• HPF M3 loses control at the strip

      We just received these videos of an HPF turbo M3 which lost control at the Irwindale 1/8 mile strip and ended up going into the wall. The car looks like it is running bias-ply slicks with radial street tires up front. This may have contributed to the driver losing control as the general rule is not to mix due to instability up top. Additionally, once the car starts to sway it looks like the owner stays in it and then tried to brake too late which causes him to lose it. Hopefully we will have more details on the car and the setup soon.





      This article was originally published in forum thread: HPF M3 loses control at the strip started by Sticky View original post
      Comments 195 Comments
      1. spdu4ea's Avatar
        spdu4ea -
        Click here to enlarge Originally Posted by Sticky Click here to enlarge
        Sorry but part of this is incorrect, an open diff applies the same torque to both wheels at all times. Each wheel does not get different amounts.
        Different ways of looking at it... Jack one wheel in the air like this wavetrac video:

        and you'll get ~0tq to both wheels despite the open diff's input shaft getting >0 power/torque from the engine. All of that energy is wasted through the lifted wheel. In this same scenerio, a spool will transfer power/torque equally to both axles, but only 1 will be able to use it.


        Click here to enlarge Originally Posted by Sticky Click here to enlarge
        Something you guys need to understand about a Limited Slip, which he was already running:
        You got that quote from a how-stuff-works article (inaccurately) describing the operation of a conventional clutch-type LSD (http://auto.howstuffworks.com/differential4.htm). That description is inaccurate in this part:

        Click here to enlarge Originally Posted by how stuff doesn't work Click here to enlarge
        the only time the clutches step in is when something happens to make one wheel spin faster than the other, as in a turn. The clutches fight this behavior, wanting both wheels to go the same speed. If one wheel wants to spin faster than the other, it must first overpower the clutch
        Again, the clutches are activated based on a minimum input torque being met. If you're coasting in neutral with an LSD, no torque is activating the clutch packs so there is no resistance to wheel speed differentiation. Think of it like a manual transmission:

        No torque going through diff = clutch pedal pressed to the floor (clutch disk pulled away from mating surfaces and freewheeling)
        Some amount less than lock-up torque going through diff = clutch pedal held partway between floor and released (clutch disk slipping between mating surfaces)
        More than lock-up torque going through diff = clutch pedal released (clutch disck forced against mating surfaces preventing slip)


        The 01+ M3 uses a 100% different, speed-based LSD design: the GKN Visco-Lok. BMW calls it the M Variable, but it is not exclusive to BMW and is also found in the '08-10 Viper... Whereas under load a clutch type differential tries to lock the axle (both wheels turning same speed), the visco-lok is reactive -- it needs a wheelspeed difference before it applies torque to the other wheel. It never, ever, ever, ever tries to lock both axles together.

        The visco-Lok's functions essentially the same as a gerotor pump-type LSD, described here by wikipedia:

        Gerotor pump
        This works by hydraulically compressing a clutch pack. The gerotor pump uses the housing to drive the outer side of the pump and one axle shaft to drive the other. When there is differential wheel rotation, the pump pressurizes its working fluid into the clutch pack area. This provides a clamp load for frictional resistance to transfer torque to the higher traction wheel. The pump based systems have a lower and upper limits on applied pressure, and internal damping to avoid hysteresis. The newest gerotor pump based system has computer regulated output for more versatility and no oscillation.
        Hmm problems with oscillation sound familiar? There are no electronics in the E46 M3 LSD (not sure about E9x)... And the internal damping and pressure limits were calibrated for 262 ft-lbs -- not 640...
      1. Sticky's Avatar
        Sticky -
        Click here to enlarge Originally Posted by spdu4ea Click here to enlarge
        and you'll get ~0tq to both wheels despite the open diff's input shaft getting >0 power/torque from the engine. All of that energy is wasted through the lifted wheel. In this same scenerio, a spool will transfer power/torque equally to both axles, but only 1 will be able to use it.
        Ok, and? Maybe different ways of looking at it but it is still a constant that an open diff provides the same torque to both wheels at all times. They must spin at the same speed, period.

