I've completed the work to replace the bearings in the front suspension arms. It was a bit of a saga.......
 |
| 'New old stock' suspension am bearings |
I'd removed the front suspension units back in mid 2021. You can read about that HERE.
With the weight of the wheel hubs removed I could move the arms independently to see how they felt. They felt a bit 'notchy' but not too bad. Others have reported very stiff arms - which cannot be good for the 'magic carpet' ride of a DS. The problem is that the arms do not rotate through much of an arc, so even though the surfaces are hardened, the bearing rollers tend to wear just a small area of the arm - eventually creating grooves which prevent the bearing rollers moving freely and causing stiffness. I think this is through 'brinelling' - which seems to be where excessive pressure in a small area causes a permanent indentation in a hardened surface. Lack of grease, or the ingress of water - or both can add to the problem....
 |
| Dried out grease (photo by Deesse Bombe) |
 |
| Photo by Deesse Bombe |
That's the science over and done with. What you need to know is that something that starts off looking like this......
 |
| Polished bearing surface (photo by Deesse Bombe) |
....ends up looking like this.
 |
| Worn bearing surfaces (photo by Bleudanube) |
To be honest, if the surfaces are that worn, new bearings and fresh grease are not going to make a lot of difference. Citrotech in Holland have a solution: they cut back the worn surface and fit modern bearings. More on that later.
Anyway, I planned to strip and examine my arms - and hopefully salvage them. I did the first part - dismantling - way back in summer 2021. The procedure for stripping and rebuilding arms (or 'front half axles' as the book calls them) is covered in Operation DX.410-3 in Citroen manual 518. That dates from 1966, but this operation is still relevant for later vehicles.
 |
| Workshop manual |
By chance, as I was reading up the procedure back then, someone called 'Deesse Bombe' posted on Facebook with their own efforts. And from the many photos included, they were VERY good efforts..... Deesse Bombe knew exactly what he was doing, how and - more importantly - why.
 |
| Fate lends a hand.... |
Having studied the manuals and all his photo, I made contact with Deesse Bombe through Facebook. He was enormously helpful in explaining which bearings are available and where to source them. And later on in this post I'll pass those details on to you.
Guided by Deesse Bombe, and with a new spring in my step and a confident swagger, I set about dismantling my units. With the weight of the arms, the whole thing is heavy and awkward, but the unit bodies are soft alloy and wouldn't cope well in a vice. So I rigged up a bracket to bolt them too - just like tool MR.3053-120 in manual 518 - and then clamped that bracket in my vice.
 |
| Suspension unit and bracket in my vice |
The nut for the bottom arm is hidden behind a protective metal cover (a 'cup'). This is part 9 in the factory photo above and is pressed into the recess. It has a threaded end to 'grab' it by. I rigged up a puller. Tightening the nut raised the bolt - which pulled the cup out. If the cup breaks or is jammed and needs to be ripped out, replacements are available.
 |
| Removing the 'cup' over the bottom nut |
 |
| The nut is folded over to replace loosening |
Clearly, to get them off, you need to bend the folded edge of the nut back. You need to be particularly careful that you do not damage the threads on the arm when you do this, and when you remove the nuts. As a final point, you need to try and make sure you salvage the nut from the top arm. At the time of writing, stocks of the bottom nut can be found, but I wasn't able to find replacement top nuts. It's possible Citrotech are re-making them but I've not seen them for sale. So look after them!
 |
| New and old bottom arm nuts. Top nuts are not available |
You will also need BIG sockets for this work: 46mm for the top arm nut and I think 38mm for the bottom one (I actually used a 1 1/2" socket that I had). With the nut's removed.......nothing happens! The arms need some 'encouragement' to leave the housing. The top arm seemed to need greater encouragement as the arm also passes through the pivot that is at the base of the suspension cylinder.
