When to shift?
 

(Not sure what happened this morning because I asked this question but the post is not here. I'll do it again...)

I'm wondering when is the best time to shift.

I think the ideal is that you shift out of gear at the RPM generating the max hp and (again, in theory) your next gear brings you to the RPM with the max torque. Is that true?

But what happens if your gearing and peak numbers don't allow for that perfect a match? Do you adjust the hp RPM shift point so that your next gear is still at peak torque or vice versa?

Take these dyno numbers that Cliff posted this morning. Let's use the first set of dyno numbers since the spread between hp and torque is wider, and let's assume an M21 wide ratio tranny to make it interesting (and more relevant to my car ;))

Do you shift out of 1st at around 5k at about peak HP even if that brings you to 2nd gear below 4k where torque is peak? Or do you hold 1st a few hundred extra RPM's to keep 2nd at 4k?

My next question is: How can I calculate RPM drop for each gear so i don't have to drive around with my nose in my tach at 5,000 RPM!

http://forums.maxperformanceinc.com/...8&d=1492765368

Since I raced a 4-speed car for a few years, I'll take a stab at this one.

You experiment with shift points, til you find those which produce the lowest ET. Without actual, on track testing, all you have are opinions. And, as we've seen on this and most every forum, opinions can differ WIDELY, on most any given subject.

There can be too many variables with engine & cam torque & power numbers, trans & rear gear ratios, etc.

THAT'S NO FUN!

I guess I was hoping for a theoretical ideal as a starting point. I live too far from the closest track to test regularly. I'm planning to hit the track this year, maybe later in the season in case I break something I don't lose the entire summer. So I was looking for starting points.

I'll be taking the car to the dyno in the next few weeks once my new fuel system is in. I have a benchmark dyno report when I first bought the car, but i later discovered that I wasn't getting WOT, so that's not much of a benchmark. I've since changed carb, intake and did a valve job with new springs (the old ones were VERY weak as per the machine shop who measured them).

I know nothing beats opinions faster than a time slip, but indulge me.

PS Hey, 2,000th post!


(I've attached that old dyno run from 3 years ago as a reference point.)

http://i300.photobucket.com/albums/n...psrzpy5uav.png

Did the Dino wind it up higher than 5800? Your photo at the bottom is cut off.

Based on the info shown, I'd shift it at 5,800.

The photo is Cliff's from another thread this morning. Look closely and you'll notice that there are two dyno sheets there, so I was suggesting we look at the first one since those numbers are closer to mine plus the spread between torque and HP is wider.

"...THAT'S NO FUN!..."


Oh yeah, it was LOTS of fun, not only the testing, but especially winning races. :D

Not too many things in this life have been more fun, to me, than yanking on the Hurst shifter, in my GTO--and especially on the track. :)

The RPM drop is easy for a manual trans but has more variables with an automatic trans.

(new trans gear ration / Old trans gear ratio) * RPM @ shift

► (1.52 / 2.52) * 5800 = 3498.4

Stan

Typically you want to maximize average power when shifting. This typically means shifting a little past peak depending on how sharply power drops off after it's peak. Each car will be different and that's why you need to experiment your shift rpm. One thing to keep in mind also is that shift rpm may be different for each gear depending on your gear set.

As an example somebody running a low gearset TH350 is going to have a larger spread from 1-2 than somebody running a standard gearset. But the spread from 2-3 should be roughly the same. As a result the person with the low gearset may need to shift later in 1st gear than the person with a standard set. Otherwise they risk dropping out of the power band in 2nd gear. This is crude however because your converter and how much slip it has plays a role. That's why as mentioned earlier doing this work on a manual is easier.

If you know which gears are in your tranmissions and the spread between them, you can take that information and cross-reference it against your dyno sheet to calculate where your shift points for best average power "should" be. That gets you a starting point, but because your engine will behave differently in the real world vs when it was on a dyno, you'll have to make adjustments based on your ET and trap speed to really dial that in.

I love him! Car nut and engineer combined. And I saw this video way back when but now I used my numbers and calculated this all out for the 1-2 shift from 4000 RPM all the way to 5500. And wouldn't you know it, shifting at 5300 RPM nets the same torque in 1st as it does in 2nd gear. Which is where most people find their best times at with big Pontiac engines.

I LOVE IT!

Now I just need to put it in excel!

