Sopwith Dolphin Retrospective No. 7 - The Engine

The engine for a Dolphin I was a geared Hispano Suiza; which I'm hoping looked very much like this:

The original generated 200hp, apparently, and I remarked to someone, in a jocular manner, that my 1/72nd scale replica would therefore develop 2.8hp. But this can't be right, can it? A model 3 inches long having the power of almost 3 horses at its disposal? That would be simply crazy.

Actually, the power might be roughly proportional to the capacity of the cylinders, or the cube of the dimensions. So that makes a more manageable 0.0005hp. But there would be other factors; other things that don't scale quite that easily, like fluid viscosity; not to mention the lack of provision for moving pistons and a crankshaft in what is after all a solid block of plastic.

On with the weary business of assembling a small thing that somewhat resembles a rather larger thing. I didn't have any proper drawings for the Hispano, so I estimated measurements from various sources (including Windsock 54 of course!) and did some more drawings in TurboCAD, as a guide for cutting and alignment. In the event, the dimensions were way out; but the front view came in useful for aligning the cylinders, when I eventually bothered to align them properly. More about that later.

I made the crank case from two pieces of solid plastic, shaped by my usual technique of sanding on a flat piece of wet and dry. The lower portions of the cylinders were discs cut from plastic rod and then sanded to a uniform thickness in situ. The cylinder blocks were a solid piece of plastic apiece.

As before, I got reasonable results with the sanding; except that when it came to the cylinder blocks I found it very difficult at this size to get the ends nice and square. By this point, my plan of inventing techniques as I went along was starting to fall into disuse. Eventually I got a reasonable result, but before doing anything like this again I'm going to contrive some planing type of tool to get a nice true planed surface; or buy something. Matt (of the Wargames Table) kindly referred me to a web-based company Antenocitis, and they stock a tool they call the True Sander that looks as though it should do the job.

After a bit of work, the main engine pieces were ready for painting:

You'll see here that I've made the prop-shaft from brass rod, for strength. The blocks at the rear approximate the magnetos. The two discs are gears for the Constantinesco interrupter mechanism for the Vickers guns. Incidentally, it seems that as a further solution to the old firing-at-the-enemy-without-shooting-the-propeller-to-bits conundrum, some Dolphins were fitted in the field with Lewis guns outboard of the propeller arc, very much in the manner of a Spitfire or a Hurricane. Ahead of their time or what?

When it came to painting, I was planning to persevere with Humbrol 27001. But after the slightly unsatisfactory results of brush-painting, I decided it was time to get the airbrush out. As I understand it, Metal Cote is designed for airbrush use in the first place. It certainly doesn't need thinning. And airbrushing gives a beautiful (slightly dull) finish. Pity it doesn't stay on if you handle it.

You'll notice in this photo a bit of copper wire, which is threaded through a hole drilled in a tiny cube representing part of the interrupter mechanism. While the paint was drying, I moved it back and forth a couple of times to stop the drilled hole being clogged with paint. Later, a new piece of wire would be threaded through to represent the hydraulic lines leading to the guns. I had learnt the hard way that copper wire is reluctant to stick to plastic; so drilling location holes saves a lot of trouble and gluey mess.

The cylinder blocks on the original were apparently finished with a black vitreous enamel, according to my sources. I used Humbrol satin black for this. I initially mixed it too thick for the airbrush and got a pebbledash kind of effect; but enamel paints are very forgiving of this sort of hamfisted technique, and flow into a nice smooth finish.

When it was all dry, I tried it for fit, with the cylinders temporarily held in place:

 This revealed two problems:

1. The engine block was slightly too long and overhung the rudder bar. The rectangular block at the lower rear of the engine block had an oil pump mounted below it, with oil pipes feeding up to the engine, and this would foul the rudder bar if I added it.
2. There was something wrong about the way that the cylinder head covers protruded from the fuselage. I had made them from simple lengths of Evergreen half-round, knowing that this wasn't accurate but hoping for the best. Now I had to admit that they didn't really work, and needed to be taller for their width.

I decided to redo the cylinder blocks. As for the engine overhang, I argued to myself, I could leave off the oil pipework; nobody would actually see it. On the other hand, I could imagine aviation modelling historians of the future marvelling as they employed futuristic miniature cameras to probe the realistic interior of this tiny Dolphin, only to be sadly disappointed and embarrassed for my sake when they witnessed the mistake. And I didn't want to look bad in the eyes of futuristic aviation modelling historians, who would evidently have a lavish grant to do this sort of thing judging by their micro camera equipment, and would be planning to do it thoroughly. Possibly the grant would be obtained by linking their work to global warming, with the argument that priceless works of styrene were in danger of melting into small sad puddles because of the cold winters, or were going to float out to sea because they had been carelessly left 1 inch above the tideline. But I don't want to speculate on the ethics of futuristic aviation modelling historians, who cannot (yet) answer back.

