Blade type ARB -- some thoughts and a design

[Guest]

Just some thoughts.

Just about all performance cars have ARBs front and rear. Noteable exceptions are a few F1 cars (although it's hard to take much from vehicles with virtually no suspension travel), monoshock suspension formula cars and hill climbers (Dallara, Gould, Megapin) which have an alternative method of roll control, Jedii (ridiculously hard springs, but it seems to work), and Rallycinq's car.  

ARBs should perform a variety of functions: in roll it acts like an extra spring, especially at the outer wheel (thus lighter springs can be used), it controls roll, it attempts to move weight from the inner wheel to the outer wheel and in combination, front and rear bars can dynamically move weight from one end of the car onto the other. More detail in the late, great, Allan Staniforth's books, but you get the idea: ARBs are a very potent suspension tuning tool. The "obvious" use of ARBs is to control which end of the car looses grip first (oversteer/understeer).

On road cars ARBs are simple steel bars (generally made out of a similar material to road springs) mounted in flexible rubber (or, if you've been at them, polyurethane) bushes. On race cars, they'll run in at least nylon or ptfe bushes or proper bearings. The issue here is that any drag or slop anywhere on the bar will impede its function.  

The aftermarket rear ARBs available for (relatively) easy fitting to Centos are adjustable to suit the 'set' of the car -- that is, for pre-load. A few are adjustable for stiffness, albeit crudely. A decent race car will have fully adjustable ARBs, either by a sliding bar (look at a photo of the rear suspension of a Lotus 49) or by a blade. Blade type ARBs have the very distinct advantage of being relatively easy to make ----pit adjustable. The advantage is that you don't have to pit to make adjustments (which on saloons may well involve lying in the rain in the paddock at Cadwell, or wherever, or, on track days, lying in the mud at some godforsaken airfield somewhere) and, of course, the adjustments you need will vary according to conditions, tyre wear, and even corner to corner. ARB controls are just one of those buttons Rubens complains about there being too many of.

Imagine a round steel bar supported at either end with a weight hanging from the middle. It will deflect equally however many degrees you turn the bar through. Replace the round bar with a peice of steel strip and it will deflect most with the thinnest section horizontal, least with the thinnest section vertical. And, most important, it will show varying degrees of deflection as the angle is changed. You can (mind your fingers!) try this out yourself with an olde style dinner knife. This is what gives blade type ARBs their adjustability. For ----pit adjustment all that's required is a crank to turn the blade linked to a control in the ----pit (perhaps by a cable or a step motor).  

An idea for a blade type adjustable rear ARB for Centos: weld two bearing carriers onto the rear beam parallel to the boot floor -- these resemble short tubes and run from front to rear and carry a pair of taper roller bearings in each -- very similar to a motorcycle headstock (but shorter). Inside the bearings sits a shaft with a bearing carrier on the end of it: it could be a whopping great rose joint, but better a simple sealed ball bearing race. Suspended from these two bearings is a peice of spring steel bar cunningly machined in the centre portion (the supported section, if you like) to form the blade. The unsuported end portions of this are linked to the ends of the suspension arms by drop arms, rose jointed at both ends. Adjustment is facilitated by a link taken off an arm mounted to one end of the bar.



This design is a bit unusual in that it uses the centre portion of the ARB as the blade element (this is why the mounts pivot in 3 dimensions). More normally the blade will be at one (or both) of the outer ends.

[Note. If this design hasn't been patented or prior use demonstrated, I hereby copyleft it. Anyone is free to use it for private use. Anyone using it commercially must send me one of the versions they use (with details of what vehicle it fits) so I can sell it. Anyone producing it must acknowledge me as designer. Contact by Private Message on Fiat Forum and Cento Rebels to fingers99.]

Space is limited (especially if you've turreted in rear coil overs) so it's vital to check that nothing fouls. Pop out the springs and use a jack to check that everything is free on full suspension compression and extension and when one side is at full compression, the other at full extension. If it isn't, fix it before you drive the car: you may not be the person injured as a result.

An inevitable consequence will be an increase in NVH, which may be ameliorated to some extent by using polyurethane bushes on one end of the drop arm, but for circuit use I'd not recommend it.

More importantly, to work properly, the chassis needs to be extremely rigid and Centos are not exacly renowned for rigidity, especially torsional rigidity. Seis are better than Cinqs, but by no means good enough.

If you can fit a cage, this can be overcome, especially if the cage incorporates the susension pick up points (something which no existing cage does, but shouldn't be too difficult). May as well pick up the pedal bar, too. Actually welding sheet metal between cage and A and B posts as on most professionally built rally cars and racing saloons will help too.

If you don't want to fit a cage (and they're a real PITA on a practical, everyday, road car), the options available include seam welding (difficult to do on an old car, or even one that's had seam sealer applied, and at least a partial respray will be required), filling box sections with structural foam (potentially very messy, but an old, old, trick used in aircraft, boats, and such for years) and using composites to stiffen the whole thing up. One trick often used in the days when cars really rusted was to "membrane" the sills. Lancia Fulvias, like Centos, have an extra horizontal section (a "membrane") which extends along the sill as far as the A post. Extend this further back as far as you can and you gain a lot of rigidity. If you can arrange to swage Vs in it along its length you gain more rigidity for no extra weight.



