Steering friction reduction

[Poweron]

While searching for possible steering rack options for FSAE teams, I couldn't find a clear way to get a sense for the friction of a certain rack until it arrives. Has anyone had luck in reducing the friction of racks that your team has purchased? How about designing the steering column to upright assembly to minimize friction? Material selection is one easier aspect but at some point the constraints and gearing appear to be the major influence. I feel our team has run into the issue of angling u-joints too near their limits and therefore minimizing feedback from the road, but that's just a hypothesis at this point. Given that I haven't had the chance to feel the steering setups from other teams maybe someone with that experience can give some advice.

[Jonny Rochester]

If you purchase a steering rack (we have at Stiletto) there will not be an issue with friction in the rack itself. Steering racks are very basic, and very effective, used in most cars, so a purchased rack can be assumed to be a well engineered item. More critical is your choice of steering rack ratio (and to choose that you need to decide your toe-base at the upright and maybe a few other things).

The more important thing is your arrangement of uni joints (or bevel box etc). Maybe you can do a sketch of your proposed steering layout including angles.

[BillCobb]

"Steering Friction" is best deigned and managed as the allocation between "upper" and "lower" steering system friction. Upper being rotary friction above the rack/pinion interface and lower being a combination of rotary and linear friction elements from tires, ball jonts, tierod ends and slipper bearing elements within the rack.

As such, friction is a low form of damping which is generally required in vehicle steering systems because it can produce undesirable effects (free control oscillation) and wheel "kick" if left unchecked. On the other side of the fence, too much friction makes a vehicle awkward to drive, reduces driving path precision and bothers the crap out of a driver. In power assisted steering ("power steering") the rack force booster acts just like a power transistor and we should all know what happens if you goof up THAT amplifier analogy: it goes unstable and you let out the designed in smoke particles.

A simulation of various steering maneuvers backed up by some test data with standard SAE or ISO test procedures can get you a very comfortable, nimble, precise and fun to drive vehicle that anyone besides a weight lifter or a prosthetic armed operator can "feel" comfortable with. It makes a great thesis project, too. A haptic based (feeling/right brained) operator is also necessary to get the most favorable response.

Results may vary, do not try this at home. If you have sore arms for more than 4 hours, you need a girl/boy friend.

Hardware parameters which usually are included in the systems engineering approach to this include tire properties (including scrub torque), castor, steer arm length, tie rod plan view and side view angles, gear C-Factor (the rack displacement per 1 rev of the input shaft), gear 'efficiency' column friction and steering wheel radius and of course the open loop gain of the vehicle in terms of g's per degree. I haven't included rack booster mechanics in the list because it requires a bit more knowledge of the hardware that is way out in Proprietary Land. Basically you would need to test a real gear on a bench.

Believe it or not, this is an optimization process that has a universal optimal solution for each class of vehicle (sports, luxury, grandma, light truck and FSAE cars).

[Poweron]

Interesting Jonny, I was under the impression that different steering rack manufacturers had different build qualities and those differences would be a major factor. Perhaps our team just got unlucky one year. We used to run a Stiletto rack but have been trying different options the past few years.

Bill, thanks for the reminder that friction is helpful. My concern stems from an episode I heard at FSAE West that at least one judge suggested that a steering wheel should be extremely easy to turn (eg. with one finger) when the wheels are off the ground. It's possible that there is more to blame than friction, but given that I have yet to see such a steering system, it's possible that either this advice is inaccurate or I'll have to look into your mention of an optimization method. In your experience, is the majority of that optimization done on a test bench setup or can it be done during the design phase without the hardware components? It appears that at the very least I'll have to get a model our entire steering linkage setup in its entirety to start analyzing it.

[BillCobb]

The optimization can be done completely with a simulation if your set of constraints and parameters (geometry and force and moment characteristics)is reasonable (ball joint friction, rack friction, vehicle dynamics [understeer, tire properties, overall steer ratio], column friction Cardan joint mechanics and friction).

Then its just plug and play. Since most cars a have manual (unassisted) steering, the catch 22 is very low speed effort vs high speed work (work being torque vs angle to generate the required lateral g's for a turn).

A Matlab Simulink program is a bit fun to produce and operate. From this, you will realize (hopefully) that maneuvering a vehicle is actually a moment balancing act, not a displacement controlling process. A steering system moment is reacted by the vehicle dynamics to counteract it. Notice that I didn't say "steering wheel moment".

[jd74914]

If you purchase a steering rack (we have at Stiletto) there will not be an issue with friction in the rack itself. Steering racks are very basic, and very effective, used in most cars, so a purchased rack can be assumed to be a well engineered item. More critical is your choice of steering rack ratio (and to choose that you need to decide your toe-base at the upright and maybe a few other things).

The more important thing is your arrangement of uni joints (or bevel box etc). Maybe you can do a sketch of your proposed steering layout including angles.

The first statement here is not always true; you cannot assume a purchased rack will function well mechanically. I have seen many COTS (commodity) racks like those made by Stiletto with varying amounts of friction and stiction; some were acceptable, others less so. Stiletto racks (along with most JR Dragster racks) also have a huge issue with mount compliance since their mounts are super close together and as such do not well counter the rack moment. One of the best things my team ever did was design our own rack with perfect ratio and well-supported ends.

I do agree that joint angles and supports are critical. The second best thing we did was remove all universals from the rack and run a straight shaft (your mileage may vary here).