When I was still racing my GTO it took awhile to find the optimum shift point. I was all over the rpm range till I found what the engine liked. Even though the advertised range of the Lunati BMII cam is 1800-5800 the engine liked a shift point of 5200. Above that and the engine pull felt flat and netted no better times or mph. This was with a Saginaw four speed with a 3.11 first gear, 3.55 rear end gears and slicks. The old engine I ran years ago with a RAIV cam liked a shift point of 5800. So different combinations vary. In my opinion it's what the engine likes.
Like ponyakr the four speed can be more challenging but a good driver can run consistent times.

My opinion. I've drag raced since 1966 and I've always found that I got the best results shifting at the point when the engine stopped pulling hard. If you lose a lot of rpm on the shift you could go another 100 rpm before shifting.

Experiment and have fun.Ideally you want a smooth transition into the next gear.Linear power as some would say..years ago when I used to use a G tech for testing..you could see the G's..You should start with a hard pull and slowly the G's drop off..if you shift and you feel a surge of power you have shifted too high,if it lags you have shifted too low,time slip increments are the best.Its the fun part of the hobby!

Don't get carried away on the 1-2 shift. You'll be past your peek numbers and stuck on the rev limiter before you know it. I don't like getting twisted all that tight in first gear. JMO.

When the optimum rpm is found, how do those of you with a four speed shift? Ill use a the term a friend keeps asking me about.
"Are you power shifting"? Keeping the accelerator to the floor and stabbing the clutch for the gear change. I don't, to me it just seems unnatural, I cant keep my foot planted while trying to shift. Now if you have a "smooth" shifting transmission, and possibly a rev limiter it may be worth trying, but I still cant bring myself to try it. I have seen a you tube video of a guy doing it in an old dodge or Plymouth.

Since I started racing four speeds in 1977, once I put the gas pedal to the floor I don't lift and I use the clutch, but you have to have near perfect reflexes and coordination. If you let off during shifts then you lose as much as 2-4 tenths. Guys like Ronnie Sox and Dave Strickler were the best on four speeds that ever was.

Like this :D Mark gettin' it done with a little 302 and a Nash 5 speed with 1:1 5th and 5.38 gears, using a clutch and standard "H" pattern type shifter. Nothing fancy.

https://youtu.be/-emhm4hw3Kc

You mean the same HP, right Adam? Engine torque is force as measured at the crank. But torque at the wheels varies by the gear ratio. So HP better represents what is getting to the ground through the gears. (Yes, another engineer here.)

I don't drag race, only test and tune nights for me and those are very infrequent, as in years in between. But I find shifting beyond peak power to give best results. How much beyond depends on the curve. If it drops off fast then not far beyond and if the curve is flat then a good ways past it. I think your math exercise is a good starting point but test passes at the track will tell all.

As for power shifting I pull back with my right foot but not all the way out. I've done the foot on the floor thing but without practice its shaky, and I never wanted to break a car practicing it. Launching a stick car with lots of power is another whole topic, and much more critical to results. I always just accept some wheel spin when playing around on the road, but that's a big no-no to getting good times at the track. For some reason my self control with a stick is not as good as with an auto trans, haha...

No, I did mean torque based on the video's recommendation. Without an engine only dyno session, what else could I do? The chassis dyno is measuring in 4th gear, which is 1:1 ratio.

And yes, absolutely, launching a stick will add more variability than anything else. I started a thread on this a few years ago of you want to search on my user name. There were some good posts in it.

To me that is what counts. These were done about 5 years ago for a thread on a different forum.

Stan

Exactly correct.

Acceleration now in G = horsepower now * 375 / (Mass lbs x MPH now)

The video is totally correct.
He states the wider the trans gear split, the more engine torque must dive to justify shifting - otherwise engine redline will be the best shift point.

Another excellent video.

https://www.youtube.com/watch?v=4BOryrhs9iA

With a 3 foot bar I can generate 600 ft.lbs of torque by hand.
That is supposedly a substantial value.

Even with four buddies no way we will generate a lowly 200 horsepower.
Just too much energy!

Last week I was playing with converting engine cylinder pressure to car acceleration. Without using torque numbers or gear ratios.

I have to watch that long video, but the Engineering Explained video says to use torque and not HP, then he goes on to calculate how torque is affected by different gears, and therefore, when you should change those gears. So when I do the math based on my own torque curve (which is very flat) I found that the torque in 1st at 5300 RPM is 704 and the torque in 2nd at 3950 RPM is 705, making that the ideal RPM to shift. I'll repeat this exercise for the other two shifts as well. I found it interesting that the answer was 5300 RPM since that's about the number most experienced Pontiac'ers tend to use.

