Here’s the new transmission tunnel part – sides are mdc, and curved surfaces are body-filler. There’s an entire wooden frame in there keeping it all in shape.

I’m using some ultra sticky double sided tape to hold the side panels of the tub in place. You can see the stuff here. It’s incredible.

Here is the polypropylene plastic sheet forming the side of the tub mould. The tape gives me 0.7mm clearance between the sheet and the side, which is a good gap when I apply the epoxy adhesive to put the final tub in place. The adhesive has a maximum structural fill gap distance of 2mm so precision is called for.

Some setbacks of late

I’ve not gone away or given up on the blog, but have had a bunch of setbacks recently with this composites journey, so I thought I’d get them out here and move on.

Wrongness 1
Firstly, my compressor died (all-in-one motor and pump unit seized). The replacement part isn’t due until tomorrow.

Wrongness 2
The moulds I made for the fuel tank are wrong. I did two things that went wrong:

  • I made a part of the mould with a 180 degree return. No worries, I thought – I’ll just fill the back of the return with expanding foam, then I’ll have a nice sensible surface to envelope-bag against. Nothing like it. Two things broke – firstly putting it under vac meant the trapped gas in the foam needed somewhere to go, so I struggled to get a full vac
  • The return was in essence being crushed by atmospheric pressure, and the mould flange cracked
  • Wrongness 3
    I can’t vac-bag a mould with a 180 degree return – every time you turn it over, the cloth moves about. I need to rethink the way I’ve moulded this part.

    Wrongness 4
    My plan to infuse the transmission tunnel as three separate thin cosmetic parts, and then put the moulds back together and do the remainder of the infusions failed. I’m using a new release agent, and it’s excellent but there way no way I could get the backing materials off the mould with out popping the part out. Then I noticed that I’d got a lot of voids on a sharp return which would have meant a lot of chipping out and filling.

    Wrongness 5
    The tunnel top mould that I had to polish didn’t fare terribly well. The edges on the tunnel top that turn down into the tunnel sides (or up rather, on the mould) chipped a bit under the polishing disk.

    So, get well programme is:

    1. fix the compressor
    2. scrap the tank moulds and make new ones
    3. scrap the transmission tunnel moulds
    4. cut out the seat squabs and footwells, and make a single composite tub

    An infusion went wrong

    So, I was finally ready to infuse the transmission tunnel top, after deciding on a compromise. The compromise was to not spray the part with clear gel-coat first, even though I had some scratches in the mould. [1]. I reasoned that the scratches will leave a positive on the part, and that can be flatted off with some wet-and-dry and a little polishing.

    IMG_0137.JPGSo, I laid up the part (one layer of 350 facing cloth, and one layer of 200 backing), found to my surprise that I got the bag to seal first time, and set about infusing the part.


    This is the first time I’ve had a stuck infusion, and as I think about it, it was a culmination of a bunch of factors all adding up together to cause issues. The factors were:


    • IMG_0138.JPGI decided to infuse along the short side and go up the full length of the part, rather than infuse across the shortest distance (I’m running low on infusion spiral)
    • it was 10C in the shed when I was mixing the resin, and I thought “It’ll be OK – it means I can use quick catalyst and it won’t go off all that fast”. I ignored the gloopy sensations.
    • I was feeling impatient, and thought I could get away with quick resin, allowing 10 mins for degassing because it was cold
    • my oven is out of action: it’s in pieces while I make it bigger for the transmission tunnel. I thought I had enough kingspan for the job, and I was short by one roof baton length. This is now bought and sitting in the boot of my car. This again was an incentive to go for the quick catalyst
    • when I was mixing it, it felt more viscous than normal, which should have been a warning that stuff was about to go wrong.
    • when I infused it, it got stuck half way along the part and the resin went off in the pot.


    So, I think the resin being too viscous meant it didn’t march along the part quick enough before it started to go off. If I’d used slow resin the infusion could have ran at the snails pace it was going at and it wouldn’t have been a problem. What’s more, I could have warmed the resin up (with the slow catalyst) and it still wouldn’t have been a problem. For the handbrake bracket, I actually had the resin at 40C to ensure it wetted the part out properly. So, this evening I will finish the oven, take the half-part out of the mould, salvage the cloth I can which I’ll keep for backing layers, and have another crack at it.

    [1] My compressor has died, and the replacement part (£160 if I can wait for a machine-mart VAT free weekend)


    Sometimes stuff goes wrong, and you just need to stop

    IMG_0120.JPGSo, as you can see in this photo below, I need to make an access hatch for the transmission tunnel. I want it to be flush fitting and reasonably strong.

    I’ve taken a mould from this, and made the flat sheet for the top plate (2 layers of 300gsm, one layer of 200gsm gives 1.05mm thickness).