I used a stout piece of timber and a hammer. Be aware that the top arm has splines which fit onto the suspension lever - the bit with the rubber bump stops on it. When you knock the top arm out, this will release the lever part AND ia thrust washer. Carefully note the order and orientation of these.
 |
| Lever, thrust washer and splined top arm |
With the arms out of the way, you then need to pull out the rubber seals on the shaft ends to get at the bearings. If you are careful these could be re-used, but you can also buy replacements. Note in the photo that the lip of the seals at the arm ends sticks out a little.
 |
| Rubber seal over the end of the bearing |
 |
| Bearing cone pressed into the unit. See the light patterns from wear |
 |
| You can just see the notch behind the cone |
The technique is to first apply a little heat to the alloy unit around where the cone is sitting. I used a hot air paint stripping gun as a heat source. This causes a little expansion and means the cone is more easily removed with a drift. I worked around the diameter of each cone, tapping as I went, and took my time so that the cone came out straight and did not snag. I added a bit more heat from time to time. When the cones fall out, you'll be tempted to catch them - careful they'll be hot!
 |
| Suspension taper roller bearing and cone |
I examined the cone face and the bearing contact face on the arm. I could feel no grooves when I ran my fingertips around, but they clearly caught the light differently - which was the telltale sign of where they had been in contact and under pressure. My grease was still wet and in place and there were no signs of rusting or significant failure. I was relieved to conclude that these were salvagable.
 |
| One side done..... |
 |
| ......one side still to go |
Sourcing Replacement Bearings
Here is a diagram showing the bearings inside an arm. In this case a lower arm. The bearing cones (that are pressed into the unit) are shown in red. The bearings are nestling inside these. Note how the bearing at the arm end contacts a tapered face on the arm, whereas the bearing at the nut arm has a lining which contacts a straight (two parallel sides) section of the arm. Part 10 is a protective rubber seal at the arm end. Part 3 is a protective metal cup over the nut end. Part 12 is a metal spacer (a shim) that sits behind the front cone. More on that later......
 |
| Cutaway of a lower suspension arm |
The top arm arrangement is pretty similar. In fact the front (arm end) bearing and rubber seal is the same as on the bottom arm. The rear (nut end) bearing is different to the nut end bearing on the lower arm and has a rubber protective seal rather than a metal cup. The main difference on the top arm, is that it also carries the lever for the suspension cylinder.
 |
| The lever fits on splines on the nut end of the upper arm |
The bearing sizes are given in the parts manual though the part numbers themselves changed over the years. Deesse Bombe kindly explained what bearings were needed for where, and what replacements were available. And weren't. I've tabulated them here. Click on the image to enlarge it.
 |
| Front arm bearings |
The 'arm side' bearings and cones (ZC 9620 051 and ZC 9620 052) are not available as modern equivalents. At the moment there seems to be a fair number of 'new old stock' Citroen originals available - though prices vary widely!
 |
| Original - but expensive |
 |
| But bargains can be found! |
MUCH harder to find are the 'nut side' upper arm bearings - ZC 9620 115U. If you can find these, they may cost you more than £150 each. For those reasons, it's perhaps best to just re-pack and re-use the ones in your arms if you can.
The bearing and cone of modern bearing 32206 are suitable replacements for the nut end bearing of the lower arm. The bearing thickness (not it's diameters) is very slightly less than the Citroen original (0.25mm) but - being the nut end bearing - this makes no significant difference to the geometry of the arms.
If you are buying new bearings I was offered the following advice:
- Buy good quality SKF J2/Q bearings- Avoid non-European manufacture (though SKF have quality control - right?)- Avoid cheap Ebay bargains - there are counterfeit SKF bearings out there....
 |
| 32206 bearings |
Painting The Parts
I next picked up this job in June 2022. I de-greased and then cleaned all the iron parts on a wire wheel and with an electric drill. The main alloy bodies are an awkward shape with many nooks and crannies and I had them vapour blasted.
Again, following a tip from Deesse Bombe - and although I couldn't feel any roughness with my fingertip - I polished the bearing contact surfaces of my arms. Judging by the way the light reflected off them, It did make a difference with the surface being more consistent.
 |
| Polishing the arms - a 'before' and an 'after' |
I was painting the steering relays and anti-roll bar at the same time. Everything was degreased with panel wipe......
........and then got a coat of epoxy primer.
 |
| It's what gardens were made for |
By now I'd sourced all the bearings, seals and nuts I planned to use. As per my tale above, the bearings were a mix of genuine Citroen bearings and new SKF bearings.
 |
| All ready |
The new bearings have a protective light oil coating but no more than that and, before the bearings can be used, they need to be packed with grease. I had thought of heating a large pot of grease up beyond melting point then dropping the bearings in until it cooled and 'set'. But that is very wasteful of grease. Especially as it's advisable to be using a good quality grease.
