optimized shift points (Engineering project)

T-bird4vr

Registered User
For my ME 415 (mechanical systems design) class, we each had to do an "optimization" project. So I used my dynochart/program to optimize shift points from acceleration between gears. I finally go the math done tonight. My presentation is Thursday and I should have it done tomorrow so I will post it then. I have a couple neat graphs to show.

But from the math:
1st hold as long as possible
2nd ~6100 rpm
3rd ~6200 rpm

The autorotor sure likes to rev high, I will probably just shift at 6 for all gears at the track.

If dynojet would have coperated more (the company releasing specific drum information), then I could have estimated 1/4 mi times, 0-60 etc. That would have been awesome, but difficult to do accuratly without some wind drag numbers. This is cool enough though.

here is a teaser pic, the second graph is what is important, you can see directly when the acceleration from the next gear is greater than the prior one. This was done using Matlab
acceleration.jpg


I think this stuff is cool, but I am nerdy too :eek:
 
Looks impressive. So, for those of us who went to school before this math was even invented, how about explaining what it means??

ira
 
Let me get the presentation put together today and I will post a link to that. Either powerpoint or I can probably pdf it. I know 99% won't have access to matlab software.

But basically where the lines cross on the bottom graph is where I should shift if the transmission would shift instaneously. The left y-axis is acceleration, the values are arbiturary becuase I couldn't get the necessary information from the dynojet company, but it doesn't matter for setting shift points. The velocity is accurate (mph) on the horizontal x-axis.

So the goal is to maximize acceleration through a high speed, say 150mph. The purple line is first gear, I want to stay in that as long as possible (engine-redline) because the acceleration in first is always higher than in second green line. But where the green line crosses the redline (3rd) is the point where the acceleration in 3rd becomes greater than in second. Same thing with 3rd to OD. And you can back out the engine rpm from the velocity to determine the shift points.

The interesting thing is I didn't directly use hp/tq to get this. Just time, speed (mph) and engine rpm from the dyno chart. Then it depends on gearing ratios, tire diameter etc.
 
So for arguments sake, if I read the chart correctly, it shows why you should finish the quarter mile pulling away in 3rd gear. Right??

ira
 
Looks like 3rd gear is good to ~140 in my car with the stock size tire height.

here is the presentation
point click
edit(need to make a couple changes first)
 
Last edited:
To me it looks like would you benefit from having second gear MUCH closer to first gear. (ie so the graph of first gear intersects the graph of second gear) just a thought.

-Darrel
 
t-bird4vr,

that is an impressive chart to me. i think it's showing more than just the optimum shift points and crossovers.

I know you said the acceleration line (vertical) on the 2nd graph were arbitrary numbers, but they actually are very close in terms of force applied at the wheel (at the end of the wheel radius on the tire tread, in thousands of lbs) to what your car with the AR would put on the ground according to my excel sheet. I have a spreadsheet that can calculate the actual wheel torque or force being applied for any car (i have plotted many SC's on sccoa and other formidable high powered cars), taking into consideration wheel size and radius, for known parameters (rpm, hp, tq, gear ratios, % of parasitic loss)- that graph closely matches my calculations- the numbers are in terms of thousands of lbs of applied force ie, 1=1,000lbs, 2=2,000lb, etc. the thousand lb figures are coming from calculating net engine torque output for a given rpm (low and high rpm's), and multiplying it by the total gear ratio (including the rear end gear) in 1st, 2nd, etc.., and factoring in the wheel radius to get the true amount of force in lbs being applied to the wheels. this spreadsheet even calculates the speed the car should be traveling at when it's generating that torque.
i know that sounds like alot or maybe even far fetched- maybe i'm completely wrong, but i don't think that anyone on this board has considered why his or her car maybe faster or slower than another other than measuring hp figures- hp is kind of a vague measurement (it's really irrelevant), if you really dig in and do the math.
not trying to start anything here, just sharing some info- take it for what it's worth or trash it. i know it's not "real world" or proof as in dyno data, but anyone can realize this by doing the math calculations from a dyno sheet.
 
Last edited:
ya, hp is the time derivative of the torque or kinetic energy. It is how quickly the kinetic energy is changing in-other words.

Force is going to be directly related to the linear acceleration since f=ma. But I needed the moment inertia of the dyno drum to correlate the acceleration on the drum to an acceleration on the road. (converting rotational K.E. to linear K.E.) The acceleration values are thus arbiturary, but there is some unknown constant to multiply them by to get accurate values (neglecting tire and wind drag).

There is a lot more stuff I could figure if I knew the geometry and mass of the drum. But I don't, and dynojet won't disclose the good stuff.
 
think i could get the MATLAB and .ppt? and permission to use your project for my class (im an aerospace major)? this is awesome.

are you running 2.73s in a 5 speed?
 
