Swapping to rack and pinion steering

From Crankshaft Coalition Wiki
Jump to: navigation, search

by: Cobalt327, Crashfarmer, Crosley, Jon, Oldguy829
(Click here to edit this page anonymously, or register a username to be credited for your work.)


[edit] Introduction

Swapping to a rack and pinion (R&P) steering system from other set ups has become a commonly performed swap and a sometimes necessary part of vehicle steering and suspension modifications to older vehicles. Often clearance is gained for engine and exhaust placement and an improved steering feel can also result.

[edit] Turning radius

Many cars suffer from poor turning radius after an R&P install. There are 2 basic reasons. Either the rack was placed too far left and interferes with the wheel, or no adjustment was made to the steering arms.

On a typical RB (recirculating ball) type steering box, the Pittman arm has a “throw” of 7 inches side to side. It is usually connected to steering arms effectively measuring 7 inches long (The 7 inches is measured from the center of the steering arm mount, where the king pin or ball joint pivots the spindle, to the center of the outer tie rod where the steering pivots). Typical Cavalier racks have 6 inches of throw and originally connected to the upper strut in a manner that represents about 5 ½ inches from center of the strut to the center of the tie rod. Connecting the rack to the original arms causes a loss of nearly 20% of steering angle. It is strongly recommended that this be dealt with before installing the rack.

[edit] Bump steer

Changes to the steering arm can affect the steering geometry and can introduce bump steer. So, if you are going to address this issue, (some people don’t) do it before the rack install.

There are at least two acceptable methods of dealing with this issue. The preferred method is to replace the steering arms with shorter ones. On the '41 Pontiac upgrade covered in the wiki 1937-1957 Buick Oldsmobile Pontiac suspension upgrade the original '65-'70 Chevy steering arms were replaced with a set from a '55-'57 Chevy.

1955-'57 Chevy arm on left; 1965 arm on right

[edit] Bending or replacing steering arms

There is a difference in mount width between the BOP arms and Chevy arms, so they won’t work on BOP without the Chevy spindle upgrade. If you are unable to find shorter arms for your application, bending the originals is the next option. You will find mixed opinions on this issue. Some will insist that heating and bending steering arms compromises their structural integrity and should never be done. Others warn you to be sure they are forged and not cast arms. Bending forged arms may be OK, bending cast is not.

Others will swear that rodders have been bending arms for decades, with no failures. Please research this issue to your own satisfaction, as to the correct procedures to use to bend the arms and whether safety will be compromised in any event.

By putting an "S" bend in the arms they are effectively made one inch shorter, pivot point to pivot point. This also made them one inch lower than the original position. To compensate for this change in geometry, the rack was positioned 1 inch lower than the original center link. A correction could also be made by fabricating the center bracket with a 1 inch drop in the inner tie rod mounting points. However you address it, by doing it first you will have 3 fixed points to work with: Inner control arm pivot point, lower ball joint pivot point and outer tie rod pivot point. This leaves you with the front/back and up/down location of the inner tie rods to deal with.

[edit] Rack height importance

We mocked up everything in the fashion described above and then designed our rack takeoff mount (inner tie rod mount) to fit. Once the inner bracket was welded up and mounted to the rack, we installed the tie rods and checked for bump steer. The final minutest adjustments were made by tapping the rack mounts on the frame with a hammer. Remember, at this point the rack was just clamped to the frame. When we found perfection, we tack welded the frame mounts, removed the complete rack assembly and welded the mounts and gussets in place.

To understand how crucial the height location is, follow this mathematical extrapolation: A 1/8" difference in height of the rack, caused a 1/16" difference in the location of the tie rod arc (at full compression or rebound), compared to the ball joint arc. 1/16" at the tie rod (6 inches from the spindle center) becomes almost 3/16" at the rear of the tire (typical 30 inch tire). This causes a reciprocal movement in the other direction at the front of the tire. Now we are dealing with 5/16". Note that BOTH wheels are doing this, so the toe in (or out) changes 5/16" x 2 = 10/16" or 5/8" total during suspension travel caused by acceleration or braking; all this while you are trying to drive in a straight line. This illustrates how height is the most crucial dimension in locating your inner tie rods.

[edit] Control arm to tie rod length

Duplicating your lower control arm length with your tie rod length is not nearly as critical. In fact many original setups were unequal length. The trick is to make the tie rod as long, or longer, than the lower control arm. When the tie rod is longer, it will have a flatter arc than the control arm. The flatter arc means the tie rod will move outside the ball joint at the extremes of suspension travel and create additional toe in. Up to 1/4" will not be noticeable in handling or tire wear. If the tie rod is shorter than the control arm, the opposite will happen. 1/8" of toe out will put you all over the road.

On the '41 Pontiac, the lower control arms are 18 inches pivot to pivot. We used the original Cavalier tie rods, which measured 21 inches pivot to pivot.

Rack with original tie rods/center take off

Using this combination, the bump steer (toe in) changes less than 1/4" at full suspension travel, and less than 1/8" in normal operating range. It's been said that additional toe in is beneficial during hard acceleration or hard braking, as it tends to stabilize the front end.

[edit] Inner bracket/inner tie rod mount

One thing not covered yet, is the angle of the mount for your rack. We chose to rotate the rack until the steering sector was pointed at the end of our steering column. This made a fairly straight-on connection for the intermediate steering column link, with no extreme angles to deal with. It also made the connecting link less than 11 inches long, including the U joints, so no supporting heim joint was necessary. It did however, require the center bracket to be built with an angle included in the mount. On the first bracket we built (we made the mistake of fabricating the bracket before we changed the steering arms), we had the mounts offset from the center of the rack. A 1 inch plate is necessary to move the bracket out beyond the bellows. Then we ground out a portion of a 1 inch square tube and welded it at an angle to the 1 inch plate.

