Sopwith Dolphin Retrospective No. 6 - More Fuselage Clutter

Next: the instrument panel and all the clutter behind it.

 I spent a lot of time studying Windsock 54 and the Cosford pictures before doing this bit. It looks as though there was a bit of variation in the layout; it probably depended on the variant. But the Cosford version looks like the Dolphin I, with its geared Hispano engine, and that's the variant I'm making, so I'm probably not too far off.

To make the instrument panel itself, I cut a piece of card about the right shape and then sanded it down until it just fitted between the fuselage sides. Then I taped it to dark-coloured card, scanned it and imported it into TurboCAD. On top of this, I drew a nice symmetrical version with the cut-outs for the guns, printed it out, taped it to plastic card, and cut out the panel proper. I painted it a reddish brown, using a slightly different shade than the framework. I don't go to a lot of trouble representing grain; at this scale, it could end up looking a bit artificial. What I do is paint it quite thin with a couple of swift strokes of a medium-sized brush, from one side to the other. This gives enough of a grain effect, I think. When it was dry, I gave it a coat of Satin Cote.

For the instruments, I cut little sections from Evergreen rod of various diameters: 0.5, 0.64 and 0.8mm; then ground them flat between sheets of wet and dry. I did quite a few of these tiny discs, and chose the best examples. I stuck them on masking tape and painted them. Then I did the dials with a very fine brush, and corrected afterwards with black. A big blob of Gloss Cote represented the glass cover. The maker's plate (?) is baking foil. This all worked pretty well, except that when I prised the instruments from the masking tape, they came off with ragged edges, so I had to spend some time neatening them up with a knife and a magnifying glass. It was just slightly less effort than redoing them.

Once all the instruments were ready, I superglued them to the panel. The procedure is to put a tiny dot of superglue in the correct location and then plonk the instrument onto it. This prevents smears of superglue. The result was pretty convincing:

Those readings should be correct for a plane sitting on the ground, I think you'll find. Though I could be wrong.

The easiest thing to do next would have been to glue the instrument panel into place on its own - this would make it very easy to line up. But because the cockpit sides have a tendency to curve inwards, they would be exerting pressure on the panel and I didn't want it to buckle. So I decided to reinforce it with the ammunition boxes before assembly.

The ammunition boxes, then. These are quite a complex shape, and I wanted the ammunition feeds to be open at the top, as in the original, so you can see the ammo feeding into the Vickers guns. To start with, I drew outlines in TurboCAD on top of the instrument panel I had already done. Then I made a couple of rubbish attempts at translating this into plastic. In the end, I cut out solid pieces for the boxes themselves and made the vertical bits as a box construction out of strip. Then I painted the completed assemblies with Humbrol 27001 and when they were dry glued them to the instrument panel with superglue. Windsock 54 shows that the panel is parallel to the cabane struts and at a slight angle to the ammunition box, so I added a bit of thin strip as a spacer at the bottom of the panel. After this, I glued the complete assembly into the fuselage.

But there was a slight complication. I had noticed when checking for fit that when the panel rested on the longerons as it should, it was not quite horizontal. The whole framework must have been pushed slightly out of true when being shoved into the fuselage skin. This would continue to have ramifications later. For now I just had to adjust the panel slightly to true it up, leaving a tiny but unnoticeable gap at one side. But this was a warning to keep a close eye on things. From experience I knew that if you once allow a feature to get out of true, then it can act as a distraction when aligning other features, and you end up with a crazy skewed out-of-kilter sort of effect. For now, the project continued, but with early signs of shakiness.

I made the link chutes as a box structure like the ammunition feeds and glued them directly to the ammunition box so that they aligned with the apertures in the fuselage. But was it really that simple? Have a guess. (Clue: No). The trouble was that when cutting out the fuselage sides, I had cut out the apertures without any thought that they would correspond to an internal arrangement of tubes and things. So I hadn't left enough room for the ammunition box between the instrument panel and the link chutes. The good thing is that I had noticed this before starting to fix everything in place. So before going any further I had the job of extending the square apertures with the point of a very sharp knife. Later on, I would have to close up with tiny fillets of plastic, but I could put that off for now.

Here's what it all looked like at this stage:

Not its good side

Starting to look a bit messy, isn't it? In retrospect, I don't think Humbrol 27001 is ideal for small pieces like this. It is not really designed to be applied with a brush, and goes lumpy if you try to overcoat it. And it comes off on your fingers. Also, I would have liked a more shiny finish, as you see in the Cosford restoration; but aluminium does oxidise, doesn't it, so in that respect it's probably quite realistic.

Fortunately, you won't see the ammunition-related clutter from this angle in the completed model, though in my heart I can't really accept this as an excuse. But, anyway, luckily it all looks quite good from the cockpit, which is the only angle it will be visible from in the finished model.

All except for one detail, that is. The trouble is, if you look closely at the photo at the top of this post, you might just make out that the starboard ammunition feed has come out distinctly wonky. I noticed quite early on that I had got this slightly wrong, simply by not taking enough care to get a good rectangular box shape. It was only when I had assembled everything that the thought began to grow on me that it wasn't good enough. In the back of my mind, I instantly evolved a stratagem that saved me a lot of annoyance and hasty ripping-out:

1. Do nothing about it. Over the next few days, gradually acknowledge the problem and:
2. Do something about it, as yet to be determined (Clue: it usually involves ripping things out and redoing them).
3. But for now, do nothing.

At this stage, then, the project was beginning to take shape, on the whole, and the techniques I had used were proving to be successful, more or less. But already there were some worrying signs of a downward trend in quality, that could become cumulative if I let it. Some lessons needed to be drawn here:

• Measure things.
• Measure things against each other.
• Keep things square
• Align things against an established datum. Don't align them against other things that are already misaligned.

Next: might as well get on with the engine, then.