Steering Racks

[mlupfer]

Looking for some feedback.

An opportunity arose to work with people in the the automotive steering sector to design and produce and FSAE rack and would like to target as many teams as possible.

What does everyone prioritize in their steering rack choice.

Weight
Ratio
Hand-wheel max angle
Rack length
Rack travel
Cost
Sensors
Mounting points

I know there aren't many options out in the market. The cheap one is stiletto. For a while lots of teams ran the M/Z rack which was light but carried a heavy price tag and wore quickly. Also for the European teams would it be beneficial to have an off the shelf system with Electric control available since semi autonomous is coming. Also how many teams have custom setups v.s. off the shelf. Any feedback would be greatly appreciated.

[DougMilliken]

Seems to me you are asking the wrong questions(?)

Whenever I've design judged, one of the most common problems is a steering system with far, far too much friction to allow precise steering by the driver.

Another very common problem is too much slop/free-play/compliance, but this is more likely outside the rack in the rest of the linkage/joints/column-support/quick-release.

If you could make a light and low cost steering wheel QR that (by design) had *zero* freeplay, including when loaded with high steering wheel torque, I think this could help over half the cars I've seen at the Michigan comp over the years.

If Z were still posting, I suspect he would suggest a Hirth joint or something related, https://en.wikipedia.org/wiki/Hirth_joint

[mlupfer]

Doug

- I think what you are talking about with too much friction is not the correct terminology. I think you are trying to describe steering effort which is generally very high on FSAE cars. Most road cars target 2 to 3 HW Nm but on an FSAE car its closer to 5 to 8. A lot of that is due to the need for very large driven wheel angles and teams not wanting to have large handwheel angles. Also the steering axis' are not conducive to low effort steering with large caster and kingpin angles. Also static steer is your worst case scenario and out on track efforts drop dramatically. Personally rack friction makes up very little in the overall steering effort.

- As for slop/lash in the system i agree most teams have a very bad designs. between columns that have too much free play to rod ends that develop compliance. That's a problem that involves a lot more work.

[DougMilliken]

Good discussion!

In design judging, I've often used a quick test that someone told me was the "Carroll Smith steering friction test". With the front tires lifted off the ground, move the steering wheel toward one lock, then give a quick flick in the other direction and remove your hand. Ideally the steering (with zero load on the tires) will bounce off the opposite stop and come back a little. I can count on my fingers the number of FSAE cars that passed this test...a very small percentage.

If you have a copy of "Learn & Compete", we put a few comments on steering in Chapter 6, p47. I've seen many FSAE cars with over-constrained rack bars and a lot of rack friction (this may change as the rack mounting distorts under different loading conditions).

[mlupfer]

[QUOTE=DougMilliken;128311]Good discussion!

With the front tires lifted off the ground, move the steering wheel toward one lock, then give a quick flick in the other direction and remove your hand.

This is a much better test than what iv'e seen a bunch of judges do with the wheels still loaded. I appreciate the clarification on your process!

As for rack over constraining i am only familiar with the internals of the M/Z rack since that is what we used when i was on a team. From what i remember it maintained 3 points of contact so it should not have been over constrained. I can't attest to the design of some other racks on the market.
One trade off generally with rack friction is also lash. The more force you put into the pinion mesh the less lash but the more friction. That's a recurring issue in the auto industry. It would be possible to make the rack bearing adjustable so teams could set their own lash and friction to their liking. I think the KAZ rack has something like this built into it.

[BillCobb]

Here are my thoughts based on 40 years in the industry:

Weight: sure. Some will believe the sprung C.G. location will benefit.

Ratio: There is no such thing as a 'steer ratio' with a rack & pinion system [technically], because the steer arm on the knuckle/upright goes into the formulation.
The industry calls the rack transformer a "C-Factor" which is mm of travel per 2pi rotation of the pinion.
Yes a 'C-Factor' is related to the desired max hand-wheel during low speed maneuvering as well as high speed dancing with cones.

Teams could benefit with a variable steer ratio setup to help with this dilemma using single Cardan joint characteristics OR a Variable C-Factor rack (ala Bishop gears).

Rack length and travel are part of this whole mess, too. How much wheel travel will a car want/need ?

Rack stops are recommended. Otherwise over-traveling a rack will usually result in busticated part(s).

Sensors ? Pinion rotation may be what you are thinking, but the HANDWHEEL is the reference for gain, understeer and effort data streams.

Mounting system types are a wide open design area. Sleeves, bolsters, tabs and clamps are the norm.

______________ Now for the bizarro: ---------->

Electric power assist. The e-Cars are loaded with juice. Maybe time to 'amp-up' the low speed maneuverability problem with a 48v booster.

How about "steer-by-hose" ? Use hydraulics or even water or GatorAid to transform handwheel rotation to king-pin rotation ? Use metal piping, so there is still a "mechanical connection"
Heck works just fine for my boat.

The 'steering system' is usually comprised of handwheel (removable), a column (shaft) down to a transformer (rack), and out to an upright (knuckle).
This is all good territory for a supplier to have designs for parts, including a CV joint type geometry crutch (NOT tripots), some nice threaded tierod ends to allow for rack centering and toe alignment.

There are also some possibilities for rear steering, too ! (go figure that 1 out ! (NOT 4WS).

And as for friction: racks and pinions are not very friendly couples. The pinion wants to push-away the rack from contact, hence the need for slipper bearings or sliders to minimize this. The more
rack force there is, the more push-away occurs, hence opportunities to 'jump a rack' (Hey, my steering wheel spokes ain't centered anymore.) The heavy pressure applied to prevent this can cause a LOT of friction to be present.
In this case, friction is not your friend. But, on the other hand (get your hand off your butt), friction may be necessary to prevent 'free-control' oscillation instability. An this also requires controlled friction over time, mileage and age.



There ya go ! I may drop in at Michigan, Doug. See you there ?

[DougMilliken]

> As for rack over constraining ...

The worst cases I've seen are often with a short rack (junior dragster?) that is extended, and then additional bushings are added to support the longer rack bar.

Another problematic setup (which looks so nice and light weight in CAD) is to integrate the outer rack bushings with the car chassis/spaceframe, the problem comes as soon as the space frame distorts a little under load and the linear bushings pinch on the rack bar (don't recall if I've ever seen a setup like that that included large self-aligning sphericals to isolate the rack bar from frame distortion).

Bill -- sorry to miss you, I'm not going to be at Michigan this year.

[BillCobb]

The general rule is that there is actually one ONE steering gear mount. This mount is used to deliver the prescribed transmissibility, understeer, durability and crash barrier intrusion.

Forbidden to use TWO structural mounts. The 'other' mount is used to hold up it's end. Gear housing is not to be used as a structural member.