Matt
Admin
Admin
| Posts: 485 |  |
|
Mitsubishi Diffs - 2005/11/07 11:05
Does anyone have a good understanding of how the Mistu diffs work? I've been put under the impression that the rear diff also takes the job of the center diff. can anyone confirm or rebuke this? 
|
|
|
The administrator has disabled public write access. |
Matt
Admin
Admin
| Posts: 485 | |
|
Re:Mitsubishi Diffs - 2005/11/08 08:17
I found this spec page for the 2004 car. It has some description of the helical LSD, but not what I was looking for. http://maximum-cars.com/Cars/Car.php?carnumber=567
|
|
|
The administrator has disabled public write access. |
Matt
Admin
Admin
| Posts: 485 | |
|
Re:Mitsubishi Diffs - 2005/11/08 08:29
Hope fully some one blows me out of the watter on this one, because I must have been full of crap. Anyway, the 2006 is pretty sexy.
Post edited by: matt, at: 2005/11/08 08:31
|
|
|
The administrator has disabled public write access. |
Micah Fedke
User
Fan
| Posts: 3 |  |
|
Re:Mitsubishi Diffs - 2005/11/08 11:02
mr. i,
first, you know this, ( http://en.wikipedia.org/wiki/Limited_slip_differential ) right?
here's a little more: ( http://auto.howstuffworks.com/differential.htm ), but not nearly enough on the topic.
alright, so to answer your question, lets go to the source - mitsu's website: ( http://www.mitsubishicars.com/MMSA/jsp/lancerevolution/features_specs.jsp?t=mechanical )
here we're seeing that the new 9's have a torque-sensing up front, an "active" center, and a rear "mechanical". there's a lot of double speak there.
the front is pretty straight forward - the torque sensing differential will act like an open diff when you don't need it (ie. driving straight down a transit road, parked outside a fat chick's house at midnight, on the trailer to the body shop, whatever), and it will quickly scale itself up to and beyond the effect of a locker or a welded diff (which is a constant 50/50 torque split) when you start to spin one of the front tires.
the center is actually mitsu's special ACD unit. that means its a clutch-type differential, but with a computer brain moving the clutches, not some cold, uncaring spring pack. that means the computer can control the torque split between the front and the back, as necessary. say you're coming around a corner in your evo, and you get on it hard - what's going to happen? the nose is going to pick up, lose traction, and start plowing. but the computer senses this. it pulls some of that torque away from the front wheels, letting them push less and regain some traction, and it puts that extra torque to the back wheels, allowing them to push the back end around, and, given enough throttle, swing the back end out. now, a regular clutch or torque sensing center diff could have done the same thing, right? right. but not until you started to slip, deep in the corner. a regular mechanical diff would offer only a recovery plan. but the computer controlled diff looked at your speed, your throttle input, your steering input, your relative wheel speeds and all sorts of other stuff, and it predicted your situation. it already began to increase the rotation potential of the car as you entered the corner, even before the front end even gave out.
you can also make "suggestions" to the ACD (gravel/tarmac/snow) with a special switch on the dash. that just gives the ACD's ECU a starting point to work from.
the rear diff may or may not be mitsu's special AYC-enabled diff. as i understand it, that would basically be a computer controlled torque sensing diff. (don't ask me what's inside THAT thing) if it's not, then it's just a regular torque sensing diff. if it IS AYC, then it's going to enhance the effect of a regular torque sensing differential with more computerized, predictive jiggery pokery. i don't know a lot about this one obviously, but i do know that it works in much the same way as the active center diff, using various inputs to predictively actuate the differential and make the car feel as if it is steering around a corner as soon as you turn the wheel, instead of playing catchup once you are in the middle of a corner.
mitsu releases japanese and european evo 8's and 9's (maybe 7's too? i'm not sure) with ACD and AYC. one computer controls both diffs, meaning that it can use the benefits of both together, to create some kind of superhero drivetrain. in america, where breasts are big and folks are scared of buying evos, we got a real "baby's first evo" in 2003. no computers anywhere in the drivetrain, with a mere open front / viscous center / torque sensing rear setup. it works, sure. but it was definitely watered down compared to what our overseas friends play with. i think the evo MR (and perhaps the regular 2005's as well) began to get the ACD. but i didn't hear anything about AYC for the 2005's. so does the new american 9 have AYC? i'm not sure, but if someone could chime in with a yes/no, that would be great.
hope this helps.
