De-powered steering pinion failure

#1

Maybe a warning of sorts to others.
Luckily the final break happened in the paddock while in tech!


Technically broken twice (torsion bar & pinion stop tabs).

The point being if your rack/steering ever feels strange or feels like it might be skipping teeth, don’t put off looking into it and getting to the bottom of it! I think our has been half broken for a while as we’ve had weird steering issues for quite some time but we couldn’t find where the problem was.

After showing this to some experts, they were quite surprised the community hasn’t seen more failures in de-powered racks as torsion bar stop tabs are not meant to take all the loads of turning/curb kickback/etc, just to limit the torsion bar movement as the power steering assist takes over.

#2

Can you elaborate on this? I don’t understand.

#3

Most of the “work” is done by the power steering fluid moving the piston in the rack body. When you twist the steering wheel the torsion bar flexes which align different ports on the outer spool valve which redirect the flow of fluid. This can be taken to extreme – think of your Grandpa’s LIncoln which you could turn with just a pinky finger – it just has a weaker torsion bar and higher pump pressure.

In a manual rack this would all be one solid piece. On the power rack the torsion bar is the only thing that connects the 2 halves axially and the pinion stops (for lack of whatever their real name is) limit the radial twist. It’s a little hard to understand without the pieces in your hand.

Maybe some of these would help:
http://auto.howstuffworks.com/steering4.htm
http://www.woodwardsteering.com/index.php?option=com_content&view=article&id=662:9a-how-power-steering-works-the-torsion-bar&catid=44:woodward-tech

#4

If I understand you correctly you are saying that the loads on the “torsion bar stop tabs” (pinion stops?) are higher because we don’t have enough fluid (transmiting force equally throughout a volume) in our depowered racks, or because we don’t have pressurized fluid (doing something beyond the above).

Situation 1 above is like a PS car with its engine off. Situation 2 is the PS car with it’s engine turning.

I’m not trying to be disagreeable, I’m just trying to understand precisely how a depowered rack might be vulnerable to accelerated wear. I’m having trouble visualizing how something that “limits radial twist” gets significantly higher loads because most of the pressurized fluid gets drained out. Are these higher loads something that happens only at the very limit of left and right full lock? Is it something that happens when one wheel takes a hard sudden lateral load like banging on a gator?

#5

This is what happens in a depowered rack:
You turn the wheel, the torsion bar flexes until it his the stops (~5deg), now that the stop tabs are engaged they can turn the bottom of the pinion that has the gear on it which then moves the rack. So every time you turn the wheel you have to engage those tabs; any kick back/curbs/tire knocking/etc… is also feedback that can also push against those tabs.

#6

What’s the fix? Is there a way to weld the torsion bar to the pinion?

#7

[quote=“jlucas” post=66661]This is what happens in a depowered rack:
You turn the wheel, the torsion bar flexes until it his the stops (~5deg), now that the stop tabs are engaged they can turn the bottom of the pinion that has the gear on it which then moves the rack. So every time you turn the wheel you have to engage those tabs; any kick back/curbs/tire knocking/etc… is also feedback that can also push against those tabs.[/quote]
And that is different from PS operation how?

#8

The forces on the pinion gear will be greater w/o PS. The hydraulic pressure on the rack unloads the pinion.

#9

Sounds like a safety issue.

Is there any other manual rack that would work that could be made to be spec?

#10

Ryan: My understanding is there is no manual E30 rack for the USA.

Scott:
This is what happens in a powered rack:
You turn the wheel, the torsion bar flexes which aligns different ports on the spool valve and directs the hydraulic pressure to one side of the rack or the other; when you stop twisting the torsion bar the hydraulic pressure is bypassed an just goes to the reservoir. If the pinion tabs touch it’s just barely contact and not any significant loading as the fluid pressure differential is what is pushing the rack from side to side.
Not a rack and pinion but it might help understand the fluid flow paths:
circa 1956: http://www.youtube.com/watch?feature=player_embedded&v=Ni1nrwU1DjM

#11

Hard to tell for sure from the pix, but it looks like the spool valve could be removed and the pinion shaft welded to the input shaft. It also looks like if you drilled a hole through the two shafts where the hole is now that captures the spool valve that you could drive a roll pin through them. Don’t know if the roll pin would have enough shear strength though.

#12

[quote=“jlucas” post=66695]Ryan: My understanding is there is no manual E30 rack for the USA.

Scott:
This is what happens in a powered rack:
You turn the wheel, the torsion bar flexes which aligns different ports on the spool valve and directs the hydraulic pressure to one side of the rack or the other; when you stop twisting the torsion bar the hydraulic pressure is bypassed an just goes to the reservoir. If the pinion tabs touch it’s just barely contact and not any significant loading as the fluid pressure differential is what is pushing the rack from side to side.
Not a rack and pinion but it might help understand the fluid flow paths:
circa 1956: http://www.youtube.com/watch?feature=player_embedded&v=Ni1nrwU1DjM[/quote]

There is an e30 manual rack available. It’s a bit slower and likely to be expensive. Finding one in the US that is used but serviceable might be a challenge.

