curve with some ramp and drop is better than a flat curve (give equal area under the curve). Josh indicated a ramp would allow the throttle to work and a torque loss at the high end to allow for upshifting. My thought was that it should not be the engine setup and map that does this for you, but the driver modulating torque through the gas pedal. The driver lets of the throttle for an instant to unload the transmission and upshift. The driver clutches and blips the gas to matchrevs on downshift.

Say you are manuevering a long increasing radius turn. The corner is such that you can accelerate at a constant rate and be on the edge of your traction envelope. If you have a flat torque curve with linear throttle response, the driver simply lays on the throttle at a constant rate as the corner opens up. Torque increases linearly with the driver input even as wheel speed/RPM increases. If throttle response is linear, but the torque curve has peaks, ramps, or undulates the the same constant throttle rate will produce variations in torque output as speed/rpm increase. If the torque ramps or jumps, the driven tires lose lateral grip as they try to overdrive the car and it oversteers. The driver would have had to vary his throttle rate throughout the corner to maintain peak grip.

[BryanH]

SEY you are on the money, too late to explain why. a morning job. Geoff you remind me of a dog on an ice rink when you get techie on engines.

[ketan]

guys be a bit more specific on topic of the post

[Tommo]

Some where between 0 and 20,000 rpm

specific enough?

what engine are you using?
na?
what are your performance targets?

you'll need to be a little bit more specific first

Tom