|
|
|
The administrator has disabled public write access. |
ryanrlamothe
Moderator
Moderator
| Posts: 221 | |
|
Re:Mitsubishi Diffs - 2005/11/08 16:08
Micah,
Hey, thanks for the reply man 
A couple of things, since I have gone through the rigorous work of learning differential design recently. There are a few things about differentials:
1) There are essentially five types of limited slip differentials
- Viscous Coupled (Passive)
- Worm Gears/TorSen/Helical (Passive)
- Mechanical Plated/Clutched (Passive)
- Hydraulically Actuated Plated/Clutched (Active)
- Mechanically Actuated Plated/Clutched (Active)
2) It is easy to confuse the terminology of all of them, but these are the correct descriptions
3) No limited slip differential can produce more binding torque than a welded differentials or mechanically locked differential. All limited slip differentials have a maximum shear torque, at which point the ability to couple %100 begins to gradually reduce. For example, most viscous limited slip differentials have a shear torque of approximately 45lb.ft. The best limited slip design will be able to deliver binding shear strength equal to or greater than the torque input (in most cases that equals engine output plus gearing ratio). But due to heat, size, weight, and other considerations that is usually not designed/achieved. AWD drivetrains are designed to deliver maximum torque needed during full dynamic conditions, but are not usually able to deliver all of the available torque to only one wheel. For example, to pull a Rally car out of a ditch using only one wheel if the other three wheels are not on the ground.
4) Worm Gears/TorSen/Helical (Passive) are an interesting design. These differentials cannot couple when any one of the shafts is able to rotate without resistance (although Quaife limited slip differentials contain a friction modifier that allows a lower torque limit before release to allow the differential to function with one wheel off the ground under special situations). Even given that limitation, these differentials are desired for their near instant ability to couple and decouple, their exceptional torque biasing ratio capabilties, and the fact that they do not cause torque bind during braking. Since they are unable to couple without both shafts having rotational resistance differences, they have not been extremely popular recently in Rally in light of newer differential technologies. One possible solution that has been used recently is the usage of the brakes to stop a wheel slipping and sending the torque back through the differential to "work against itself" and cause the Worm Gears/TorSen/Helical (Passive) to work as though all shafts have resistance.
5) A Mechanically Plated/Clutch (Passive) differential can be shimmed to achieve high torque shear strengths and ramp/preloaded to give extremely responsive torque input throttle reaction. But, Mechanically Plated/Clutch (Passive) differentials are statically set, so too high of a torque shear and it will act exactly like a welded/locked differential, which is undesired under constant operation.
6) A viscously coupled limited slip differential couples on both acceleration and deceleration, as it couples due to any differential in shaft rotation, and is not dependent on shaft rotation direction. This is why viscously coupled differentials are not used as the front differential, because the differential would be coupled during deceleration causing understeer. This is also why Mechanically Plated/Clutched (Passive) differentials come in 1, 1.5, and 2 way configurations.
7) The time required for full-lockup before shear torque shear is extremely important as well. For instance, a hydraulically actuated plated differential can take approximately 300ms to couple, whereas a mechanically actuated plated differential can take approximately 24ms to couple. A viscous coupling differential can take up to 3 seconds, although that number is an average, and the exact calculation is completely dependent on the differential in shaft rotation speed and existing fluid temperature levels.
8) Most active differentials work with a VSS (Vehicle Speed Sensor) or for better description a wheel speed indicator. In this case, the TCU (Transmission Control Unit) works in combination with the ECU (Engine Control Unit) to conclude which wheel is slipping and attempt to move torque to the wheel(s) which are not slipping.
9) In the case of such technologies as AYC (Active Yaw Control) the ECU also uses sensors such as the steering wheel input sensor to determine steering wheel angle, accelerometer, and lateral g-meter to determine where the driver is intending the car to go in relation to where it is actually going. For example, in a left hand turn, if the vehicle is understeering the ECU will reduce the front passenger side coupling and increase the rear drivers side coupling to induce the front to swing in alignment with the rear in the path of travel.
10) To continue #8, experience has shown that the desired setup for an AWD drivetrain on gravel is to have all differentials coupled either partially or fully during acceleration and decoupled fully during deceleration. This is due to a control issue, where a coupled front and center differential will produce undesired understeer and a coupled rear differential will produce undersired oversteer. With a completely "active" drivetrain, those decisions are made within the ECU and delivered to the differentials to give the best desired performance. With a partial "active" drivetrain, the best that the ECU can do is guess as to how the front and rear differentials are going to react to a given input. This is why you see many Rally cars with Mechanically Plated/Clutch (Passive) front and rear differentials that are either 1 or 1.5 way, so that they do not couple during deceleration.
11) The US-spec Mitsubishi Lancer Evolution (RS, IX, MR) has the following configuration
- Worm Gears/TorSen/Helical FRONT (Passive)
- Hydraulically Actuated Plated/Clutched CENTER (Active)
- Mechanical Plated/Clutch 1.5 REAR (Passive)
Post edited by: ryanrlamothe, at: 2005/11/08 17:11
|
|
|
The administrator has disabled public write access. |
|
Forums
