• TR4A - Rebuilding the Seats

    TR4A – Rebuilding the Seats - Part 1

    In January 2013 I finally got round to the seats, the last item in the total rebuild of my 1966 TR4A.

    The car is now painted in Royal Blue, code 56 or close to it and the trim colours were chosen by my wife, who's favourite handbag is light grey leather. I'd seen this combination on a Jaguar XK and knew it would work.

    I did the door-cards and wheel-arches in Vinyl, but my son persuaded my to do the seats in leather. The trim was supplied by John Skinner and his leather and Vinyl are an extremely close colour match.

    I had no previous experience of upholstery work on car seats and so this article has a certain amount of “dont do it this way” about it. If you are a beginner like me you may find this quite helpful!

    I am a precision engineer by trade. I like to work to drawings and a tolerance of 0.1mm or 10 thou if I'm not bothered much. With this soft seat stuff you don't have any drawings and 1/8in is about as close as you get most times.

    However there are a few places where quite a small size alteration makes a big difference to the end result.

    Anyone doing this needs to know that the 4A seats are, like many things you find on a TR, quirky and anachronistic and actually must have been quite hard to make in the first place.

    To me this is a sufficient reason to rebuild them rather than switch to some mass-produced item from a automatic factory in the far-east. I just can't see the point in bothering with a classic British car if you do this. Just buy a Japanese car if thats the kind of seat you want to sit on.

    There is no symmetry about these seats. The side nearest the the doors is cut back, probably to allow the seat-belts, a new thing at the time, to pass over.

    They don't use traditional, vintage style, upholstery techniques like the side-screen cars do. There are no springs, no listings, hog-rings or traditional webbing.

    They used Pirelli-webbing, borrowed from the Ercol furniture current at the time. The seat-bases were Vitafoam and the seat-backs expanded styrene.

    Under the seat-base is a rubber diaphragm.

    None of these materials will be serviceable after 50 years. Both of the foams were made by filling into a mould and the back particularly is an awkward and critical shape.

    If you can get reproduction moulded foam you will save a lot of work and get a better result. I didn't and had to build up the backs from standard flat parts. The seat-base foam is not so critical and worked OK more or less as bought.

    The seat-covers, leather in my case, are not just covers that slip over the foam. Both the base and back have them bonded into the foam.

    This joint gets put under a lot of tension and must be done accurately and very strongly or the whole job will look awful and come to bits.

    I know someone who moans bitterly that a professional upholstery shop charged him 600GBP to fit his seat covers even though he supplied the covers. He didn't like the fit they got either. When I told him that I had spent 60GBP just on glue for these seats, he began to see the other side.

    Just in case you are feeling a bit put off, here is a bit of encouragement:

    Attachment 28222

    At this stage I still have not worked out how to secure the bottom corners of the cover for the seat-back. Leather is not compatible with the round clips used for Vinyl at this position, its too thick.

    So this is still a work in progress and a few flaws are visible to an experienced eye.

    Stripping Down

    You will need an old blunt screwdriver, a yard of thick strong string and if the rubber feet that they sit on are badly rusted up, a blowlamp!

    I told myself to strip down one and leave the other alone so I could see how it needed to go back. Of course I didn't do this. Actually, I don't think it would help much. It would help a bit, but I like to see progress.

    Begin by removing the seat-backs which have three fixings down each side and three across the top.

    The ones down the side are similar to the kind of fittings that hold on the door-cards. They go in a round hole in the metal seat-frame. You lever the seat-back away from the seat-frame and they pop out, hopefully.

    The top three are different. They are an ususual S-shaped clip and are removed by sliding the seat-back downwards, after of course you have released the side-fixings.

    Break the side-fixings and you can find a replacement fairly easily, replacing the top fixings will be harder.

    The covers are held to the frame with small U-clips and these can be levered carefully off and most of them can be re-used. Some will break or be bent so you will need at least some new ones. Not hard to get.

    There are two big round clips that hold the two bottom flaps. These go round the frame-tubes. Again, these are unusual so keep them for reuse.

    Actually they didn't work out for me so I have some spares left.

    There is a big strong spring that holds the base of the cover for the seat-back. Its hooked into the frame-tubes. It will take all your strength to get these out. Put a loop of string under one of the hooked ends. I pulled upwards with the seat on its side.

    The seat-foam may be glued to the frame in several places but the glue will be weak by now and easy to free off.

    Unscrew the rubber feet and the covers will lift off complete with the foam inside. Note the position of the packing-washers because these have been chosen so that the seat sits equally on both feet.

    There should be funny spring clips which engage the end of the runner. The retention provided by these is minimal and I planned to use a better method.

    I had to cut the cover on one seat because the feet were rusted solid, but otherwise I got the covers and foams off without damaging them.

    They do contain a lot of clues that you will need when you refit the new ones so DONT discard them.

    I pulled the covers out of the foam so that I could see how they were bonded together. Doing this was OK on the seat-base because the foam was completely beyond any further use.