Membraning is possible on just about any box section, but if there are no convenient spot welds to drill out it can involve a lot of work. As an aside, a cheat used in Superbike racing on ancient Kawasakis was to pull off the swing arm, chop down the outer sections with a bandsaw and weld in a steel section. After carefully grinding and filling the welds the result (which was a cheat, albeit a necessary one) would generally pass under the scrutineers noses.

There's no real point in extending any stiffening past the suspension pick-up points: all you'll do is strengthen the parts of the car which are intended be weak: to crumple.    

Consider buying some unidirectional carbon fibre cloth and using this to re-inforce the shell along the obvious chassis joins -- where the sills meet the floor, where the floor meets the "transmission tunnel", where the rear crossmember joins the floor, etc. The cloth needn't extend more than an inch in the short directions (but the surface must be taken back to bare metal) and should be built up in the "V" of the join. If you like a fancy finish you could pop a layer of biderectional cloth over the top, but it's not going to impart much extra strength. Consider also the possibility of filling in the roof channels (where the rubber strips lie) with unidirectional carbon fibre, and of casting Imola type side sills in unidirectional carbon fibre and bonding them to the body with epoxy. The direction of the fibres, of course, should be along the longest run. This stuff is absurdly strong -- I cast some round bar a little more than 1cm in diameter years ago using scrap uniderectional carbon cloth and polyester resin. I'd estimate that it has at least the strength of 8mm steel bar for less than a third of the weight.  



Obviously, do any welding before you foam or use any resin!      

The rear bar is part of the job. To augment it an adjustable front ARB is really called for. Alfaholics have some on their website which look apropriate, but, naturally, being designed for Alfas, you'd not expect the dimensions to be spot on for a Cento!





  

[fingers99]

Just some thoughts.

Just about all performance cars have ARBs front and rear. Noteable exceptions are a few F1 cars (although it's hard to take much from vehicles with virtually no suspension travel), monoshock suspension formula cars and hill climbers (Dallara, Gould, Megapin) which have an alternative method of roll control, Jedii (ridiculously hard springs, but it seems to work), and Rallycinq's car.  

ARBs should perform a variety of functions: in roll it acts like an extra spring, especially at the outer wheel (thus lighter springs can be used), it controls roll, it attempts to move weight from the inner wheel to the outer wheel and in combination, front and rear bars can dynamically move weight from one end of the car onto the other. More detail in the late, great, Allan Staniforth's books, but you get the idea: ARBs are a very potent suspension tuning tool. The "obvious" use of ARBs is to control which end of the car looses grip first (oversteer/understeer).

On road cars ARBs are simple steel bars (generally made out of a similar material to road springs) mounted in flexible rubber (or, if you've been at them, polyurethane) bushes. On race cars, they'll run in at least nylon or ptfe bushes or proper bearings. The issue here is that any drag or slop anywhere on the bar will impede its function.  

The aftermarket rear ARBs available for (relatively) easy fitting to Centos are adjustable to suit the 'set' of the car -- that is, for pre-load. A few are adjustable for stiffness, albeit crudely. A decent race car will have fully adjustable ARBs, either by a sliding bar (look at a photo of the rear suspension of a Lotus 49) or by a blade. Blade type ARBs have the very distinct advantage of being relatively easy to make ----pit adjustable. The advantage is that you don't have to pit to make adjustments (which on saloons may well involve lying in the rain in the paddock at Cadwell, or wherever, or, on track days, lying in the mud at some godforsaken airfield somewhere) and, of course, the adjustments you need will vary according to conditions, tyre wear, and even corner to corner. ARB controls are just one of those buttons Rubens complains about there being too many of.

Imagine a round steel bar supported at either end with a weight hanging from the middle. It will deflect equally however many degrees you turn the bar through. Replace the round bar with a peice of steel strip and it will deflect most with the thinnest section horizontal, least with the thinnest section vertical. And, most important, it will show varying degrees of deflection as the angle is changed. You can (mind your fingers!) try this out yourself with an olde style dinner knife. This is what gives blade type ARBs their adjustability. For ----pit adjustment all that's required is a crank to turn the blade linked to a control in the ----pit (perhaps by a cable or a step motor).  

An idea for a blade type adjustable rear ARB for Centos: weld two bearing carriers onto the rear beam parallel to the boot floor -- these resemble short tubes and run from front to rear and carry a pair of taper roller bearings in each -- very similar to a motorcycle headstock (but shorter). Inside the bearings sits a shaft with a bearing carrier on the end of it: it could be a whopping great rose joint, but better a simple sealed ball bearing race. Suspended from these two bearings is a peice of spring steel bar cunningly machined in the centre portion (the supported section, if you like) to form the blade. The unsuported end portions of this are linked to the ends of the suspension arms by drop arms, rose jointed at both ends. Adjustment is facilitated by a link taken off an arm mounted to one end of the bar.



