Below is a graph from my recent dyno test. What if anything can I learn from the graph about acceleration and shift points?

I seem to recall you want to be accelerating through the highest torque band. For my car this would seem to be between about 3900 and 4900 RPM. Since there's about a 1500 RPM difference between gears, suggesting a shift point somewhere between 5200 and 5400 RPM? Or am I reading too much into this? The car was in 4th gear for this test. Does this only tell me something about this one gear or can I generalize anything from this?

Generally on the track I shift around 6500 RPMs in 2nd and 3rd gear then run 4th right up to the rev limit which is set at 6900 before shifting into 5th. Since most of my torque is at the lower end of the RPM range, would I be any better off shifting sooner in each gear then staying in 5th longer? Am I right that top speed is going to be in 5th and not 6th? (I have an '04 MCS). Any help appreciated. Chart shows torque and horsepower at the wheels.

The more I learn, the less I know.

 

good question. i don't have dyno info for my car. on my last track weekend, however, i found that by shifting to 5th well before redline (maybe 6k), i had an extra 5mph at the end of the longest straight. anyone w/ more experience or knowledge care to chime in?

 

To answer this question accurately, you have to generate a "thrust chart", which takes into account both the torque curve from the dyno *and* the gear ratios from your transmission/differential.

If you're past the peak torque point for your engine, upshifting can drop the RPM back down closer to peak torque, but the upshift will cause you to lose torque multiplication in the gearbox/differential because of the higher gear ratio after the upshift. As a result, if you shift right after reaching peak torque, you'll lose more from the loss of torque multiplication than you'll gain from dropping the engine back into the peak torque range.

Optimizing shift points involves looking at each gear and figuring out at what RPM you have more to gain from upshifting and dropping the engine RPM than you'll lose from the loss of torque multiplication. It varies depending on the engine/tranny combination, and will even vary from gear to gear, but for most combinations, the "sweet spot" occurs near the engine's peak horsepower point. For the MINI, this is usually near or at redline.

 

ruquiscott: i'm so old that i never thought that i would have to generate a "thrust chart" again. however, i might be able to finallly use my degree in physics. thanks for the response. i'll now do a more scientific analysis of what i leared intuitively.

 

gnhovis,

Here's an example using the numbers from your dyno chart. First, here are the overall gear ratios (including both transmission and final drive) for the pre-2005 Getrag 6-speed:

1st: 11.425 : 1
2nd 7.181 : 1
3rd 5.397 : 1
4th 4.407 : 1
5th 3.656 : 1
6th 2.986 : 1

Now, let's look at a hypothetical shift point. Let's say that you're trying to figure out when to shift from 4th gear into 5th. We'll use 6500rpm as the "test" shift point.

At 6500 rpm, the engine is producing about 145 lbf-ft of torque. The tranny/differential combo multiplies that to 639 lbf-ft (145 lbf-ft X 4.407).

After the upshift to 5th gear, the new RPM will be 5392 RPM (6500 X 3.656 / 4.407). At 5392 RPM, the new engine torque will be about 160 lbf-ft, so the upshift will get you 15 more pound-feet of torque from the engine. But, will it be enough to overcome the loss in multiplication from the transmission in 5th gear?

The new torque value for 5th gear after multiplication in the gearbox is about 585 lbf-ft (160 lbf-ft X 3.656). So you see, even though the engine is making 15 more pound-feet of torque at the new lower RPM, the actual multiplied output from the transmission in 5th gear at 5392 RPM is almost 55 lb-ft LOWER than the 4th gear output at 6500 RPM.

So, 6500 is too early to shift from 4th to 5th, if maximum acceleration is your goal. How much longer should you wait? Well, using 6900 RPM as the test shift point gives us similar results - 625 lbf-ft @ 6900 RPM in 4th gear before the shift, but only 567 lbf-ft @ 5725 RPM in 5th gear after the shift.

Even though 6900 RPM is still "too early" to shift from 4th to 5th, you have to shift anyway, assuming that's the engine's redline. I didn't do the calculations for the rest of the gears, but I suspect the result will be the same. Your particular engine only loses about 25 lbf-ft of torque from its peak at 4100 RPM all the way to redline at 6900 RPM. This means that you'll never gain more than 25 lbf-ft of torque at the engine as the result of an upshift. This gain is likely going to be more than offset by the loss in multiplication in the higher gear.

The short answer to your question is "For an engine where the torque doesn't drop off significantly between the peak torque point and the engine redline, just use the engine's redline as your shift point".

EDIT - I forgot to mention that if you give your tuner the overall gear ratio for each gear (tranny and final drive combined), his dyno software may be able to generate the thrust chart automatically, showing you visually what the best engine speed for each upshift will be.

 

holy cow that is a very intense answer - and probably correct also!

as a drag racer and now track fiend - I always had the same information from any car I was racing.