    The foam for the seat-back was just about reusable. Thinking I had new foam, I separated the cover from the foam without much care. Subsequently I had a lot of trouble with the new foam for the seat-back not fitting properly and had I taken more care, may have chosen to reuse it.

    Assess the foam for re-use carefully and cut the cover free, rather than just pulling it apart like I did, if you think you may need it.

    I hope you have got over removing the big spring because now you do the same string trick on the diaphragm. Except there are eight hooks to do. Only the first few will put up a fight. Note how the rear two are upside-down.

    More work with the string-pulling will release the hooks on the Pirelli straps. These will be time expired and won't put up much of a fight.

    The rubber feet should be held with counter-sunk screws. Mine had a mixture of counter-sunk and round-head and I couldn't undo two of them.

    I ground off the end that was sticking out of the captive nut and applied a bit of heat. Then quench with cold water and do this a few times. Try not to cook the rubber foot. This is better than other methods that might spoil the captive nut.

    Your seat-frames should now look like this:

    Attachment 28282

    Fixing the Frames

    The frames were blast-cleaned. There's a local chap does this cheaply. If you do this, scrape away any adhesive and grind off any really rusty bits. This ensures they will come back really clean.

    I did mine in February, sprayed with oil and wrapped them in cling-film until I felt like welding them.

    Close inspecion after blasting showed a crack in the drivers seat, on the curved tube, where the base turns into the back.

    I always grind a crack out so I can “weld in the V”. On doing this I found that these curved corners are reinforced by another internal tube. This helps a lot if you are welding around here because it reduces the chance of blowing a hole and allows use of a higher current.

    Attachment 28226

    I welded over the crack by about 3/8in each side, following the axis of the frame-tube, rather than following the crack around. It took a few passes to get right around the tube.

    If you do a repair like this, somewhere that can't be seen later, then resist the temptation to linish the welds flat. Just leave 'em ugly.

    Attachment 28227

    Whenever I weld a crack, in a place subjected to cyclic loading, I always add something to try to stop another crack forming. Remember that a weld upsets the carbon/iron balance in the steel and this means harder areas meet with soft areas along the weld line. This makes welded repairs likely to fail again.

    I added triangular reinforcing plates to both sides of the round tubes. Four plates on each seat. Later on I filled the gap with closed-cell polyethylene foam to make a soft edge for the seat-base foam to sit on. I also added two braces from the seat-back to the plates where the rubber feet fit.

    Attachment 28228

    Working out how to squeeze all this extra structure in without spoiling the fit of the foams took a while.

    However it transformed the way the seats feel when you sit in them. Before, although you may not have noticed, the seat back flexes backwards when you put your weight against it. Afterwards they feel rigid and I found this extra support reassuring and comfortable.

    Note the quarter-circle cut-back on the corner of the plate. This is to allow the hook on the diaphragm to reach the rail and the diaphragm to move down when you sit on the seat.

    The plates have Dutch-bends on the top edge to make this smooth and round. You just do this by making an ordinary bend first, take it a bit past 90 degrees and the crush it flat in a vice.

    These frames have the gap, between the lower rail and the rail that holds the seat-base foam, just bridged with Hessian, except for straps in a couple of places. This is a recipe for a lumpy finish around there and I chose to fill this in using 0.8mm sheet steel.

    This needed two pieces running from the existing bridge piece on the sides of the frames and ending at the rear where the tubes and rails all come together.

    These need to curve around to follow the curve of the front corners. And a flat, well almost flat, piece in the front that meets the side pieces. This is the most important piece from an appearance point of view. The joint is on the bracket that holds the seat to the runners.

    I ran these through my wheel joddling-machine to stiffen them a bit an get a nice fit to the rails. This isn't vital, you could leave them flat.

    Make a patterns for these first using card. My favourite material is a breakfast-cereal packet. Note that the two side are not quite the same shape at the thin end.

    Attachment 28245

    Note the cut-out that goes around the old strap that joines the two rails together. I retained this strap just to keep the whole thing in its original alignment.

    If you decide to do this too, be aware that you will need to be able to weld very thin sheet to the much thicker rail. I did this by running a short bead on the thin stuff to build it up a bit. This allows use of enough current to get into the thick rails.

    I also had to fit a new lower rails and rebuild the corners because these had rusted thin. Actually they were pretty thin to start with because the factory cut away too much metal.

    With leather you need to put in a fair bit of tension to get a smooth finish and these rails need to be in good order. I used 10mm angle from my local hardware store on the front and sides.

    At the corners I made round repair pieces using 1.2 mm sheet folded and gusseted so it would form round a curve. I left the original corners pieces in situ and welded the repair pieces to them.

    Attachment 28246

    The hooks for the diaphragm had worn the rails of the drivers seat almost through to the edge of the rail! These required welding up and re-drilling. If I were doing this again I'd drill them outer bigger because I found filling a 1/8in slot with weld to be quite tricky.

    I was searching the web for images of TR seats and found a picture of an early TR6 frame. I noticed that it had five Pirelli-webbing straps instead of the four I had on my frames.

