Recently installed 28mm torsion bars to our 75 V6 in lieu of the existing (standard) 25mm. Pretty happy with the result - reduced roll and dive etc. Also installed Pace knuckle risers to help manage camber change in roll.
I'd like to experiment with increasing front roll stiffness a tad more. Alfaholics do a 29mm dia front anti-roll bar. The existing / standard unit is 25mm. Has anyone had experience with a 29mm front bar on a 28mm torsion bar equipped 75 3.0?
For clarification sake the 3 litre V6 75s came out with:
22.8mm TBs
24mm front anti roll bar.
21mm rear anti roll bar.
I can't coment on the end results of your setup, but 28mm TB are 2.27 times stiffer than standard.
And a 29mm ARB 2.132 times stiffer than standard.
Prolly really needs to be matched with dampers/shock absorbers with a lot more rebound damping and a suitable increase in bump damping.
Hi Colin, I currently have a 29mm Swaybar with 29mm torsion bars. (GTV6 with short torsion bars, same as 75) Great improvement and works well even on the road. I'd reccomend talking to Vin about addressing the rear end to compliment your front end modifications. (Lower roll centre and some camber and toe) the roll centre change will also help reduce the screwed roll where the car wants to really lean on the outside front wheel and jacks up at the rear inside wheel.
Thanks for the replies.
Should have mentioned Koni Yellows installed so damping will be tuned once I sort the spring rates.
Good to learn 29mm anti-roll bar and torsion bars work on the road. The car is not likely to do a lot of track work.
With regard to the rear end I'd prefer to limit mods initially to spring rates and roll stiffness. Agree camber and toe changes to the rear would help but this is not a race car and at present I'm trying to resist cutting metal.
Car has been lowered about 20mm all round - Eibach "sport" springs in the rear. I figure the combination of knuckle risers and custom lower ball joints at the front has re-set / raised the front roll centre to some degree.
I'll lash out and get the 29mm front anti-roll bar. Plan to back to back test it against the "standard" front bar at Winton in November. I'll be using an impartial driver with a lot more motorsport experience than myself!
Are you going to be running R spec tyres at Winton? If not, I suspect you will get better times with the smaller AR bar but would be curious to see the result.
Hi GTVeloce,
I won't be running "R" spec tires. Just fitted the new Yokohama Advan V701 road tires all round so these will be used. Not looking for lap times. More interested in what increased front roll stiffness does for general balance. Car is quite "pointy" at present so a minor increase in corner entry understeer may not hurt. Running more caster than factory specification has improved high speed stability.
Doubling front roll stiffness may not work but playing around with setup - that's half the fun!
Well Winton last Sunday has been run and done.
A week or two prior to the event I replaced all rear anti-roll bar bushes (looks like they were the originals) with nylon (actually looks and feels like delrin) and polyurethane bushes. Rear end noticeably improved - definitely more linear and predictable. The old bushes were really bad. Car now is fully poly bushed throughout. End result was a more balanced chassis due to what feels like a marginal increase in mid corner and corner exit rear grip.
After timed practice and two heats on the long track in the morning, we installed the 29mm front bar over lunch. Discussion with driver prior to lunch suggested the front end was at its limit. Tire condition supported this. The comparison was however not ideal given the short track was used in the afternoon. However after lunch two heats with the stiffer 29mm bar were run.
From the side of the track (at the out-grid gate), the roll angle seemed very similar however tire squeal was more consistent and driver consistently found the apexes. Car "looked" a tad quicker. After two session we called it a day as the brakes had been VERY marginal all event - significant fade and slightly softening pedal. Given revs were limited to 4K for the day, lap times were good enough to suggest chassis (and tires) were right on the limit.
Driver reported front end measurably improved - certainly more grip and less road used per lap. Car now rear end and tire limited. From this its fair to assume the increased front roll stiffness allowed better use of the relatively low grip road tires. Close inspection of the tires confirmed less outer edge feathering. Tire pressure increase dropped a bit as well.
Next project is to improve brakes and replace the tired Konis. Also would be interesting to test in the wet!
Interesting, but also confusing and logic says, a bit condradictory.
If the car is well set up and your original post specs says that it is (apart from the dampers), then a fatter front bar should push the car more into understeer.
