Torque Vectoring Goes Modular

When engineers at Ricardo (www.ricardo.com) first looked into torque vectoring 5½ years ago, they weren’t sure it was enough of a leap forward. “That’s why we combined it with a hybrid drive system to giver more value for the money,” says Jim Hey, chief program engineer, Transmission and Driveline, Ricardo. The proof-of-concept vehicle—a modified BMW X5 SUV—used an electric motor to drive a gearset in the transfer case and send torque to the rear. Says Hey: “We were able to apply a brake torque to the gearset to drive it to the front via a 20 kW e-machine that could be configured to provide a zero emissions mode, regenerative braking, launch assist, and more in a permanent all-wheel-drive vehicle.” The one problem was that the market wasn’t ready, which left the team with a highly compact, highly efficient mechanical gearset and no customers.

“We soon realized that if we could apply a brake or acceleration torque to one element of the gearset,” says Hey, “we had the ability to create torque and wheel speed differences across two output shafts.” That is exactly what OEMs wanted, which led to the engineering team swapping the e-machine for a pair of low-capacity wet clutches. “We can reduce, by a factor of seven, the amount of torque we have to transfer through the clutch. As a result,” he claims, “we can control a torque delta of about 1,400 Nm across the axles with a clutch capacity of just 200 Nm.” That’s because lots of short bursts instead of large force applications over long periods keep thermal loads and actuation forces low, and allows the device to spend 80% of its life at 20% of its full capacity.

Hey doesn’t deny that the functionality of Ricardo’s system isn’t unique, Honda (Super Handling All-Wheel-Drive) and Mitsubishi (Active Yaw Control) have comparable systems in production, but he does claim Ricardo’s unit is less costly, heavy, and complex. “What we did was create a module that plugs into an existing differential or axle unit so it can be retrofit into existing applications to give OEMs a range of options.” And the Ricardo unit takes up less space, especially in terms of the width between the differential’s output shafts, and allows the retention of standard-length halfshafts for optimal articulation angles. “We envisage a range of modules sized for various torque ranges that can be used across platforms and by multiple OEMs,” says Hey. This is especially true of rear-drive and transfer case units, though front-drive modules either could bolt on as a secondary device or be package protected in the gearbox design. The only other items are deciding whether to provide the clamping load on the clutch packs through hydraulic, electro-hydraulic, or electro-mechanical means, and how to tie the control unit into the vehicle’s electrical architecture. “Part of our strategy is to license this technology to an OEM or Tier 1 supplier, and we’ve had several discussions with suppliers and talks with an OEM that wants to license it directly,” says Hey. “We’re pretty sure it will be on a car by 2010.”

Ricardo’s torque vectoring device uses joined sun gears that require the use of additional annulus gears. However, since each planet carries the same load, overall size can be smaller and low precision—and lower cost—components can be used.