Cylinder Heads – 12/08/14
Building the Prototype AMD Series Engines
The blanks for the cylinder heads were all cut from the bar stock and machined square and to size some time ago. The cylinder heads are the most complex components and the most difficult to machine, with little scope for error.
I am making SOHC and DOHC versions of both the 5cc and 10cc singles, but I am making two interchangeable cylinder heads for the 20cc in-line 4, one SOHC and one DOHC.
On completion of the blanks, the first job was to machine the underside of the cylinder heads and to drill the bolt holes, which I did using the coordinate method with the DRO. There was also a job for the lathe with machining the recess in the under side.
The main profile of the valve banks and other features was then machined on the mill. There was quite a bit of metal to remove and it got a bit tedious, but all went well.
To machine the angled features on the cylinder heads, a fixture was made up. The rotary table was mounted vertically on the mill, with an angle plate bolted to the rotary table to form a sort of shelf. The rotary table was zeroed and the angle plate set roughly horizontal in the Y axis before tightening the mounting bolts. A slab of aluminium was bolted to the angle plate and a strip of aluminium bolted to the slab, to form a fence aligned with the X axis. The aluminium slab was skimmed flat with an end mill and the fence skimmed to align with the X axis. Bolting the cylinder heads to this fixture then allowed the angles to be set accurately using the rotary table scale and handwheel dial.
As a reminder, I am building five prototype engines as follows, to test the design in its basic configurations:
Engine 1. AMD5S, air cooled SOHC, glow ignition, lubricated by oil mixed with fuel
- Engine 2. AMD5S, water cooled, DOHC, spark ignition, oil pump
- Engine 3. AMD10S, air cooled SOHC, glow ignition, lubricated by oil mixed with fuel
- Engine 4. AMD10S, water cooled, DOHC, spark ignition, oil pump
- Engine 5. AMD20IL4, water cooled, DOHC, ignition method TBD.