On a basically standard car you can see how a fat front bar can help reduce understeer, but with the higher front roll centre height and better camber curve characteristics, plus the decent TB upgrade, you've addressed most of the issues.
But based on the way you've said the tyres wear, it sounds like attention to camber, caster and toe angles would help a lot more.
The other thing is, with the long top ball joints the car will have toe in with bump travel.
Pretty much everything you read about bump steer says that toe in with bump travel should be avoided at all costs.
Because toe in with bump travel causes 'roll over steer'. And by that it means that the outside wheel will turn more into the corner because of the body roll, than what the driver intentended.
So the steering's behavior is inconsistant.
It's great that you've shared your thoughts on the effectiveness of the 29mm bar.
Hi Duk,
This car now runs 16" wheels in lieu of the original 14's. Given the change in wheel dia, recent installation of knuckle risers and about 20mm reduction in ride height when 28mm TB's installed, the front roll centre is highly unlikely to be as manufactured. Given the improvement in front grip the stiffer front bar seems to have realized, its likely the front roll centre is lower than standard. The front COG is probably at tad lower or close to standard. Increasing front roll stiffness and marginally lowering the COG has probably reduced to some degree the front roll centre vertical movement. The Rear roll centre is probably around 15 < 20mm higher than standard due to the larger diameter wheels / lower profile tires. The resultant rear COG would be close to standard or a bit lower given reduced ride height Eibach springs.
The resultant change in roll axis is far from ideal and although it can be improved I don't plan to relocate the Watts Link Pivot any time soon. Installing drop spindles is a low priority, however I'd like to improve bump steer a bit. This is a road car only and unlikely to be fitted with anything other than road tires. The car however needs new dampers as a priority.
The front end has some minor static negative camber on account of the knuckle risers. I agree more positive caster is probably worth a try. I may see if more caster can be added when I replace the Koni's, however overloading and binding of the control arm is a bit of a concern.
The object of Sundays exercise was to see what impact more roll stiffness had on the car and the test drivers feeling for the theoretical reduction in front roll centre movement. Could he quantify in real terms the improvement in handling a less dynamic roll centre should provide? All within the relative safety of a race track! Reading the physical condition of the front tires, the improvement in temperature spread across the tread and lower pressure increase suggests improved tire usage.
Changing the track layout for the afternoon added an unwanted variable however we did have a fun day!
For a "road" 75 the car corners relatively flat. This combined with "standard" road tires and the pilot driving well within his capabilities may have reduced to some degree the impact of roll over steer. The driver claimed with the stiffer bar far less steering correction was required. The consistency of tire squeal supports his opinion.
Colin
With the standard geometry and with a parallel lower control arm, the front suspension's roll centre height is well below ground level.
The long top ball joints do an excellent job of raising the RC.
http://vsusp.com is a great tool to help you plot the front suspension and see what it wants to do and how the long top ball joints effect the RC and camber curve.
Interesting discussion. If I can add a couple of thoughts from my 116 experience. You mention replacing the Koni dampers. Why and what with? I'd suggest getting them revalved to suit your new spring rates. Unless you want dual adjustable ones that it, but that seems overkill for road use.
I know you don't want to cut metal but that rear roll centre eventually becomes a real limitation. If you don't want to cut up your current de dion there are plenty of them around. And while you're at it an adjustable rear bar is really. really good to have.
When we used to build track Alfetta's eventually you'd get to the point where the front sway bar was almost irrelevant. I know everyone puts on stiff ones, but my view is that's just countering all the other issues with the design. Remember a sway bar works by limiting grip on one side, and transferring it to the other side. Great for turn in, opposite of what you want mid corner. You only have 4 tyres, you want all of them working. By fixing the roll centres, destabilizing the rear end a little you largely do away with turn in issues and remove a lot of the need for a sway bar at the front.
Hi aggie57,
The existing Konis were installed by the OP more than 25 years ago - I have all his receipts! I guess I could have them rebuilt and revalved, however I have adjustable Yellows on the 159 and the ability to tune without disassembly is a great feature. The Koni Yellows on the 75 are not externally adjustable.
I noted on Sunday many of the transaxle Alfas had the Watts pivot level with the base of the De Dion tube or thereabouts. Seems a popular mod. I'd certainly consider it but I'd like to do a bit of Roll Axis simulation / modeling prior to that. I've seen the adjustable rear bar and the tunability it allows looks attractive. Looks like major disassembly is required to swap rear bars however?
