Lee200, on Dec 06 2010 - 08:26 AM, said:
Here are some notes on torque converters and how to simulate them in GPL.
The "stall speed" is defined as being V=0 and really has nothing to do with engine speed. It is simply the point where the turbine is not turning. Above the stall speed, the turbine turns and transmits torque to the gears. Passenger car torque converters are designed so that the engine can idle and not produce car motion; especially with the brakes applied. Racing torque converters are designed to have a stall speed that matches the engine's maximum torque rpm to produce maximum car acceleration.
At stall speed, the torque converter multiplies input torque by a factor of around 1.5 to 2.5 times. This multiplication factor decreases in a linear fashion with V and can go below 1.0 when V=1.00. Obviously, this is inefficient and passenger car torque converters have a mechanical clutch that activates around V=.90 which physically locks the impeller and turbine together so they rotate at the same speed. This immediately causes V to be 1.00 and torque multiplication to remain constant at 1.0.
The efficiency of a torque converter is measured as the ratio of output to input torque taking into account the torque multiplier effect. It is parabolically shaped, beginning at 0% at stall speed, rising to a maximum around 80% when V=.80 and decreasing as V increases to 1.00. If a mechanical clutch is fitted and engaged, the efficiency increases linearly from that point until it reaches 97% to 100% when V=1.00.
My research has turned up very little information on maximum torque converter efficiency before clutch engagement, but the few data I've seen show maximum efficiency between 75% to 90%. Close to maximum efficiency occurs over a fairly wide V range of .40 to .90.
The multiplier effect and efficiency probably vary depending on the input torque and impeller speed which are directly related to the engine. However, the only data I've seen and this discussion are based on a constant input torque and impeller speed.
For GPL, we don't have code to create a torque converter so the only way to simulate it is to modify the engine BMEP/torque curve and/or the engine's capacity so that less torque gets to the gearbox. Assuming a racing engine is usually accelerating or decelerating, we could also assume that the torque converter's V will typically be somewhere around .50 where efficiency is only slighlty less than maximum due to the relatively level shape of the parabolic curve at that point.
For the Chapparal 7.0L engine, Richard used an average of about 81% efficiency in the high rpm range where a racing engine normally operates which seems reasonable.
The Chaparral had what was considered at the time to be a "mystery" transmission. From my research today, it seems this was a torque converter coupled to a normal three speed manual transmission. I've found no data on the torque converter's efficiency or whether Chaparral used a clutch to couple the impeller and turbine rotors. I doubt they bothered with an impeller/turbine clutch as a racing torque converter is rarely at V=1.00 where the clutch could enhance efficiency. Also, one source says that Chevrolet didn't introduce the lockup clutch on their torque converters until 1979 so it's doubtful that the Chaparral had one.
Although less efficient than a manual clutch, the torque converter allowed Chaparral to do away with the clutch pedal and thus gave the driver the ability to operate the movable wing with his left foot instead. Obviously, Jim Hall thought the gain from the wing outweighed the loss from the torque converter.
First, this post was only to display some infos.Not putting a doubt on the mod team work on the physics , and I am sure that 's the way you feel it.
I took for granted readers knew about the way a torque converter runs in my post, but thank really to explain it clearly, what I should have done first !
We have some figure for the Chaparral input torque factor.The ratio was 1/2.41.But as you say, it can't be a fixed number and depend on the flow efficiency on various conditions during competition use on different tracks.
For the engine power and torque curve, the graph is the one provided for the 427 engine R & D Chevrolet department.It is reproduced on the engineer Van Valkenburg book " Chevrolet.... racing ? ".But on all the infos we have on the Chaparral 2F, it is well the 525 Bhp that were claimed.Hey, seen what the Midland/Detroit bridge men racing philosophy, they could hide as well some numbers, here...
For the 2G, Jim Hall used the 427 but with higher power figures, and that can be this engine numbers.
Anyway, it won't change much the power curve, as you can compare on the 460 put on the later Chaparral 2 H.
From what I have learned from the Torque converters different architectures, a way to improve the ratio efficiency between Impeller and Turbine, was lighter turbines with less inertia;And a way to get that, was putting two turbines that received the oil flooding .
One thing they could do, too, was machining special shaped Stator. As you know, the stator as an influence on the oil flowing speed, thus on the torque transmitted to the turbine, and to the output shaft.Without any later lock up clutch.
Beyond all this, and I wanted to illustrate with few pics of Tarpaulin covering the gearbox affair, we can wonder now why Jim Hall put such importance on keeping the casing from the eyesights of specialists and competitors .There was not manys things to learn from the casing shape, apart, as Kurleto says the absence of the planetary geartrain.
And the torque converter wasn't a component with problems. Anyway, the tools and the time necessary to work on it, wld have put the car first back to the Midland or Germany garage. No one could know what was really inside in the pits .the funny thing is Van Valkenburg point out some ignorant journalists that couldn't deduct anything from seeing what was on the casing, but nevetheless told stories in races reports....
Maybe Jim Hall had simply to respect a gentleman contract with the R & D guys?
Edited by M Needforspeed, Dec 06 2010 - 09:29 AM.