I decided to do nothing for now about the engine. But I mulled over the steps that would theoretically ensue. It would involve removing the magnetos etc. from the rear, slicing a bit off the engine block, then rebuilding the magnetos etc. And to avoid a big silvery mess, it would be best to clean off the paint. Clearly that was going to be a lot of trouble for something that nobody apart from imaginary futuristic historians would ever see.

Engine having its bath

So I cleaned the engine in a bath of paint thinners, hacked off the magnetos etc., shortened the engine and redid the magnetos. As it happened, the thinners weakened the structure at the back, so that it more or less fell off, while leaving the rest unharmed. That was lucky.

I also redid the cylinder blocks, and spent a long time squaring them up, still lacking basic tools for planing true. While I was at it, I ripped out the wonky starboard ammunition feed and built a new one. But here I hit a baffling problem. I was sure that I had used 0.75mm strip the first time; but when I tried again I found that 0.75mm was too wide and 0.5mm strip was too narrow. And I had nothing in between. Baffling. So I ended up sanding down some 0.75mm strip, hardly believing what I was doing. It eventually came out much better though, so I was relieved about that.

Having built the new cylinder blocks and painted everything again and tried for fit again, it was time to assemble the engine. I used Revell non-liquid polystyrene cement for this, because it takes a few minutes to set and gives you time to check the cylinder alignment from all angles. I then left everything overnight to avoid squashing it out of true while trying to get it into place.

So, next day, did it fit? No it didn't. It was only now that the cylinders were exactly in place that I could know for certain whether it would all fit, and it didn't. It went in, but slightly skewed, because the port cylinder block was pressing against the fuselage side at the back. I ended up having to shave half a mm from the back of the cylinder head covers and the fuselage sides.

After this, now that I had taken the trouble to make space for the pipework, I tried to add it from a single piece of brass rod. It should feed up from a little disc-shaped pump below the engine to the rear of each cylinder block. But this proved overambitious. The brass rod was not flexible enough to go where I wanted it to go, and I removed a lot of paint by scratching and handling. This was quite demoralising. After brooding on the matter for a while, I decided to give up some of the more fancy detailing and just add the fuel pump and a short length of pipework that would be visible below the instrument panel and ammunition boxes. I had intended to add ignition leads from prepainted stretched sprue, but now felt disinclined to take so much trouble. The future historians with their questionable funding would take a pragmatic view.

But once I had retouched the damaged paintwork and added the somewhat compromised pipework, I started to feel more confident again, and decided that my engine could have some ignition leads after all. I guessed that the leads might have had some kind of braided insulation, and so painted the stretched sprue matt black. You can see the very same sprue excitingly drying in a previous post. I had already drilled very fine location holes in the cylinder blocks that were almost impossible to see even with a magnifying glass. The procedure was to glue lengths of stretched sprue into the location holes, wait a minute or two, then carefully bend each one at right angles. After this, I gathered them all up in a bundle on top of each corresponding magneto. This would have been a lot easier if I hadn't been using slow-drying superglue, that allowed the leads to keep springing away when I thought they were fixed.

Ignition leads

After all this, I could finally glue the engine into place. By the way, I had to do a bit of adjustment to get the height right, and ripped out the horizontal engine bearings a couple of times, but that's enough tedious detail for one post.

Here's how it looked now:

All that invisible detail was worth it, wasn't it?

Unfortunately, for any reader who's stayed this far as well as for me, that wasn't really the end of the story with the engine. When I happened to look at the whole thing from the front, after the glue had dried, I noticed that the engine was canted to one side. It was too late to realign it, but even if I could, this would throw the cylinder blocks out of alignment.  I realised that when gluing the cylinder heads in place, I must have got them slightly misaligned.

I started thinking about what I would need to do to rectify this situation, but this time I had to admit that it was plainly a mad and forlorn hope. Firstly, I would have to rip out the engine without being able to reach the points where it was glued to the framework, therefore running the risk of pulling the fragile framework to bits. Then I would have to break the cylinder heads off, probably causing further damage in the process.

The other alternative was to put it down to experience and finish the project as best I could with a wonky engine, with results that the futuristic historians would quickly pass over.

So what happened next? The by-now predictable answer is in the next post.