End view of a bracket with height and angle offset from original mounting position
Front view of the offset bracket

In the final install, we needed the mounts to be nearly in plane with the original mounting holes. This changed the design completely. In the offset bracket, the 1 inch plate can be mounted to the rack with the original bolts (12mm 1.25 {fine} thread bolts) in the second version, longer bolts were required. Rather than weld the flat bar at an angle, we chose to drill the mounting holes at an angle and then weld in some grade 8 12mm bolts for the cavalier inner mount to fasten too.

Nearly straight through, but at an adjusted angle. Note that the mount compensates for the rack being about 1/2 inch off center in the car.

Rather than use the original Cavalier tie rods, you may decide to build the tie rods out of tubing with adjustable tie rods on each end. You will need a tapered reamer for this operation.

Power rack installed, rear view
Power rack installed front view
Installation using original Cavalier inner tie rods and fabricated bracket

[edit] Steering linkage

The original steering ('37-'57 BOP) consists of a manual box and the steering sector is a solid shaft all the way to the steering wheel. If you shorten the original column, be sure to relocate the end centering bearing in the shortened portion. If you cut off an early BOP column, there will be no centering bearing to use. Since the original column outer diameter is only 1-1/2" you may need to fabricate the centering bearing holder.

[edit] Tilt column

We took this opportunity to convert to a tilt column. Late '70s to early '80s GM vans have a nice clean column, 2 inch diameter. No ignition lock tumbler assembly (it was on the dash) so it looks right at home in most old cars. This column has self canceling turn signals and in our case, (automatic trans) a lighted gear shift indicator and neutral safety switch. We were lucky enough to find a tilt column, but they are getting rare (after market columns are available from Flaming River, etc.).

[edit] Other design considerations

A few general observations: The straighter and more direct you can make your steering linkage, the better. There is an excellent article on steering linkage- Steering shaft assemblies from junkyard parts, or you can order up the pieces from Borgeson et. al.

Different columns have different output shafts, so check carefully before ordering. There are almost too many options here to discuss.

It would be prudent to look at the steering linkage hookup, before deciding on the final rack location. You may want to align the steering sector on the rack specifically to clear headers, etc. Multiple intermediate steering links can be used, just follow the manufacturers guidelines on maximum angles, center supports, joint alignment, etc.

New linkage.jpg

[edit] Power assisted steering

[edit] steering pump

If you are going to install a power assist unit, there are a few more issues. The power steering pump for the Cavalier rack is designed to fit a transverse mounted front wheel drive 4-cylinder or V6; no attempt was made to fit it to an early '80s SBC engine for obvious reasons. Instead, an original type pump was used that fit the original brackets. Piece of cake. The rack will have a tube and O-ring style connector for the metric high pressure hose. There are adapters available that convert the metric fitting to a flare fitting, similar to the early model pump.

[edit] Hoses

Because the pump and the rack sector end up so close together, you may want custom hoses made. Once everything is hooked up and functioning, you can test drive your creation. It may not feel just right. A rack is more sensitive than a RB box. Higher caster settings will help keep it centered.

[edit] Steering "feel"

If it is simply too “light”, “soft”, or “over responsive”, you may want to address the flow and pressure differentials between RB and R&P.

[edit] Pressure relief valving

Most older GM pumps will have a pressure rating as high as 1350 psi. The '82–'94 Chevy S10 has the lowest rating at around 1100 psi. The Cavalier rack was designed for a pressure of ~1,000 psi. You can adjust the pressure with a simple shim kit from Borgeson.

The pressure relief valve is located behind the high pressure output fitting on the back of the pump. You may need a small magnet to pull it out of the recess. The shim kit comes with a tool to help with removing the end nut, a new O-ring for the outer fitting, and several shims with a guide on how many shims to use to attain the desired pressures.

Basically, adding shims reduces the pressure on the spring, allowing the bypass to open sooner and recirculate the fluid, rather than force it to the rack- thus making the pressure lower. The kit allows for reduction to about 750 psi, suitable for a Mustang II rack, which allows you to get well below the normal range for a Cavalier rack. Reduce the sensitivity to what suits you. In most cases, the pressure reduction valve can be removed and replaced with the pump in the car. Yes, you have to drain and refill the pump each time, but that’s pretty minor.

[edit] Flow rate

The second adjustment available is flow rate. The flow rate is determined by the size of the hole in the high pressure fitting (the one you took out to get to the pressure relief valve) on the back of the pump. The earlier pumps had an output hole of 5/32". The Cavalier pump has an output hole of 1/8" (approximately 20% smaller). I could not find a fitting with the smaller orifice to fit the older pump (newer GM pumps use metric fittings). Perhaps if you start with a later S10 pump, metric might not be an issue- but this needs to be confirmed first. As an alternative, it was fairly simple to weld shut the orifice in the original fitting and re-drill it to 1/8".

Note: The experts contend that reducing the pressure will reduce the amount of assist provided, which can be reduced below factory specs to give it a heavier feel. The system is designed with a variable ratio, that is, it is designed to give more assist the further you turn the wheel, like for parking maneuvers, etc. You can't harm it by providing less than factory pressure. The flow rate seems to be more of a factor in the sensitivity over center, where you really don't need any assist. In any event, both changes were noticeable from the original test drive.

High pressure hose fitting from the back of the power steering pump
Weld or braze, then re-drill the center hole to change flow rate, not the side hole
You can edit this article right now.

Just click the "edit" tab at the top of this page, or click one of the [edit] links to the right of an article section.

For more info, see Help:Editing pages.

Personal tools