This winter I plan to get started on my two-stroke horizontally opposed engine projects. My quest is basically to see how simple an engine can be built and if simply scaling up model airplane engines (my main source of information) can be done. My goal this winter is to build a twin cylinder engine that produces 30hp and eventually build a 4-cylinder engine for use in an Ultralight and possibly larger engines in the future (the engines are made modular so adding additional cylinders should be easy). I plan to start by trying to build a 100cc twin cylinder steam engine and if successful move on to a 100cc gas engine, then a larger volume 30hp gas engine.
Construction of this engine is a little different from most engines. I plan to build the cylinder head, the cylinder and half the crankcase as one piece so the only bolts needed are to join the two halves together. Although this is harder to machine it should result in good power: weight. I plan to use as much commercial parts as possible. I plan to use 2-stoke piston (40mm diameter for the 100cc engine and 80-90mm diameter for the 30hp engine). I favour large bore short stroke engines as I think that is the easiest way to get enough power and have the engine breath properly. The engine I'm trying to build is similar to the McCulloch 4318 (nothing new). I've though about purchasing a McCulloch 4318 or a couple Chotia 460 to salvage parts and copy the timing but the way I plan to build the engine is different and results in lower primary (crank-case) compression so I'm not sure how transferable the timing of these engines would be so I decided to start with typical timing and do trial and error adjustments. For the intake I'm planning on using commercial reed valves and carburetor mounted on the crankcase. I had plan to use piston ported intake but though the reed valve was better since it give more equal fuel and oil distribution since I'm planning on using one carburetor for both cylinders. I plan to use model airplane/Hall sensor ignition initially.
I had originally planned to use a highly accurate wood model that is repeatedly dipped in wax to build a wax model. See the pictures at http://groups.yahoo.com/group/DIY_2S/. The idea was that with precision wax casting the only equipment need to build the engine is a drill press and a rigid hone. This method works nice in that the outer dimension of the wood model can be accurately made and then form the inner dimensions of the wax model/engine which are most important for the engine. One of the biggest problems with this method is that the heat fins need to be added to the wax model. I used compressed air to remove the wax model from the wood mould but had some problems getting the wax model off the wood mould and got frustrated so gave up on this method. I now plan to use lost foam casting and an engine lathe. I'll build a solid foam model of the engine and use it to cast a solid block of aluminium that is then machined. The biggest anticipated problem with this is getting a good enough bore for the engine to run. The problem is that the cylinder head and cylinder is one piece so the bore is closed on one end, which makes boring more difficult.
The crankshaft is a simple affair with no counter weights and is made from bar stock (pins) and 1/2" plate stock (arms). I plan to use industrial needle roller bearings and seals.
Another major undertaking and problem is the propeller. Since I want 30hp at 4000rpm I probably need a custom prop. I plan to make a composite propeller with hotwired high density PVC foam core that is then covered with fiberglass.