Peter Rooke starts a new restoration project, beginning with stripping the paint and making new wheels for a Fairbanks-Morse Plugoscillator engine.
The origins of the Fairbanks Morse Co. date back to 1823, when the company made cast iron plows and heating stoves. The early company, E&T Fairbanks, moved into producing scales, becoming a leading U.S. manufacturer. Employee Charles Hosmer Morse was instrumental in acquiring the Eclipse Wind Engine Co., and later became a partner in E&T Fairbanks, which became Fairbanks-Morse and Co. F-M started producing oil and naphtha engines in the 1890s before moving into kerosene and diesel engines, eventually becoming a dominant force in the fast-growing engine industry.
The Model Z was one of F-M’s most successful engines. First produced in 1916, more than a half a million Z engines were produced over the next 30 years. The Plugoscillator engine was made between June 1917 and 1919, but spark plug and high-tension magneto engines started to appear in 1918. This engine shows serial number 386013, produced in 1919.
The previous owner had painted the engine blue and red. These colors were not to my taste, so the engine would be stripped, cleaned and repainted green to get a closer match to its original color. Stripping the engine would also allow me to examine its parts more closely, fixing and refurbishing parts as necessary. There were one or two broken items, plus other replacement parts that did not look original.
The first step was to remove the igniter and magneto before taking off the flywheels. This meant using a scraper to remove the red paint from the ends of the crankshaft, finishing off with emery cloth before lightly oiling the clean surface. The flywheel clamp bolts were slackened and the keys were removed. One side came out easily, but the other proved to be a little more difficult and it was discovered it had some rust inside that effectively glued it to the flywheel. Eventually it gave, and with the flywheels removed the piston and crankshaft were then removed, along with the cam gear, pushrod and mixer.
A priority was to get the sub-base free, so that it could be used to check measurements for making a new cart. With the sub-base free, it was clear that the gas tank had been painted at some stage, and this paint was now peeling. Fortunately, the galvanizing under the peeling paint still looked good. Once the gas tank was removed the peeling paint was cleaned off with a scraper and then the inside of the tank was cleaned and flushed.
After researching carts for this engine, a plan was drawn up using the dimensions of an original cart on a friend’s 3 hp Z. The first stage was to make the four wheels, 12 inches in diameter with 2-inch-wide faces and five spokes. Some 2.5-inch-diameter steel bar was obtained for a bargain price from a friend who was clearing out his workshop.
The steel for the hub was cut to length, 2.5 inches, and then drilled through before opening up the hole with a boring tool to 1.35 inches inside diameter so the nominal 1-inch pipe axle would fit. The hub was then profiled to leave a raised centre section. After profiling, the hub was mounted on the piding tool so that holes for each of the five spokes could be drilled before being opened up with the mill to provide a recess to be filled with weld.
To prepare the rim of each wheel, some 2-inch-wide iron was cut to a length of 37.5 inches, resulting in a wheel nearly 12 inches in diameter once formed round. Before forming the strip into a rim, five 0.125-inch holes were drilled, spaced at intervals of 7.5 inches, the first holes being drilled 3.75 inches from one end. This was done so the join in the rim would not be near a spoke. Small holes were drilled at this stage to prevent weakening the metal and make it difficult to achieve a clean bend. The ends of the rim strips were also chamfered, ready for welding them together. The rim strips were too thick to fit my small rollers, so the vise and a bending jig were used, with a scrap piece of curved cast iron used to impart a smooth bend.
After the first sequence of bending, the rim was compared to an old wheel that was also 12 inches in diameter. A white crayon was used to mark the rim where it needed more or less curve and it was put back in the vise with the jig and adjusted accordingly.
After completing the forming of a rim, one end was held in the vise and pressure applied with one hand at the other to hold the ends together so that they could be tack welded together. Once satisfied with the alignment the ends were fully welded, the excess weld ground off to leave a smooth surface.
