Determining Upshifting and Downshifting lines for AMT

[NV7891]

I am modeling a vehicle transmission. I have the transmission ratios, Torque vs RPM, sfc vs RPM curves. I would like to know how the up- and down- shifting lines are determined. It would be nice if I could get some help with it.

Thank you.

 

[GregLocock]

Hopefully the full throttle shift points are obvious.

If you have a traditional torque converter with a lockup clutch then you need to consider the torque multiplier characteristics, and the torque capacity of the LUC.

then it gets hard

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[NV7891]

Thank you. I also have a velocity vs RPM curve with earliest upshifting and downshifting positions marked. Could you please direct me to a source where I could read-up on it and understand better. The map I have is as attached. ( source: Automotive Transmission: Fundamentals )

 

[NV7891]

Sorry, hadn't attached the image.

 

[GregLocock]

That's for a manual trans or the very first pass at an AT, for a truck.

Sorry, the little I know is from proprietary material.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[NV7891]

AT for a truck ? Did you mean Automatic transmission for Trucks?

Thank you and please bear with me if the question seems stupid.

 

[GregLocock]

Yes for a truck or even a tractor. Ford made a 10 speed auto for tractors in the 60s.

http://66.49.166.42/VFTR/trans.htm

when they worked they were brillant

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[BrianPetersen]

Clarify your question. What is it that you don't understand?

In the chart that you linked, the engine RPM follows the lines to the right side of the graph. Imagine the driver starting off from a stop and having a fixed upshift RPM of 2100 as per the graph. The revs follow the 1st-gear line to the right as the road speed increases until the upshift RPM is reached, at which point the revs drop (to the left) while road speed remains the same until it finds the 2nd-gear RPM at that road speed (just over 1500 rpm), and so forth, until the maximum speed.

If the driver slows down (braking), it slows to a minimum speed in each gear (evidently around 800 rpm), at which point the revs have to rise to match whatever it will be in the next gear lower (about 1100 rpm), and so on down the graph.

Is that the understanding that you were looking for?

Or do you get that, but you are looking for the *actual* logic needed to establish the upshift and downshift conditions?

*That*, is exceptionally difficult. The upshift RPM will generally depend on throttle position, and if the driver is cruising at low revs in a given gear but then asks for acceleration, when engine load reaches full torque and the driver is still asking for more, it will have to do a downshift (if it is permissible to do so).

Simple logic like that is how old-school hydraulically controlled automatics worked, and even how older-generation electronically-controlled automatics worked. But it leads to an annoying hunting between upshifts and downshifts when going up hills. A lot of newer automatics infer the amount of slope that you are driving on, and establish a minimum gear that it will not upshift beyond, in order to combat hunting. A lot of newer powertrains are "torque-request" based; they infer how much torque the driver is requesting to be put to the wheels, and pick the right gear and throttle position to achieve that without hunting.

A lot of newer automatics will automatically downshift when going down hills, to make use of engine braking, so that the vehicle doesn't tend to run away from the driver (and it reduces the chance of overheating the brakes).

On an auto tranny with lots of ratios (the new 8- and 9-speed automatics, for example), if you always pick the absolute optimum gear at all times, you will drive the drivers nuts with all the shifting. Every tiny uphill and downhill will call for multiple shifts. So there has to be a "stay in the current gear if it is reasonably possible to do so" factor also ...

 

[1gibson]

Find a tuning forum (HP tuners, or specific vehicle forums, start with LS1 stuff 4L60 / 4L80 transmissions.) There will be screen shots of the transmission tables, and if you get lucky you might find free software to open and view different .bin files. A lot of the discussion will be based on just getting swaps to work, but there should also be more details for fine tuning.

 

[Tmoose]

There was a time when smooth, even undetectable, auto trans shifts were prized highly especially at part throttle. One definition was equal rear wheel torque before and after the shift. What that really means is Equal HP before and after the shift. Having the throttle at one position tends to help enforce that condition somewhat. In those precomputer days other mecho-hydraulic days tricks were required, like slightly engaging and progressively applying both gears at once.

 

[BrianPetersen]

and massive torque converter slippage. It's also why 1st gear in the old school traditional 3 speed automatic used a one-way clutch or sprag to handle reaction torque, and a big accumulator and orifice on the 2nd gear clutch. When the 1-2 upshift valve shifted, the accumulator gradually started filling through the orifice and slowly applied pressure to the 2nd gear clutch, and only when that clutch actually transmitted enough torque to take up the load would the 1st-gear overrunning clutch let go.

Nowadays, the drive-by-wire takes care of it ...