GlenL

I've been thinking about adding an option to my spreadhseet to allow different shift points in each gear and to calculate what those shift points should be for maximum acceleration. When I've got time this winter I'll get to that.

Playing with shift points does make something clear, though, those last 200-400 RPMs are more exciting that excitable. That is, it's fun to hear the engine rev but short shifting a bit doesn't negatively impact speed much at all.

ArthurPE

depends on the ratio
if the gap is large, hold the gear as long as possible
as the ratios get closer, it has less impact

if the thrust (T x ratio / r) drops below the level of thrust a shift would put you at, then shift early
this is easily determine by graphing each gear, thrust vs rpm (or speed), I did both...

on the RS4 graph, no where does the next gears thrust exceed the lower gears (gets close in 5 and 6)
so run to HP peak/redline

in the 'example', you can see that thrust in 1st drops below thrust in 2nd
so you would want to shift at the horizontal intersection of 2nd's thrust and 1st's
which is below the redline(>5200), 4600+
3rd can be run to redline (no intersection)
and 5th to 6th, may want to shift ~4300


cars, especially hipo are seldom set-up like the example, economy sometimes (where short shifting is desireable) and trucks

EspritS4s

I believe you. What I meant was, based on this thread, I'll bet you did correct him on his article....point....by....point.

mark kibort

Hey, what can I say, I call them how I see them.

IcemanG17

Arthur
To make this argument more 928 focused...could you do it using a 928 powerband & gearing....I have attached the dyno sheet from the 928 estate and its gearing is stock 4 speed automatic with 2.20 so the ratios (overall):
1st: 8.0872
2nd: 5.3064
3rd: 3.1592
4th: 2.20

Could you make the actual force graph using this info?

Looking at the Audi's its clear that the lowest gear possible always will accelerate better than the next higher gear....I am wondering if this is also true for my 928, since the powerband is dying so bad past 5500rpm

mark kibort

No, im sorry Arthur, the only joke is someone that projects as much as you do!

Again, for someone that claims to "get it" you still are not seeing the light.

Again, you still have not given me the shift points. a few more posts of technical self pleasure and no answer! Im not demanding, but asking.
Can you provide an answer and give an intellegent reason for it?

Now, suddenly, you are so skilled that you an "Tell" when to shift. that is rich too! how funny! Trust me, i have a few racer friend with the same self proclaimed powers. one of which was a friend who I was beating regularly, becuase his torque curve fell off and he "Felt" like it was senseless to continue to rev. WE had this exact same discussion. now he is a real force to recon with.

You and your "LOL"s you are a funny guy! joke status, really!

Now, look, i already asked you on how we can quanitify the HP-seconds, or Kibort or time periods that seem to skew a hp curve to the left. as it turns out, it really isnt a correct way or even a meaningful way to do it. I am giving you that . A bone if you will! Ill stick with average HP for now, and have for many years of doing this.

Here is where you are all wet. Your now tossing in "FEELINGS". I was talkin about maximizing acceleration. if you think that its OK to lose 10Hp to short shift, and that is not "worth" it to shft, expecially at the top end, then two things.
in your example, if you short shifted you would be down near 300lbs of thrust at the rear wheels for the 500rpm you short shifted!! that is quite a bit of accelerative force. dont you think. (assuming a 24" diameter tire)
1. you dont race
2. you have changed the discussion to feelings , rather than physics.

people live and die on the track for 5-10hp differential. as far as acceleration goes for our race cars (and street cars for that matter) 10hp is like putting in 100extra lbs in "most cases". LOL too!

Now for REALITY.

If the HP curve is such that the post shift RPM is lower than the pre-shift RPM THEN, it pays to stay in that gear longer! so, with the example of YOUR car, where the shift from 3-4th is .82% RPM drop, 8000rpm drops to near 6600rpm. is the HP less at 6600rpm, than 7800rpm? yes it is, so the corresponding thrust will be less by shifting. in your case, a lot less. prove me wrong there. you cant. why?
acceleration = power/(mass x velocity) which says, at any time you have more power at the same speed, you have greater acceleration!!!! its the law. prove it wrong, just try!

Now you go on with the insults. Hmm, more projection huh Arthur. YOU are now in reality and you are now posting MIS INFORMATION. Anyone want proof, just give me the thrust forces of any point on your hp curve that shows equal or greater value post shift at or before 7800rpm. On YOUR car and YOUR HP / torque curves!

mark kibort

Yes, Glen. You usually are the voice of reason. you are correct.

mark kibort

Arthur, you then make it easy to make my point. every single gear thrust value is greater than the next gear. So, that means redline shift. does anyone disagree? I was going to poist a couple of graphs, one that really hits what we are saying to the tee. notice this is a peakie engine vs a 928.
notice how there is an intersection in the later gears where you get a shift point. this was very important for me to work out, due to oiling concerns, but a desire to wring every single bit of accelerative force out of my car.