        Click here to enlarge Originally Posted by spdu4ea Click here to enlarge
        Again, the clutches are activated based on a minimum input torque being met. If you're coasting in neutral with an LSD, no torque is activating the clutch packs so there is no resistance to wheel speed differentiation. Think of it like a manual transmission:

        No torque going through diff = clutch pedal pressed to the floor (clutch disk pulled away from mating surfaces and freewheeling)
        Some amount less than lock-up torque going through diff = clutch pedal held partway between floor and released (clutch disk slipping between mating surfaces)
        More than lock-up torque going through diff = clutch pedal released (clutch disck forced against mating surfaces preventing slip)


        The 01+ M3 uses a 100% different, speed-based LSD design: the GKN Visco-Lok. BMW calls it the M Variable, but it is not exclusive to BMW and is also found in the '08-10 Viper... Whereas under load a clutch type differential tries to lock the axle (both wheels turning same speed), the visco-lok is reactive -- it needs a wheelspeed difference before it applies torque to the other wheel. It never, ever, ever, ever tries to lock both axles together.

        The visco-Lok's functions essentially the same as a gerotor pump-type LSD, described here by wikipedia:
        I think we are getting away from the point here.

        The OS Gilken is a clutch type isn't it? How would that help exactly? It wouldn't.

        The M3's diff is a limited slip (speed sensing design) and it didn't save the car, so the differential discussion is really pointless.
      1. spdu4ea's Avatar
        spdu4ea -
        Click here to enlarge Originally Posted by Sticky Click here to enlarge
        an open diff provides the same torque to both wheels at all times. They must spin at the same speed, period.




        I think we are getting away from the point here.
        thread split?

        The OS Gilken is a clutch type isn't it? How would that help exactly? It wouldn't.
        The OS Giken is a clutch type and it would lock on launch forcing both axles to turn the same speed. It would have helped prevent the rear end instability that caused the rear end to start veering to the right @ 1:10 -- which lead to the tankslapping overcorrections after the 2-3 shift (all of course still ultimately caused by driver staying in it beyond his skill)

        The M3's diff is a limited slip (speed sensing design) and it didn't save the car, so the differential discussion is really pointless.
        The M3's LSD doesn't react until after there is a wheelspeed difference.


        Listen closely during the 1st second, you'll hear the LR start squabbling for traction & spinning (and see the rear of the car hunkering down) before the right rear wheel even moves. This react-compensate speed-based design results in differing wheel speeds the entire time traction is low -- totally different from a clutch type...


        Don't get hung up on the term "LSD." It encompasses a wide range of different designs, features, and operation. Pump action, double-barrel, and semi-automatic are all shotgun types -- but they differ greatly.
      1. Sticky's Avatar
        Sticky -
        Click here to enlarge Originally Posted by spdu4ea Click here to enlarge
        The OS Giken is a clutch type and it would lock on launch forcing both axles to turn the same speed. It would have helped prevent the rear end instability that caused the rear end to start veering to the right @ 1:10 -- which lead to the tankslapping overcorrections after the 2-3 shift (all of course still ultimately caused by driver staying in it beyond his skill)
        I see absolutely no reason why the clutch type LSD would make any difference. The Gilken LSD is not a 100% locking diff (like in offroad cars). Cars with clutch type LSD's go into the wall just as easily with a poor setup.

        Click here to enlarge Originally Posted by spdu4ea Click here to enlarge
        The M3's LSD doesn't react until after there is a wheelspeed difference.
        That is how an LSD works, when the wheels are going different speeds, like an outer or inner wheel in a turn. A clutch type LSD also senses different speeds it just goes about it slightly differently with the clutch, instead of the fluid friction. Still the same principle.

        Click here to enlarge Originally Posted by spdu4ea Click here to enlarge
        thread split?
        Maybe, not really necessary. Regardless, a peg leg (one wheel spinning) is the result of the power going to where the is the least amount of resistance. Since it has no traction, it will just keep spinning until it gets traction and then torque will be sent to both wheels. An open diff can't send different torque to different wheels, still a moot point.

        Click here to enlarge Originally Posted by spdu4ea Click here to enlarge
        on't get hung up on the term "LSD." It encompasses a wide range of different designs, features, and operation. Pump action, double-barrel, and semi-automatic are all shotgun types -- but they differ greatly.
        Not really, they are all designed to do a similar thing in a similar application. RWD limited slips aren't going to differ all that much on the strip, even vs. an open diff. A C63 with an open diff puts down great launches, same with 335's and 135's.

        The LSD's benefit comes mostly on the roadcourse, not in a straight line.

