Transfer case on AWD models: was it meant to be reliable ?

[dgorbenko]

yeah, I saw this video couple days ago, nice one. I wish Hyundai did the same and install a fork there (in Stinger's transfer case) to engage the clutches. I'm close to get back my stinger to the garage again and continue looking into problem. Still haven't found a clutch pack and still I have some knocking sound happening twice per wheel rotation. Need to lift the car and start searching...

 

[St1nger Guy]

So if the fuse is pulled to shut off the oil pump aka drift mode would this also help the clutch plates last longer since they wouldn't be used? Most of the time I don't need AWD and if knew I could extend the life of my transfer case by pulling the fuse I would go to the trouble of installing a drift switch to shut-off the oil pump. I imagine I'd probably kill more rear tires but that sounds like a perk to me.

 

[D.J.]

So if the fuse is pulled to shut off the oil pump aka drift mode would this also help the clutch plates last longer since they wouldn't be used? Most of the time I don't need AWD and if knew I could extend the life of my transfer case by pulling the fuse I would go to the trouble of installing a drift switch to shut-off the oil pump. I imagine I'd probably kill more rear tires but that sounds like a perk to me.

if most the time you don't need AWD, most of the time the front and rear shafts will be at the same RPM - As in, no wear of the clutches at all. the input and output would be the same RPM - so essentially no wear of the cutch pack would occur.

 

[St1nger Guy]

So fun mode wouldn't also save the clutches?

 

[oddball]

So fun mode wouldn't also save the clutches?

Specifically the opposite.
Now the clutches will always be disengaged. But a wet clutch pack is never *really* disengaged like a manual clutch. The steels and frictions always rub against each other. So now they're spinning wildly against each other all the time (except for perfect straight line cruising) and there's NO cooling or lubricating flow. The needle bearings on the shafts are also probably not getting the lubrication they need.

Short bouts are fine, but that fluid does a bunch of jobs.

 

[LordKOTL]

Bumping out of curiosity--in case anyone knows since this thread taught me a ton of how the transfer case works.

Regarding the way the transfer case splits out power according to the marketing rags, (i.e. 50/50 f/r, 40/60, 20/80), am I reasonable in thinking that the amount of pressure that piston puts on the clutch basket controls how much power can be allocated to the front wheels? Like, if the piston is fully engaged it splits 50/50. If the piston is disengaged the clutch pack is not completely freewheeling, so you get something like 20/80 or whatever the "drift mode" power split is because there still is some friction between the clutch plates.

Thus, could we assume that the piston is not simply on/off, but rather it could have gradations to the amount of pressure it applies to the clutch pack, and thus that's how it can alter how much power the front wheels get?, so if the piston is only partially engaged it would allow a controlled amount of slippage so the front clutches "break away" at a given power split (i.e. 20%, 40%, etc.)?

Again, just curious. Thanks in advance if anyone knows.

 

[dgorbenko]

Probably, yes. I don't see any other option beside that.

Just out of curiosity, again, I don't really understand how the torque can ever be split to anything but 50/50 for a prolonged time period. Let's say you wanna 30/70 f/r. How do you achieve that ? By allowing frictions and steels to slip against each other ? OK, it makes sense until the car starts moving, because at that exact moment the front and rear wheels are going to rotate with the same angle speed, meaning, there will be no slippage between frictions and steels in transfer case.

Or, take at look at a transfer case from another point of view: you drive at 60mph, the pump doesn't compress clutch pack, and since the front and rear wheels rotate with the same speed does it mean the torque is split 50/50 without engaging compression ?

 

[Submarines]

Probably, yes. I don't see any other option beside that.

Just out of curiosity, again, I don't really understand how the torque can ever be split to anything but 50/50 for a prolonged time period. Let's say you wanna 30/70 f/r. How do you achieve that ? By allowing frictions and steels to slip against each other ? OK, it makes sense until the car starts moving, because at that exact moment the front and rear wheels are going to rotate with the same angle speed, meaning, there will be no slippage between frictions and steels in transfer case.

Or, take at look at a transfer case from another point of view: you drive at 60mph, the pump doesn't compress clutch pack, and since the front and rear wheels rotate with the same speed does it mean the torque is split 50/50 without engaging compression ?

You're forgetting transfer of torque to the wheels and road.

The shafts are rotating at the same speed, but power isn't being transmitted to the front, the shaft is moving from the front wheels being pushed. Torque is the force to rotate with resistance, if there is no resistance than it's free wheeling hence no torque and pure RWD.

