A bracing experience

[johnl]

Recently fitted new front and rear dampers. I couldn't justify ($) Koni or Bilstein, so new stock items were sourced from Vlad at Arese.

The old rear dampers were totally shot so the new ones were a huge improvement. The fronts were not great but not nearly as far gone as the rears, so I was expecting less improvement. Having said that, I was quite disappointed, the new front dampers didn't seem much better than the old ones. The front end was too soft and while the dampers were reasonably OK over long amplitude undulations etc. they seemed unable to cope with rough surfaces, allowing the wheels to patter and judder over broken / irregular road (especially corrugations).

I was then at the point of tracking down best pricing I could find on some B4 Bilsteins (hang the expense, the car was not nice to drive). Then, on an impulse I made a simple front 'strut brace' (tower top to tower top) and fitted it. I wasn't expecting a lot, at best maybe a slight improvement hard to not dismiss as possible placebo...

The thing is that a 'strut brace' really shouldn't do much on this car. The front towers are directly attached to the bulkhead which should provide at least adequate lateral rigidity between them (and the rest of the chassis). So I was surprised when the pattering and juddering lessened quite substantially. The dampers now feel at least adequate for the job, as opposed to previously feeling irritatingly inadequate. I'm now quite happy to keep the stock spec dampers.

But wait there's more. The steering now feels a little more responsive and a tad more linear. The car generally feels a bit more 'robust' in a difficult to describe way. The handling is somewhat 'better', with improved stability. The slight rough road creaking from the base of the windscreen has gone.

I'm very happy (and surprised) with the affects of this brace, and very sure it isn't a placebo perception. Next step is a lower subframe brace.

Regards,
John.

[pasey25]

I actually just bought a brace with a set of new KW coilovers which are on their way from Europe. Glad to hear they worked well on your car.

[cc]

Guys
Could we have some pics and prices? please
cheers

[johnl]

Haven't had time to organise photos, or figure out how to post them...

Cost? About $10 worth of steel tube, which is about 2mm wall thickness and a about 25mm OD (didn't actually measure it, just bought some tube that looked 'about right'). The brace is very simple, with each end of the tube flattened with a hammer on an anvil (how neatly dependent on skill at hitting things). Each end was then drilled with one hole, which fits onto one of the four 'strut' studs on each side of the chassis. The studs I used are the ones furthest forward and innermost.

The flattened ends need to be angled to account for the tops of the 'strut towers' not being horizontal, but angled downward toward the middle of the chassis. The brace holes need to be elongated with a rats tail file because the studs are angled inward / upward. This hole elongation doesn't affect the location of the brace because the brace isn't prevented from moving by the strength of the studs (in shear), but by the clamping force of the studs / nuts.

The engine mount bracket (right side 'strut tower') needs modification because the raised edge of it gets in the way. This involved flattening the raised edge (and part of a raised ridge) on the anvil. The brace is still more than strong enough to do what it needs to do.

My $10 cost doesn't include the cost of paint.

Haven't yet had time to make a lower brace, but I think it will be a little more complex than the upper brace, or the commercially available lower braces on the market (seen in interweb). This is to minimise ground clearance issues, since the engine / gearbox force a lower brace substantially downward. Where clearance is tight I think my lower brace will use thick / wide (and somewhat heavy) flat strip welded to upwardly angled steel tubes each side of the engine / gearbox, then bolted to the forward bolts on the subframe. This should reduce the vertical depth of the bracing member below the engine / gearbox, so that it's about 8mm thick rather than about 20mm or so (as might be needed with a tube). Looking under the car, I don't think any (more or less) longitudinally oriented bracing tubes are needed (or likeley to do anything particularly useful), just a cross brace at the very front of the subframe.

Regards,
John.

[johnl]

Despite fitting the upper tower brace having improved the perceived performance of the new TRW 'standard' front dampers to the point I thought they were at least acceptably OK, more k's down the track and I've changed my mind. The damper performance does feel at least improved with the brace fitted (at least the front end feels more under control than without it), but it's still not good enough, just better than it was. Also, the damper performance is significantly better the colder the ambient temperature is, and worse the warmer it is, which must be due to fluid viscosity being affected by temperature?

It's odd that the rear TRW dampers are fine, no complaints at all.