If you are saying to use HP instead of torque, then aren't you challenging the validity of his video entirely?

Where is 704 and 705 coming from?
What are the gear ratios?

What hits the rear wheels is the important thing.
If you provide your specifics I can show you both methods that end up with the same final result.

If you are referring to calculated torque at the wheels being the same then you are right. I always think of it in terms of engine HP at resulting engine RPMs though. Mathematically it all works out either way. And we don't much talk about wheel torque when discussing cars.


Sent from my iPhone using Tapatalk

Using the formula from EE's video and my own torque curve from when I first dyno'd the car on the chassis dyno 3 years ago, I got the following math:

2.2 (1st gear) x 320 lb.ft = 704lb.ft @ 5300 RPM
1.64 (2nd gear) x 430 lb.ft = 705 lb.ft @ 3950 RPM

and to get the RPM drop of 3950 it was:

2.2 5300
---- * ------ = 3950
1.64 x


At a 5300 shift point, the torque was the same before and after the shift.

From that you have more horsepower at the lower RPM

320 lb.ft @ 5300 RPM = 322.9 horsepower

430 lb.ft @ 3950 RPM = 323.4 horsepower

Engine torque is dropping like crazy up top. Agree you would not shift higher.

Yes. (remember that this is in 1st and 2nd gears, not 4th gear). But yes, that's my point.

If you do this math at 5400 RPM shift the resulting numbers are 660lb-ft in 1st @ 5400 and 705lb-ft in 2nd @ ~ 4025 RPM.

If shifting at 5200 RPM the final torque numbers are 748 in 1st and 705 in 2nd. Oh wait, does that mean I should shift at 5200 and not 5300 (theoretically of course, I know that 100 RPMs difference in 1st gear would be hard to vary at this level of performance)?

Adam track rental is friday, we can get it sorted out, I will be willing to test your car if you want!! :)

I'm finally having my new 1/2" fuel system put in this week (from pickup to carb). I don't think it'll be done for Friday which is why I didn't sign up for track day. But I do want to go this year for sure!

And thanks for the offer to drive my car. I've got it covered :D

Yup, slightly higher torque at the wheels and corresponding slightly higher HP after the shift according to the math.

Glad your car is a stick, 'cause otherwise we would have to get into the effects of a torque converter!!

I have to go out now but I'll run these numbers down the RPM range more because my peak RPM is 450 at 3000 RPM (well, it was before I changed my intake and carb).

So if I look to be at 3000 after I shift, that means I'm shifting at 4024 RPM. Ironically, one of my best ever 0-60 times was when I shifted at 4000 because I was losing traction in 1st gear.

NOW I really need some track time!

That dyno sheet your using is from a car dyno?

I would think the torque readings are what is after the gear ratios?

:confused:

Try the dyno chart that was posted for Cliff's engine?

See what it shows?

:confused:

this is the chart I'm using

Chassis dyno

Shifting near 4000 RPM would kill your power average down the track.

http://performancetrends.com/blog/?p=7

From your info, from around 4000 through to 5300 your engine is making near 320 horsepower or more.

450 torque @ 3000 is only 257 horsepower.

Acceleration now in G = horsepower now * 375 / (Mass lbs x MPH now)

So imagine "now" you are at 3950 RPM making 323.4 horsepower, then you shift causing RPM to drop to 3000, now your putting 257 horsepower to the ground.
Better off instead to make about 320 HP at 4100, 4200 etc and continue on.

I think you are getting confused Adam, its torque at the wheels that matters. Holding first to get more torque to the wheels is more important than the engine running at peak torque after the shift. You got it right the first time.

One of the reasons I prefer to think of it in terms of HP. Less confusing and more intuitive.

That link was an interesting read. So I laid all this math out in a spreadsheet. Based on the original video's theory of shifting for peak torque, I'd be shifting around 5200-5300, but when I added a conversion to HP for each of those figures, it raises the shift points a few hundred RPM, especially in 1st gear.

Take a look at this spreadsheet:

https://www.dropbox.com/s/xxeg08ql8x...COPY.xlsx?dl=0

Remember HP is not a function of gear ratio so calculating it like you are doing is not correct. Hence the 700+ HP numbers which are way more than your engine makes. Just read the HP right off the dyno chart. So for 1-2 your car will be making 320 HP at a 5300 RPM shift which matches nicely with the 320 HP at 3950 after the shift. And so on for the other shift points. Doing it this way is a lot easier than calculating wheel torque and you end up with the same results.

