Peter Rooke begins the restoration of a John Smyth engine.
1914 John Smyth
John M. Smyth Co. General Merchandise of Chicago, Illinois, sold all manner of goods for home and farm via mail order, including gas engines. This catalog business was not as big as Sears or Wards, but it was predominant in the Chicago area.
The engines sold under the John Smyth name were manufactured by the Waterloo Gas Engine Company, Waterloo, Iowa, which supplied engines to some 67 other companies including Eaton, Jackson, Majestic and Sandow, to name but a few. When Waterloo was sold to John Deere in 1918, supplies of these contract engines ceased.
The Waterloo Gas Engine Co. added serial numbers to the engines they produced, and for some brands of contract engines a letter prefix was added. Plain serial numbers were used on the Smyth engines, and this engine, serial number 106,203, would have been manufactured in 1914, when numbers ran from 91,673 to 113,017.
This engine weighed too much for the usual mobile workbench, so it was stripped before being lifted onto it. The first step was to remove the cylinder head. After removing the four nuts securing the head, and the rocker arm pivot, the head was still stuck tight. The piston and connecting rod were then removed, along with the governor and pushrod. To try and knock the head free, some 3-inch square lumber was inserted into the cylinder from the back and hit with a dead-blow hammer, but without success. However, a 3-foot long length of 3-inch diameter aluminum acquired for another project was inserted in the bore, and used like a battering ram the head eventually started to move.
The flywheel pulley had to be removed next, but the securing bolt had no head. Fortunately, there was some thread left, so the appropriate nut was screwed on and then welded to the remains of the bolt. After applying penetrating fluid the bolt eventually yielded and was removed.
The visible end of the crankshaft was lightly sanded with emery cloth to remove the worst visible rust, then penetrating fluid was applied. The hydraulic puller was set up, but with pressure applied there was no sign of movement. The puller was left under tension, but after a couple of days, during which more penetrating fluid was applied and the pulley hub was occasionally shocked with a hammer blow, the pulley started to move and was then gradually eased off.
Examining the pulley-side flywheel key, it was clear the head had been filed off at some stage. A puller was made using some angle iron welded to the remains of the key. By tightening a bolt threaded through the angle iron and pushing against the flywheel, the key was moved. The crankshaft was cleaned and oiled before a block of wood was put under the crankshaft so that the flywheel could be bumped to turn it on the crankshaft to start its removal. Once the flywheel had slid near the end of the crankshaft, tapered blocks of wood were put between the flywheel rim and the floor to support the heavy flywheel. A curved puller made short work of removing the key for the other flywheel.
I decided to remove the water hopper to reduce weight, and also provide another clean surface. This might be needed when the cylinder was bored out to fit a sleeve due to deep rust pitting. Removal was not easy as the nuts and bolts had rusted and there was little room to work. One method would have been to use a hacksaw and carefully cut through the thick gasket and the bolt stems. The danger with this is the mating surfaces might be scored, so a nut splitter was used instead. It took time, but the nuts were eventually removed.
The collar on the crankshaft for the governor was rusted in place so the hydraulic puller was used to ease it off, with a bearing puller attachment set against the collar so the pressure on the collar’s thin walls was evenly spread. The same method was used to remove the timing gear wheel and key.
Once the engine had been stripped to its basic components it was taken outside and lightly pressure washed, taking care that any remaining paint was not destroyed. Full pressure was used inside the water jacket and around the cylinder to blast out any accumulated dirt.
The big-end bearing on the connecting rod was loose even with all shims removed. The grooves to channel grease across the bearing were only faintly visible. It would need to be rebabbited. To begin, the connecting rod was held over a lead pot and the old bearing heated with a gas torch and melted out. After cooling the bearing shell was cleaned, especially the four anchor holes to secure the bearing metal.
A 0.375-inch length of 2.5-inch diameter steel was trued before drilling a 1-inch hole in the middle and cutting a recess to form the outside edges of the bearing. The last 0.25 inches of some 1.5-inch diameter steel was then turned down to 1-inch to fit inside the cap.
Two temporary spacer bars were cut from 0.1875-inch square bar to take the place of the shims. These were inserted before tightening the bearing cap using temporary bolts to hold them in position. The end cap and the core were assembled and set to ensure the core was in the middle of the bearing shell. Clamps were used to hold the core and plate in position until fireclay was applied to seal and hold them in position. This was heated with the gas torch to set the fireclay.
Two fire bricks were then put on the workbench to support the connecting rod. Fireclay was used to seal all the gaps around the upper part of the bearing shell and build a dam on the top to contain the poured bearing metal. The gas torch was again used to dry out the fireclay and remove all moisture.
The lead pot was then heated and some new bearing metal melted, with heat applied until the metal singed a pine spill. The core of the mold was had already been blacked using a candle and was put in place before heating the mold and the connecting rod.
A cloth was prepared by first soaking it in water then wringing it out so it was damp and not dripping wet. This could be used to seal any holes in the mold and stop any molten metal spilling across the workbench. The hot metal was then poured and left to cool naturally before breaking off the fireclay, removing the mold plate and pushing out the core.
After pouring, a hacksaw was used to cut between the bearing halves, with surplus bearing metal filed off to leave a smooth surface for new shims. Two sets of shims were cut from assorted sizes of stock to 0.15-inch and securely clamped in place.
A short length of 1-inch diameter steel rod was drilled through with the ends faced off so it could be clamped upright to the milling table. The connecting rod little or piston end was then fixed to this and the rod itself was secured to the table, packed with steel bar and shims so that it was not twisted and the bearing metal was clear of the table.
