As the title describes, this thread is intended as a technical discussion on the 932 suspension, handling and tuning. This is a continuation of the conversation that evolved from the rear lateral control arm mod described by Johnl here:
http://www.alfaclubvic.org.au/forum/index.php?topic=17233.0
156/147/GT etc suspension mods and tuning can be discussed here
By the way, there's alternative to johnl rear suspension arms - Hardrace make adjustable arms for 156/147 which can come with rubber bushings or even with ball joints. It's an option for people like me who don't have the skills. They're quite pricey at around $650 for a set of 4.
https://www.hardracesuspension.com.au/find-parts?car_make=629&car_model=630&car_type=631
As for the front, there's a set of eccentric bushes for front upper arms which allows for camber adjustment. Cost 280 USD. These guys also do rear arms similar to Hardrace but even more expensive.
http://www.highpower-racing.com/index.php?route=product/product&product_id=59
Now, I believe that a shock/spring kit will yield much noticeable difference to handling. I can recommend ST sport suspension kit:
https://www.cheap-coilovers.com/shop/st-sport-suspension-alfa-romeo-156-8196c1.html
It cost me about $750 delivered and I absolutely love it. It's stiff but not bumpy. I find it comfortable for daily driving and at the same time performing really well on a track.
Quote from: warsch on March 06, 2018, 08:58:44 AM
By the way, there's alternative to johnl rear suspension arms - Hardrace make adjustable arms for 156/147 which can come with rubber bushings or even with ball joints. It's an option for people like me who don't have the skills. They're quite pricey at around $650 for a set of 4.
Both DV and myself have spent approximately 1/10th of that (and some time to put it all together)...
Having said that, I would suggest that if you can't home bake it, then the Hardrace arms would be worth the price for the improvement they will make (if you can justify the expense, which at present I'd struggle to do...).
Quote from: warsch on March 06, 2018, 08:58:44 AMNow, I believe that a shock/spring kit will yield much noticeable difference to handling. I can recommend ST sport suspension kit:
https://www.cheap-coilovers.com/shop/st-sport-suspension-alfa-romeo-156-8196c1.html
It cost me about $750 delivered and I absolutely love it. It's stiff but not bumpy. I find it comfortable for daily driving and at the same time performing really well on a track.
Springs and dampers do what they do, but if the geometry is unstable then the full potential is not realised. An analogy might be; would you fit an uprated front ARB if a ball joint was worn out, or would you fix the ball joint first? The overly soft Alfa rear control arm bushes are kinda sorta like worn ball joints...
ST 'coilovers' are I think made by KW, a Geman company with some reputation. I think they are probably KW 'coilovers' with all the adjustments stripped off to produce a 'budget' alternative.
Regards,
John.
If all 932 series cars drove like mine I can't understand how they could be enjoyable (or fast) to drive in sprints/motorsport. From my experience, i believe that it would be reasonable for these bushes/arms to be an allowable modification in the AROCA competition rules for standard class, and in the TS cup technical regulation (not that I intend to sprint mine), much like the exemption for 105 trailing arm/trunion bushes and 116/75 castor arm bushes. It would certainly make them more competitive. However I will leave it to the competitors to petition the committee on this subject...
Wheel alignment today showed that the front end is out of spec... Camber is about -1.5 and caster 2deg on the left and 2.5 on the right...
I dont mind the camber but would really like some more caster to give the steering some more weight.
The camber didnt look right so I measured it with a large square and ruler. Its quite a bit different to the wheel alignment report... is this common for commercial alignment equipment to get it wrong? Also, as caster isnt adjustable on the 147, what sort of results can be obtained using a proper stringline technique? The way i did the stringline last time was wrong - lots of toe out, but it was symmetrical. I suspect I got the camber ok given how wrong this commercial rig seems to be...
When you drive it does it track straight with the steering wheel centred properly?
Not really enough to be a problem bazz... the caster is different side to side by 0.5 degrees,as is camber, but it tracks Ok.
Ill have a go at stringing the car properly but the caster and camber being so far out concerns me. Could this be due to a poor accident repair?
Never thought of that, but yes definitely, or you have a bent arm or mount somewhere.
