Printable View
In terms of cost and complexity.
Provided is a sectional view from our Titan R&P.
Attachment 864.
Total Cost -If i do remember- was $1100, and we were running a low budget.
We didn't even consider PA with GB in our decision matrix last year. Only we were comparing between designing and manufacturing R&P or Buying one.
I may agree with Z that we were copying what the steering system in a FSAE car have to be.
Three factors to consider:
1) In a Formula car I do not want the driver to be crossing hand-over-hand or shuffling the wheel around. Check out onboard videos of the faster FSAE cars - you'll rarely see more than 90 degrees wheel movement, no matter what steering system is used.
2) Pitman arm steering is nonlinear, as you'll find if you play with a kart on a stand or make some drawings (when you draw it, make the steering arm longer than the pitman arm). Even better, it's falling-rate, so it can be fast on-center when effort is low and slow well off-center when effort increases.
3) This is a sprint race. The driver can apply more force to a steering wheel for an 11-minute enduro than he or she can for a 3-hour Le Mans, Grand Prix, or Indianapolis race or for a NASA spaceflight. If your best driver's a big Texan SOB, you can reasonably demand that he be strong. If your best driver's a small lady, the weight on the front tires (and thus the force required to keep the car balanced) is reduced, cutting down steering effort (1). Draw "reasonable-looking" steering geometry, find inputs (KPI, trail, steering arm length) that work for the tires, then calculate how much steering force you'll need both near-center (fast-corner) and near-lock (slow corner) - then USE that as an input to the driver's strength training program!
(1) In most FSAE cars, it's not a bad assumption to say that the CG of the driver's torso, head, and arms are roughly at the x-location of the CG of the car/driver combo and the driver's legs have a CG near their knees. If your car has a mass of 160 kg, then your driver's mass of 80 kg is 1/3 of the combined mass, and 40 kg of that is the driver's legs...
So you think a pitman arm straight from the column is a realistic option do you Charles? I know the suitability of this option shouldn't be dictated by the answer to these questions alone, but have you ever seen it done on an FSAE car? And is this what your car is fitted with?
Sure you don't want to cross hands, but you can easily get to ~150 deg before that has to be an issue for the driver. Yes you can technically achieve 90 degrees with a pitman arm, but you have to 1) make sure your tie rods won't hit your column, 2) protect yourself against over-centring, 3) deal with how incredibly non-linear this system is at 90 degrees. Which is why realistically you aren't really going to be able to use more than 70 degrees.
So your effort levels and response are already twice what they need to be. Ignoring effort levels, maybe the driver CAN control the car with this level of response, but would they 1) prefer less sensitivity, 2) perform better with less sensitivity? I guess the only real proof is in the testing, and I've never tested a car with this much response so unfortunately can't comment further. But I would be very surprised to hear that this would be favoured.
A variety of steering gears are shown on this page, http://what-when-how.com/automobile/...ts-automobile/ Brainstorming (I've never seen one of these), take a look at "Fig. 27.53. Screw and nut re-circulating ball gear mechanism".
Instead of generating motion with "pin-in-slot and rocker arm", what if the motion of the ball nut pushes-pulls on a tube that is concentric with the steering column? This in turn pushes on one arm of a bellcrank--and the other end of the bellcrank would be the Pittman arm (called "drop arm" in the figure).
Obviously many details to work out, but this could be a lightweight solution that also allows a lot of packaging flexibility? To position the inner tie-rod ball joints correctly use a relay rod and idler arm, see http://what-when-how.com/automobile/...ms-automobile/ see "Fig. 27.49. Split track-rod with relay-rod and idler steering linkage layout".
Ball screws and matching nuts are commercially available in a wide variety of sizes and thread pitch...
Take a look thru a junkyard at all the sophisticated steering setups used in today's finest lawn mowers. Heck, my John Deere F935 is rear wheel steer by hydraulic hose. Plenty of good used and new replacement parts available for these machines, too.
Wondering why so much trailing (rear) steer since pulling on the high force tierod is easier that pushing on it. I'm used to seeing front steer on rear drive high performance cars because of the N/S motor and there's no transverse transmission inconveniently in the way. There's some relief from the mass compounding effect with this, too.
As you can see, pushing the steer axle on this Advance Steam tractor probably wouldn't cut it. It was worm gear, though, too.
And another example of creative steering geometry, the FSAE Ratrod Class, also chain steer so to speak...
CWA - It depends on what sort of car you're building! Karts have response like that and I cannot drive well WITHOUT it. I'm pretty awful at autocrossing passenger cars. Motorcycles are even more sensitive.
My wing-kart, 375 mm steering wheel, mass 80kg, zero trail, with a wheel 200 mm outboard from the kingpin axis and 15 deg kingpin inclination, doesn't have problems with excessive (> 200N) steering effort at 3+ g. It has a 50ish mm pitman arm and 125ish mm steering arms. If you build a 250 kg understeering dump-truck with 80 mm mechanical trail and a 50/50 weight distribution, this could be painful. If your ideal FSAE car is close to my ideal (150 kg, 40/60 F/R with a driver, <200 mm CG height, diff open at low torque, 25 mm mechanical trail) the numbers might work out for you. It was a close decision for A&M's -13 car.
Bill, really odd steering & suspension geometries can be found all over The Henry Ford.
That's a nice discussion going on here. Last year we considered multiple "unconventional" approaches on steering; pitman arm (exactly like Z described) was the one chosen. Angular gearbox is NEEDED because of the implications described prior on the thread by others, in our case the ratio was close to 4.5:1 to make it work. That being said, upright pickups were constrained in being almost identical with last years' so we could reside to using an RnP as emergency, if pitman steering failed us. The guy who designed the system got it to a point that ackerman and steering response was almost "perfect" (far better than any RnP system I have seen so far). However we really struggled to get a reasonable bump/roll-steer response (whether this is an issue or not is up to you). Now the really "crappy excuses" part starts; the guy responsible for the steering ditched the team, and that let our transmission guy struggling to manufacture the bevel gearbox, housing, mounts etc. in less than a month, while also fixing his issues with the transmission. Bottom end; a 250$ junior drag rack, modified to suit our needs in less than 2 days.
ECU's steering is a little different using a planetary gearbox driving a pitman arm, then to rockers. Really great for getting good control of ratios throughout motion, and allows for a super low nose (which was the main motivation). If it wasn't for the front suspension the rockers could be ditched and run it just using the planetary. The planetary was designed and built in house using off the shelf gears. Cheaper than a rack and super low slop due to the high number of contacts.
I will say that it is worth being careful with alternative steering systems as the loads can get quite large in the mechanism and mounts. Much of this is dealt with internally in a R&P.
Some tuning the guys have done this year has been quite eye opening with regards to the steering. Improved control over dynamic toe (ackermann) and rate can have some interesting effects that cannot be achieved with a conventional R&P. These cars can be super-sensitive to steering so it may be worth allowing for a little weight to have a better system.
Kev