After centering the boring head over the new bearing, the metal was bored out to the diameter of the big-end before trimming the surplus bearing metal on the sides. The bearing was then fitted to the crankshaft, where engineers blue had already been applied to the journal.
A scraper was used to remove any high spots and to shape the edges. This was repeated until there was a good fit, then the grease hole was drilled out and a “V” groove was cut in the metal to just short of the edges to disperse grease across the bearing. Finally, the shims were adjusted so the connecting rod turned freely when the securing bolts were fully tightened.
The pre-purchase photographs showed the gear wheels looked reasonable, but on turning the flywheels extreme wear was found in a short section of the eccentric gear, the teeth worn to a knife edge halfway around the crank gear.
It would be easy to replace the crank gear by making a new gear, but the eccentric gear would have to be repaired. The gears were found to be 10 diametral pitch (the number of teeth per inch) with the pressure angle believed to be 14.5 degrees. I did not have the gear cutters to cut the teeth, but a tool dealer was found who had some used ones in reasonable condition at a fair price. Unfortunately, these were different shaft diameters to the holders in the workshop so the first task was to make a new mandrel.
Cutting the crank gear was a straightforward but slow process, using the small mill with a makeshift dividing setup. Clearances and table travel were tight and could only be arranged one way, so it was all hand feed as there was no auto feed in that direction. Yet another mandrel was needed to hold the gear blank. Once the gear was finished the gear wheel was held in the lathe chuck so that the keyway could be cut.
Attention then turned to the eccentric gear. The hole for the mounting mandrel was badly worn, so the gear was centered on the faceplate and the mounting hole was bored out 0.0625-inch oversize so the gear could be accurately mounted on a mandrel for machining. This also meant that a new shaft had to be made to mount the gear on the engine. This was made with a step-up to the diameter of the gear so that no adjustment would have to be made to the pivot post of the governor latch that is also mounted on the same mandrel.
One option considered to repair the gear involved cutting out the damaged section and to dovetail in a shaped piece of cast iron, before recutting the gear. However, it was decided to first try welding up the damaged teeth and then recut them.
This should be possible if 99 percent pure nickel rods were used, welding in short bursts and allowing the weld to cool before continuing. Before starting, a test blob of weld was applied to the gear to ensure it would remain soft and could be machined and there would be no adverse reaction with the cast iron. After passing this test, the bad teeth were carefully welded and at the same time the badly worn area of the cam was built up.
After welding was completed, the gear was mounted on the lathe and the weld trimmed to the outside dimensions of the gear.
The dividing plates for the rotary table did not include a 64 teeth spacing, which presented a small problem. To get around this, the 32 division plate was used and the starting point for counting set by fitting the cutter in a “good”gear recess. Sticking with this setting the alternate teeth were cut before resetting the start point to another original gear recess and cutting the remainder.
Welding in stages at low amperage, there was concern about the degree of weld penetration. A short length of steel was therefore inserted between two new teeth and twisted to put the new teeth under pressure, but they held firm and should be resilient enough for the light load of operating the pushrod.
The pushrod was worn, bent and twisted and had already been welded in several places. Rather than fix these problems, it was simpler to cut a new piece of stock steel 0.375-inch by 0.750-inch and fit the various attachments.
The follower roller was removed first and it was immediately clear that its screw needed replacing as it already been given a makeshift repair using some braze. A new screw was machined on the lathe.
Measuring against the old pushrod, holes were drilled in the new one to match the original so that the catch plate, roller and spring retaining claw could be refitted. The new pushrod was a very sloppy fit in its rear guide, with the retaining plate having been nearly cut through with wear. The pushrod would still work, but it would twist around in the guide and wear quicker so the guide was repaired. The guide was set up in the mill table and the top and bottom of the slot were trued up with the bare minimum taken off the side to clean it up. The same amount was taken off the side for the bearing plate so the new pushrod was an easy sliding fit when the new plate was fitted. To compensate for the metal removed from the top and bottom of the guide a strip of steel was cut and shaped to shim it. Finally, a new cover plate was cut from steel sheet 0.0625-inch thick.
The igniter contact points had been worn out long ago, and the main spring was broken. On stripping the igniter it was also clear that the moveable electrode shaft required replacement as it was both rusted and bent. The hole in the igniter for the moving electrode shaft was not badly worn and did not need
bushing, even after cleaning out any carbon. A piece of 0.375-inch diameter silver steel was trimmed to length before getting a small block of steel that was drilled with a similar diameter hole. There then followed a period of sawing and grinding to produce a similar shape to the original igniter head before brazing the two together. When cool, a slight taper was cut on the head on the side that bears on the igniter before using grinding paste to evenly match the two. Finally, a 0.0938-inch cross-hole was drilled for the peg to hold the torsion spring.
A replacement torsion spring had to be made. A length of alloy was tapered to match the old spring, but some 0.15 inches smaller diameter. Holes and slots were drilled to hold the spring wire before using the lathe to wind the spring. Tension on the feeding wire was maintained by passing it through a brass block mounted on the tool post with a screw pressing on it. When the number of turns had been reached, the end was bent round a temporary post before placing the spring in a simmering oven overnight to set.
New contacts were made from an old nail and the remains of the worn contacts were drilled out before fitting the replacements through the holes and peening over the shaft. The igniter was assembled with new mica washers and tension spring as well as a new fahnestock clip. The points were adjusted to give a gap of 0.625-inch when open, a small screw on the right side of the igniter used to adjust the gap.