This design is a bit unusual in that it uses the centre portion of the ARB as the blade element (this is why the mounts pivot in 3 dimensions). More normally the blade will be at one (or both) of the outer ends.

[Note. If this design hasn't been patented or prior use demonstrated, I hereby copyleft it. Anyone is free to use it for private use. Anyone using it commercially must send me one of the versions they use (with details of what vehicle it fits) so I can sell it. Anyone producing it must acknowledge me as designer. Contact by Private Message on Fiat Forum and Cento Rebels to fingers99.]

Space is limited (especially if you've turreted in rear coil overs) so it's vital to check that nothing fouls. Pop out the springs and use a jack to check that everything is free on full suspension compression and extension and when one side is at full compression, the other at full extension. If it isn't, fix it before you drive the car: you may not be the person injured as a result.

An inevitable consequence will be an increase in NVH, which may be ameliorated to some extent by using polyurethane bushes on one end of the drop arm, but for circuit use I'd not recommend it.

More importantly, to work properly, the chassis needs to be extremely rigid and Centos are not exacly renowned for rigidity, especially torsional rigidity. Seis are better than Cinqs, but by no means good enough.

If you can fit a cage, this can be overcome, especially if the cage incorporates the susension pick up points (something which no existing cage does, but shouldn't be too difficult). May as well pick up the pedal bar, too. Actually welding sheet metal between cage and A and B posts as on most professionally built rally cars and racing saloons will help too.

If you don't want to fit a cage (and they're a real PITA on a practical, everyday, road car), the options available include seam welding (difficult to do on an old car, or even one that's had seam sealer applied, and at least a partial respray will be required), filling box sections with structural foam (potentially very messy, but an old, old, trick used in aircraft, boats, and such for years) and using composites to stiffen the whole thing up. One trick often used in the days when cars really rusted was to "membrane" the sills. Lancia Fulvias, like Centos, have an extra horizontal section (a "membrane") which extends along the sill as far as the A post. Extend this further back as far as you can and you gain a lot of rigidity. If you can arrange to swage Vs in it along its length you gain more rigidity for no extra weight.



Membraning is possible on just about any box section, but if there are no convenient spot welds to drill out it can involve a lot of work. As an aside, a cheat used in Superbike racing on ancient Kawasakis was to pull off the swing arm, chop down the outer sections with a bandsaw and weld in a steel section. After carefully grinding and filling the welds the result (which was a cheat, albeit a necessary one) would generally pass under the scrutineers noses.

There's no real point in extending any stiffening past the suspension pick-up points: all you'll do is strengthen the parts of the car which are intended be weak: to crumple.    

Consider buying some unidirectional carbon fibre cloth and using this to re-inforce the shell along the obvious chassis joins -- where the sills meet the floor, where the floor meets the "transmission tunnel", where the rear crossmember joins the floor, etc. The cloth needn't extend more than an inch in the short directions (but the surface must be taken back to bare metal) and should be built up in the "V" of the join. If you like a fancy finish you could pop a layer of biderectional cloth over the top, but it's not going to impart much extra strength. Consider also the possibility of filling in the roof channels (where the rubber strips lie) with unidirectional carbon fibre, and of casting Imola type side sills in unidirectional carbon fibre and bonding them to the body with epoxy. The direction of the fibres, of course, should be along the longest run. This stuff is absurdly strong -- I cast some round bar a little more than 1cm in diameter years ago using scrap uniderectional carbon cloth and polyester resin. I'd estimate that it has at least the strength of 8mm steel bar for less than a third of the weight.  



Obviously, do any welding before you foam or use any resin!      

The rear bar is part of the job. To augment it an adjustable front ARB is really called for. Alfaholics have some on their website which look apropriate, but, naturally, being designed for Alfas, you'd not expect the dimensions to be spot on for a Cento!





  

[purf]

I'd imagine a rear arb would only be worth it if you had front coilovers and a pretty har core cento? Strut braces etc...

Or am i wrong - should it be the first thing to imrpove?

I have seen J33evos (i think) rear arb on the silver 1242 16v cinq which looked good.

Could standard alfa rear arb parts be used to make up something a bit simler than a blade design for a cinq?

[fingers99]

I'd imagine a rear arb would only be worth it if you had front coilovers and a pretty har core cento? Strut braces etc...

Or am i wrong - should it be the first thing to imrpove?

I have seen J33evos (i think) rear arb on the silver 1242 16v cinq which looked good.

Could standard alfa rear arb parts be used to make up something a bit simler than a blade design for a cinq?

Probably, but you'd loose easy adjustability. Aaron used the ARB off an MG, IIRC. The ARB is actually much more powerful in terms of handling than the springs/shocks. Unless you were going for best quality suspension, I'd be inclined to go for something like the Koni shock/spring kit, but the Sei one, rather than the Cinq one, which is really too stiff.