The curves in your dyno chart are very flat which would mean that the shift points should not be that critical at the track. But then again with carb/fuel flow issues resolved the curves may look a lot different. And real world never works out quite like theoretical due to other factors in play.

Sounds good. Thanks for all the chatter. As always, very informative and fun.

I hope to have the new fuel delivery system completed this week. I'll drive it a bit, then off to the dyno. THEN, I'll plan a trip to the track for some "put up or shut up" real testing! My goals at the track are to:
1) NOT break anything
2) not break anything too expensive
3) find a consistent launch pattern (traction is not easy with this car)
4) run 12.50's or better (car weights 3700lbs with me in it)
5) have fun! (which means goal #1 & 2 must also be met)

I figure between the new carb, intake, valve job and new springs, the timing/tuning improvements and the fact that now I have WOT (which helps!) I'd LOVE to see 500 crank HP but will settle for 450. (So 400+ at the rear wheels I guess using 15% drivetrain loss factor for manual.)

X1000

I do not know anything about Adam's car or setup but playing with a computer simulated run shifting at 5300 5200 5200 and saying that this is the best then 5000 5000 5000 and 5500 5500 5500 are both only about .03 worse. You would be much better off working on getting good 60 foot times than playing with getting the perfect shift rpm.

Stan

I don't think there is any getting around the fact that you should go faster if you maximize the average power under the curve. Physics won't let you get away from that. However, as has been suggested elsewhere in this thread, a lot of these Pontiac builds have really flat power curves. For my money, if I could run 11 flat shifting at 6300 RPM or 11.1 shifting at 5500 RPM, I'd lose the tenth and keep my engine together for a long time and enjoy racing. So as I see it, you can probably shift a good bit earlier than would be theoretically optimal, still run really close to your theoretically optimum et/MPH, and have an engine that lives a lot longer. Not to mention it will help keep the rest of your drivetrain intact.

For the big 455 street/strip engines most woln't care all that much about rear gear used or exactly where you shift them at. This assumes you've done your homework and built a pretty stout one that makes peak power past 5000rpms and has a broad/flat torque curve.

For my engine I can shift anywhere from 5000rpm's to 5800rpm's and it will only vary about .04 seconds for the entire spread. For this reason I short shift at 5000rpm's for most runs vs "grinding" up the engine more just to put up a little better time slip.

Many years ago I owned and raced a 1965 Nova SS with a high-winding SBC in it. I shifted it around 7000rpms and went thru lights around 7500rpm's. If you short shifted that little bugger you'd get your BUTT handed to you in a bracket race.......FWIW.......Cliff

Stan,
You are correct. I was just enjoying learning about the intricacies of bench racing, partly because I will likely have limited track time when I go so I'd like to at least start off with a theoretical baseline good shift points.

As for the 60' times, I've experimented with different launch techniques by timing 0-60 runs with a phone app that seems accurate. I've gotten three bests of 3.6 (in a row actually). Not sure how that converts to 60' times, but when I search all other cars' 0-60 times, cars with times of 3.6 seconds are running about 11.90 ET. And the car pulls strong up top, so who knows what it'll do!

AGREED 100%! See goals #1 & 2 above. LOL

Adam. What phone app do you use?

Tapatalk. Why do you ask?

I thought you had a phone app for measuring your car performance? Maybe I misread that.

DUH! Sorry about that. Yes, you read correctly - it was me who read incorrectly when you asked -- I do have a phone app for measuring the following: 0-60, 1/4 mile, as well as a rolling time from any starting to any stopping speeds. (ie, 5-60 MPH).

It's called CT Speedometer Systems PRO. This is an Android app but I'm sure there are Apple versions or similar models. I find it to be accurate AFAIK. But I've only done 0-60 because I'm not very comfortable taking my 64 to 110MPH or more screaming on the streets of LI. Too risky. 0-60 is as I get onto the highway is practically legal. ;)

Here's the results of my experimenting at the track just yesterday. The first timeslip is me shifting manually at 5000. The second is with the trans. in DRIVE shifting it's self at 43-4400. My car is on the right on both slips. The pics aren't super clear, but they read 15.301 and 15.328