I suppose a set of front Bilsteins is still on the wish list. Anyone want to offer experience with B4 vs B6, and maybe vs OK condition stock? My heart says B6, my bank account may insist on B4...

Something I found interesting, disconnecting the front ARB (as part of my quest to diagnose the slight steering pull to the left) caused the front dampers to work even less well. With the ARB disconnected the dampers struggled even more to cope on rough surfaces, the wheels were bouncing from bump to bump and generally the dampers just couldn't cope (notably better with the ARB reconnected). I wonder if this might be because the ARB (a stiffer GTA item) is connecting the left and right suspensions, so at least some one-wheel bump / rebound force might be transferred via the ARB to the damper on the other side, with the 'other' damper adding some resistance to oscillation on the side experiencing the bumps?? If so then in some instances the damping action is mostly provided by the damper on the side encountering the bumps, but also to some lesser degree by the damper on the opposite side...?

Or, maybe, the spring rate (frequency of oscillation) of the ARB is significantly and strongly enough 'out of phase' with the spring rate of the coil springs, so the two differing spring rates oscillate oppositely (to some degree) and thus acts to damp one-wheel oscillation...? Or some combination of all these things?

Regards,
John.

[johnl]

Hmm,
had a thought about the way the front dampers behave (poorly controlled over rough surfaces where the 'height' of the irregularities is relatively small, but better controlled in undulations etc). If you remove the little plastic caps on top of the 'struts', you can see the damper rod moving (quite a lot) up and down in its rubber mount when pushing down on and lifting the chassis. I don't think the mounts are significantly worn, no apparent cracks or undue clearances when I had them out for the damper change. The rubber part of the mount that attaches the damper rod to the rest of the mount (and then to the chassis) seem fairly soft, and I suspect this is just how they are from the factory.

The rubber in the damper mount is in effect a short stroke and fairly soft elastomer spring (connecting the damper rod to the chassis), and to make things worse it is an undamped spring. I'm wondering how much this might affect the ability of the damper to control suspension motion over short sharp road irregularities, i.e. fail to control the suspension oscillation adequately because the dampers are not properly damping the springs in the first cm or so (of suspension travel) each side of ride height (or whatever suspension height exists when the wheel rolls over the rough surface and excites the spring).

I'm thinking a much stiffer 'strut' mount might be beneficial (that part of the mount that connects the damper rod to the chassis). Can't find such a thing online...

Regards,
John.

[Colin Edwards]

Hi John,
The problem may compounded be damper stiction.  The "perfect" damper or whole suspension system for that matter should feature zero stiction.  As soon as any load is put on the suspension / damper it should deflect to some degree.  If the damper rod seal features high stiction some other suspension component may deflect in lieu of the damper rod / piston.  When this happens the suspension will be largely undamped and ring at its natural frequency. 
Maybe the elastomer bush at the top of the strut has gone soft over time.  This bush when deflected will heat up a bit so it has some damping properties and is not a "pure" spring.  This combined with high damper stiction may cause your wheel / tyre chatter.

Fair chance the significantly more expensive Bilstein or Koni product will feature less stiction.  I recall Bilstein reduce stiction by use of a lubricant whereas the Koni does it by very clever guide / bush / seal design.  Either way it comes at a $ cost!

Quadrant Suspensions used to have a damper dynamometer.  This would reveal things like seal stiction and fluid cavitation etc.  Maybe some suspension specialist out there has a dynamometer to test dampers.

Colin

[johnl]

Hi Colin,
The rubber in the mount will see 'sticktion' as a momentary additional resistance at the damping rod, it will 'think' the damper has some weird initial stiffness followed by a softening, and I agree that this would cause more initial flexure in the rubber in the rod mount. However, the rubber would also 'see' the added (intentional) stiffness of something such as a Bilstein B6 (or whatever stiffer than stock damper), so I suspect this would also cause a similar problem in some degree. If so, then fitting stiffer dampers may not eliminate this 'short stroke' lack of damping effect?