 

[oddball]

Wet clutches can be partially engaged. It's not the best for them and you wouldn't want to do it under high power all the time, but they certainly support it. That's literally how automatic transmission shifts work - the outgoing and incoming clutch packs both slip while transitioning, allowing the shift to be smooth.
That's why that clutch pack has so many clutches in it. A normal automatic trans clutch, even of that size, has maybe four clutches. Sometimes more - sometimes less! The overrun clutch in the 200-4r has *two* clutches of that size! Anyway, having so many clutches lets it slip without burning. That's the main risk. As long as there's enough fluid moving through the pack to cool it, it can be partially engaged and slip via the fluid. Once there's too much heat then the steels warp, the lining bakes, and things go bad fast. The large number of clutches spreads out the heat.

So, yes, for variable torque split, the pump will be pushing enough fluid to partially engage the clutches. For 50/50, the clutches are fully engaged. Note: it might not be 50/50 depending on gearing, full driveline setup, etc etc. E.g. if some company says the most they can do is 60/40 (FWD cars with an AWD option typically can only send less than 50% to the rear) that's because of how the system is built. The clutches in the transfer case are fully engaged for that 60/40 split, just the physics of the driveline make the split 60/40.

Like Submarines said, there's *0* torque split when the clutches aren't engaged (RWD mode only) and just cruising along. Everything attached to the front wheels is just spinning - no power transfer.

 

[LordKOTL]

Wet clutches can be partially engaged. It's not the best for them and you wouldn't want to do it under high power all the time, but they certainly support it. That's literally how automatic transmission shifts work - the outgoing and incoming clutch packs both slip while transitioning, allowing the shift to be smooth.
That's why that clutch pack has so many clutches in it. A normal automatic trans clutch, even of that size, has maybe four clutches. Sometimes more - sometimes less! The overrun clutch in the 200-4r has *two* clutches of that size! Anyway, having so many clutches lets it slip without burning. That's the main risk. As long as there's enough fluid moving through the pack to cool it, it can be partially engaged and slip via the fluid. Once there's too much heat then the steels warp, the lining bakes, and things go bad fast. The large number of clutches spreads out the heat.

So, yes, for variable torque split, the pump will be pushing enough fluid to partially engage the clutches. For 50/50, the clutches are fully engaged. Note: it might not be 50/50 depending on gearing, full driveline setup, etc etc. E.g. if some company says the most they can do is 60/40 (FWD cars with an AWD option typically can only send less than 50% to the rear) that's because of how the system is built. The clutches in the transfer case are fully engaged for that 60/40 split, just the physics of the driveline make the split 60/40.

Like Submarines said, there's *0* torque split when the clutches aren't engaged (RWD mode only) and just cruising along. Everything attached to the front wheels is just spinning - no power transfer.

That's about what I was thinking: what you and @Submarines said. Just driving in a straight line the wheels should all be turning at the same rate; no slippage happens even though slippage can happen.

So if I'm understanding this right, if I'm driving into a curve: Until I enter the curve all wheels are spinning the same speed No slippage happens. Once the front and rear wheels start spinning different speeds then the clutches can slip. The amount of slippage determines the power split. So in sport at ~20f/80r power distribution the clutches can slip more, but since there is still some pressure on the clutch pack some power does get transferred up front ~20%. This would make it feel a lot more rear-biased and the rear end would "push" the front around a curve. But if it's in a 50/50 balance (as the marketing bumph says for Comfort mode*) then the clutches are fully locked together and any slippage that does happen, it means that there is still a lot of power transferred up front so they should be getting something as close to 50% of the power with respect to the design of everything, so it would feel more even.

* - All of the marketing material I've read says, "up to 50-50." I'm pretty sure that there has to be some electronic logic which would update the balance of power continuously so that, in say eco mode which is advertised at ~40f/60r it would try to maintain that but it can vary based on driving conditions.

Thanks to you, @Submarines and @dgorbenko . Very informative!

 

[D.J.]

from this source

Kia Stinger - 4 Wheel Drive (AWD) System

Description and operation​

Description

–​	This system is to optimize driving performance through distributing the driving force to the front and rear wheels actively depending on the road condition and driving condition.

① Distribution of electronically controlled driving force

•​	The magnitude of the driving force transmitted to the front wheel is variably controlled by fine-tuning the assembling torque on the electronic AWD transfer case (Multi disk) through recognizing the road condition and driving state from the sensor signals input to the AWD ECU.

② Mutual controlling with the related driving control system

•​	It acquires the best performance in car driving through mutual controlling with brake system such as ABS and VDC.

