Until recently my 147 was affected by the steering issue that I have lengthily described below. IMO this is a design problem, probably arising from an attempt to reduce NVH, but it also has very undesirable side effects on the steering. As far as I can tell, all 147s / 156s et al will almost certainly be afflicted with this issue.
One issue I’ve had with the 147 steering is that for a rack with such a ‘fast’ or ‘quick’ steering ratio, my steering never felt as precise nor as responsive, nor as ‘fast’ as I expected it ought to. Rather, I’ve always found the steering a bit ‘mushy’, a bit lacking in on centre feel, and not feeling quite ‘connected’ with the front wheels (almost as though there might be squidgy rubber parts somewhere in the steering system...).
Anyway, this is now no longer a problem, the cause has been identified, and rectified...
---------------
The other week I was performing an alignment on the toe settings (stringlines etc), and I was experiencing difficulty with the front toe angles not responding as expected to tie-rod length adjustments. The problem was that the front toe angles kept randomly changing during the alignment procedure, and also randomly NOT changing as a tie rod length was significantly changed. When attempting to make only a very small toe adjustment the tie-rod could be rotated quite a lot but cause zero toe change, and then a tiny rod rotation would suddenly see a substantial change in the toe angle.
I had noticed this difficulty during previous alignment sessions (making the alignment procedure a tedious PITA), but this time I decided to try and find the root cause (not immediately obvious). To start I gave the front suspension / steering the usual 'health check' by jiggling, wiggling and prying all the ball joints, tie-rod ends, rack ends, wishbone bushes etc, and as expected all were good.
However, with the front wheels jacked off the ground and the steering wheel ‘locked’ at the straight ahead (immobilising the rack shaft and isolating any potential left side issues from right side issues), manually applying inward / outward force to each front wheel (at 3 o’clock and 9 o’clock), it was easily possible to cause either front wheel to move around the steering axis, to a quite significant degree.
This movement was not a ‘knocking’ free play, but a ‘squidgy’ kind of rubbery vagueness, occurring at both wheels. This could be readily seen as substantial toe change when force was applied to the wheel. The tie-rod could be seen moving axially relative to the ’locked’ rack shaft, but there was no ‘loose’ or obviously worn component that I could find.
But, I did manage to identify the source of this unfortunate rubbery ‘plunging’ movement. It occurs in the tie-rods, though not in the rack-ends nor tie-rod-ends as might be expected, but WITHIN the tie-rods themselves...
-------------------
So how is this even possible:
In my experience, every tie-rod I’ve previously ever seen was just simple steel rod connecting two articulated joints (i.e. the ‘rack-end’ axial joint and the tie-rod-end). However the 147 / 156 (et al) tie-rod is not a simple steel rod, but something more complicated, i.e. the inclusion of an unusual steel cylinder located near the inner axial joint, hidden inside the rack bellows.
I’d noticed these rather unusual cylinders on the 147 tie-rods some time ago, but never given it much thought and then just forgot about it. But with my attention now focussed laser-like on these strange tie-rods, closer inspection revealed that lurking within each of these cylinders is a soft rubber bush that obviously must permit some degree of axial plunge...
This means that these tie-rods are not solid fixed length rods, but two metal components connected by soft rubber. When loaded the rubber must permit movement between the two metal parts, so causing the tie-rod length to randomly change depending on the strength and / or direction of axial load passing through it at any given transient moment.
This cannot be a ‘good thing’ in a tie-rod, because it will cause the toe settings to be significantly random, i.e. toe will not be stable (among other issues). The steering will have less ‘fidelity’, be vaguer than it otherwise would be, and also less predictable. This would explain the numerous occasions in which I could have sworn that I had felt the car momentarily self steer, most often only slightly, but sometimes more than that (nah, must have imagined it...).
This is a novel form of bump steer, deliberately allowed just to reduce NVH issues. It isn't too dissimilar to problems caused by a rack housing attached to the chassis using overly soft rubber bushes. But, I suspect that it is probably worse since each front wheel will be able to randomly change its’ toe angle individually (a softly mounted rack will at least allow both wheels to retain their relative toe angles, regardless of bushing compliance).
Also, there is no way to know what the toe setting actually is when driving the car. Just because the toe was set to X during alignment, does not mean that it will be X when the car is moving. Even when driving in a straight line, any road force acting around the steering axis will cause these tie-rod bushes to deform, so causing the tie-rod length to increase or decrease, and so causing toe to randomly change, and probably differently on each side of the car...
--------------------
Finding the problem:
With the steering wheel locked in position, and with road wheel and the steering bellows removed (visually exposing the entire tie-rod), when a firm axial force is manually applied to the tie-rod (i.e. pushing or pulling hard against the end of the steering arm), the tie-rod can be seen to become significantly shorter and longer as the tie-rod bush rubber distorts.
