Design Approved, Mould in Production: An FG Coil Cover Update

Once a carbon fibre design is approved, there are still four stages between the CAD file and a part you can bolt on: the mould gets cut, a first sample comes off that mould, the sample gets test fitted on a real car, and only then does the production run start. The FG coil cover has just cleared the first gate. The design is signed off and the mould is in production now.

I get asked for updates on this one more than anything else in the FG range, so rather than drip feed it through DMs, here is the whole picture with the actual scan and CAD data.

Where this part is up to right now

The design phase is done. I have seen the final CAD with my changes applied, signed it off, and the job has moved to tooling. That sign off is a hard gate I hold on every new part, because once a mould is cut, any change means cutting it again.

The part is not buyable yet and I am not pretending otherwise. There is no stock, no first sample, and no confirmed release. What there is, is a locked design and a tool being made against it.

Why this one needed a brand new mould

The previous FG coil cover was a wet carbon part. Wet carbon is made with VARTM, also called vacuum infusion, where dry fabric is stacked in the mould, sealed under a vacuum bag, and liquid resin is pulled through the whole layup in one shot. It cures at room temperature or with mild heat. Worth knowing: most wet carbon on the market is a carbon surface ply over fibreglass structural plies, which is the same VARTM process with a cheaper stack behind the cosmetic layer. Anything I release in wet carbon would be full carbon on every ply.

This version is dry carbon, which is a completely different animal. Prepreg fabric already has the resin in the weave at a measured ratio, it is kept cold until use, and it cures in an autoclave at roughly 120 to 180 degrees under about 6 bar of pressure. A mould built for room temperature infusion cannot survive that cycle. So a new high temperature tool was the price of doing this properly, and the tooling cost was the single biggest line item on the whole job.

The scan is the part nobody sees

Before any of the design work happened, the mounting data had to come off real parts. I 3D scanned my own rocker cover and the factory coil cover that sits on it, and the raw files were not small, 119MB for the 3mf and 486MB for the STL. The feedback I got back was blunt: without scanned install data there was a high risk of a part that looks right and fits badly.

The image below is the baseline. Teal is the 3D scan of the factory rocker cover. Purple is the 3D scan of the factory coil cover, sitting exactly where it lives on the engine. That pairing is the reference every mounting point on the new part gets measured against.

Next is the new one in position. Same teal rocker cover scan underneath, but the white part on top is the redesigned coil cover dropped onto it. This is the check that has to pass before a tool gets cut: does the new design sit where the factory part sat, and does every hole still land on the same boss.

The approved design on its own

These two are the new coil cover by itself. Cleaner crown, the factory clutter taken out of the top surface, and every hole and boss carried straight across from the scan rather than eyeballed.

If you are weighing up whether the upgrade is worth doing at all, I wrote separately about what a carbon coil cover actually changes on a Barra, including the honest answer on weight.

What happens from here

1. The mould gets cut

This is the stage the job is in now, and it is the longest single stage. A high temperature tool for a part this size is measured in weeks of machining and finishing, not days. Nothing visible happens during it, which is exactly why it feels like the project has gone quiet when it has not.

2. First sample off the tool

Once the mould is finished, a single sample gets made from it. One piece, not twenty. This is the first time anybody sees the design as a physical object, and it is where surface finish, edge quality and wall thickness get judged.

One thing worth knowing about samples: a single first article is trimmed by hand, while bulk parts are CNC cut. So a sample can have edge quirks that will never show up in production, and I have to be careful not to condemn a good tool over a hand trimming artefact.

3. Real world fitment check

The sample comes to me and goes onto an actual car. Not a jig, not a bench, not a render sitting inside a scan. A real engine bay, with the harness, the breather and everything else that lives up there fighting for the same space.

What I am checking: do the mounting holes line up without persuasion, does it sit flat on the rocker cover without pulling on the fasteners, is there clearance around the coil packs and the breather fitting, and does it look right from a normal standing height rather than from a render angle. If any of that fails, the tool gets adjusted and a second sample gets made. That loop is annoying and it is also the entire point.

4. Production run

Only once a sample passes on a real car does the full run get made. The minimum is 20 pieces for a design like this, which is why one mould has to serve every finish. The finish split, how many gloss twill, matte twill, gloss forged and matte forged, gets decided right at the cut stage rather than up front, so the run can follow what people are actually asking for instead of a guess I made months earlier.

5. QC and freight

Photos before dispatch, then air freight, then customs. The freight leg is short compared to production but it is the one that has caught me before. If you want the longer version of why this whole cycle takes as long as it does, I broke down why carbon parts sell out for months at a time in an earlier post.

Why I am not giving you a date

Because I would be making it up. On an earlier FG run the ready date moved from the tenth of a month, to the twentieth, to the following month. Nothing dramatic went wrong, that is just what tooling and production do. If I publish a date now, the only thing I guarantee is that someone builds a plan around it and gets let down.

What I will do instead is post at each milestone: mould finished, sample made, sample fitted, run finished, shipped. If you want that in your inbox rather than having to check, get on the waitlist on the Ford Falcon FG dry carbon coil cover page and you will hear at the same time I do.

What this means if you are waiting

Nothing you need to do. No deposit, no commitment, no risk of missing out quietly, because the waitlist gets notified before anything goes on general sale. If you are building the rest of the car in the meantime, the carbon fibre parts for the FG Falcon that are ready now are all listed, and you can browse the full range by make and model if you have got something else in the shed.

FAQ

Is the FG coil cover available to buy yet?

No. The design is approved and the mould is in production, but there is no stock and no sample yet. The product page is live so you can see the finishes and pricing and join the waitlist, but nothing can be ordered until a sample has been fitted to a real car and the production run is done.

Why does a new carbon part need its own mould?

Because the mould is the part, in reverse. Every surface, radius and hole position comes off the tool, so a design change means a tool change. Dry carbon adds another constraint: prepreg cures under high heat and pressure in an autoclave, so the tool has to be built to survive that cycle. A mould made for room temperature infusion cannot be reused for it.

Why scan the rocker cover as well as the coil cover?

Because the coil cover does not mount to thin air, it mounts to the rocker cover. Scanning both, and scanning them together in position, is what proves the mounting points on the new design land where the factory ones do. Design a cover from the cover alone and you get a part that looks correct and fights you on the car.

What is a first article sample and why only make one?

A first article is a single part pulled off the finished mould to prove the tool before committing to a full run. Making one costs a few weeks. Making twenty from an unproven tool costs the whole batch if something is out. It is the cheapest insurance in the process.

What happens if the sample does not fit?

The tool gets corrected and another sample gets made. It adds time and it is far better than the alternative, which is twenty parts that need a drill to go on. A sample that needs a tweak is a normal outcome, not a failure.

Why is the finish split decided so late?

One mould makes every finish, so the split only matters at the moment the fabric goes into the tool. Holding that decision until the run starts means the batch can follow real demand instead of a forecast, which is how you avoid a shelf full of one finish and a sold out sign on another.

Following along

I will update this properly at each milestone rather than posting noise in between. If you are on the waitlist you will get it first, and if you have a question about fitment on your specific car, send it through, because those questions are exactly what shapes the fitment check when the sample lands.

Riley Baginski, founder, RB Innovations.