I'd like to try a say 27.5mm front bar - not seen one though! The 29mm does work on the track but I'd prefer a tad less roll stiffness for the road. Agree the stiff front bar is a bit of a band aid approach however its a road car and I'm a bit concerned lowered control arms as a result of drop spindles may compromise ground clearance. Thanks for the advice.
Colin
Now you mention it I do recall R&R of the rear bar needs the transaxle to be dropped somewhat. It's been years since I did that though, others will be more able to comment on the details.
You are right about clearance of the lowered watts linkage pivot. Ok on a track car of course.
Somethings that might help, in case it wasn't clear.
The rear roll centre is (rather obviously) defined by the centre pivot point of the Watt's linkage bellcrank. That pivot point remains largely consistant with its distance from the ground.
The front roll centre height is defined by the intersecting angles of the lines projected into space via thru the pivot points of the top and bottom control arms.
However, the important thing to understand is the distance between the front and rear centres of gravity and the front and rear roll centres.
These distances are defined as the 'Roll Couple(s)'. There is a front roll couple and a rear roll couple and is basically the LENGTH of the LEVER that the centrifugal force has to pull on the car and effect its lateral weight transfer and body roll.
The longer the lever, the greater the lateral weight transfer.
Actual body roll causes very little additional lateral weight transfer. Body roll is the byproduct of lateral weight transfer.
What body roll does to a tyre's contact patch, that's a different box of frogs. But really only effects independant suspension.
A Watt's linkage located rear suspension maintains a basically consistant roll centre height relative to the ground.
But due to longitudinal weight transfer from accelerating and braking, the rear centre of gravity moves down and up (repsectively).
This changes the length of the Roll Couple and effects lateral weight transfer behavior during acceleration and braking.
So the car trying to corner while accelerating will squat at the rear (depending of the spring rates and damping behavior) and have a slightly shorter rear roll couple and so so suffer slightly less lateral weight transfer onto the outside rear tyre.
Unfortunately there is more to add to the thinking and understanding department.
Because the roll centre height of the front suspension of every road going car is always lower than rear's roll centre height, during cornering, there is actually a frontal weight transfer. Weight that was otherwise supported by the outside rear tyre is transferred onto the outside front tyre.
This is referred to as the Roll Axis Inclination and is the angle from the rear roll centre height down to the front roll centre height.
The steeper this angle, the more dramatic the effect of longitudinal weight transfer during cornering.
While during braking, the rear of the car can rise slightly (again somewhat dependant on the damping behavior), the roll couple will get longer and during cornering, more lateral weight will be transfered onto the outside rear tyre.
There is a big fat 'BUT' here, tho. And again it is effected by spring rates and damping. But during braking, tho longitudinal weight transfer also results in a lot of nose dive. This changes the front suspension's roll centre height a significant amount and pushes the front roll centre height even lower.
Unfortunately there's another 'BUT'! Becuase the (standard geometry) front suspension's roll centre height will go down further than the front of the car's centre of gravity, the front roll couple gets longer. This, along with the nature of all front suspension roll centre heights being lower than the rear roll centre height, results in an even steeper Roll Axis Inclination and results in even more weight being transferred onto the outside front tyre......................
Incidentally, the conventional Watt's linkage pivot point (sort of) saga is the reason why the 'Project Blueprint' era V8 Supertaxis had an (not sure what you'd call it) inverted Watt's Linkage.
The bellcrank was attached to the chassis (via the adjustable height mechanism) and the arms attached to the rear axle.
By doing this, the rear roll centre, that is defined by the bellcrank's pivot point, maintained a consistant distance from the car's rear centre of gravity. Obviously as the fuel level changed (and these cars had their fuel tanks in the boot, above the rear suspension), changing the rear roll centre height often made improvements.
But the key thing to note here is that the roll centre's height relative to road is irellevant. It is the distance from the (vertical) centre of gravity that is the important factor.
An finally(!)............. Weight transfer occurs regardless of spring rates and antirol bar rates.
It is effected by track width, height and mass of the centres of gravity (front and rear) and the 2 roll centre heights.
