Bentley fabrication 1

This article follows on from the short series of articles by Mike Sayers on machining ‘castings’ for his 1:3 scale 3 Litre Bentley starting here. He now turns attention to fabricating the large cylinder block.

The following two photos show a 3 Litre cylinder block, dragged out of a friend’s garage purely to illustrate what had to be made. It looks grubby, but it is about 100 years old, and it had a very hard life. It was in a racing Bentley in the late 1950s / 1960s, and had been tuned up to a high degree.

The thickness of the foot , shows that at least 1/4” has been machined off the base to raise the compression, and a number of pieces have been welded in the side to increase the swing of the con-rod (because it had a long stroke crank in it). The bores have also been opened out.

It is a miracle that it has survived.

This is the item that had to be recreated in its standard form as a 1:3 scale cylinder block. It has been the component that has given the most head scratching and self-doubt of any of the components made.

The previous drawing is a vertical section through the Number 1 cylinder, and shows what needs to be recreated. As can be seen from the first two photos, the block is ‘open decked’ all the way around, with water plates screwed on, so getting inside was fairly easy. The previous components ‘ were fairly complex shapes on the outside; this one has the disadvantage of being a complex shape on the inside.

The valves are angled as seen, which creates a problem because you can’t get a cutter up the bore at the correct angle. The valves can’t even be removed without removing the guides, as when they are pulled down, the valve hits the side of the cylinder wall. Every time a valve is removed or replaced, the guide has to be pulled out beforehand. It’s a really tedious process. If the valves had been vertical, the machining could have been carried out from below.

This sketch shows the projected area across the valve heads in relation to the inner bore of the cylinder. There are actually four valves per cylinder, and the section shows the valve heads ‘overlapping’ the inner bore. In fact, as they are tilted, they miss the cylinder wall when opening. But what this shows is how difficult it is to get in and machine the valves.

Why was it made like this? In the 1920s, this was the standard arrangement for high performance engines. It is a carry-over from the Edwardian racing cars and World War I in-line water cooled aircraft engines. It is quite a light arrangement.

They also had the benefit of very skilled moulders and foundrymen to produce these complicated castings. A lot of these skills have been lost.

The original idea was to try and make it out of a block of cast iron. The outside was fairly straightforward to machine, but how to machine the inside? The bore is 28mm and is 80 mm deep. Getting cutters down seemed impossible.

Four ideas were considered:

• Machining from solid, which was a failure.

• Make a detachable head. The whole component, however, is designed in such a way that there is no way bolts can go through it, and besides, by doing that the final component would not be representative of the original.

• Make the liners separate and screw them into the top of the head. That idea was discarded.

• The last idea, which was considered the only way to go, was to prefabricate the whole component with pre-machined subcomponents of cast iron.

That was the method finally chosen. It took nearly three years to decide, during which the rest of the engine was being made. The crankcase and the top assembly had been made, but doubts about the cylinder block nearly ‘killed’ the project.

These 11 subcomponents make up the cylinder block:

•The cylinder head (red).
•The four cylinder barrels (black).
•The bolted down foot (green).
•The base of the water jacket (dark blue).
•Four corner pieces for the block.

•

The first subcomponents to be made were the cylinder barrels. These are fairly short tubes machined from continuous cast bar. The inside bore was honed to a finished size, and the outside was machined. The outside is parallel over approximately half of its length with a 1° taper over the rest. Triangular features are incorporated at the bottom which made production a little awkward. The top of the barrel had to be a fixed size to spigot into the head, and the bottom had to be a fixed size to spigot into the base plate. Both these fits needed to be reasonably tight.

This photo shows the bottom section, with a tool post milling head finishing off the tapered section around the triangular features.

This shows the machining of the base plate of the block, shown green. All the bolt holes have been drilled, and the holes for the liners have been bored, with a decent fit onto the liners.

Beginning the machining of the inlet side of the cylinder head. Note the block of cast iron is located against the parallel bar and the end stop to allow repositioning. This is a solid block of cast iron cut over length. There are slots cut both ways at each end, so that whichever face it sits on, there would be some gripping faces to clamp onto. This leaves the whole of the top surface clear for machining operations.

The parallel bar which spigots into the table provides a parallel edge to mount against. The end stop, which the job is pushed up against, allows a zero datum to be set up as described in the previous talk. The zero datum is in the centre so that whichever face the component is turned around on, the datum is fixed and co-ordinate drilling of details can be carried out on each face.

The spark plug holes have been bored, and the ‘siamesed’ inlet ports machined, each of which serves two cylinders and four separate ports to the valves. At this stage, there is still work to do on the inlet ports. When the ports for the valves are drilled, they will be deepened, opened up and blended together.

Setting up to drill and ream the valve guide holes using a roller of the same radius as the tappet movement.

A challenging setup because if the valve guide holes are drilled from the combustion chamber side, there is no datum for drilling. The original drawings show the positioning of the holes relative to datum point A, which is dimensioned to the base as shown. The valves are canted over at 15°15’ to the vertical. When it comes to machining though, the difficulty is spatially determining point A, so a solution had to be found.

What was known was where the rocker shafts were, and how high they were. So these became fixed points. The radius of operation of the rockers were also known. These facts allowed the valves to be positioned from the opposite side to the combustion chamber.

The tooling blocks shown in the photo above, were made to bolt to the top of the head using the tapped holes which will eventually secure the cam chest. The rocker shafts can then be positioned exactly where they will be when the cam chest is bolted on. The roller represents the radius that the rocker works at.

The table has been set at 15°15’ (the angle of the valve to the vertical) using a sine bar. The datum point is known because the block is set up against the end stop, so the valve positions are known longitudinally.

By using the 6mm round rod in the spindle, the rod can touch on the outer radius of the roller (at X), the roller removed, and then the spindle moved back half the diameter of the rod to position it directly over the valve.

Once the spindle has been set up, all the tooling can be removed. The table is now positioned longitudinally over the first valve position, which is then drilled and reamed. The spindle is then moved along to the next valve position which is drilled and reamed, until all eight valves on that side of the block are completed.

The set-up is then reversed with the table tilted the other way, and the roller is placed in the other side of the tooling blocks. The machining of the eight valves on the other side of the block are then completed in the same manner as the first eight.

The whole tooling set up ensures that the tip of the valve on the top is exactly where it should be when the rocker hits it, and the valve head is exactly where it should be inside the combustion chamber. This tooling system worked very successfully.

Lapping in the valves. The valves have been produced accurately relative to the combustion chamber and the spark plugs. All the inside of the combustion chamber is finished, with the valves ground in, and the seats checked before assembly.

Trial assembly of the base, barrels and the bottom of the water jacket.

A reasonably tight spigot fit was preferred into the base plate, and into the cylinder head at the top. The intention was to have a self-jigging assembly, which once put together, would keep itself in alignment. This reasonably tight fit, however, is against silver soldering convention.

Some test pieces were, therefore, made in cast iron, with a similar tight fit, and silver soldered together. These test pieces were then sawn in half and polished. They were then inspected for full penetration. Full penetration was achieved.

Here the combustion head has been added for a trial fit.

Continued in part two here