•​	The driving force transmitted from the engine and transmission to the transfer case is transmitted to the rear differential through the propeller shaft. It is transmitted to the front differential through the propeller shaft when the operation of AWD is required.

1.

Structure of transfer case –​ The AWD ECU distributes the driving force to the front/rear wheel through controlling the multi plate clutch on the AWD transfer case by analyzing the input information, i.e. the wheel speed, accelerator and steering angle depending on the road condition and driving state.

 

[D.J.]

And notably from here

Kia Stinger - AWD Control System - 4 Wheel Drive (AWD) System

When a vehicle runs at normal speed higher than 60KPH on regular roads, it is controlled under the 2WD conditions.

 

[LordKOTL]

And notably from here

Kia Stinger - AWD Control System - 4 Wheel Drive (AWD) System

When a vehicle runs at normal speed higher than 60KPH on regular roads, it is controlled under the 2WD conditions.

Thank you, great info! Like I said I was curious as to how our rides work. I assumed that it was like a typical AWD where it had a center differential, then I saw we had a transfer case but none of the typical selector knobs or settings like on a typical 4wd car I'm used to seeing, and then I came across this thread and the internals didn't look like the innards of a typical transfer case (i.e. no chain).

I think that does make sense for controlling the car under 2wd (I'm assuming RWD) above 60kph (37.28 in freedom units), but truth be told I've never taken a sharp enough curve at ~40mph which I would be able to tell the difference in handling between the marketed power balance between the drive modes, and given that it just rained today I'd want to wait. . Most of my experiences is if I take a corner aggressively, but not much more than say, 25-30mph. Sport mode it tends to push the rear out like a RWD car, while in comfort and in Eco it tends to be a lot more balanced like I'd expect from an AWD car. So, I just wondered how the car did it and how it worked.

Again, thank you for the info!

 

[vadelavicolte]

from Kia's service manual (I have access), ECU commands oil pump's output to control engagement of the clutch pack. if it's viscous coupling, it won't need a pump, just splash action is enough for lubrication. High duty cycle=clutch engagement for 4wd, low or no duty cycle = low torque transfer or no engagement, I hope it helps

 

[Thomby]

That's about what I was thinking: what you and @Submarines said. Just driving in a straight line the wheels should all be turning at the same rate; no slippage happens even though slippage can happen.

So if I'm understanding this right, if I'm driving into a curve: Until I enter the curve all wheels are spinning the same speed No slippage happens. Once the front and rear wheels start spinning different speeds then the clutches can slip. The amount of slippage determines the power split. So in sport at ~20f/80r power distribution the clutches can slip more, but since there is still some pressure on the clutch pack some power does get transferred up front ~20%. This would make it feel a lot more rear-biased and the rear end would "push" the front around a curve. But if it's in a 50/50 balance (as the marketing bumph says for Comfort mode*) then the clutches are fully locked together and any slippage that does happen, it means that there is still a lot of power transferred up front so they should be getting something as close to 50% of the power with respect to the design of everything, so it would feel more even.

* - All of the marketing material I've read says, "up to 50-50." I'm pretty sure that there has to be some electronic logic which would update the balance of power continuously so that, in say eco mode which is advertised at ~40f/60r it would try to maintain that but it can vary based on driving conditions.

Thanks to you, @Submarines and @dgorbenko . Very informative!

If the transfer case/center diff can truly lock up, then front and rear should spin at the same speed, and if your rears are on ice, your fronts should get 100% of torque. If we're capped at 50/50, to me that implies that the transfer case cannot lock up hard enough to handle all the engine torque, and once it slips, your front torque is limited. If you have no grip at all in the rear, you won't be able to put anything down up front.

But my guess is that 50% is just an estimate based on a specific condition, because the transfer case breakaway torque isn't going to vary across your torque curve to always be 50% (if it can hold 50% of 376 lbs-ft, it can hold 100% any time your output is 188 or less). Maybe someone calculated that across a certain rpm range at partial throttle you're going to be around 200 lbs-ft and the transfer case can handle 100 lbs-ft, so it's "up to 50/50" but if you're trying to get out of a snow bank at low speed/throttle, you'll be well under that 100 lbs-ft limit and therefore can send everything to the fronts.

Thoughts?

 

[D.J.]

Sorry a didn't capture this data log at a more eventful drive ... I wish I did this on snow/Ice.

But it should be interesting regardless.

Car was in ECO mode during this video
Can see the clutch pressure increase with any acceleration.