You could literally have bricks instead of springs and the latteral weight transfer will still happen during cornering.
http://www.carrollsmith.com/books/#tune2win-detail
As relevant today as it was in 1978
Back in 1976 I bought "How To make Your Car Handle" from The Technical Book Shop in Swanston Street, Melbourne. Still have the book but Technical Books is sadly long gone! A good book for its day and most of the content still relevant - as long as your not tuning a FWD car!
Anyway, while under the 75 the other day bleeding brakes, I got to thinking about "Roll Stiffness" and how its applied to the TB front end of the 75. Given how close to the centre line of the wheels the "springs" are determines effective roll stiffness, where do we assume the effective "spring" location is on an Alfa TB equipped car? Methinks the effective location is somewhat inboard compared to a traditionally coli sprung front end. This location would tend to reduce the "efficiency" of increasing spring rate. Thoughts?
Front suspension verticle forces are put into the chassis via the lower control arms.
So efectively the 2 chassis rails are pushed upwards where the LCAs attach. This verticle movement is some what resisted by the diagonal box sections that go from down near the engine mounts to A pillars at the base of the windscreen.
It does look like the chassis is subjected to a lot of torsional loading being a torsion bar front suspension and plenty of people have said that the chassis isn't particularly torsionally rigid.
Fatter front antiroll bars will increase the twisting forces being applied to the chassis.
I was thinking some time ago that a horizontal brace that looked basically like ladder laying on its side, attached to the chassis, in front engine, but as close as possible, could help.
Basically the front of the chassis does very little to resist torsional loading. By adding a brace that could help there, the front could become more effective.
We found the increased roll stiffness and increased spring rate due to thicker torsion and anti-roll bars exacerbated the tired damper issue. Kissing kerbs to straighten a corner unsettled the car a tad more with the stiffer front anti-roll bar. The poor old Koni's were now not providing enough rebound damping.
A closer look at the dampers revealed "adjustable" on the tube - so out they came! Seems they were set around 50% stiff - similar setting to the yellows on the 159. I have now reset the adjustment one half turn from full stiff on the rear and one third turn from full stiff on the front. A HUGE improvement.
Not tested on the track yet, however on the road the car is SO much more settled over bumps. A LOT less crashing about and general pitching and yawing. A quiet bit of "test track" close to home suggests car now a lot happier riding kerbs mid-corner.
Cant wait to try this at Broadford shortly.
Debrief with driver post Broadford confirms the damper adjustment is close to ideal. Significantly improved control and spring ringing / oscillations reduced. Recent brake upgrade allowed driver to attack corners a bit more aggressively. Respectable lap times given self imposed rev limit of 4750. A tendency for subtle oversteer on corner exit. Maybe I will need to consider altering the roll axis!?! Thoughts?
Driver suggested a very minor increase in front rebound. Re-set front dampers 1/8 turn stiffer. Ride on the road appears firmer but still very ok for general use. A punt up and down my local test track confirms front end stability seems better.
Are you using the same offset wheels front and rear?
If you are, increasing the rear track width using suitable spacers, to match the front could be an easier initial option than changing the rear roll centre height.
Making the rear bar adjustable so you could slightly soften it could help aswell.
What are your rear spring rates?
Around 175/185lb/in should give a rear natural frequency about 10% higher than the 28.7mm TB in a full weight 3 litre (heavy bumper bars) and that's right about where you want to be.
What bumper bars are you using?
If you're still using the impact type, get rid of them and use the lighter ones (especially at the front). That'll bring the rear pollar moment a bit further forward and may help a touch.
Have you sorted out the bump steer yet? Inconsistent steering behaviour could be misleading the driver.
Hi Duk,
Yes - same offset wheels front and rear. To match the front track I'd need to fit 14mm spacers to each rear corner. A not insignificant amount. The Simmons wheels are pretty fat in the hub so longer studs would need to be fitted. I've seen studs shear and that worries me! Still a very minor reduction in lateral weight transfer due to the wider track may be worth testing. Anyone know where longer (and tougher) studs can be sourced?
OKP do a nice looking rear bar with 4 settings. That could also be worth testing. This particular bar features rose jointed drop links. Ok for a race car - high maintenance on a road car.