    The webbing on mine had “expired” and I had reached the conclusion that these had been put in with a lot of tension.

    The TR6 seat also had a square section of tube at the top to accept a headrest. My wife suffers a bit with back and neck problems so I decided to make provision for fitting headrests by broadly following the TR6 method.

    You can see the new rectangular tube,its support brace and the bracing/reinforcing at the base.

    Attachment 28248

    Making New Straps I soon worked out that if I drilled a few extra holes, on the vertical sides of the frame, I could fit in six Pirelli-straps and set them up with a bit less tension. Exactly how much tension and the calculations I used to establish this appears below. The hole pattern shown uses all the existing holes and in most cases a strap goes into one old and one new hole.

    Note that the lowest strap is still in its original position and the clips that hold it in place are “upside down” and “on the other side”compared to the rest of the straps. This happens because the frames “change direction” at this point becoming closer together as you go downwards rather than farther apart as they do on the rest of them.

    These clips differ in shape on the sloping side of the seat compared the more vertical side. So that there are three different types of clips. Having decided to add straps I had to make extra clips of course.

    The top three clips, one of each type, are the originals the lower three are my “copies”. They should be spring steel really but mine are just “coat-hanger” wire which compensates for not being springy by being a bit thicker. They work just fine. The exact shape is not that critical.

    Attachment 28284

    These clips were originally held into the ends of the straps by rivets passing through two oblong fish-plates. The modern replacement for this is a “staple” and these are easily available from suppliers of the Pirelli-webbing.

    However when you use a staple all the tension in the strap is concentrated at the point where the staple passes through the strap. The fish-plates tightly grip the strap, when correctly rivetted and this spreads the load.

    Attachment 28278

    Its not hard to drill out the rivets and recover the fish-plates. I was fortunate that a “Model Engineering” supplier nearby stocked the right kind of rivet. And I had the rivet-snap to set them.

    All I had to do was to work out an efficient means of making the fish-plates required for my extra straps. Normally I would cut the plates out clamp them together and drill through so as to make the holes line up.

    But these plates are too small to work this way. So a piece of plate was drilled by hand using a mini-drill with a 1mm drill. I had a piece of 0.1in matrix board, used to build electronic prototypes and used this as a drilling guide.

    Attachment 28252

    The fish-plates were then “liberated” from the sheet with shears. This produced a nice crop of little plates most of which would pass a pair of rivets without much trouble.

    Attachment 28253

    The rivets used were 3/16in long, dome-headed and made from soft iron.

    Attachment 28251

    The tool needed to “set” them, called a rivet-snap, has an indentation to fit the dome-heads. The rivet-head must sit in the snap so that there is a tiny bit left above the face of the tool. This is vital or the rivets will set loose.

    Hold the snap in a vice using soft jaws. Put a rivet through a fish-plate, two layers of strap and the other plate. Hold everything together with a small tool-makers clamp and peen over the exposed rivet-end, whilst supporting the dome-head in the indentation in the snap.

    Its not just a matter of smacking the rivet hard. I used the smallest hammer I have and coaxed the rivet to flatten out and tighten up evenly. Done right the rivet head will look like the photo above.

    So now the tricky business of cutting the Pirelli-webbing to the required lengths. There were no clues to be had by measuring the old straps.

    Some of these were so stretched that they were longer than the width of the seat. Those at the extreme top and bottom were still offering some resistance and had to be stretched to remove them. But not by much.

    I set up a test with some new webbing by trapping a 1m length under my kitchen door and making marks on the floor to show how much I could stretch it. I could easily pull a 5% increase but to get 10% took a determined show of strength.

    I measured the widths of the seat-frames, at the places where the fixings go and set up a spread-sheet to allow for the length of the clips and the 30mm folded over for the rivetting operation.

    I decided on using a stretch ratio of 6-8%. Lower at the bottom, that does not take much load, and higher at the top. I was worried that the tube I had added, for the head-rests, might give a feeling of having a “lump in your back”. Trust me this will not happen, straps or not. I know this now. The foam provides plenty of resistance to avoid this.

    If you used four straps as original, you would want to cut the following lengths of Pirelli-webbing, starting from the top and working down, lengths in mm:
    378 394 415 415

    I based this on 8% for the top two and 7% stretch for the lower two.

    But I added straps and moved them about. The spread-sheet had the power to allow for this and I actually cut them:

    378 392 407 418 426 418

    I got my webbing from John Skinner with the seat-covers. The roll had nearly enough for four seats! But it was 1.5in wide and the original, and therefore the clips, was 1.25in. Having a bit of extra width means that I get the same resistance with less tension. One of the reasons I did not cut for a 10% stretch.

    I just trimmed the ends on a slope to suit the clips, marked the lengths and the position of the folds.

    Attachment 28279

    Note the trial fit being carried out on newly-blasted frames, wrapped in cling-film!

    Attachment 28280

    Contined in Part 2 so as to overcome the limit of 15 images per posting.

    Alan J Turner


    July 2013