On many cars, you can shuffle the subframe a skerrick, this might help even it up.
In a heavy front hit, the longerons crease slightly fairly close to the firewall (approx. plumbed down from the rear battery tray bolt). Hard to see with the motor in there. I've measured a couple of unbent or lightly damaged shells and the build always seems pretty true on diagonal checks.
Quote from: Divano Veloce on March 08, 2018, 07:11:49 PM
The camber didnt look right so I measured it with a large square and ruler. Its quite a bit different to the wheel alignment report... is this common for commercial alignment equipment to get it wrong? Also, as caster isnt adjustable on the 147, what sort of results can be obtained using a proper stringline technique? The way i did the stringline last time was wrong - lots of toe out, but it was symmetrical. I suspect I got the camber ok given how wrong this commercial rig seems to be...
For stringlining the toe- measurements it is important to note that the front track is nominally 16mm wider than the rear track. This has an affect on the toe measurements (front and rear) if the stringlines each side are set to be the same distance from the wheel centres front and rear (to place the lines I measure to the centre of the little plastic 'hub caps', assuming each wheel to be CAD CAM machined equally and the caps to be pretty much identical in dimension). To get the lines parallel to each other and the car centreline, the lines each have to be placed so that they are 8mm farther from the rear wheel centres than from the front wheel centres.
Note that these are only nominal numbers for relative track widths, and the reality may not be exactly the same as the nominal widths. Ideally the actual track widths should be measured. I did measure mine when I first aligned my 147, and they were very close to the nominal, so I now just use the nominal numbers, because I'm a bit lazy (though I really should re-measure with the 'new' Toyota based arms now fitted...). Note that for this purpose the 'track' should be measured at hub height, not at contact patch level.
So, how much can this affect the toe alignment? If we assume that the difference in track widths has not been accounted for and toe has been set with the lines placed at the same distance from the front and rear wheels (meaning the lines will converge toward the rear of the car due to the narrower rear track), and the toe has been set to 'zero' (according to the lines), then this will result in each wheel being in reality at 2mm of toe-out. This times 2 for an axle pair, and the total toe-out will equal 4mm, which is quite a lot of error. For example, if you though you had toe set to say zero, it would actually be 4mm toe-out.
These numbers were found using 'full size' drawings in CorelDraw. The shorter the wheelbase the worse this issue is, and the 147 has a fairly short wheelbase. Note too that the wheels should be checked for run-out at the rims, as bent rim edges will (of course) affect any measurements made using the rim edges as data points.
As CB already suggested, caster can be changed a bit by moving the sub-frame forward on one side and backward on the other. There isn't much scope here, the sub-frame can't be 'nudged' more than a few millimeters. Note that 1cm of sub-frame movement (you won't get that much...) on a given side will move the lower ball joint longitudinally by about the same amount, and change caster by about 1°.
Caster can also be changed by longitudinally slotting the four holes at the top of the 'strut' tower, i.e. the four holes through which the studs in the casting that is the upper wishbone mounting protrude. There is also a round 'lump' of aluminium on top of these castings, which will need to be cut away so that it can move in the hole in the strut tower. Again, longitudinally moving the upper wishbone mount casting by 1cm will move the upper ball joint by 1cm and give very near 1° of caster change (and you can probably get that much change here).
Note that with this suspension design, the caster angle changes with suspension movement (increasing as ride height lowers / wheel goes into 'bump'), so ideally caster should be measured with the car on a properly level surface. It is hard to measure caster directly, there are no convenient machined surfaces from which to do so. I have made a jig that has two 'arms' which pass through the wheel spokes and 'lock into' the stub threads protruding through the nuts on the upper and lower ball joints, the 'arms' being welded to a square section steel tube connecting the two arms. I measure caster angle from this tube using a digital inclinometer. Sorry, hard to describe this thing...
I'd also like a fair bit more caster, but It's a real pity that it would be very difficult to achieve substantial caster increase. The architecture of this suspension is not conducive without major modification...