I don't think the rubber in the mount is in poor condition, I've seen bad ones and they develop cracks with age as they soften (not seemingly the case here). Many suspensions use rubber to attach the damper rod to the chassis (mostly Mac Struts, and I don't mean rubber 'pancake' or Silentbloc style bushes, though these must suffer in some small degree too, probably not enough to be noticeable).  I have noticed other cars with such a set up can have a minor undamped oscillation over smaller bumps / less than smooth surfaces that I've often thought might be due to this means of attaching the damper rod (I could be wrong, but I think there is a good reason why you wouldn't use this design on any sort of competition vehicle). With some cars the rubber rod mounting may be quite stiff, with others it may be quite soft (because a soft rod mounting absorbs vibration rather than it being transferred more directly into the chassis), which might be why the short stroke lack of damping caused by it seems noticable with some cars and not others...?

Thinking why it only seems to be the front suspension that is giving me trouble, could it be that the rods in the rear dampers are better located by the addition of the 'hats' that prevent the top of the rear struts from detaching from the chassis? These 'hats' may prevent excessive vertical motion in the rod as it loads and unloads, at least in rebound...?

Yes the bush will heat up in use, and this will tend to soften it. I said earlier that the rubber in the mount was effectively an undamped spring, which is convenient for descriptive purposes (near enough), but actually it isn't true. The rubber has an 'internal' damping property as the molecules move past each other as the rubber deflects. This creates an internal friction in the rubber (or any elastometric spring, and actually to a really tiny degree in any metal spring), heating the rubber in a not entirely disimilar manner as the oil in a 'shock absorber' (damper) gets hotter in use. However, I do think that the effect is minute in the scheme of things, otherwise the rubber would get so hot that it wouldn't last long.

BMC Minis (et al) suffered problems due to the rubber cone springs (replacing the more 'normal' steel springs) being significantly softer in warm ambient conditions than cold, and since the rubber was moving quite a lot (carrying all the load of the cars weight as the suspension moved) also tended to soften in use as they warmed up (then harden up again when they cooled). You can often feel a similar effect with rubber bushes in many suspensions, the rubber bushes are harder when cold and soften when warm, at least one reason why many cars in my experience handle better on cold days than hot days.

Another may be that the damper fluid is just a little bit thicker when cold than hot, though better dampers probably use better fluid that doesn't suffer so much. Thinner damper fluids tend to be better than thicker, because thinner fluids are already thin and don't get much thinner as the temperature increases, but thicker fluids do thin out significantly.

Regards,
John.

[Colin Edwards]

G'day John,

The missus had a 2007 147 Ti 2 door selespeed from new.  Great car!  Given the low profile tyres fitted as standard, reduced suspension travel and great handling / grip, I don't recall ever noticing poor front end suspension control.  Is it possible the Ti spec front dampers fitted to "Series II" 147's feature less static friction / stick-slip?  Are the "later" dampers a different beast to the "earlier" model dampers?  Would Ti spec dampers be better value for money than the Bilsteins or Konis?

Colin

[johnl]

Colin,
I don't think the "TI spec" dampers are any different to the dampers fitted to non TI cars, but I could easily be wrong (listings I've seen seem to specify the same dampers for all 147s, other than GTA). At any rate the dampers fitted are not the Boge or Sachs that I understand are the OE manufacturer, but TRW supposedly made to the same spec (i.e. damping rates). The TRW brand dampers work quite acceptably on the rear end of my car, just not on the front.

It might be Alzheimers, but yesterday I couldn't think of the name commonly used for strut mounts that eliminate stroke compliance in the damper rod attachment. This morning I remembered that they are known as "pillow ball" mounts. These provide a metal to metal attachment that doesn't allow the rod to move (relative to the chassis, or rather, any rod movement must be into / out of the damper body and not at the rod mount) with little or zero damping effect on the suspension motion. With a rubbery rod mounting, the damper must be unable to fully control both sprung and unsprung mass motion until the compliance in the rubber has been 'taken up' (this also has implications for the 'speed' at which weight transfer occurs, and consequently on steering and handling responsiveness, i.e. make the steering and handling less 'sharp'). This must be the case because the damper rod is in effect attached to the chassis with a 'spring', and consequently must mean that the suspension is effectively undamped (or at best poorly damped) for a short stroke length each side of the static ride height (not actually ride height since all motion doesn't start at static ride height height, but I think you know what I mean).