The profile of the wheels was again checked and one or two adjustments were made using the vise before opening up the five holes in each rim to 0.5 inch, the outer face side also being countersunk so it could be filled with the weld. All that remained was to make 20 spokes to join the four hubs to the rims. These were made longer than necessary to make it easier to fit the rim prior to welding.
To help align the component parts of the wheel for welding, a metal disc was screwed into the middle of a piece of board and a 12-inch-diameter circle was scribed from its center. Some pieces of 0.25-inch-thick flat steel taken from the stock rack were then placed under the rim to raise it up and hold it at the right height given the greater width of the hub.
The spokes were then inserted and the alignment of both the rim and hub checked by measuring the distance from the rim to the hole for the axle with a ruler. Minor adjustments were made by moving the rim, tapping it with a hammer to nudge it into position. The spokes were then held in place with some tack welds before removing the now-forming wheel from the jig. The extra length of the spokes sticking out from the rim were sawn off before completing the welding. Any surplus weld was then removed with a disc sander and grinding wheels to finish off the wheels.
When the four wheels had been completed, attention turned to making the axles and their supports. The bolsters to hold the rear axle were made from some 0.188-inch-thick steel strip that was 1.5 inches across at the widest point. A template of the shape of the bracket was marked out on a piece of heavy card stock and this was used to check progress while bending the metal.
The rear bolsters were formed by part cutting through the steel then widening this to create a “V” at the bend points. Steel plate 1.5 inches wide was used, with a 0.5-inch hole drilled first in the middle of each piece for the engine rear securing bolt. The two pieces of steel were profiled to taper them slightly at the lower bend so that the brackets would meet up with the tube to hold the axle.
The axle support tube was cut from some 1.7-inch-outside-diameter thick-wall tube, 1.35-inch-inside-diameter, and then welded to the bolster. To ensure the axle was correctly aligned, a support tube was tack welded to one bracket. The two brackets were then clamped to the workbench so they were square and the axle was inserted. A spirit level was used to ensure the axle was level with the workbench before the second bolster was tack welded in place. Measurements were then taken to check that everything was square and in alignment before the bolsters were fully welded. Any cuts made for bending were also welded to strengthen them. Finally, to complete the bolsters a flat piece of steel was welded to the underside of the support tubes, shaped with the grinder and then drilled and threaded for the 0.375-inch axle clamp bolt.
The bottom part of the axle pivot was made by welding a disc of steel to some steel tube bored out to fit over the axle. The underside of this steel ring was also reinforced with the addition of a small plate with a hole threaded though both this and the ring for a 0.375-inch clamp bolt to hold the axle in place.
A length of steel 2.5 inches wide was bent to form the support for the pivot, with a further piece cut 10.75 inches long as a support between the two pieces of angle iron. A template was first cut from thin plywood to check the profile of the bends. As before, the steel was partially cut through at the bend point with a cutting disc to create crisp bends without using a heavyweight press. To bend the steel, a long piece of 1-inch square steel was clamped to the steel plate to act as a lever and to keep the steel straight, the rest being held in the vice at the bend point. The steel was then bent to form, progress being checked against the template.
The bottom two bends of the pivot support were formed, then a hole was drilled for the pivot pin. Each end of the support was then bent flat to join the main rails, with any surplus steel cut off with a hacksaw. The fitting was next either clamped to the bench or held in the vice to keep its shape while a thin bead of weld was run along each of the four cuts at the bend points to strengthen them. A 2-inch-diameter washer was then made to fit over the pivot pin, with a smaller washer being used under the securing bolt.
The side rails for the cart were cut from 2-inch-wide angle iron, each piece being 30.5 inches long. The positions for the holes that would need to be drilled to secure the front pivot, rear bolster and the engine to the angle iron were marked, and then center punched before being drilled.
Wheel stops on the axles were made from four washers, one for each wheel, each held in place by split pins passing through holes drilled in the axles. Four more washers were made for the outer ends of the axles, these being wider before cutting a groove for the split pin.