Measuring camber, it's important that the car is on a level surface, of course. Note that with only driver weight in the car, that camber changes more on the passenger side of the car than on the driver side. This is because the driver weight causes the body to lean measurably toward the driver side. This body 'roll' adds negative camber to the passenger side wheels and positive camber to the driver side wheels. However, the driver side gains very little net camber change because there is a negative camber gain associated with the vertical suspension movement on the driver side, which counteracts the positive camber change caused by the body leaning over...
Hope that isn't too clumsily explained...
Regards,
John.
Quote from: Divano Veloce on March 08, 2018, 07:11:49 PM
The camber didnt look right so I measured it with a large square and ruler. Its quite a bit different to the wheel alignment report... is this common for commercial alignment equipment to get it wrong? Also, as caster isnt adjustable on the 147, what sort of results can be obtained using a proper stringline technique? The way i did the stringline last time was wrong - lots of toe out, but it was symmetrical. I suspect I got the camber ok given how wrong this commercial rig seems to be...
"Large square and ruler" is going to be pretty rough, IMO. I use a digital inclinometer, and a simple square tube jig that spans the wheel rim edge to rim edge (measure camber off the jig face).
Commercial equipment can easily be out of spec. It costs money and / or time for the owner to calibrate or get it calibrated, and if no-one is complaining...
Stringlines can't really measure caster, at least if it is possible I have no idea how to do it.
Regards,
John.
The HPR poly eccentric bush kit mentioned in an previous post in this thread should provide some increase caster by allowing the upper ball joint to be moved rearwards a tad.
Bushes like that will adjust caster, and camber. They will change both at once because rotating an eccentric cam will cause the upper ball joint to move both laterally and longitudinally. This might be OK or might not be, i.e. you may be able to achieve a desired caster angle and a desired camber angle at the same time, or may not be able to. At two different combinations of eccentric cam rotations you may get say the X caster you want, but quite possibly not the Y camber you want (i.e. it would be Y+ camber at one combination of cam rotations giving X caster, and Y- camber at the other combination giving X caster). The affects of either unequal caster or unequal camber side to side might be significant, ot might not be, depending.
Also, poly bushes do not like being misaligned, and only to operate in single axis of rotation. Lets assume as a starting point, that both eccentric cam bushes are fitted in the 'wishbone' at the same eccentric rotation. The wishbone will freely articulate around a single axis because the bushes are perfectly aligned with the pivot bolt and free to rotate around this axis with no binding. If we now rotate one bush, then the pivot bolt is moved at that bush, the bolt becomes 'kinked' relative to the bush and the axis of rotation in that bush no longer matches the orientation of the pivot bolt. Nothing actually gets 'kinked' because the clearances are quite tight and the bolt pretty rigid, but what happens instead is that the poly bushing material becomes compressed and decompressed in different parts of the bush.
This condition may now exist at both bushes in the wishbone, with some degree of bind in both bushes. Where a part of the bush is compressed the friction in that part of the bush increases, possibly substantially, depending. On the opposite 'uncompressed' side of the bush bore a gap will appear between the poly and the bush sleeve, which may allow ingress of dirt.
Any binding may be very minor if any bushing misalignment is quite small, or may be substantial enough to be problematic if misalignment is substanial (i.e. if a lot of the potential adjustment has been used). I do know from experience (with ARBs mounted in poly bushes) that it takes only a very small amount of 'tightness' in the bush to cause quite a lot of binding.
This kind of eccentric adjustment is more suited to use with rubber bushes, that will tolerate substantial misalignment with no issues, or with spherical bearings, and less suited to use with plastic bushes.
Plastic bushes with eccentric adjusters (or anything that causes some degree of internal misalignment) may work acceptably in a given application at a given adjustement (or below a certain amount of adjustment), or, loads caused by misalignment may squeeze out lubricant, allow dirt to enter, cause the bushing to become noisy and / or to wear prematurely.
Regards,
John.
Divano Veloce wrote:
"I fitted the GTA bar last night and test drove it this afternoon expecting another quantum leap in handling. Alas no... I suspect that my string line wheel alignment is not quite perfect and with the GTA bar the rear end seems to dart about with weight transfer.... This was hardly noticeable with the 14mm bar.
DV,
I've moved this reply into your 'suspension tuning' thread.
