Torque converter model for simulation 2

[FredD]

Does anybody have informations about a torque converter model for simulation:<br>
- equations<br>
- valid parameters<br>
<br>
I'd like to simulate an automatic gearbox.

 

[gutz]

do you want to have a physical or computer type

 

[chevy]

I used an exponential buildup function for the ?three? gears. I used the exponential to model the output torque of the converter. Basically, you need to decide how much torque is required to move the vehicle and set the paramenters of the exponential so that this torque is produced at around 1000 rpms. This number very dependent on the engine characteristics, so if you just want to model a generic AT, 1000 is probably as good as anything else. You can basically decide for yourself at what rpm you want the convertor to "lock up". Car manufacturers seem to have varied the lock up setting over the years. A more realistic model that accounts for heat, friction and etc. is difficult without actual experimentation with a specific AT.<br>
I might be able to help more if you described exactly what you want the model to accomplish.<br>
<br>
I am not an automotive engineer, so hopefully if I have given you erroneous informantion someone will correct it. <br>
<br>

 

[brad3378]

Any of you guys find anything since your last post here?

 

[klarin]

Good starting point cann be LuK library (

http://www.luk.de).

 

[cni76]

iam righ now working on torque redcution box
for a two wheeler , if u can tell whether u are going stepup
or down may be i can help u

 

[GregLocock]

Here's some characteristics of a torque converter that we use (approximate):

Stall speed 2000 rpm
Torque multiplication max 2.2 - more like 2 for practical use
Speed at which input speed=output speed (as near as makes no difference) 2600 rpm

We also have a lockup clutch which independently locks the torque converter out.

Having a dig around on the web I can find no open source data that really gives more than that. It is very complex - mapping a torque converter gives a whole bunch of graphs.





Cheers

Greg Locock

 

[MaxRaceSoftware]

Does anybody have informations about a torque converter model for simulation:
- equations
- valid parameters

I'd like to simulate an automatic gearbox.
==================================================
I've developed a couple of computer programs that
have a &quot;Torque Converter Model&quot; in them .
One program..you input HP, ETC and get ET/MPH GForces, ETC each and every foot distance with torque converter
Slip/LockUp % , and torque multiplication values
...the other program goes the opposite way,
you input 60,660,1320 Ft ET and get HP/Torque , ET, MPH, GForces,Aero, Rolling Resistance Losses, ETC each and every foot down 1/4 mile

The &quot;Torque Converter Model&quot; must be very close to real-world because it models in accuracy to .001 thousandths or better most of the time ...howvere , one program version
requires the user to know the max-torque converter multiplication value ..the other program version,
the program &quot;guesstimates&quot; that for you very accurately Larry Meaux (meauxrace2@aol.com)
Meaux Racing Heads
MaxRace Software
ET_Analyst for DragRacers

 

[MaxRaceSoftware]

Engine Speed RPM
(Impeller) = 1800 rpm

Turbine Shaft Torque LockUP
Speed RPM Multiplication Efficiency %
zero 2.25:1 0.0%
175 2.1:1 20.0%
350 2:1 40.0%
600 1.8:1 60.0%
950 1.5:1 80.0%
1350 1.1:1 85.0%
1500 1.05:1 90.0%
1800 1:1 100.0%

you can use a Quadratic Curve Fit equation like
Y = A + B * X + C * X ^ 2
where Y = Torque Converter multiplication
and X = LockUp percent in whole numbers
A=2.226194744185
B=-.0009012975
C=-.000122745
I use different Curve Fit equations in my program but this will get you within .01 to .001 thousandths of a second
accuracy in acceleration simulation .

The above data was a keypart in getting my program to correlate with real world acceleration simulation..
some of the other key parts were to compute Shift Time
and add to Elapsed Time ,,,then you have to compute losses from rotational inertia of engine and drivetrain components,
, tires/wheels ..those 3 areas finally will let your simulation correlate to real world data to within thousandths of a second accuracy


Larry Meaux (meauxrace2@aol.com)
Meaux Racing Heads
MaxRace Software
ET_Analyst for DragRacers

 

[MaxRaceSoftware]

QBasic 1.0, 1.1, or QuickBASIC 4.5 to VBDOS source code
can use in Visual Basic 1.0 to 6.0
------------------------------------------

'CCLock(Ft) is Clutch or Converter Lockup percent
' at each Ft distance

SELECT CASE CCLock(Ft)
CASE 80 TO 100
'*****Quadratic Curve Fit equation for LockUp % from 80% -to- 100%
aquad = 3.465
bquad = -.046071429#
cquad = .00021428571#
Yquad = aquad + (bquad * CCLock(Ft)) + (cquad * CCLock(Ft) ^ 2)
TorqMultiply = Yquad
CASE 79 TO 80
'Linear Fit: y = a + bx
ALinear = 30.591
BLinear = -.368
Ylinear = ALinear + BLinear * CCLock(Ft)
TorqMultiply = Ylinear
CASE ELSE
'*****Linear Curve Fit equation for LockUp % from 0.00% -to- 79.00%
ALinear = 3
BLinear = -.01875
Ylinear = ALinear + BLinear * CCLock(Ft)
TorqMultiply = Ylinear
END SELECT

IF TorqMultiply > ConvMultiply THEN TorqMultiply = ConvMultiply
IF LockUpp = 100 THEN TorqMultiply = 1!


'---TorqMutiply will be times engine torque value
'---ConvMultiply is what the user enters for
'---maximum torque multiplication

Larry Meaux (meauxrace2@aol.com)
Meaux Racing Heads
MaxRace Software
ET_Analyst for DragRacers

 

[tcad]

You can refer to the manual about the gear_box model in the saber. There are a lot of mechanical models.

 

[cshami]

That question can't really be answered so easily. There probably isn't a hardrive big enough to create a model that accounts for all of the various parameters in a TC. For example, changing any one of the 20 or so blade angles can completely change effeciency. Research SAE papers for a good chunk of info. Authors include Marathe, By, Anderson, Lakshminarayana. CFD is the most useful tool for estimating TC performance.