Bearing Grease
SKF 'VKG 1/1' was recommended to me. It's an extreme pressure grease suitable for high load, low speed applications.
 |
| SKF 'VKG1' Grease |
However I'm not convinced SKF still market this but it is still available here and there. Possibly old stock? I eyed up 'LGEP 2' - which is also an extreme pressure grease but it seems aimed at VERY big machinery and I was worried it would be very thick and heavy - which is certainly not like the grease I removed from the bearings. In the end, and on advice from my bearing supplier, I used SKF 'LGMT 2' grease, which is said to be suitable for automotive applications. I think that's what Deesse Bombe used too. If I was doing this again (and who says I won't be!) I might go for LGEP2 as 'extreme pressure' should be a major consideration......Remember 'brinelling' at the start of this post?
Packing bearings involves having a large blob of grease in your palm and then, with the other hand, pressing the bearing down and across the grease blob. The idea is to force the grease up between the rollers of the bearing.Rotate the bearing so that you pack all the way around. It's quite messy. The grease in your hand melts (especially in a heatwave!) and I couldn't take any photos of the technique, but there is a short and helpful video on Youtube HERE.
Refitting The Bearing Cones
The trick was going to be replacing the cones in the arm unit. From his photos on Facebook, Deesse Bombe used a puller set to refit the cones. It can also help with the removal of one or two, but not all.
 |
| Bearing puller set (photo by Deesse Bombe) |
I'll confess now and say I followed his lead so closely, I even got EXACTLY the same set....it was a June 2021 birthday gift, so even though I didn't use half of the pieces in the kit, I didn't feel so guilty - and it certainly made the job a lot simpler.
The set offers a range of discs of different diameter which can be used to pull the cones gently into place. Obviously you want to make sure your disc diameter is wide enough to contact the edges of the cone - but not so wide that the disc gets jammed in the arm unit..... The shape of the alloy unit means that you will need to vary the puller set up according to which cone you are replacing.
 |
| Fitting a top arm arm-side cone |
The cone of the 'nut side' bearing of the top arm is recessed inside the alloy unit. Fitting the cone means choosing a plate that fits inside the unit.
 |
| Choosing the right tool for the job |
 |
| Just the right size to go inside the alloy unit. |
 |
| Winding in a top arm nut-side cone |
As with removal of the cones, applying some heat around the circumference of the alloy unit helps to expand things and making refitting easier.
Refitting The Top Arm
The idea is that you refit and tighten the TOP arm first, then move on to the bottom arm - checking the castor and adjusting the shims as need be. More on that later.... I loaded the bearing and seal with grease before fitting.
 |
| Arm side bearing and seal ready to fit |
There are special tools for fitting the seals to the correct depth(!).
 |
| Tools for setting the depth of the seals |
I don't ever expect to see those tools in my lifetime - let alone own a pair. Measurements/ gaps are given in manual 518, but I think the key thing to note when rebuilding the top arm is that the body of the seal on the arm side is recessed into the alloy unit slightly (see 'r' above), whereas for the same seal on the lower arm, the body is fitted flush to the unit. What you don't want to do is squash the seal in and up against the roller bearing. There should be a small gap between the two to allow grease to gather there and feed the end of the bearing if need be
 |
| Leave a gap between the seal and the bearing |
Extra grease (the manual says about 50 grammes) was added into the recess and all around the bearings and seals.
 |
| I added extra grease inside the arm unit |
 |
| 6 = suspension lever, 7 = thrust washer, 8 - seal |
There is a right and a a wrong way to fit the spacer (aka 'thrust washer'). The manual says it should be fitted so that 'the flat face is against the roller bearing". I couldn't work out what that meant but, the spline section does not go all the way through the thrust washer. On my car, when I removed the parts, the splined side was against the lever - so that's how I chose to refit it. I cleaned up the splines with a wire brush and applied a small yellow mark to the first tooth of the splines on the lever and spacer, and to the equivalent groove on the end of the arm. The thrust washer is meant to sit inside the rim of the seal - so that it presses on the bearing beneath. It was still VERY fiddly to aline the parts up and took several trial-and-error-attempts. The sticky grease helped.
 |
| Upper arm thrust washer in place. You can just see the splines on the side facing the camera (Photo by Deesse Bombe) |
When the splines DO line up, firstly the thrust washer slid onto the arm, and then the arm end started to slide into the lever. Then it all stopped. At this point, the fatter machined part of the arm had met the bearing and the arm was a tight fit. A bit more shoving got the arm a little further through the bearing, spacer and lever - but not quite enough of the way through the lever to get the nut on the end to pull it through.
 |
| Threads emerging through the Lever (Photo by Deesse Bombe) |
There is a trick you can use here though. Although the nuts and socket sizes for the top and bottom arms are different, the thread sizes are the same. The bottom nut has an extended, thinner end. You can use a bottom arm nut, turned over, to begin to get some purchase on the threads of the top arm. (I'm assuming here you didn't bugger up the bottom arm nut removing it, or have a new bottom nut to hand). Once you've managed to turn the nut a few times, you can fit the proper top nut and continue from there. The torque settings for the nut are given in Operation DX.410-3 in Manual 518. When you are sure you are happy with the fit, torque and movement of the arm, don't forget to bash the edge of the nut over to stop loosening up.
 |
| A top arm refitted |
Refitting The Lower Arm And Checking And Setting The Castor Angle
Refitting the bottom arm is pretty similar to the top arm - but without the added complication of needing to fit the arm through the suspension lever. However it is important to pay attention to the bottom arms as the suspension geometry is partly dependent on the fitting of the lower arm bearings. The Citroen DS steering is set so that it has a positive 'castor' (or 'caster') angle'. In short, the bottom hub ball joint is slightly forward of the top hub ball joint. 'Positive castor' improves straight line stability at high speed and improves cornering. Both desirable attributes in a DS. Ask President de Gaulle.
 |
| Caster angle (image by 'Summit Racing') |
A DS is meant to meant to have a castor angle of 1.5 degrees. it's not a lot.
 |
| DS caster angle given in manual 518 (1966)..... |
 |
| .....and the same from manual 583 (1975) |
Note how manual 518 says this is 'adjustable'.... Adjustment is made by putting a shim of the right thickness behind the cone of the bearing on the arm side of the lower suspension arm. A thicker shim pushes the out arm out further and so creates a greater 'positive castor' angle.
 |
| A caster adjustment shim |
Now.....given that with their hydraulics, Citroen have a reputation for high quality, high precision engineering, I can't understand why different shims are options for the lower arm. Why couldn't they/ didn't they machine all their alloy blocks and the arms/ eyelets so that there was a 'standard fitting'? I will need to leave that question there and just accept that Citroen expect you to need a shim of at least 4.6mm but no greater than 7.0mm.
 |
| Shim parts from the Parts manual |
The 'quantity needed' is '?' presumably to make sure you don't assume you need one of each. These are an 'either/ or' thing.....I made a note of the thickness of the shim that was fitted to my suspension unit. Out of curiosity as much as anything.
 |
| I'm taking this to be a 6.0mm shim - part D.412-84G |
I think its reasonable to assume that when the cars are first assembled, the appropriate shim is fitted for each arm to give the right castor angle. So if you take it apart and reassemble it with genuine Citroen parts, it should all go back together and retain the castor angle. Shouldn't it? I will also need to leave that question there and just accept that, when re-assembling arms, Citroen garages were expected to check and adjust the shims to maintain the castor angle.
You will see in the procedure at Operation 410-3 in manual 518, that there are a pair of special tools (1865-T and 1866-T) that serves as proxies for bearings and cones - meaning that you don't need to keep pulling out and and replacing the cone if you find need a different shim behind it.
 |
| Tools for checking and setting the caster angle |
Once you've selected your shim, then there is then another special tool (tool 2321-T) that has two parts that fit into the ball joint eyelets of the two arms and so is used to check what castor angle it gives. The tool does this by measuring the offset of the ball pin eyelets. The correct value needs to be between 24.75mm and 25.25mm. That's just 0.5mm tolerance.
 |
| Using 2321-T to confirm the offset, and so the castor angle |
Most DS hobbyists aren't going to have the tools to set and check the castor angle and will just put everything back together. That's what I was expecting to do. I'd seen tool 2321-T (the one for checking the angle) for sale on Facebook back in April 2021 but passed it up. I made enquiries a week or so later and it had gone of course. However I knew an old friend and DS expert had tool 2321-T and I asked to borrow it. Not only did they send it to me, but told me I could keep it!
 |
| Tool 2321-T |
 |
| The two parts slot together |
Having reassembled and torqued the lower arm, I nervously fitted tool 2321-T. I brought the two arms together......