I've been wondering why your experience with a stiffer rear ARB seems so different to mine. I don't have an answer, just a speculation...
Substantially increasing the rear roll stiffness means that for any degree of lateral aceleration there will be an increase in lateral weight transfer across the rear axle line (and a corresponding decrease in lateral weight transfer across the front axle line, which is why increasing rear roll stiffness decreases understeer, i.e. it causes a reduction in rear grip and an increase in front grip).
This means that more of the lateral force will be resisted by the outside rear tyre and less by the inside rear tyre, so more force is acting on the outside tyres' sidewall, which will then laterally deflect more. If the sidewall is relatively soft then you might be getting a lot of lateral deflection in the outside tyre, that you weren't getting to the same degree when both rear tyres were more equally sharing the lateral load.
The sidewall may deflect fairly easily until it reaches X deflection, and then abrupty reach the limit to which it will easily deflect. An abrupt effective stiffening of the sidewall under lateral loading might cause a momentary reduction in grip, at which point the sidewall might 'relax' and grip then be regained, and then the sidewall will load up again, etc. This non linearity might cause a 'darty' behaviour, or I might be speculating myself up a garden path...
I don't really know, but perhaps a stiffer tyre, or increasing rear pressure (which effectively stiffens the tyre case) might help. I'm running a lot of pressure in my rear tyres (way more than the front tyres), and while the handling is sharp and responsive, I wouldn't really describe it as 'darty'.
Regards,
John.
Hi Johnl,
Although different in design to the 147, installation of eccentric poly bushes to my 159 upper control arms has been very successful. Their is enough compliance in the polyurethane to cope with the off axis alignment when setting the respective eccentrics. This plus the grease supplied wth the bush kit eliminates any stiction. Very simple to dial in camber and caster adjustments. A very cost effective mod that improves the cars' ability to use the tyres more efficiently. Had them in the car for around 3 years without any dramas whatsoever.
Hi Colin,
I'm pleased this is working for you.
You say; "Their is enough compliance in the polyurethane to cope with the off axis alignment when setting the respective eccentrics." Which suggests these bushes are not the harder 'racing' grade?
I suspect that softer grades of poly may have enough give that some misalignment isn't a significant problem. My experience with poly bushes has been with harder 'racing' grade plastic. I suspect my misgivings would likely be valid for such harder polyurethane, but I could possibly be wrong, and it may also be very case specific.
What I have experienced with hard poly 'D' bushes (for ARB mounting), is that only a very small amount of 'crush' between the bush and the ARB causes a great deal of friction, and in a fairly short amount of time the grease squeezes out and causes the bush to become noisy ('creaky' / 'groany').
Regards,
John.
I had some time last night to investigate. It looks like the front sub-frame is a replacement spare part item, along with the lower engine mount rubber as they have spare parts type labels on them. I removed some of the sub-frame bolts and it looks like there's some scope for moving the sub-frame forward and to the drivers side but to get the caster back in spec it suspect it would need some slotting. I would like to get it back in spec and maybe a bit more, see if this improves the super light and un-communicative steering that is now the main cause of my discontent with the car. As the steering rack moves with the sub-frame, and that the steering arms are fairly low down on the spindles, what sort of bump steer impact would moving the sub-frame have compared to moving the strut tops?
Some stuff seems to be ex-inventory on all 147s, look at the label date if it can be read.
CB, I cant find a date but there is "Aust" on the label which might be a reference to Australia. The label is pretty deteriorated from exposure.... I couldnt see any evidence of repair or damage underneath the car.
Quote from: Divano Veloce on March 13, 2018, 01:03:41 PM
I removed some of the sub-frame bolts and it looks like there's some scope for moving the sub-frame forward and to the drivers side but to get the caster back in spec it suspect it would need some slotting. I would like to get it back in spec and maybe a bit more, see if this improves the super light and un-communicative steering that is now the main cause of my discontent with the car.
DV,
the lower sub-frame bolts insert into significantly long tubes (in the frame) that do not lend themselves to being 'slotted'. I don't think you'll achieve any significant caster adjustment increase by such a means, let alone any caster increase worth having....