To complete the cart, a metal handle was made from 0.5-inch-diameter iron rod made up from several pieces welded together to give the desired length. Rings for attaching either end of the handle to the cart were made, each with an internal diameter of approximately 1.75 inches, formed at the end by heating then bending around metal formers. The rings were then welded to the rest of the handle for an overall length of 45 inches.
The Plugoscillator is a low-tension magneto fitted on the igniter in a similar fashion to Webster igniter brackets. In the case of the Fairbanks, the mounting spigot to the engine is tapered, enabling the bracket to be held in position by a clamp without any need for a gasket.
A Sumter Number 14 magneto was fitted to these engines, this magneto being specifically made for the Plugoscillator. Around the time that this engine was made Sumter was taken over by but continued to manufacture magnetos as a pision of the Splitdorf Electrical Company.
The magneto was removed from the bracket and then bench tested to see if there was a spark. The magnetic pull was strong, but there was no evidence of any spark. The points did not appear to be closing properly, indicating a weak spring on the moving electrode arm. The magneto was then stripped and cleaned, and the points leveled with a small sharpening stone. The mica tube around the fixed electrode appeared good, but there was little insulation left around the washer and securing nut for the wire from the magneto. After renewing this insulation, the magneto was bench tested again and it produced a fat blue spark.
You can replace the insulator that the shaft runs through with mica tube and washers on the end, but it is an odd size so you need to start with a bigger, machine-able piece of mica and turn it down to size. Be careful if you take one apart because the old insulator may be brittle or already in pieces.
If the magneto needs to be checked, remove it from the bracket and use a cordless drill with a socket adapter on the magneto shaft/nut to run it. Connect a voltmeter from the igniter wire and to ground. With the drill on high you should see 18-24 volts DC if it is hot. If it’s much less than this then the magneto needs to be charged or rebuilt.
The fuel mixer appeared to have been rebuilt previously as it had a new needle and reservoir overflow valve. However, the needle knob was made from stainless steel and looked out of place. The original needle valve handle would have had a sculptured edge. To replicate this, a 0.125-inch-thick disc was cut from 1.5-inch-diameter steel, with 0.3125-inch-diameter hole then drilled through the center.
To get an accurate profile for the sculptured edge, the design was drawn using computer software, then printed and glued to the steel disc. The shaping was done by removing surplus metal with a small hacksaw followed by small files for final shaping. Using the lathe, the head on the needle was reduced in diameter to 0.3125 inch to match a hole drilled in the disc. The new disc was then brazed on, enabling the old needle to be retained.
The reservoir overflow valve is closed when you fill the mixer with fuel to start the engine before switching on the fuel pump to draw fuel from the tank. A replacement head had been made with bright stainless steel, which I wanted to “tone down” so that it looked more like ordinary steel. I did this by heating the head until it was red hot, then quenching it in old oil to it a dull and patchy black finish.
The mixer itself was stripped and all parts cleaned before painting. Attention was paid to the pump ram and the check ball and its seat to ensure they all fitted and worked as intended. It was later found that there was a very coarse adjustment of the needle valve, with a fraction of a turn being sufficient to fully open it. It was decided to lap the valve needle seat, as the needle itself looked good, with a smooth, gradual taper.
To do this, the needle was held in the lathe chuck so that a dial indicator could be used to measure its taper, with the compound slide then adjusted to this setting. A piece of steel was then turned to this taper and then used with grinding paste to clean up the needle seat. After a period of grinding the seat was cleaned out with kerosene and the needle fitted. By blowing through the inlet nozzle it was possible to tell that the needle now had a range of adjustment rather than being either open or closed.
When refitting the fuel pipes PTFE gas tape was used to seal the threads, which were all well-worn. When fitting the mixer to the cylinder head, make sure the jet pipe does not obstruct the movement of the throttle plate. This engine had mounting studs fitted, making it impossible to fit the mixer to the cylinder head without bending the jet pipe and or/the fuel feed/overflow pipe when trying to locate them in the fuel tank. I removed the studs and made cap-head screws to be used in their place. One stud proved particularly stubborn to move, even after heating, so I welded a nut to the stud and used a wrench to unscrew it.