D-bird said:
To me it looks like would you benefit from having second gear MUCH closer to first gear. (ie so the graph of first gear intersects the graph of second gear) just a thought.
-Darrel

Give up top speed for closer ratios. Everyone knows that makes your car quicker, which is somewhat reflected on the graph. You could play around with the ratio's and optimize them for a minimum time/max acceleration as well. You would need to set an rpm limit and top speed limit.
 
imo, what you have just disclosed is the proper way to build up an SC. instead of throwing on bolt ons and hoping for the best, set what top speed you want, figure out maximum gear ratios allowable based on that top speed, and then add the parts necessary to make the engine put out as much big power as possible from the low end up to the rpm required for that top speed. optimize and tuning is the last step before you rip.

btw, from what i see from your AR setup and gearing, 330rwhp generates enough torque on the top end to kick 90% of the so called fast cars out there. SC's weight and drag is still a factor though.

i have found that the ideal power level for the SC (street driven) is about 400rwhp- this pretty much matches all the new king of the hill cars, including the Z06. torque to weight ratios are very close at this level- the sc has to shift more though which makes it slower 0-60.
 
Last edited:
matteo said:
think i could get the MATLAB and .ppt? and permission to use your project for my class (im an aerospace major)? this is awesome.
are you running 2.73s in a 5 speed?

Isn't the .pdf good enough? I will think about sending it all. Nothing I did was too difficult though. I just took a while to figure out all of the omega's (tire, motor, drum). Do you have a copy of mat-lab? And what school/year are you?

2.73's and a 5spd? That comment worries me a little, I guess I figured most people know my car and/or check my member page, and then there is item of only 4 gears on the graph. It is a 95 auto with one of dalke's AR's. If I had a 5spd I would have plotted all gears. ;)
 
hytorksc said:
imo, what you have just disclosed is the proper way to build up an SC. instead of throwing on bolt ons and hoping for the best, set what top speed you want, figure out maximum gear ratios allowable based on that top speed, and then add the parts necessary to make the engine put out as much big power as possible from the low end up to the rpm required for that top speed. optimize and tuning is the last step before you rip.

btw, from what i see from your AR setup and gearing, 330rwhp generates enough torque on the top end to kick 90% of the so called fast cars out there. SC's weight and drag is still a factor though.

ya, weight is what is going to kill acceleration. Nothing here is new lol, it just shows some of the physics behind it. From the plot, my car will go over 200 mph. :rolleyes: but that is at 6500rpm, if only the gearing is limiting it. There is only so much gear you can put into a car and still live with it. More gear = more acceleration = more shear force to road = traction issues, not to mention less fuel economy and more engine wear.

I do agree it would be nice to be able to change the internal gearing of the 4r70w. But that neither sounds cheap or easy ;) . A 5spd would have a prettier plot.
 
T-bird4vr said:
Isn't the .pdf good enough? I will think about sending it all. Nothing I did was too difficult though. I just took a while to figure out all of the omega's (tire, motor, drum). Do you have a copy of mat-lab? And what school/year are you?

2.73's and a 5spd? That comment worries me a little, I guess I figured most people know my car and/or check my member page, and then there is item of only 4 gears on the graph. It is a 95 auto with one of dalke's AR's. If I had a 5spd I would have plotted all gears. ;)
Yeah I have MATLAB. I'm a junior at Arizona State. What year are you, and what are your plans for after college? There is a huge aerospace industry out here and you could get a good job no problem with a mechanical degree. Plus then I’ll have someone out here who appreciates SCs :) .

I figured you left 5th off since its so far out there, and not important in a quarter mile. But i guess you're right, it would be much more complete if you did all 5. And no, I don’t really read peoples member pages.

Maybe I’ll do a drag analysis on my SC for my class instead. Since I’m in the aerospace program, that would make more sense.
 
matteo said:
Maybe I’ll do a drag analysis on my SC for my class instead. Since I’m in the aerospace program, that would make more sense.

It would be pretty cool if you could get your sc in a wind tunnel. Seems like I ran into some sort of drag coefficient for our cars on the internet, but it has been a while. That information along with a quick rolling friction test (push your car with a scale between you arms and the bumper), and then if we could score the moment inertia of a dyno-drum you could predict almost anything.

I'll graduate in December of this year, so I am getting pretty close. As far as plans, don't really have any. I want to stay with a small-medium sized company though, no Lockheid Martin or Boehing. I have a contact at edlebrock so that is really the ultimate goal. Small-medium size performance car parts company, as a test/design engineer. Location, I wouldn't even care.

send an email to julian@iastate.edu and I will send you that stuff. It will probably be tomorrow though.
 
XR7 Dave said:
Good lord!
So are you going to burn me a new chip or what david? Not only that, but can you double check the temp my radiator fans come on?

Besides, I know you want me working for edlebrock anyway ;)
 
T-bird4vr said:
So are you going to burn me a new chip or what david? Not only that, but can you double check the temp my radiator fans come on?

Besides, I know you want me working for edlebrock anyway ;)

Dude, you are never going to work for Edelbrock if you don't learn how to spell the name. Let me guess, your intake manifold is made out of aluninum? :p

What are you talking about "chip"? :confused: I don't know anything about any "chip". :confused:
 
Back
Top