If you were extremely enthusisatic it might be possible to move all of the tubes in the sub-frame, i.e. cut the tubes out, elongate the resulting holes in the sheet metal, weld the tubes back in (with gussets to cover where the holes had been slotted). Of course this would be a big project, requiring the sub-frame to be removed, possibly requiring the rack to be reloacted on the sub-frame, plus other unknown etceteras....
Of course this is assuming it is even possible to move the subframe forward by moving the mounting tubes in the frame, it may not be practically possible at all...
Quote from: Divano Veloce on March 13, 2018, 01:03:41 PM
As the steering rack moves with the sub-frame, and that the steering arms are fairly low down on the spindles, what sort of bump steer impact would moving the sub-frame have compared to moving the strut tops?
I don't think that moving the subframe forward would directly affect bump steer (?). There may well be some indirect affect caused by any consequent change in caster angle.
An
increase change in bump steer will occur with any change to the caster angle. This is so whether the change in caster has been created by moving the upper ball joint longitudinally, or the lower ball joint. For example, if caster were to be increased (by whatever means), then the end of the steering arm (in side view) will become lower to the ground. This changes the horizontal angle of the steering tie-rods (in front view), which become angled more downward from the rack end to the steering arm.
As a result of this change to the static tie-rod angle, the outer end of the tie-rod will now move (in bump or rebound) through a different arc relative to the lower and upper ball joints, than it previously did. In this example of increased caster angle, this change in relative arcs as seen in front view will cause the front wheels to gain more toe-in with 'bump' motion and more toe-out with 'droop' motion. As the chassis leans with body roll the outside wheel goes into bump and gains toe-in, the inside wheel goes into 'droop' and gains toe-out. This means that with body roll both front wheels dynamically 'toe' toward the cornering direction, so there is a 'roll oversteer' moment introduced, i.e. the front wheels 'auto-steer' in the direction of the corner, more than can be accounted for by the angle of the steering wheel, and it is all rather 'non-linear'...
This may be very minor, or it may be problematic. When I increased the caster angle on my CB7 Accord (more than doubled it, by moving the lower ball joint about 3cm forward), there was a significant increase in roll oversteer. This manifested as the chassis turning into the corner very well, then as roll motion occurred (even though roll motion was small due to large ARBs and stiffer springs / dampers) it would turn in more than the driver had intended, so the driver had to back off the steering as the car approached mid corner. At first this characteristic was entertaining, but the novelty soon wore thin.
I fixed this problem by lowering the Accords' steering rack, but I can't see this being feasible with the 147...
Regards,
John
Thanks John for another detailed response to my predicament..!
Another option to correct caster may be to move the lower control arm mounts about on the subframe.
I notice on the UK forum that the "sports kit" suspension alignment settings are more like -1.6 deg camber and 4.3 deg caster. Is this the result of lowering springs or does the sports suspension have different components to achieve these alignments?
First thing though is to verify the caster. I don't trust the results from the alignment shop as the camber alignment is way out side to side. John, can you tell me more about your caster measuring system?
I believe that's because when you lower 156 that results in more negative camber (and may be caster too). I put -40mm suspension on my car and camber is visibly different without any adjustment. Not that you can adjust it anyway. Sportpack suspension is lowered too so I assume that's why factory settings differ just to reflect that.
Quote from: Divano Veloce on March 14, 2018, 07:20:36 AM
Thanks John for another detailed response to my predicament..!
Another option to correct caster may be to move the lower control arm mounts about on the subframe.
DV,
Having given this some thought myself, I doubt that it's practical to try and achieve significant caster change at the lower subframe. Yes, it is possible to make a small change by moving one side of the subframe forward and the other backward.
You might also get a mm or two by loosening the bolts holding the lower wisbones and then levering the wishbones backward or forward while re-tightening the bolts.
All up, moving the subframe and the wishbones on the subframe, I would guess you might move the lower ball joints longitudinally by maybe 3 or 4mm, certainly not a lot.
I think that slotting the upper holes in the 'strut' tower offers more scope for changing caster, and will achieve X caster change with less collateral affect on wheelbase length. This is because the upper ball joint is significantly farther from the hub centre than is the lower ball joint, so moving the upper ball joint longitudinally by X has less affect on wheelbase than moving the lower ball joint by X. This is what I did, i.e. slotted the four holes in one of the towers with a rats tail file, elongating them by 5mm to increase caster by 0.5° to equalise caster side to side.