        Click here to enlarge Originally Posted by spdu4ea Click here to enlarge
        This react-compensate speed-based design results in differing wheel speeds the entire time traction is low -- totally different from a clutch type...
        How is it any different from a clutch type that is designed to do the same thing? React based on the speed of one wheel vs. the other, it just does it with a different design.
      1. spdu4ea's Avatar
        spdu4ea -
        Click here to enlarge Originally Posted by Sticky Click here to enlarge
        I see absolutely no reason why the clutch type LSD would make any difference. The Gilken LSD is not a 100% locking diff (like in offroad cars).
        OS Giken (no L). No it isn't a rigid spool, and no it isn't an offroad-style locker... But it is next closest thing and it can still lock 100% which makes it very different from other LSDs. Try one and you'll understand
      1. GG///M3's Avatar
        GG///M3 -
        All these posts trying to blame it on the diff...... Click here to enlarge
      1. spdu4ea's Avatar
        spdu4ea -
        I've said many times that driver skill was the underlying cause. I was just responding to those who said a diff upgrade would make no difference.
      1. TaZaM3's Avatar
        TaZaM3 -
        I will say it again, in my 7+ years of owning this car and abusing the hell out of it Ive never ever had one of my rear tires spinning without the other. Ive never had diff problems.... I did run bias plys once (never went to the track) I drove it on the street and they were very unstable.
      1. Sticky's Avatar
        Sticky -
        Click here to enlarge Originally Posted by spdu4ea Click here to enlarge
        OS Giken (no L). No it isn't a rigid spool, and no it isn't an offroad-style locker... But it is next closest thing and it can still lock 100% which makes it very different from other LSDs. Try one and you'll understand
        Actually, I switched to a different diff in my E92 M3. I went with a higher locking clutch type diff from dan@diffsonline, I went with a 30-90% variable lock differential which should offer the same characteristics as the OS Gilken. I tried it, didn't like it, went back (more for the gear ratios) but I did not feel an advantage on the strip vs. the LSD I already had. I did feel it in the corners though, especially tight and low speed ones.
      1. GG///M3's Avatar
        GG///M3 -
        Click here to enlarge Originally Posted by TaZaM3 Click here to enlarge
        I will say it again, in my 7+ years of owning this car and abusing the hell out of it Ive never ever had one of my rear tires spinning without the other. Ive never had diff problems.... I did run bias plys once (never went to the track) I drove it on the street and they were very unstable.

        BAMMMMMM. Good info, and we have all seen many of your video's raping the streets. The person who went to the track in the video lacked understanding of what mixing true slicks with radials can do on the track. Everyone seems to want to blame it on the diff instead of thinking.
      1. Sticky's Avatar
        Sticky -
        Click here to enlarge Originally Posted by GG///M3 Click here to enlarge
        Everyone seems to want to blame it on the diff instead of thinking.
        I think more info on the diff isn't a bad thing but let's be honest here, a different diff won't make up for driver error and won't make bias-ply more compatible with radials.

        The problem was the driver stayed in it once it started swaying. A diff needs to put the power down to make use so even with a different diff you would have to stay in it to get benefit. Staying in it is what caused the accident because once the moving started then he got on the brakes way too late only exacerbating the rotation.
      1. ogshooter's Avatar
        ogshooter -
        Sorry if this has been covered already of if it's a stupid question but... On the topic of mixing the two types of tires being a bad idea. That would not apply to say a drag radial like the NT05r, R888 or BFG would it? For instance if You were running a high performance summer radial up front such as a NT05 and then some NT05rs on the rear. That would be ok right since they are both "Radial"? Thanks in advance guys.
      1. Sticky's Avatar
        Sticky -
        Click here to enlarge Originally Posted by ogshooter Click here to enlarge
        That would not apply to say a drag radial like the NT05r, R888 or BFG would it? For instance if You were running a high performance summer radial up front such as a NT05 and then some NT05rs on the rear.
        Nope, they are both radials, no issue.
      1. ogshooter's Avatar
        ogshooter -
        Click here to enlarge Originally Posted by Sticky Click here to enlarge
        Nope, they are both radials, no issue.
        Thanks Sticky! Just wanted to make sure before ordering tires.
      1. Sticky's Avatar
        Sticky -
        Click here to enlarge Originally Posted by ogshooter Click here to enlarge
        Thanks Sticky! Just wanted to make sure before ordering tires.
        No worries, definitely better safe than sorry.