Re. video. Thanks, but I’m probably not going to understand the insides of the steering box well enough to judge whether or not this is a safety issue. Maybe I’ll find something that shows the insides of a very similar steering box at work. Certainly we’ve all seen lots of suspension components fail so we accept risk every time we leave the grid. The important thing is to address the most dangerous and most likely risks of failure, and to be in tune with signs of impending failure.

OP, you made a reference early on to your steering feeling “odd” or something like that, prior to it’s failure. Can you elaborate on that?

#13

Scott, Jeremy can correct me if this is wrong but I think an easy and valid way to think about this is that the tabs Jeremy shows broken off experience the same amount of force that your arms and hands do at the wheel. Once the hydraulics are disabled the assist you would normally get to steering inputs is gone. Feedback from gators etc is also similarly affected (no hydraulic buffering). So under most circumstances there’s not a lot of increase in the forces on the tab, it’s still pretty easy to turn the steering wheel. I can think of two instances where the effect is much larger, wheel drops/gator jounces, and driving around the paddock. Hence Jeremy’s failed at tech. But it is definitely an eye opener and something to think about.

Another harder way to think about it is the whole spool valve, torsion shaft, pump, piston system is nothing more than a hydraulic servo motor with the steering input shaft as the control input and the rest serving as both gain and feedback blocks. Smallish control inputs generate largish outputs. Remove the hydraulics and the gain goes to one. The weak point in the system is the connection between the input shaft and the pinion drive. Which in this case are the tabs.

A third and much simpler way to think about it is in a powered rack, the steering input is used ONLY to twist the torsion shaft (and displace the spool valve openings), it doesn’t actually have to move the rack. In a depowered rack the input shaft and tabs have to move everything.

#14

[quote=“jlucas” post=66695]Ryan: My understanding is there is no manual E30 rack for the USA.

Scott:
This is what happens in a powered rack:
You turn the wheel, the torsion bar flexes which aligns different ports on the spool valve and directs the hydraulic pressure to one side of the rack or the other; when you stop twisting the torsion bar the hydraulic pressure is bypassed an just goes to the reservoir. If the pinion tabs touch it’s just barely contact and not any significant loading as the fluid pressure differential is what is pushing the rack from side to side.
Not a rack and pinion but it might help understand the fluid flow paths:
circa 1956: http://www.youtube.com/watch?feature=player_embedded&v=Ni1nrwU1DjM[/quote]

I was just takling about any BMW rack that might fit our cars. I understand that our racks could be modified to be a bit more robust, and I’m not sure this is an actual issue yet considering only one person has had a failure.

Just imo, I wouldn’t think that taking steering racks apart and modifying them would be a good idea from a new racer perspective. Was there any manual rack made for a US car from BMW? Even if it was an older rack…

Just a thought. Throwing ideas out there.

#15

@Rhodes. That was helpful, thx. Where the hell did you learn this?

@Foghorn. I did some poking around on manual racks. Apparently there were 2 flavors, the early manual rack and the very early manual rack. The former is supposed to have come out in '85, has a low ratio and is perceived as unsuitable for track use. The very early rack is supposed to be a reasonable track solution. Some time on RealOEM looking at part #'s would be needed for confirmation.

#16

Interesting. Don’t want to jump the gun too quick on racks failing, but it may become an issue. I don’t have any issues with mine…yet.

Would be cool to find a universal solution. Don’t want to replace something that isn’t broke, but having a proper manual rack would be nice too.

Can the guys from Bimmerparts help?

#17

The easy ways to think about it are courtesy of Jeremy’s link:
www.woodwardsteering.com/index.php?optio...tid=44:woodward-tech

The hard way is courtesy of Va Tech & SMU.

I’m an engineer. Sorry.

#18

[quote=“Ranger” post=66709]
OP, you made a reference early on to your steering feeling “odd” or something like that, prior to it’s failure. Can you elaborate on that?[/quote]

Since at least back through the 2011 season (and I think the 2010 as well, prior to that we still had the power steering connected) we’ve had some steering issues that we could not find the answer to. Main symptom was we would align the car (steering straight, and set the toe) and then later on track, typically after hitting a bump or curb the steering wheel straight ahead position would jump or skip to the side so that to top of the wheel would be at the 1 or 2 o’clock position, however when the toe was checked it was still as it was set before. My theory after examining the broken pinion pieces is that one of the broken pieces would get wedged between the torsion bar and the outer housing causing the steering to “stick” in that position and then getting released by bump input which then unwound the torsion bar (or vice versa). You can see that one tab is particularly deformed/parallelogramed. I’ll be able to tell better if the problem is fixed with the new rack at MidOhio at the end of the month. It’s possible that our rack has been partially broken since we bought the car already built 5 years ago.

Over it’s race life our rack has taken at least 2 hard nocks into the tires (most recently this April, which was 3 race weekends ago) so that probably, of course, contributed to it’s failure. The problem of course being there is no easy way to inspection pinion for damage.

I’m not saying the sky is falling, just to be aware that this can happen. The only thing I can think of worse than complete steering separation is 100% brake failure. Neither of which I hope any of us have the misfortune to experience on track.

#19

Jeremy, I’ve recently disassembled a rack in order to de-power it. Can you tell me how the pinion shaft is taken apart? Also, I understand you have since welded the shaft to the torsion bar. Can you go into a little more detail on where you welded?

#20

My new fav movie. i was ready to “duck and cover” for fallout protection when it started up.:wink: