Removing Underseal From The Underside Of The Chassis

I finally got round to starting (that's 'starting', not 'completing') a job I've been putting off for ages: scraping the underseal off the floor of the car. 

Underseal removed from the underside of the car

As part of the prep. for replacing the outer sills, I had tided the inner sills/ box sections - which meant scraping some underseal from the sill/ floor join area. Rather than just underseal those areas again, I felt I needed to go all the way and scrape the whole floor area, de-rust and apply paint. before re-applying underseal in one uniform coat over the whole area. (So, another of those small jobs that grows and grows.....).

Replacing the outer sills

Overall, the underseal didn't look that bad, but there were tell-tale signs of it lifting - mostly on joins and corners. When it lifts, it traps water and that can lead to rusting unseen beneath the underseal. I already had signs of rust along the edges of the reinforcing strips along the length of the sills.....

 .......and around the seatbelt mounting points. Also, on closer inspection, there were other signs of bubbling on the floor pan  and so I wanted to investigate for peace of mind.

 The car was properly supported on four proper axle stands and what had been putting me off was the hours of back breaking working lying on the floor with the car only about 50cm off the ground. The underseal is bitumen based and so quite rubbery. You have to put quite a lot of effort to get beneath the surface to lift it. Originally I crawled under the car on a padded dolly, but found that when I applied pressure to the scraper - the dolly moved backwards! That didn't work!

Entering Limbo Land

I started on the floor pan beneath the rear passenger area and attacked it with a gasket scraper that I had sharpened. Some technique was needed but I found that with the scraper at just the right angle and my elbow wedged on the ground, the underseal came off in long, satisfying,  leather-y curls. I had to frequently re-sharpen the scraper. Occasionally it's edge dug into the metal where the strengthening ribs were pressed. Not enough to make a hole, but enough to stop me in my tracks and waste my energies. Cleaning around those pressed ribs was fiddly and time-consuming.

Working around the 'ribs'

I moved on to the area beneath the front passenger area. That proved to be areal slog..... Whereas the underseal had previously come off in curls. THIS STUFF DID NOT WANT TO BUDGE. It was far harder and even with considerable effort, only came off very reluctantly in tiny, dry flakes. The flakes splinted off in shards that got in my eyes, mouth and nose. I don't know if this difference is due to proximity to the engine and engine heat?

Underseal from the back and underseal from the front.
Spot the difference?

Another complication in this area is the domed recess where the front exhaust silencer sits. being domed, I couldn't get my scraper in at the right angle. Even at the time of writing I have not removed the underseal form the dome ends. I'm thinking I will need to use a wire brush on a grinder or drill.

Front floor and silencer recess

I also scraped the underseal from the area under the tank. On the plus side, the underseal was softer and came off the bigger areas in curls again. On the down side, there are pressed ventilation holes in this area so some fiddly detail work was needed.

Under the petrol tank. Actually very happy with what I found!

Although hard work, I think it was a wise move to remove the underseal. I found the floor covered in 'liver spots' of rust. Nothing more serious than surface rust but the start of what could have been a significant problem. Looking at some of the strips of removed underseal, what I found was tiny pin-prick holes on the outside. Almost too small to see and certainly not obvious when hidden away under a car.

Pin prick holes in the underseal........

Those tiny holes tell a different story on the other side of the underseal and you can see where water and rust are starting to get in - the 'liver spots' I found. Given that my car has been off the road for 20 years, all this occurred before 2000. If the car had been used in that time and I'd not bothered to tackle this job, that rust would be much, much worse.

......And the rust forming beneath

Because of the placement of the axle stands, I still have to do the area beneath the rear arm mounts and the boot floor area.

Still to do....

At the front, I still have to do the area beneath the front suspension mounts.

Front 'tusks'

For all my efforts, this is only the start of the 'first pass'. There is a still a sticky patchwork of grey covering the metal. It has the texture of linseed putty. I need to scrape as much of that off as I can with a flat scraper, as I think trying to remove it with an abrasive pad at this stage will soon see the pad clogged up.

Still work to do

I realise that there has been water ingress between the strengthening strips and the floor pan. I think it would be very destructive to remove and replace the strengthening strips based on what I've found - which is a solid floor. My aim is ti neutralise and convert what rust there is, and then seal the edges to prevent any more moisture getting in. 

So what have I learnt so far?

1. My garage floor is very hard

2. I don't want to do this job again

3. The original underseal was applied to bare, unpainted metal

4. That, by and large, it had held up very, very well

5. That it's hard to remove

6. Underseal is only as strong as it's 'weakest link'. It doesn't matter that 90% is good, if 10% fails and allows the car to rot.

 

Inside Every Small Job........

.......is a BIGGER job waiting to get out....... 

I'm getting closer to finishing some unplanned 'rectification' work. Not a big job, but something I could have done without.

Whilst re-undersealing the sills on my car, I decided to remove the long (2.17m) control rod that goes from the ride height lever in the cabin, to the height corrector in the rear. 

Tidying up the sills.

Removing the rod would give better access for cleaning, prep-ing and painting etc. More importantly, I could spray on gooey underseal and not worry about gumming-up the movement of the rod. I'm planning to rebuild the rear height corrector at some point, so would probably need to disconnect the rod anyway. It all sounded like a good idea.

I had already taken the cover off the front wheel arch to remove the piping to the engine bay and removing the control rod should have been a simple matter of unscrewing the rear end joint, then sliding the rod out of the sill through the front wheel arch.....

Rod at the bottom, slides through hole at the top...

First things first. The rear end of the rod is threaded (to make it adjustable for length) and fits into an articulated knuckle that operates a secondary height corrector rod. Although the area around the height corrector in the rear wheel arch wasn't rusty (due to some prudent use of Waxoyl back in the early 90s), the tip of the control rod was brown and slightly crusty. Okay - it was rusty. 

Rear height corrector. With disconnected rod in bottom corner

With the front linkage already disconnected, I removed the split pin to separate the second control rod at the rear height corrector and expected the threaded end piece on the long rod to then simply unscrew........It wouldn't budge. I tried penetrating oil - no luck. Unfortunately that bulbous end piece prevents the rod being withdrawn - so having committed to removing the rod (well - i'd taken the pin out already, hadn't I), it had to come off. 

There was no room to apply heat to the end piece in situ, for fear of melting or damaging something else. So I went down the caveman route and applied mole grips - to both the rod and the end piece. Sure enough, I felt some rotation! I gave it a slight wiggle and tried again. More rotation! Shortly before the end effortlessly snapped off.........

I was in shock. I sat for several seconds holding the two broken pieces as they were for several seconds, with a sad, cavernous air gap between them. How had I got to this point? HOW HAD I LET THIS HAPPEN?!? 

A problem I didn't want

I pulled out the (now) 2.16m piece of rod through the hole in the front wheel arch and inspected it. The vast majority of it that followed the line of the sill was absolutely perfect. Only the last 10cm that sat in the rear wheel arch were corroded. Of those 10cm, approximately 1cm was now snapped off and sitting in the knuckle. That would have to come out. 

Now losing 1cm off the end of a 2.17m rod doesn't sound like a disaster. It's true, I might have been able to simply have a shorter rod - though I wasn't sure whether that would then mess up the height corrector operation. 

I made some initial enquiries to find another but options seemed limited. With the rod being over 2m long, it wasn't the kind of thing that would easily go in the post from abroad.  With a dead DS sunken on it's haunches, it's not the easiest piece to remove from a hulk and so, with a fair amount of work needed to remove one from a dead DS, I didn't think anyone would be in a hurry to pull one for me either. Also, it was very likely that the end pieces might be seized anyway - just like mine. My best bet was to find one already removed and available in the UK but I had no luck. I was told the usual fix for this was just to braze on a new threaded end - and so that is the route I went down. Which meant that, as well as repairing the rod, I would need to salvage that joint.......

Luckily, there was still a few millimetres of thread still sticking out of the fitting.

Broken end stuck in joint....

I tried welding a nut to the end. But there was not really enough surface area to get a strong join and that didn't work. I tried gripping that end piece with mole grips - and it snapped straight of. Damn! The threaded end of the rod was acting like serrations on a sheet of paper and providing nice, clean tear lines - thwarting my attempts. How was I going to get it out now? 

Going backwards......

I parked that concern for the moment and went back to the rod......

I'd had no joy with brazing previously, having tried with MAP gas. Okay for household plumbing but I couldn't get the heat needed for brazing. I'd been moaning to my friend Peter Bremner ('Badabec') about this. Hearing my control rod woes, he offered to shut me up by giving me an old TIG welder 'buzz box' plus an arc brazing torch - just the job! I picked the 'buzz box' up at the D rally at Little Horwood in July. With most brazing done nowadays using gas, arc welding/ brazing is old technology. But it works. 

Ye Olde Arc Welder

The two carbon electrodes are slid together like chopsticks until they touch and 'arc'. The arc produces significant heat and light. The light is enough to give you sunburn and 'arc eye' - temporary blindness.......

Arc eye. No - really

.....and the heat is hot enough to, well, you could braze with it, for one thing. The heat from the arc is used to melt a filler rod to join the two pieces of metal you are working on. While exceptionally hot, the heat needed to melt the filler material is less than that which would melt your metal pieces.

An electric arc (photo by Achim Grochowski)

For the repair, I could have bought some steel rod and applied a thread myself. But it was just as easy to buy a pack of long M5 set screws. (As opposed to bolts that are threaded their full length, 'set screws' are threaded at the end but are unthreaded for part of their length). 

Now, because of the age of my DS, the M5 threaded parts have a 0.75mm pitch. Modern M5 bolts/ set screws will typically have a 0.8mm pitch. However, as I was going to have to tap and re-thread the fitting, I could re-tap it to 0.8mm pitch to match the set screw I was using. 

Hacking the heads off two screws, I experimented. I practiced extending the threaded section just a little. 

Adding more thread

Yep - that worked.

More thread added

How to join the set screw to the rod? It wouldn't be a strong repair to simply butt two ends together and 'grow' a blob of filler material around the join. Also, it wouldn't be fit for purpose: I would need to insert the rod through it's small guide eyes when I refitted it. For the repair to be useful, the diameter of the rod needed to stay as it was. The filler needed to be between the pieces - not around them. I used a grinding wheel and file to create a 'staggered offset' joint. (I don't know what else to call it - a step joint?). This meant that the surface area between the pieces would be much greater - giving a stronger join.

I planned to cut an offset joint in both pieces

I took out enough material from each part to enable the filler material (a silver alloy in my case) to flow between the two pieces. 

Lining up the joint. I left an air gap

After cleaning the buzz box up, I gave it a go. I needed both hands free. After carefully clamping the two pieces in position, I used the buzz box to heat the pieces until the filler rod melted when it contacted them. It didn't take long to get used to the technique. By varying the distance between the two carbon rods, you can alter the size of the flame/ heat. If i'm honest I also applied a bit of heat directly to the filler rod to encourage it to melt and flow. 

Very quickly I could see (through my welding helmet) that the joint was made. Once it had cooled a little I gently bashed off the slag with a hammer. After a bit of cleaning up and dressing with a file, it looked very respectable. The diameter was within tolerances  and the two pieces were aligned and straight.

Can you see the join?

This just might work! Could I do that a second time?

Yes. Yes I could. The real repair went just as well.

The actual control rod. Actually finished. Actually

If I'd thought about it, I'd have paid more attention to the offset and made it a snugger fit, but it was fine as it was: the rod fits through the nylon eyes that will go back in the sill, and the joint looks tidy and strong. Job done.

Before and after

I still need to drill out and re-tap the knuckle joint that's got the snapped rod piece stuck in it. Rather than do it by hand/ eye, Peter has suggested doing it on his lathe. I think that's the way to go: I can't afford to mess up the drilling and end up with an over-sized hole that the rod doesn't fit. Like they say: inside every small job.......

Scuttle Ends - Welding Repairs

More welding repairs. This time to the panels at each end of the scuttle. To give them their proper name they are (apparently) “lower front pillar closing panels”. 

"Lower Front Pilar Closing Panels"

To me they are the 'slippers'. Not sure who first called them ‘slippers’ - but it stuck with me. In fact each ‘lower front pillar closing panel’ has two pieces. The other pieces I'm going to christen ‘epaulettes’. Still not sure where I mean? I mean the areas where the bonnet hinges are mounted.

Scuttle end areas as they should be. No dirt here.....

As well as holding the bonnet hinges, these are also run-off area for any rain getting pushed up the bonnet to the screen and, when you operate your wipers and screen wash, all the grit and debris gets swept to the sides of the screen and down into these areas. Being horizontal, any excess water can just sit there until it evaporates. Consequently they can fill up with damp dirt and mud and are either unseen/ un-noticed with the bonnet down or, with the bonnet up are un-noticed simply because the bonnet and hinges are in the way! The short version of all of this is that they are damp little rust traps. 

Scuttle ends as they usually are. Plenty of dirt!

The ‘epaulettes’ make the problem worse. Once any seam-sealer fails, water gets behind them - again unseen - and rust out scuttle behind. from there water is able to drip down into your footwell behind any carpet or trim and rust out the floor and box sections…If you find that problem, you might blame a leaky roof (and you might be right to) but also think about these areas. Also, (unless you are Alberto Jansen) the next time someone tells you they have a  "totally rust free car" - remember this photo:

Beauty is only skin deep......

I’d removed the bonnet and hinges from my car and this revealed a thin layer of dried sludge in each ‘slipper’. Scooping this out showed the surface underneath was pitted. I took a wire wheel to the 'slippers' and also to the ‘epaulettes’ as their edges looked suspiciously crusty and compromised. Sure enough - more pitting. Prodding with a screw driver revealed holes: down through the slipper into the ‘A’ posts where the doors hang and (bonus!), on the RHS, also into the scuttle and behind the dashboard. This is how I tackled that RHS. I already had the screen and dashboard out. Well that was two fewer jobs to do. With a wire wheel I revealed where the spot welds were at the top of the A posts, along the scuttle panel join, and around the epaulettes. I marked the centres of the spot welds with a centre punch to give my spot drill something to work with.

Too many holes.....Spot welds marked.

I also found two bits of brazing. One right at the tip of the slipper (the ’toe’?) and another where the epaulatte fold over the top of the dash edge near the ‘A’ post. I carefully cut these out with an air cut-off tool but a Dremel would have done the job.

Brazing on the 'toe' of the slipper

With a spot weld drill, you need to plan ahead. Ideally you want to drill your holes through the pieces/ panels you intend to replace, and leave the ‘mating panel’ or under-surface intact to give you a good base to plug weld back on to.

A spot weld drill

Unfortunately the scuttle area is quite tight and difficult to get a spot drill in. On the outer edge of the panels, it means drilling through the lip of the front pillar - the panel that the door hinges are fixed-to. Couldn’t be helped.  

Drilling out the spot welds

I couldn’t get the spot drill perpendicular on the welds on the inner edge either. Drilling at an angle means the drill is likely to skip around, but also, that if you try to fully drill through the top layer of metal to release the spot, you will find that you have drilled out a lot of the layer underneath as well. In the case of my RHS repair, that wasn’t going to be a problem as I needed to cut out and replace that pitted piece of scuttle anyway. After some careful drilling and chiselling, it all came off........

Removing an 'epaulette'.

.......reasonably well. Behind the epaulette revealed the lurking creeping rust.

And removing a 'slipper'.

I couldn’t see how the back end of the slipper was held in place but, as it was corroded, was able to snap the slipper off at that point to remove it.

On my car there had been a previous repair to the screen frame and I found the base point to just be a mass of weld. 

Signs of previous repair 1

Signs of previous repair 2

With the part off, it was clear that the replacement was significantly longer and I couldn’t see how all the pieces were meant to fit together where the slipper meets the ‘A’ post. Did the slipper sit on top of the screen channel or did it erm…’slip’ up behind the back? Did the slipper body go inside the screen channel end or outside it?

How was this meant to fit?

The answers to those questions would need to wait, and I turned my attention back to repairs - replacing the pitted metal on the scuttle. I marked out the area I needed to replace. You can see how the rot was caused by the slipper and epaulette. I tried to mark out the smallest area possible. 

Getting ready to cut out the rot

No going back now......

Too hole-y for my liking

I used state-of-the-art K.A.D. software (Kelloggs Aided Design) to make a template for a patch. This needed to be about 1mm smaller all round to allow me to weld it in place and then grind off the excess. 

KAD - Kelloggs Aided Design

I used 1.2mm sheet. After a few trial fits, a bit of shaping to match the scuttle curve and fettling for size, I was good to go. 

Having clamped the patch in place I started by tacking one end and, when I was confident the patch had it’s 1mm gap all round and was level to the surface of the scuttle, tacked it in a few other places and then gradually filled the gaps in - re-checking for fit as I went along.

Welding the patch

I had a couple of incidents of blow-through - where the welder melts, rather than welds, the metal leaving a nice hole to deal with. I think this is probably down to my inexperience. I should have cut out a bigger area to give me good clean thicker metal to weld to. Welding thin metal is bad enough but welding when there is pitting is almost impossible. I did some welding from inside the dash area of the car to make sure the patch was sealed all the way round.

Getting there....

While I could, I took a good peek lower down into the hollow ‘A’ box sections - the bits with the door hinges on. These showed very little, if any, rust protection by Citroen. That said, there was some delicate surface rust there but nothing at all serious.

Inside the 'A' pillar

I scrubbed what I could with a wire brush and applied a rust remover (phosphoric acid), a water-based rust treater (a Jenolite product) and finally a rust sealer (a Frosts product). I was very suspicious of, (and disappointed-in)  the Jenolite rust treater. Although I’d scrubbed off visible rust, after application, it seemed to work in places but to have created rust in others? What’s that about? Was it because it's water-based?Given that that area had survived 50 years with little rust, I reasoned that coating it with the Frosts product plus a generous squirt of Dinitrol through a drain hole at the bottom once it was all together, would have to do.

Several weekends went by while I trawled parts and repairs books looking for photos and clues for how it all fitted back together. In the end, Jamie at DS Workshop was kind enough to let me visit him to look at the many cars and hulks he had there and take some photos. With lock-down easing, I also arranged to meet Peter ‘Badabec’ Bremner there too. While it wasn’t exactly a party, it was nice to meet other people and talk Ds. The trip to Jamies revealed that the slipper was meant to sit inside and behind the ‘A’ post channel.........

Can you see how it fits together?

A photo Peter Bremner sent me seemed to confirm it....

Slipper behind the 'A' post (photo - Peter Bremner)

.....and a timely photo of Stephen Leech's car on Facebook decided it for me.

'A' post on a 1967 car (photo by Gergely Papp)

All of which meant I needed to get rid of and clean up the mass of weld that I had found and get behind the channel....... I came up with a drastic plan to enable me to do that: I would have to cut through the ‘A’ post. The A post is actually made up of an inner and outer channel. So that I did not weaken the ‘A’ post, I decided to remove only the inner channel. Looking back on a couple of reference photos I had managed to find, I noticed that this is exactly what some others had done.

Another split 'A' post (photo by Mark Lister)

Again, unfortunately, this meant drilling through the part I was keen to retain - the other channel. I identified the spot weld points and drilled through and opened the outer channel up. With the inner and outer channels split, I just needed to get a short section of inner channel out. I couldn’t get a cut-off tool into the inner screen channel, so drilled a row of small holes and got it out that way.

Removing a section of inner screen channel

I went back to tidying my patch. I  couldn’t get a cut off wheel in the gap under the slipper, so had to build up a stash of slightly smaller, part worn disks for the job. With these I was able to get at the excess weld around most of my patch except for the very bottom edge - which was too tight to the ‘A’ post (one reason for NOT cutting out too big a patch). Luckily those welds were below the slipper so did not interfere with the fit and so I left them there.

Grinding down the spots......

With the patch in place and the area tidied, I readied myself for a test fit of the replacement parts. The Epaulettes came directly from Alberto Jansen and ‘Citrorevanche’ as I couldn't find them in the UK during lockdown. Alberto also re-makes the slippers I used  - though I bought those through a parts supplier.

Original and after-market panels

The slipper didn’t fit. In all fairness, it’s not realistic to expect after market parts to just fit like a glove (after all, slippers for feet - not for hands!). Some fettling is always to be expected. In this case, I found the replacement panel was too wide for the gap and so needed some re-shaping to make it narrower. I also found that it didn’t have the right degree of curve as it didn’t follow the lines of the panels of my car. Again, some gentle and gradual re-shaping solved that problem. I also aded a few more welds to the captive nuts. 

The replacement epaulette panels also needed some fettling to get them to fit. Not least because I found that, as supplied, the edge that warps over the dash/ scuttle has a fixing piece welded on and so is different to my simpler ‘epaulettes’.

Replacement 'epaulette' was different to mine

I cut off the different piece and welded on a simpler edge made using a template from the one I’d removed from my car. 

Not bad...

That will do

Other than that, the repro panels were perfect! Now it was time to fit them.

At the end of lockdown I sold the first MIG welder I had bought (see post HERE), and rashly bought a proper spot welder. I’m not sure how much use I will get from it but there are certain DS jobs and other projects that I felt would be be better completed with a spot-welder. Even though I had drilled out spots that would then need to be plugged with a MIG, I decided to complete the majority of this repair with the spot welder - partly as an experiment, and partly because the welds would be in fairly visible areas.

An resistance spot welder.....

I reasoned that spot welds would look neater and more original. There would be less tidying up after welding (no plugs to grind) but there would be more prep. work. Firstly in setting up the spot welder, but also In order to get the necessary current to flow. A spot welder pinches the layers together and then welds them together. This meant removing paint, zinc and any rust from both sides of both panels to be joined - so inside the car behind the dash, outside the car on the chassis and on the replacement panels. 

When I drilled out the spot weld I noticed that there was one that held the epaulette and the slipper together, but which did not also weld them to the scuttle. That's because there was a small air gap behind. When I was confident that both slipper and epaulette would fit together on my car, I marked their relative positions and welded those two pieces together. I used a weld-through primer on those surfaces that would end up hidden. I wanted to try to avoid (or delay) creating new rust traps.

Slipper and epaulette - joined

With the assembly slipped in place. I marked where my main spot welds would be. I removed the assembly (yet again) and cleaned those area to ensure good electrical contact. The spot welder was very heavy and difficult to manoeuvre but was satisfying to use and gave good results. 

Spot welding the panels in place

Coming together nicely

I went from one end to the other, checking and re-adjusting fit as I went. I would be a liar though if I did not confess to thinking that, while different, it would have been no harder overall, for a confident welder to do the job with a MIG. But I'm still pleased I used the spot welder. It all came together very well and, apart from a few drilled-out holes that still needed to be plug-welded, looked very neat.

The final job was to repair the 'A' post screen channels of course. Rather than reuse the short sections I had cut out, I made up some some new lengths using 1.4mm sheet. I think the originals were possibly 1.2mm but 1.4mm would add-back some strength.

Something to note here is that the inner channels narrow on one side at the bottom. As such they are ‘handed’. The sections I removed seem to narrow on the inner face - in order to accommodate the epaulette.

Screen channel narrows on the side that meets
the epaulette

The Citroen parts books confirm the channels are handed, however the replacement inner channels sold by some sites seem to show that narrowing on the outer face?

I can’t see any benefit or purpose in them narrowing on the outer edge - but it makes perfect sense for the channel to narrow on the inner edge to accommodate the 'epaulette'. Anyway, if you come to do this job - or work on the ‘A’ post screen channels, check how/ where yours narrow before you order a replacement.

I used an old pair of mole grips to make a tool to help me straightened the channel out before welding. Surfaces were cleaned up for spot welding and zinc primer was applied. I used the spot welder to fix the new channel sections and then finished off with some plug-welding with the MIG. All-in-all that should make for a strong repair. All edges and welds were tidied so that the screen rubber would fit well.

Same job - but on the left screen channel

With all the welding done I applied some extra primer. Even though I applied weld-through primer, and even though I will apply seam sealer to this repair before the chassis is painted, I thinned some red oxide primer and painted that over all the joins - with the intention that it would wick into gaps and provide added rust protection. I will clean all this up when I come to properly priming the scuttle area for painting. 

Finished off with a temporary coat of primer

The same 'slipper repair was carried out on the left side of the car - with the exception that the rust was not too bad and I didn't' need to patch the scuttle. 

Similar repair on the left hand side of the car

My last welding repair before chassis cleaning and painting will be to finish the replacement of the outer sills - a job I started then left. With the spot-welder, that should hopefully be quick and easy. Remind me that I wrote that……..