    I then made the adhesive to bond the two – infusion resin mixed with glass microbeads to make a paste. All good.


    I filled the bottom panel with adhesive, clagged the two together and clamped them. After that, I popped it in the oven at 60C for about 30 mins to kick the exotherm off (with fast resin) and I left it alone for a while. I even remembered to cover the facing side with bagging film to stop it getting mucked up when the adhesive ran out as the parts were compressed.



    IMG_0131.JPGIt’s not easy to see from this, but what I think has happened is that the heat from the oven has started to set the adhesive from the outside in, sealing it. Then this has kicked off the exotherm inside which has caused the adhesive to expand, but because the outside edges are setting and gripping, and the adhesive wants to expand and can’t. So, it warps.


    I now need a different and better method, which I’m cooking up right now. Either a much slower cure, or cross drill it (where the hatch needs to be drilled anyway, and leave a few extra drill holes in the base only in the middle. Then bolt it through with a couple of plates on the outside help reinforce the shape and a tube inset at each bolt through hole to stop over clamping. I will then inject the adhesive in through the spare holes and let it go off. The open holes will give the expanding adhesive somewhere to go.

    Current Plan

    I have  bunch of stuff to do, and am feeling a little overwhelmed about which job to tackle next (at 19:30 on a Sunday night). I’m now getting over the disappointment, time and materials lost, and have a plan. The plan is:

    • buy beer
    • make savoury bacon cupcakes with my daughter
    • watch The Blues Brothers with the kids.
    • Sleep.

    Access Hatches – it’s not easy to do it well

    IMG_0121.JPGHere’s the transmission tunnel buck, and I have the moulds from it, after some finishing. It’s a split mould, one top, and two sides. Now I have to start thinking really hard about just how many access panels I put in, and how to make them.

    I’m replicating the recess and access panels so wherever I need to fit them on the tunnel, they will all fit to each other and be interchangeable.



    IMG_0120.JPGHere is the closer up view of the access hatch, or at least the recess for it. I want the finished panel to both fit flush, and be bolted through. This recess on the right will be cut out, apart from leaving a 12mm rebate all the way around to allow the panel to fit an bolt through (into the rebate).

    I will reinforce back of the rebate, and use 5mm nutsert flush fitting thread things to receive the socket-headed cap screws. The recess is 5mm deep, as is the head of a 5mm cap-screw. I’ll bond the nutserts in from the back, and use my friendly neighbourhood cad package to get the spacings right.

    Today, I’ve made a mould from the positive mould part made from the recess (gel-coat and three layers of 450, poly resin). This means I’ve now replicated the recess above, and I’m using that to make a trial CF part.

    The plan to make the flush fitting panel is:

    • 3 layers of 350gsm CF to capture the back of the panel, and this should be a perfect fit into the recess
    • drill the panel for the mounting holes.
    • Bond some 1mm washers to the inside of the base panel to spread the load of the cap screw. I wish I had better washers that were a larger washer size with still a 5mm hole. Araldyte will be fine for this.
    • cut some flat CF sheet I have spare to the profile of the top of the part. It should sit totally flush with recess above.
    • mix some epoxy with glass micro bubbles to act as an adhesive and core material all in one go. The micro bubbles reduce density by adding air, in essence. When they’re in the box before they’re mixed, they’re like an ultra-lightweight fine powder. Definitely not something of which one wants a lung-full.
    • put the panel-back in the mould, and fill it with the epoxy/glass bubble solution and put the top on. Let it set.
    • Crack it back out of the mould again, and drill through from the back right through the front.
    • Drill through from the front with a drill the same diameter as the cap screw with a mm or two to spare. This needs to be done carefully to countersink the cap screw head. I think I’ll be setting the depth gauge on the pillar drill.

    There we have it – one flush panel, with recessed bolt heads and some structural strength. I’ve spent quite a while today thinking about how this would work and I think it’s a good plan. A bit of a faff mind.


    BOM for access hatch, lower part; positive lessons learned

    Bill of Materials

    • 3 layers of 300gsm cloth
    • 80g of mixed resin consumed (120g wasted)
    • mixed with fast hardener

    Lessons Learned

    • I mixed too much resin (195g mixed: 150g of resin, 45g of hardener) and had 130 left in the pot afterwards. When the exotherm kicked in, the resin in the pot got to 82C. I should not to be thrown in the bin until it’s cold again, else there will be fire. I always leave pots like this out overnight.
    • I’d not want to just mix 80g because it would be dragging bubbles through as it ran out in the pot, so some waste is inevitable. I’ll go more for 120g mixed next time.
    • Degassing the resin makes a difference to how much air bubbles over the part. I used the catch-pot which doesn’t have a huge capacity and the lid is as splattered as a plaster’s radio so it’s not easy to see how much it bubbles. However, it did bubble a lot
    • 5 mins degassing (not quite enough) and a slow infusion (10m or so) was just enough before I could feel the cup start to warm up. It was a couple of minutes from gelling, so I turned the resin feed off.
    • There was much less resin bubbling up the exit tube, so that tube wasn’t scrap (yeayy).

