[Top] [All Lists]

Shift Points

Subject: Shift Points
Date: Thu, 12 Dec 96 12:56:24 +0100
     I'm sorry, but the 'shift to optimum torque point' is just total BS.  
     It may be an approximate rule of thumb, but to apply equations and 
     comment on how difficult it is to apply is just beyond reality.
     So what is reality: well as any engineer will tell you, it is power 
     that counts -- power at the road wheels. Not strictly BHP, since we 
     have a dynamic situation and one has to consider not only the inertia 
     of the car, but also the inertia of the engine (note that in a higher 
     gear, less energy is used accellerating the engine than in a lower 
     gear, when accellerating the car between the same 2 speeds). Also, 
     losses through in the gearbox may be different in different gears 
     (remember that BHP is usually quoted at the flywheel). 
     There is a peak power point for the engine -- unless you have a CVT 
     gearbox (continuously variable transmission) -- and when accellerating 
     the car, you have to change the engine revs, so usually, you will want 
     to start below the optimum power point and take it past the optimum 
     power point  -- spanning the peak. 
     The Saab turbo example just shows that the advice not difficult to 
     apply to engines with flat torque curves, it shows that it is WRONG!
     What does all this say about when to change up: that the point can 
     only be determined by a detailed study of the engine's power curve, 
     the gear ratios, losses in the gearbox, etc. There is no easy 
     *accurate* solution, although an *approximation* may *perhaps* (I 
     don't personally know) by changing up at the point that will give 
     optimum torque in the higher gear. But since this is only *rule of 
     thumb*, *don't* try to use it to obtain  accurate, definitive numbers.
     Finally, one is rarely driving on a straight endless track, without 
     other traffic, so the optimum shift point may be varied for 
     drivability and control reasons. 

<Prev in Thread] Current Thread [Next in Thread>