At the other tower I slotted the holes laterally (10mm), to change camber by 1°, in order to equalise the front camber. I would avoid slotting any individual tower hole both longitudinally and laterally, which I expect might weaken the tower top excessively. This is why I have adjusted only caster at one tower and only camber at the other tower, in both cases just equalise.
Quote from: Divano Veloce on March 14, 2018, 07:20:36 AM
I notice on the UK forum that the "sports kit" suspension alignment settings are more like -1.6 deg camber and 4.3 deg caster. Is this the result of lowering springs or does the sports suspension have different components to achieve these alignments?
My understanding is that the nominal caster angle for the 'normal' 147 is 3° (plus / minus 30'). I don't know for sure where the "sport kit" (whatever that is) finds another 1.3°. It might have some special parts, but I tend to doubt it.
I do know that this suspension design must gain additional caster with bump motion. This is because the upper wishbone pivot axis is not parallel with the chassis centreline (actually, not parallel to the lower wishbone pivot axis, which happens to be parallel to the chassis centreline). Because the upper wishbone pivot axis is at a substantial angle to the centerline, the upper ball joint moves in an arc that causes it to move not only laterally (affecting camber) but also longitudinally (affecting caster). The arc of motion of the lower ball joint differs because the lower wishbone pivot axis
is parallel to the chassis centreline, the lower ball joint moves in an arc that only causes the ball joint to move laterally (and not longitudinally).
This (arc of the upper ball joint having
some longitudinal moevement, and the lower ball joint
not having any longitudinal movement) means that lowering the ride height will cause an increase in caster as well as camber, but I don't know by how much per unit of lowering (though I could figure it out if my life depended on it...).
Quote from: Divano Veloce on March 14, 2018, 07:20:36 AMFirst thing though is to verify the caster. I don't trust the results from the alignment shop as the camber alignment is way out side to side. John, can you tell me more about your caster measuring system?
The thing with commercial alignmnent rigs, in my understanding, FWIW, is that they don't measure caster directly, rather they infer it from measuring the camber changes that occur when the steering is turned through a specific range of wheel angle (road wheels turned X° to the left and to the right). If the operator is very careful and knows what he / she is doing, and the rig is in good condition and properly calibrated then all should be reasonably accurate. If one of these things is not happening, then there is plenty of scope for the caster 'measurement' to be less than accurate...
My 'caster gauge' is very simple, but a bit hard to describe. Look at the wheel in side view. You'll see the upper ball joint above the height of the top of the tyre, and the lower ball joint through the spokes in the wheel. Now at each ball joint, can you see the ball joint post thread protruding through the nut? Keep these threads in mind when reading the following.
My 'gauge' consists of a square section tube that is at least as long as the distance between the threads in the upper and lower ball joints (in this case my tube is 520mm). Welded to each end of the square tube is an arm made from a flat piece of steel (25 X 2.5mm), one arm is 215mm long and the other is 150mm long. Each of these arms is at 90° to the tube axis. At the end of each arm is a semi circular notch (made with a rat tail file). To muse the jig, these two notches are each pushed up against the threads in the ends of the ball joint posts, so that the square tube is to the outside of the wheels, with one arm protruding through the wheel spokes and the other arm passing above the top of the tyre.
The notches at the end of each arm fit closely around the threads at each ball joint post, and should be a good match for the OD of each thread, with the lower ball joint thread being of larger diameter. The jig is held and pushed firmly against the ball joint post threads, and a digital inclimometer is held against a forward or rearward face of the square tube. An angle can then be read from this. Practice required to get repeatable measurements.
The lower arm is shorter and the upper arm is longer because the lower ball joint is farther outboard than the upper ball joint, so with the lower arm being shorter the square tube will be near vertical outside the car. Care must be taken to ensure that the axis of the two notches matches quite closely to the axis of the tube, otherwise the tool won't measure quite the same when measuring left side vs right side caster.
Regards,
John.