The rear springs are Eibach EW1004002HA. Couldn't find a published spring rate but a quick measure and calculation puts them at around 178lb. These springs reportedly lower the rear by about 30mm. Given the 16" wheels and lowered ride height I figure the rear COG is near stock but the roll centre is say an inch higher. Reduced rear roll couple. A subtle reduction in rear roll stiffness is worth a shot.
I'd love the 75 to be 100kg lighter and going to the lighter bumpers is a step in the right direction, however this is primarily a road car so we'll have to compromise there.
Haven't looked at bump steer at all. Given the relatively high front roll stiffness, tie rod vertical deflection is relatively low. Longer tie rods and shimming the steering rack is a low priority for the moment.
We'll see what the slightly stiffer front damper setting does for us on the track. I'll also investigate longer wheel studs at the rear. If nothing else, the wider rear track would look cosmetically better!
The OKP adjustable rear bar is tempting.................................!
Perhaps getting different offset outer rims for the Simmons could be done?
Maybe less work and cost than longer studs, because as far as I can tell, you can't remove the studs from the hubs without removing the hubs from their bearing housings.............. And I believe that would equal new bearings.
And don't be put off by rodends in road cars, just do what should be done and get rubber boots for them. Car manufacturers have been using tierod end style antiroll bar links for well over 20 years. Low and behold they have rubber boots on them. ;)
A closer look at the 75 workshop manual suggests only 10mm difference between front and rear track - 5mm per side. I'd prefer to have all wheels identical given I usually rotate tyres.
Appears Alfaholics do titanium studs for the 75 (and lots of other Alfas!) so a selection of relatively thin spacers may be an option. Removing the hub to get the studs out does look like a fair old job!!
The OKP adjustable rear bar does look attractive - good advice re the boots, thanks!
First thing though is to determine a way of defining what is actually the root cause of the very mild corner exit oversteer. Although inherent in most sedans, clearly and reliably defining its cause can be a challenge. Too little weight transfer or to much? Excessive camber change causing insufficient rubber on the road - very likely! A slightly wider rear track may reduce the positive camber just enough. Decisions decisions!
I do miss karts!
The live axle does an excellent job of keeping the rear wheels perpendicular to the road surface despite body roll.
Requirements of mild camber and toe angle changes are to deal with tyre distortion than trying to compensate the wheel getting pushed into positive camber.
Perhaps a slight increase in tyre pressure to help reduce tyre distortion?
What are you using for the main dedion pivot, rubber, polyurethane or spherical bearing?
Same for the Watt's linkage?
Spherical bearings everywhere will provide the best possible location of the rear suspension and also allows it to move thru its travel more freely than bushes do.
Thought being that, potentially, the main bush could be distorting slighly, especially with a tight LSD. The rear wheels will want to go straight ahead, potentially distorting the main bush in a sideways manor. For all intents and purposes, the outside wheel could be developing a toe out and the inside wheel toeing in. If it was happening like this, it would only be small amounts, but that's all it can take to upset the balance when you're near the limits.
Excuse the Solid Works drawing............... :-[ But I hope it helps get it across what could be happening if you have a rubber main bush.
We found 44 cold < 47 hot the best tyre pressure compromise. Car definitely sensitive to pressures!
Not sure what the main DeDion bush is - haven't got that far under the car........yet!
The existing Watts linkage bushes are definitely all polyurethane - probably around 90 duro. Installing the spherical SZ like Watts crank bearing is definitely on the to do list!
Replacing the main DeDion bush looks like major work. If its still the stock rubber item I'll very seriously consider going poly. If its the original - out it goes! Centreline Alfa do an SZ like spherical DeDion bearing. The SZ certainly handles well. A plain spherical / uniball bearing could need regular lubricant if it doesn't feature some form of polymer interface though!
The "feeling" of the rear end on corner exit is certainly not unlike an over tight LSD.
The front DeDion bush was rubber but not any more! Installed a polyurethane replacement - the existing one was probably the original. Also fitted 15mm spacers to rear hubs, longer wheel studs and new bearings.
The resultant 30mm wider rear track seems to have improved corner exit traction a bit. Car definately more stable at speed. Full polyurethane rear axle location via the front DeDion bush and Watts Linkage seems very effective for a road car.
Needs a run at Broadford or Winton to properly evaluate.