I play around on maintained grippy surfaces with with big sticky tires from time to time on 100+ deg days, pretty much worse case for the locker. It very well could be experience overheat conditions but it has never thrown a code for me and hasn't shown to be an issue during datalogging. My car is low power so maybe that is why. Would honestly be surprised if it was experiencing thermal runaway because its a very similar unit to what is found in a mk7r which naturally demands more out of the unit. There is no noticeable delay in diff action, it is applying current to the clutch packs well before slip is detected, also no complaints reviewing data logs. Would love empirical data from folks who have played with various diffs but I don't know if that data exists. My biggest complaint is there isn't a particularly good rain map--in the rain it seems to lock too hard.
Just checked out your guide. Question about something you wrote:
"What about the “preload” problem?: The worst case scenario for any LSD or the FDL unit is that of "zero load". This could be described as having the car on level ground with all the wheels pointing straight ahead & all on ice, meaning there is no grip/torque difference between the driven wheels."
What does this mean? If there is no grip/torque difference between the driven wheels an open diff = welded diff = VAQ = any differential. There is nothing a diff can do here to help you out here no matter how much it is preloaded.
Obviously you haven't fully read or understood that guide that I wrote...
For starters the motorsports VAQ gets 3 maps with different preloads, has a rain map & doesn't use the XDS systems/ or have ABS, only uses the ABS as a passthrough for the sensors
QUOTE:-
The “Electronic modules” specification lists the “ABS/ESP” unit as “Continental” (actually the MK100 unit), & “Software” for the unit as “Not active”. This proves that although the ABS unit is fitted, it is “inactive”. The reason for this is that the ABS unit is required as a “pass through” for sensors:- G44, G45, G46, G47 (ABS wheel sensors), G200 (lateral acceleration sender), G202 (Yaw rate sender), G251 (longitudinal acceleration sender), as the FDL unit requires the data from these to decide to activate or not! Without these is cannot work!
The 2015 Seat Leon Cup racer lists 3 switchable maps available in the control unit of the FDL unit:-
Map 1:- Base mode, no over slip, only yaw damping above 110km/hr, 300Nm preload during braking. For high grip (new tyres).
Map 2:- Like map 1, less yaw damping, 200Nm preload during braking. Medium grip (used tyres).
Map 3:- Preload dependant on engine torque, 200Nm preload during braking (releasing earlier than Map 2). Low grip/rain.
Please note that this engine has a maximum specified torque of 410Nm, so the FDL unit is applying half, or more, of the engines torque figure as “preload” during braking.
END QUOTE
Regarding the Preload problem you missed this bit in my guide, VAQ/FDL unit first then mechanical LSD:-
QUOTE
How the FDL unit deals with it:-
The FDL unit is not supposed to have this problem as its an electrohydraulic clutch pack diff lock. Unfortunately, it still has a “preload” type problem as it still requires a minimum amount of torque imbalance in Nm between the driven wheels to lock up. If the driven wheels cannot provide enough grip for this then the cars existing XDS+ systems within the ABS unit will be used to apply pressure to the brakes & create a torque imbalance between the driven wheels. This is why it draws heavily on the existing CAN system & sensors.
The FDL unit fitted to the Seat TCR car has an externally adjustable preload setting, with the recommended range being 50Nm (37lbft) to 100Nm (74lbft). Cold measured is 15% higher than warm measure, & preload decreases approx. 15% after 50kms of running. (info from the Seat Leon Cup Racer TCR owners manual)
An Automotive Engineering thesis was conducted in conjunction with Borg Warner AB in Landskrona (Sweden) in 2016. The aim was to develop real world driving scenarios on a test rig so that accurate computer algorithms can be made (instead of having to spend weeks sliding around a test track to get real world data). They did various driveline set ups that the Borg Warner Gen.V unit is used for (AWD, FDL, hang on front, transfer case etc). Unfortunately, they hit problems with the FDL (FXD) set up, quote:-
“The FXD system is the most difficult to recreate due to extreme variations in driveline behaviour - The FDL system will not be implemented due to far to rapid driveline behaviour". They also found on the simulated hill tests that, quote:-
"the torque that is applied upon the clutch pack coupling is varying a lot when the system oscillates & goes from zero to maximum several times. This affects the whole driveline behaviour". Basically, the unit in this configuration is very “reactive” but also “proactive” i.e. it reacts before. The driveline was oscillating like mad, as the unit tried to figure out which wheel to send the torque to! This resulted in huge differences in the recorded torque & angular velocities. This is all because it takes data from many sources, so is constantly adjusting, more so than in the other configurations & backs up my reasoning as to why they added a “bleed container” to remove air from the oil due to aeration.
Ultimately the FDL unit has no direct mechanical way of activating the clutches, so in effect it has a huge “preload” problem. The FDL cannot “sense” torque through the drive shafts & thus cannot use this to activate the pump to activate the clutch pack. It has to be “told” what the wheel speed is from the ABS sensors as it does not have its own sensors. Therefore, even if you leave the FDL unit powered but disconnect it from the CAN bus or just the ABS unit, it will not work as it is not a “stand alone” unit, & thus NOT a mechanical LSD!
How mechanical LSDs deal with it:-
Mechanical LSDs require a small amount of mechanical force/torque imbalance between the driven wheels to overcome the internal resistance in the LSD unit & start the “lock up” & transfer of torque to the wheel with the most grip. Most modern LSD units get around this with various “bump wave” cogs/ramps, or by altering the end friction plates or the gear oil “W” viscosity rating.
Quaife LSDs rely on several "Bellville" washers (coned) which are set up in an alternating pattern & the friction of gears on the end plate to provide the pre-load. The amount of preload in zero load situations can be adjusted by altering the number/type of washers, or by using a different oil viscosity. According to test results in an engineering thesis paper written on the Quaife LSD, it has
"new fresh out of box" preload values of between 13.5Nm to 19Nm (10lbft to 14lbft), with an actual bias of 72% to 95%. Once worn in the Bellville washers give a preload of 5.4Nm to 6.7Nm (4lbft to 5lbft), with an actual bias of 60% to 77%, these values being across all the ranges of acceleration & braking.
Another company called Darkside Developments has an NXG plated LSD with a preload of between 40lbft & max 70lbft, which is similar to the recommended range for the FDL unit.
Wavetrac LSDs have a device in the centre of the diff which responds in zero load situations, quote:-
"Precisely engineered wave profiles are placed on one side gear and its mating preload hub. As the two side gears rotate relative to each other, each wave surface climbs the other, causing them to move apart. Very quickly, this creates enough internal load within the LSD to stop the zero axle-load condition." The amount of preload can be adjusted by changing the standard carbon fibre bias plates for ones made from a different material & thus having a different friction coefficient, quote:-
"These bias plates provide a mechanism to tune the response of the differential as a function of applied torque load. The applied torque load manifests itself as an axial load from the differential pinions into the housing. This axial force is then considered a normal force into the bias plate, and as a function of the effective coefficient of friction, provide a resistive torque to the rotational motion of the differential pinions. The resistive torque will add to the resistance of relative rotation of all components within the differential. The resistive force, however, is non-uniform since it is a function of the axial load from the differential pinions. The unbalance of the resistive torque will manifest as non-uniform energy absorption within the differential causing a bias ratio.".
I also found a report by someone who had a Quaife LSD in a FWD car & worked with Wavetrac to develop an LSD for his make of car (i.e. used his as a test car). He found that the Quaife was more "on -off" in its transitions whereas the Wavetrac was more neutral/smoother in its transitions.
END QUOTE.