So, do you admit you are wrong about the RS4? Peaky hp as well?
You did say, it is true for all cars to shift at "max HP" remember? you even said in the later gears , "before max HP". hmmmmm
no where does it pay to shift early, except for saving the engine and that is not part of the discussion.

as a side note. Yes, you came out of no where to fix the "wrong" thinking, but remember you jumped in well after the real points were being made and argued about things that really were not in question. However, in this example, do you see how easy it is to find shift points by looking at the HP curve? For get about the HP-seconds thing. yes, I know comparitively, using it can be an indication, but a quantity of HP-seconds if not contained by a time, is "useless". (I know) but wanted your opinion for a way to factor in the "extra" time element because of a lower rate of acceleration at max HP ranges and more time spent there vs at near max torque range. Maybe it just doesnt matter. what does matter, is that using HP curves makes it very easy to determine shift points.

Mark

mark kibort

Brian, you and I have had this discussion. since your shift range is .72, you would then try and find a shift point that strattles the arc of the HP curve.
THIS is exactly the point of the discussion. your shift points are going to be around 5700rpm pretty similar to my car now, but slightly lower. my Holbert machine was more peaky and paid to shift at redline, or higher if possible. similar to Arthur's car. (but he wont admit it, even though the force curves show this undeniably)
Again, this is the point of this discussion. using HP you an see that the hp at 5700rpm, shifts down to a rpm at 4100rpm post shift at near the same hp. meaning that shifting any higher spends time working down the HP curve, which directly is proportional to thrust force. I would shift maybe a tick longer, to take into account shift times. thats the answer, and it took about 20 seconds. to figure out. at 70% rpm drop, a little higher. 5800, and if it was 75% drop, that would allow you to shift a little earlier, maybe near 5500rpm. pretty simple!

Dont ask Arthur, he might tell you how he feels when the torque is falling.
LOL too!

littleball_s4

Not only shift points are easier to calculate.

If you happen to know the mean torque over the rpm of interest, as in post number 284, you know the average power. If you know the average power, and the transmission losses, you know the time it takes for a certain mass from v1 to v2, under the assumption of post n284.

You don't need:

1) wheel radius
2) gear ratio
3) diff ratio

t=0,5*m*(v2*v2-v1*v1)/(mean power*losses)

Compare equations. Quite easier.



New abbreviation :

AFK=(away from kiborts)

littleball_s4

For you spreadsheet gurus:

In real life, adding the shifting delay is not difficult (I assume you used, as me, a montecarlo simulation) and adds lots of accuracy. Just make the active row check which gear was in x rows before and make the power zero until both active row and x rows above were in the same gear.

And, obviously, don't let anybody believe one can assume aero drag is zero in any car and get decent results.

Aero drag is quite easy: 1/2*ro*A*Cx*v*v

where ro is density, A is front area (including that under the bumper between the wheels) Cx is the drag factor and v the relative speed.

I can't upload my excel because my company paid me for it, but it's really not rocket science and it has 0,1s accuracy over 30 seconds for the kind of vehicles I use.

GlenL

"It's in there."

GlenL

Of course it's easier but it's also wrong. Go back to post 123 (not really) and recall the discussion of integrating versus RPM as opposed to integrating versus time. And the time varies with each gear as wind resistance changes.

There's eyeballing, napkins, good simulation and real numbers.

littleball_s4

I think you're right Glen, it makes no sense to argue around false hypothesis.

The complete problem is quickly solved with a proper monte carlo simulation. (quickly IF you already have it uploaded, I mean) One including shifting delay, tires drag, aero drag, proper transmission losses and clutch drop behaviour (last two ones are tricky, later is only needed for 0-x times). If possible, also with tranny inertia, but that's not important.

Then, all disagreements can be numerically assesed with the official excel.

That would be the end of HP vs torque discussion. But I don't think I'll see it in this universe, not before aliens and teletransport.

ArthurPE

there is nothing 'easy' about your 'point'
it requires supension of the laws of physics and sound practice

never said the RS4 was peaky, you dissect, parse and distort to suit your incorrect assmuptions, I said it had a wide torque band...actually, best of both
>90% torque 2500 to redline
high power at high RPM 420 ~ 7800

YOU made MY point, thank you
the Honda does not benefit from a short shift, thrust does not overlap
the 928 does
and no where do 'kiborts' factor in

again, you are looking at this all wrong...HP x sec is a meaningless quantity
there already is a quantity used for this ft lb/sec x sec = ft lb
note, it is not time relative...so therefore useless for our intended purpose