Imagine the rubber has fully deflected in bump (probably about 1cm of rod travel, but maybe more), load is then removed and the rubber is 'free' to drop back down by that 1cm, plus about another 1cm into rebound (and of course the wheel is also 'free' to rise and fall in a correspondingly undamped manner). The same issue exists in reverse, i.e. if the wheel has drooped in rebound then there will be about 2cm of undamped rod travel before the bush compliance is 'taken up' in bump motion. If this back of the napkin thought experiment is correct, then there is about 2cm of very poorly damped movement at the rod, which because of motion ratios will be significantly greater at the wheel...

'Pillow balls' are most commonly available and used for Mac strut front ends, but due to the design of the rod attachment the same issue must exist for the suspension on our 147s (and I assume 156 et al). I had a similar (but lesser) issue on my old Accord (double wishbone front end, fairly similar to the 147). On this car the damper rod was attached with a rubber 'pancake' or 'puck' style mount (two rubber 'doughnut pucks' with a crush tube and retaining disc), where each 'puck' was quite soft (though compliance was way less obvious than I'm seeing in the Alfa rod attachment). To stiffen this up I machined some washers that compressed the rubber pucks substantially before the retaining nuts tightened onto the crush tube, resulting in visible compliance (as the car is moved up and down at rest) being reduced to near zero.

The problem I'm having may be largely exacerbated by the poor quality of roads I have to drive on. On smooth roads it's not a big deal, the dampers are reasonably well behaved. Having said that, the little Alfa is suffering more than any other car I've driven on the same bad roads (other than those with worn out dampers...). If the above is more or less correct, then fitting stiffer dampers won't really fix this problem. The compliance in the rod mount would still exist, and while suspension behaviour may be improved, the full potential benefit of any superior quality dampers wouldn't be fully realised...

Regards,
John.

[johnl]

Looking around the interweb it seems that 'pillow ball' mounts are available for the 147 if one were to buy complete 'coilover' style suspension systems, the mounts included as part of the package. If on the other hand you just wanted to fit uprated dampers to existing or other aftermarket springs, then I can't find anything...

Regards,
John.

[Bobulon]

Going back to your first post, is your strut brace only attached to one bolt on each tower? If so aren't you risking twisting the tower under heavy stresses by making it pivot on that point? My brace attaches to three bolts per side for this reason I think

[johnl]

Bob,
Yes it's attached with one bolt per side. No, I don't think the forces involved will be anywhere near great enough to twist or otherwise deform the towers. Please note though that this is a data-less assumption based on experience with observing, designing and making other structures (I'd be amazed if I were wrong though...).

Keep in mind that 'strut' braces (I think they really should be called 'tower braces') only add stiffness to the chassis in compression and / or tension (despite what many people seem to think), and are not very resistant at all to any bending loads that may be applied to them. Try this, detach your brace on one side, then lift that end upward and somewhat away from the top of the tower. How hard is that to do? Pretty easy is my bet, with the end of the brace still attached flexing substantially with only a relatively small upward load applied at the detached end. This experiment demonstrates the lack of bend stiffness in the brace.

A brace attached at two points per side will (IMO) not be significantly more (or less) effective than a brace attached with only one bolt. But, for arguments sake, if a particular chassis did tend to twist it towers under load (I doubt this is likely, to any significant degree), the fact that the brace was attached at two points per side would tend to impart a bending load into the brace, which could flex into a curve.

Tower braces with any bends or curves are less rigid in compression and tension, so could be compromised to some degree (though bends / curves are often required for clearance reasons). Tower braces (or any bracing members really) that are curved or bent will be significantly less rigid in compression and tension, and so are inherently less effective than straight braces. To make up for this such braces need to be of much more substantial cross section and / or wall thickness (assuming a tubular design). Such a brace will of necessity be heavier.

There seems to be a lot of people 'out there' who disparage tower braces that use separate mounts with a tube bolted between them, assuming the bolted connections will be points of articulation that weaken the brace (which could only be the case if tower braces worked by adding stiffness in bend, which they don't, at least not to any significant degree). The assumption is that a fully welded construction will be stiffer, but I beg to differ. Due to the brace working only in compression and tension, it makes no difference whether the actual bracing member is welded or bolted to it's mounting points. The only issue with bolts is that they adequately tightened.

An often overlooked part of the design of tower braces is the attachment brackets, which need to be quite robust. It doesn't matter how rigid the bracing member may be if its attachments are flexy. Many braces I've seen have very flimsy brackets...

Regards,
John.