The bush rubber is soft enough that I could observe the length of the tie-rod change just by force of my human muscles axially loading the tie-rod. Now I am fairly strong, but I am very weak relative to the forces that will be acting on the tie-rods in use. If I can apply enough force myself to change the tie-rod length by Xmm, then it must change a lot more than Xmm in use...
It could possibly be the case that the rubber in my cars’ tie-rods had become softened with age. I tend doubt it as the rubber visually appeared to be in very good condition.
I also checked the rubber compliance in another way. Off the car I used a drill press to impart an axial load into the tie-rod (i.e. orienting the tie-rod vertically in the drill press with the chuck pushing down on the end of the rod). It took very little force on the drill press handle to considerably distort the bush rubber.
Admittedly this was a very rough and ready ‘test’ providing no actual data, but I did have had four tie-rods to check. All four looked to have perfect rubber bushes, and all four bushes seemed to be of a very similar stiffness, the bushes all defelcting to a very similar degree with only a light pull on the press handle.
------------------
Resolution:
I modified the tie-rods to effectively eliminate the rubber bushes, and they are now effectively ‘solid’ (the rubber still exists, but enclosed inside a rigid welded-up ‘canister’). The tie-rods now have zero compliance, the length is stable and does not change at all. This is a huge improvement, the steering is now much, much, much improved, in a number of ways.
There is no longer anywhere near as much ‘soft’ spongy movement manifested at the steering wheel. In other words, the ‘steering gain’ is much steeper, i.e. the rate at which the steering ‘weight’ loads up per degree of initial steering wheel rotation each side of the straight ahead position is increased (i.e. the steering ‘weight’ gets heavier faster per degree of initial steering wheel rotation, creating a feeling that there is very little ‘lost’ motion at, near and around the straight ahead position).
The steering now has a much more immediate affect even with only a very small steered input, much more so that it previously did. Understeer is significantly decreased. Steering feel and response are substantially improved. The steering wheel feels far more ‘connected’ to the front wheels, and I can better feel what the front tyres are doing.
Since modifying the tie-rods I haven’t yet noticed any instances of significant unexpected ‘self steering’, which happened reasonably often prior to modifying the tie-rods. The steering is overall just quite a lot ‘nicer’ than it was before.
This is one of the best modifications I have done to this car. The standard tie-rods really do sabotage the steering, best to get rid of them for solid tie-rods (from some other Alfa?), or modify them to the same effect.
There is a small increase in road noise on coarser sealed surfaces (not absorbed in rubber). There is also a noticeable increase in fine grained vibration felt through the steering wheel, also on coarser surfaces. These are not particularly objectionable, a price well worth paying for less ‘corrupted’ steering...
Regards,
John.
One issue I’ve had with the 147 steering is that for a rack with such a ‘fast’ or ‘quick’ steering ratio, my steering never felt as precise nor as responsive, nor as ‘fast’ as I expected it ought to. Rather, I’ve always found the steering a bit ‘mushy’, a bit lacking in on centre feel, and not feeling quite ‘connected’ with the front wheels (almost as though there might be squidgy rubber parts somewhere in the steering system...).
Anyway, this is now no longer a problem, the cause has been identified, and rectified...
---------------
The other week I was performing an alignment on the toe settings (stringlines etc), and I was experiencing difficulty with the front toe angles not responding as expected to tie-rod length adjustments. The problem was that the front toe angles kept randomly changing during the alignment procedure, and also randomly NOT changing as a tie rod length was significantly changed. When attempting to make only a very small toe adjustment the tie-rod could be rotated quite a lot but cause zero toe change, and then a tiny rod rotation would suddenly see a substantial change in the toe angle.
I had noticed this difficulty during previous alignment sessions (making the alignment procedure a tedious PITA), but this time I decided to try and find the root cause (not immediately obvious). To start I gave the front suspension / steering the usual 'health check' by jiggling, wiggling and prying all the ball joints, tie-rod ends, rack ends, wishbone bushes etc, and as expected all were good.
However, with the front wheels jacked off the ground and the steering wheel ‘locked’ at the straight ahead (immobilising the rack shaft and isolating any potential left side issues from right side issues), manually applying inward / outward force to each front wheel (at 3 o’clock and 9 o’clock), it was easily possible to cause either front wheel to move around the steering axis, to a quite significant degree.
This movement was not a ‘knocking’ free play, but a ‘squidgy’ kind of rubbery vagueness, occurring at both wheels. This could be readily seen as substantial toe change when force was applied to the wheel. The tie-rod could be seen moving axially relative to the ’locked’ rack shaft, but there was no ‘loose’ or obviously worn component that I could find.
But, I did manage to identify the source of this unfortunate rubbery ‘plunging’ movement. It occurs in the tie-rods, though not in the rack-ends nor tie-rod-ends as might be expected, but WITHIN the tie-rods themselves...
-------------------
So how is this even possible:
In my experience, every tie-rod I’ve previously ever seen was just simple steel rod connecting two articulated joints (i.e. the ‘rack-end’ axial joint and the tie-rod-end). However the 147 / 156 (et al) tie-rod is not a simple steel rod, but something more complicated, i.e. the inclusion of an unusual steel cylinder located near the inner axial joint, hidden inside the rack bellows.