Drive in a gear until you feel the car fall on it's face slightly - I mean full acceleration - when you feel the car not pulling you'll know it - mark that as your shift point, back it off 100 or 200 RPM's and you'll be fine.

The mini usually pulls to redline - especially if you have a pulley on your set up, so anything just shy is usually good enough

I have an 04 and my gearing is way different from the 05-06 cars, I ride out my gears way longer then Beaner - his method of shifting to 5th earlier was a great move with his gearing, but in my car it bogged down way too much as I am in my gears much longer then he is.

 

Thanks riquiscott, that's just what I was looking for. I'll check if they can generate a thrust chart from the Dyno info they already have.

 

Can you update us if LDG does this? I am very interested in the results. Thanks!!

 

In addition the combined gear ratios for each gear, the only other information they should need is the tire diameter. After that, it's just a matter of whether or not their dyno software will generate the chart.

 

nominated for a STICKY - any 2nds?

 

I'd second that, but maybe that doesn't count since I started the thread.

Anyway, I haven't been able as yet to get the actual data file, but I started to play with the data I do have. I built a spreadsheet to try to estimate the best shift point for each upshift. It's not exact, but I think it's a good approximation based on the available data.

I made a couple of assumptions (always dangerous). Since my data points were 250 RPM apart, I averaged adjacent points to estimate a value in the middle. When the formula calculates the RPM after the upshift, it grabs the engine torque for the closest RPM value that doesn't exceed the calculated amount, rounded down to the nearest 125 RPM increment. (I'm sure someone will be more than happy to tell me if this approach is fundamentally flawed.)

The spreadsheet suggests the following RPMs as shift points based on this method:
1-2: 6750
2-3: 6750
3-4: 6500
4-5: 6625
5-6: 6500

Margin of error is probably +/- 125 RPM due to my estimating.

Example from the spreadsheet:
Calculate net gain(loss) in wheel torque due to upshift
Upshift RPM = (lower gear RPM x higher gear ratio)/ lower gear ratio
Wh torque=eng torque x gear ratio
Gain (Loss) = difference between lower gear wheel torque at given RPM compared to higher gear wheel torque at new RPM after upshift
Current RPM 6625 New RPM 5496
Lower Gear 4 Higher Gear 5
Gear Ratio 4.407 Gear Ratio 3.656
Eng Torque 143.7 Eng Torque 159.2
Wh Torque 633.5 Wh Torque 582.1
Wh Torque (Loss) Gain = (51.4)

Here's a link to the spreadsheet.

If anyone else has RPM-Torque and gearing data, you should be able to plug it in and derive the same calculations yourself. Start in the cell with 2750 RPM since anything below that is just estimated in this spreadsheet.

Thoughts? Am I way off base?

 

not sure I buy that 6500 shift from 3rd to 4th - have a peek...

I stole your data, precomputed the Wheel Torque values for each gear/rpm combination, and graphed it

visually the question can be stated as follows...

where on any of these lines, starting with the top one (1st gear), can one get a higher Y axis value by dropping down to the next lower line (changing up)

 

gnhovis,
cmt is right - your spreadsheet gives the correct answers, but it's answering the wrong question.

To use the 4 -> 5 upshift as an example, your spreadsheet is correct that shifting at 6500 RPM produces the smallest drop in wheel torque after shifting into fifth, but why settle for *any* drop at all? Continuing to rev past 6500, all the way to redline, will result in more wheel torque for a longer period of time. Shifting at 6500 gives you 582 lbf-ft of torque at the wheel once you're in fifth, but remaining in fourth until 6750 gives you 628 lbf-ft prior to the shift, which is significantly more. So, for the few extra seconds you remain in fourth gear, you'll be putting a lot more torque to the wheels than if you had shifted early into fifth.
Acceleration is directly proportional to wheel torque, and velocity is acceleration multiplied by time. So, the longer you can maximize the torque, the higher your velocity will be after a given period of time.
Cmt's graph is a perfect example of an engine whose torque curve doesn't fall off much past the torque peak. From his lines, you can see that even the *highest* wheel torque available in fifth gear is lower than the *lowest* point on the fourth-gear curve. The same is true of all the other gears as well. So, the correct answer for your particular engine is to delay shifting as long as you're comfortable with.
If you looked at the torque curve for a "peakier" engine, the graph that cmt made would actually show some of the wheel-torque curves for adjacent gears *crossing*. In that case, the correct shift point would be where the lines cross. Depending on the gearing, not all adjacent lines will have a crossing point, and for adjacent gears that *do* have crossing points, they may not occur at the same engine speed. This is why the most sophisticated shift lights can actually be programmed with a different shift point for each gear transition.
I'm still going to save your spreadsheet, though - that's a nice piece of work, and will come in handy when there *are* optimum shift points that fall below redline.

 

Ahhh. Stated that way, now I get it. Hence the original comment that for a car with a relatively flat torque curve, shift at redline....