I have noticed that the clutch pressure is way higher and more frequent when in sport mode, but I don't have any recordings of that yet, only my own observations.

I used the music to sync the screen record overlay with the dashcam footage, if you don't like the music, just click mute.

App used for data = Car scanner (Android)
OBDII reader is a cheap $10 Chinese version.
Dashcam is a Viofo A119 V2 from 2016 (1080p @60fps)


1. Clutch pressure is likely accurate in PSI
2. Clutch torque is probably a calculated value (probably not accurate)
3. Clutch duty % probably accurate
4. Clutch stress - not sure what that is or how it's measured
5. Front rear delta speed rpm - easily measured with wheel speed sensors.
6. Front axle ratio - easily measured with wheel speed sensors
7. Accelerator pedal = true value
8. Speed - GPS speed

 

[Thomby]

2. Clutch torque is probably a calculated value (probably not accurate)

Haha that's too bad, couple wild values at the beginning there:

 

[D.J.]

Haha that's too bad, couple wild values at the beginning there:

View attachment 85759

It is in the realm of possibility, that those values are true.. and they are taken after the torque multiplication of the transmission gear ratios.

Transmission Gear Ratios: 8-speed automatic transmission
First 3.665:1
Second 2.396:1
Third 1.610:1
Fourth 1.190:1
Fifth 1.000:1
Sixth 0.826:1
Seventh 0.643:1
Eighth 0.556:1
Reverse 2.273:1
Final gear ratio 3.538:1


The amount of shaft torque at the transfer case, in 1st gear is substantially higher than engine torque at the engine, about 3.665 times higher.

I don't know where these torque values come from, there isn't a "sensor" for this.

 

[Thomby]

Oh that's a good point, wasn't thinking about being downstream of the transmission. This further convinces me that the "50/50 split" is an estimate based on what that clutch can handle before it slips.

 

[oddball]

I think there's some misunderstanding about what "split" means when talking about a transfer case and how advertising for transfer cases and differentials works.
The stinger is a RWD platform where a transfer case can engage a front driveshaft in addition to the rear driveshaft. The rear cannot disengage. Other platforms will behave a little differently, like FWD cars that can engage a rear driveshaft, or the rare bird where the transfer case really can variably engage both outputs.
When the clutches are loose, 100% of power goes to rear wheels and the front freely spins. When the clutches are fully engaged, the power is split evenly between the front and rear driveshafts because of how the gearing is set up. The rear cannot disengage - the rear driveshaft is turned 1:1 with the transfer case input because it's a solid shaft.
If the rears hit ice, then it's still a 50/50 split when the transfer case is engaged. 50% of engine power goes to rear wheels which will slip on the ice, and 50% will go to the front which, hopefully, have traction. If this was an advertisement then we would say "100% of torque goes to the front wheels!" because the rear wheels aren't doing anything. Ads for a limited slip differential will say "the wheel with traction gets 100%" while the other wheel is just along for the ride. Both phrases are correct, it just depends on what, precisely, is being discussed. The usage scenarios for a transfer case are just different than a differential. Kia won't say "100% torque to front wheels!" because the rear cannot be disengaged. But, if the rear wheels are up in the air, and the transfer case clutch is locked, then sure, 100% torque will go to the front. That's not a normal scenario though.

If the clutch in the transfer case slips, then the front gets less than 50%. A solidly locked clutch means 50% split.

The ECU calculates torque demand and assumes the engine is actually producing that output. The manufacturers spend a lot of time figuring out those tables, so it's likely pretty close at least with a stock configuration. A wide array of other values in the ECU are just calculated. e.g. there's a turbine speed value in the ECU! Some turbos actually have turbine speed sensors, but not the Stinger.

The clutch in the transfer case is pretty small, but there are several frictions. Wet clutches are generally quite strong once they are locked, but the periods between fully locked and fully open generate a lot of heat. The clutch material will burn and flake off and the steels will overheat and warp if the transitions aren't handled right. The ECU will vary the line pressure because more pressure than necessary to keep the clutch locked (or slipping just the amount they want) is just wasted energy. We're at the point where someone can spend a few hours playing with a table and reduce energy usage by a tiny fraction of a percent, so they might as well. Just like how line pressure in a transmission varies based on conditions. A race transmission will usually be maximum line pressure all the time - that's fine for racing, but it takes a lot of power to run the fluid pump and the high pressure is harder on parts.
So when just cruising down the highway, a low line pressure is enough to keep the clutch locked. When you lean into it, the pressure has to go up to keep it locked. You'll see the same thing on the line pressure of the transmission.