I’d noticed these rather unusual cylinders on the 147 tie-rods some time ago, but never given it much thought and then just forgot about it. But with my attention now focussed laser-like on these strange tie-rods, closer inspection revealed that lurking within each of these cylinders is a soft rubber bush that obviously must permit some degree of axial plunge...
This means that these tie-rods are not solid fixed length rods, but two metal components connected by soft rubber. When loaded the rubber must permit movement between the two metal parts, so causing the tie-rod length to randomly change depending on the strength and / or direction of axial load passing through it at any given transient moment.
This cannot be a ‘good thing’ in a tie-rod, because it will cause the toe settings to be significantly random, i.e. toe will not be stable (among other issues). The steering will have less ‘fidelity’, be vaguer than it otherwise would be, and also less predictable. This would explain the numerous occasions in which I could have sworn that I had felt the car momentarily self steer, most often only slightly, but sometimes more than that (nah, must have imagined it...).
This is a novel form of bump steer, deliberately allowed just to reduce NVH issues. It isn't too dissimilar to problems caused by a rack housing attached to the chassis using overly soft rubber bushes. But, I suspect that it is probably worse since each front wheel will be able to randomly change its’ toe angle individually (a softly mounted rack will at least allow both wheels to retain their relative toe angles, regardless of bushing compliance).
Also, there is no way to know what the toe setting actually is when driving the car. Just because the toe was set to X during alignment, does not mean that it will be X when the car is moving. Even when driving in a straight line, any road force acting around the steering axis will cause these tie-rod bushes to deform, so causing the tie-rod length to increase or decrease, and so causing toe to randomly change, and probably differently on each side of the car...
--------------------
Finding the problem:
With the steering wheel locked in position, and with road wheel and the steering bellows removed (visually exposing the entire tie-rod), when a firm axial force is manually applied to the tie-rod (i.e. pushing or pulling hard against the end of the steering arm), the tie-rod can be seen to become significantly shorter and longer as the tie-rod bush rubber distorts.
The bush rubber is soft enough that I could observe the length of the tie-rod change just by force of my human muscles axially loading the tie-rod. Now I am fairly strong, but I am very weak relative to the forces that will be acting on the tie-rods in use. If I can apply enough force myself to change the tie-rod length by Xmm, then it must change a lot more than Xmm in use...
It could possibly be the case that the rubber in my cars’ tie-rods had become softened with age. I tend doubt it as the rubber visually appeared to be in very good condition.
I also checked the rubber compliance in another way. Off the car I used a drill press to impart an axial load into the tie-rod (i.e. orienting the tie-rod vertically in the drill press with the chuck pushing down on the end of the rod). It took very little force on the drill press handle to considerably distort the bush rubber.
Admittedly this was a very rough and ready ‘test’ providing no actual data, but I did have had four tie-rods to check. All four looked to have perfect rubber bushes, and all four bushes seemed to be of a very similar stiffness, the bushes all defelcting to a very similar degree with only a light pull on the press handle.
------------------
Resolution:
I modified the tie-rods to effectively eliminate the rubber bushes, and they are now effectively ‘solid’ (the rubber still exists, but enclosed inside a rigid welded-up ‘canister’). The tie-rods now have zero compliance, the length is stable and does not change at all. This is a huge improvement, the steering is now much, much, much improved, in a number of ways.
There is no longer anywhere near as much ‘soft’ spongy movement manifested at the steering wheel. In other words, the ‘steering gain’ is much steeper, i.e. the rate at which the steering ‘weight’ loads up per degree of initial steering wheel rotation each side of the straight ahead position is increased (i.e. the steering ‘weight’ gets heavier faster per degree of initial steering wheel rotation, creating a feeling that there is very little ‘lost’ motion at, near and around the straight ahead position).
The steering now has a much more immediate affect even with only a very small steered input, much more so that it previously did. Understeer is significantly decreased. Steering feel and response are substantially improved. The steering wheel feels far more ‘connected’ to the front wheels, and I can better feel what the front tyres are doing.
Since modifying the tie-rods I haven’t yet noticed any instances of significant unexpected ‘self steering’, which happened reasonably often prior to modifying the tie-rods. The steering is overall just quite a lot ‘nicer’ than it was before.
This is one of the best modifications I have done to this car. The standard tie-rods really do sabotage the steering, best to get rid of them for solid tie-rods (from some other Alfa?), or modify them to the same effect.
There is a small increase in road noise on coarser sealed surfaces (not absorbed in rubber). There is also a noticeable increase in fine grained vibration felt through the steering wheel, also on coarser surfaces. These are not particularly objectionable, a price well worth paying for less ‘corrupted’ steering...
Regards,
John.
Unfortunately FIAT ePer dives off into parts substitutions for the complete rack, with no details about rack tie rod construction. Nor are tie rods listed separately and I am none the wiser.