    Cure Time

    I’m giving it 1hr at 40C, then 6hr at 60C. This will give it the maximum theoretical strength.

    Shiny transmission tunnel mould

    Fixing a Dodgy Mould

    So, I cut a corner and had someone do some finishing on my mould because I was pushed for time. It was a mistake – the mould had a lot of resin dripped on it, and the surface gel-coat was very orange-peely.

    It was a perfectly servicable mould, but the transmission tunnel is something everyone will see when they look into the car, so I want it to be facing quality.

    In order to sort the mould, I tried 600 grit on a sanding block, and it was making an impact, but not flatting it out as fast as I wanted (i.e. in one lifetime). Then I remembered that I had bought some 400 grit disks from East Coast Fibreglass’s clearance page. Add the disk to the heavy duty polisher I have, and off we go. They really chew through the gel-coat. I finished that with a 1000 disk and polish. After polishing, I saw that I had some fine scratches, so I flatted it by hand (600 and then 122 grit), and polished it back. Again, some scratches. I don’t know if they were being introduced by the polishing sponge, but I don’t have the will to keep trying.

    photoWhat I have got is a mirror finish on the surface now, with some minor scratches which you can just see at the bottom. I’m hoping that a few layers of release wax will help fill the scratches, and I’m going to use a clear gel-coat before I infuse, which will then end up with some slight positives where the scratches are. These will be easy to flat and polish by hand.

    I don’t want to flat back again in case I end up going through the black gel-coat.

    I’m not sure if I’ve contaminated my polishing sponge or not here – more experimentation is required I think.

    Handbrakes – I’m taking a different approach

    I’ve been thinking and have two major thoughts about handbrake arrangements.

    • the current option doesn’t work well – My fury has cosworth brakes on the back, with a sierra handbrake cable and escort handbrake – all very standard and familiar to most fury and Se7en builders. Similarly (I expect) to most on here, I’ve always struggled to get the handbrake to bind well – I think the handbrake cable has too many corners to go around.
    • my composite tunnel will need hard-points – in order to mount the original handbrake, my tunnel will need some major hard-points fitting in for both the lever mount and the cable pass-through.

    So, I got to thinking, rather than run a hydraulic handbrake which will not pass MOT, are there any other options. I have a few ideas, two of which assume an IRS and one doesn’t:

    • solid mechanical lock – This one is inspired by a rallying mate of mine with a hydraulic handbrake, who would stick a screwdriver through the transmission tunnel into the propshaft UJ to prove to the scrutineers that he had a way of mechanically locking it. So, taking this idea to its logical extreme, one could mount a pair of solenoids on the transmission tunnel, pointing into the UJ. Hit a button, have the solenoids go out, plunger locks the prop. If you use a laser and a photoreceptor shining through the UJ, you accurately know where the prop is, you can guarantee the solenoids will go in. Advantages – simple, but the ally panelling in the transmission tunnel we all tend to go for isn’t strong enough to support the brackets. A weld-in cradle on the other hand would do it. Disadvantages – no use if you’re drifting . Also, if you applied it at any speed above about zero, it would probably eat itself for lunch.
    • swap calipers – use a pair of electric handbrake calipers, as per many modern cars. It seems simple to make a bracket, bolt them on and give it 12v when it wants it. There are advantages to throwing away the handbrake cable and routing, bracketing, handbrake, etc. and the disadvantage is that you’re adding to the unsprung weight. I think the overall weight would be quite similar. However, these things would open up a can(bus) of horrors in control
    • Linear Actuator – use a linear actuator instead of a handbrake lever. These guys come in many pull/push forces (easily upto 400kg) and various throw lengths, from 50mm upwards. 50mm is all I’d need. They’re simple to control, basically being a DC motor and lots of quality gearing – +ve voltage sends them one-way, negative the other. What’s more, they can be adjusted for start and stop positions, and can be specced with a potentiometer for position management. Add the whole lot to a micro-controller and you have a solution.

    So, after much pondering and chin-stroking, I went for this e-stop kit from Speedway Motors in the states. The cost of a roll-your-own solution in the UK would be about £250, and the cost of theirs including shipping taking advantage of the strong pound, is about £350. Neither option includes cables, which I’ll need custom making at about £40. The cables will be Ford disk to m6 thread. Using their kit does mean I’ll save days of engineering a solution (much as I would enjoy doing it).