I took several measurements. Phew! The arms cam up in spec!
 |
| In spec! |
Buoyed-up by this, I went on and rebuilt the second left hand arm unit. Except this time when I took the measurement, it wasn't in spec. Not by a (relative) long shot....... In fact when I moved the arms together, the two parts of the tool didn't engage in the slot - they overlapped.....
 |
| Out of spec! |
I worked out that the arms were about 6.8mm out of alignment - which by an annoying quirk of coincidence.....
 |
| A different adjusting shim |
........was the same thickness as the adjusting shim on that side.
Analysing the situation, the two halves still aligned, so I ruled out one or other ball pin eye being twisted. I dismantled the arm and re-assembled it several times using alternative new bearings and even trying the old. The measurement still came out the same. I ruled out heating and bending the arm for fear it would weaken it. I ruled out getting another second hand arm as that could also be bent. I ruled out getting a p/x arm as that represented significant additional cost. 6.8mm is not a lot and I could have just lived with it. But knowing the first arm was in spec. made wee want to try a little harder with this second arm.
The shims on the two sides of the car were different so presumably were chosen to make the car 'in spec' when it was manufactured. I had to conclude that at some point in it's past, the car had taken a knock or hit a bump heavily. I'll never know. What I did know, was that to get my arms back into spec, either the top arm needed to come out further (adding a shim where there shouldn't be one), or the bottom arm needed to have the shim completely removed - which should not be the case.
 |
| Choices...which arm to reset? |
I did try completely removing the 6.8mm adjusting shim from the bottom arm - but the trouble then was that too much of the arm stuck out on the nut end and the threads were not cut long enough for the nut to tighten to the bearing. Perhaps that is why Citroen expect a shim of at least 4.6mm to be used? I considered adding 6.8mm of shim to the top arm - but then there might not be enough thread sticking out to screw the nut home. In the end I decided to do a bit of both. Using the Citroen shim as a template I had my local machine shop make me a selection of alternative shims.
 |
| A selection of shims |
Splitting the difference, I decided to reduce the shim on the bottom arm by about 3mm (which took it just below the minimum shim size but allowed the nut to be fully tightened), and add a shim of about 3.5 mm to the top arm (which still left plenty of thread for the nut).
I fitted the cones, bearings and seals for the umpteenth time and once again grappled with the splines on the suspension lever. With both nuts torqued up I took a few measurements......the average was 24.30mm - *just* below the bottom spec. threshold of 24.75mm. but considerably better than before.
 |
| Close enough! |
Final Thoughts
Tool 2321-T is part of the standard DS workshop tool board - which I take to include those tools that, (of all the many, many available) are used most often. So if checking was a frequent or routine activity, then maybe it was quite common to find that the castor angles of cars was out of spec? Maybe lots of Ds driving around have bent arms?
 |
| Workshop tool sets |
All said and done, I've ended up with arms that are in a better state than when I removed them but it was quite a journey, not without expense and with a few worrying times. Had I had the benefit of tool 2321-T available when I first removed my hubs, I'd have checked the arms for 'true' there and then. I assume I would have found the left hand unit wanting and, needing to obtain an alternative, would probably have gone straight down this replacement route. Which would have saved me a lot of effort, and saved you the time taken to read this post.
Footnote:
I said earlier in this post that there was an alternative to all this work. If you don't have the facilities or the confidence to do this - or simply can't be bothered and have deep pockets - then you can get replacement arm units on a 'part-exchange' basis, whereby you swap your old arm units for ones that have already been rebuilt (and hand over money of course). In this case, the units are rebuilt with modern, easily obtained bearings. This necessitates machine both the arm part and the alloy body part to take these new bearings. The usual part suppliers offer these replacements. They are not at all cheap and prices vary. That might mean that there are several sources for these rebuilt arms and that attention to detail/ quality varies. Citrotech is a respectable and trusted Dutch company not only selling but rebuilding parts to 'as new' condition.
 |
| These arms have new, modern bearings fitted |
They have a short video showing the process of re-cutting the suspension parts for new bearings. You can see that the tapered face of the arm is cut back. The video is available HERE.
My friend Peter went down the replacement route and - as seekers of DS knowledge - we both agreed it would be interesting to see how his arms measured up for spec. using tool 2321-T. And the answer is that not only were they 'notch-free' they also both came up on spec. I should have looked more carefully to see whether Citrotech are re-using old top arm nuts or - as I suspect - have had some more made up.
