In this second installment of the series, Peter Rooke builds a cart for his Economy engine with help from a Glenn Karch article.
Peter Rooke's circa-1923 Hercules Model F engine.
This is part 2 of Peter Rooke's series on restoring a circa-1923 Hercules Economy Model F engine. Start at the beginning with part 1.
Looking for information about the carts used on Economy engines, I found Glenn Karch’s article in the May/June 2003 issue of Gas Engine Magazine about the styles of carts fitted to these engines. This article mentioned that carts made from 1920 to 1928 had a frame 26 inches long, axles 18 inches wide and wheels 9 inches in diameter, with rims 2 inches wide front and rear. The front wheels do not turn under the cart frame. This information, together with photographs found online, enabled me to draw up plans for a cart that would be a fairly close replica to an original.
The two rails of the cart were cut from angle iron 1.5 inches wide salvaged from a scrapped axle. These fit the bill after welding short lengths together and filling in unwanted bolt holes. The axle rods were cut from some 3/4-inch nominal pipe giving a true OD of 1.05 inches.
Starting with the rear axle, angle iron was cut to three lengths before drilling holes for the axle rod. A section of the angle near the corner on the uprights was removed and any edges to be welded were chamfered to give a larger surface for the weld to bond to. The dimensions of the rear axle assembly were marked out on some scrap board as a guide when welding. With the axle rod put in place to ensure alignment, the three sections were clamped to the board and the bench to hold them firmly in position for welding.
To produce the rounded top corners for the axle bracket, flat 0.25-inch thick bar was bent to shape. First, a piece of scrap rod turned to the diameter of the desired bend was welded to a piece of angle iron. The end of the steel bar was heated red hot with a propane torch, then bent to shape using this temporary jig. Once cool, it was cut to fit. The corners were welded together, then cleaned up with a grinder.
To make the wheels I made four rims 9 inches OD and 2 inches wide. Fortunately, I had a piece of an old lamp post with a 0.25-inch wall thickness that would be perfect. This was set up on blocks and cut into sections using a disc grinder, cutting slightly oversize so the rims could be trimmed to more accurate dimensions on the lathe.
The spokes and hub for each wheel can be made in two pieces. The hub was made by first drilling, then boring a hole for the axle rod in some 2-inch diameter steel. The hub was left on the bar so that a tidying cut could be made before both ends were tapered down to a diameter of 1.5 inches.
Some 0.25-inch thick, 8-inch square plate was used for the wheel spokes, the gaps for the necessary 8.5-inch finished diameter filled with off cuts. To start, the center point was marked and a 0.25-inch hole drilled. The rotary table was fixed on the milling machine. One of the central pins for it was 0.25-inch diameter. Using this would provide a consistent and accurate locating stub for each plate.
A digital readout fitted on the mill was zeroed on the center point of the plate and then the unwanted metal between the spokes was cut out using an end mill. Coordinate points had already been calculated using a plan drawn on the computer, but having said this, it might have been quicker and only slightly less accurate to just cut within lines scribed on the metal.
Once the spokes had been cut out the outer diameter was cut and finally the central hole for the hub of the wheel. This left four flats that had to be built up to reach the 8.5 inches diameter. Scrap pieces were used, being welded in place. The spokes were then cleaned up with a grinder and files, first leveling out the weld, then giving clean edges to the spokes before they were partially rounded.
To fix the hubs to the spokes, an axle rod was held level in a vise, with the spokes pressed onto the hub and the hub slipped over the axle. This was then spun by hand and eyeballed to check that it ran true. After the odd tap as necessary with a hammer this was achieved, and then some spots of weld applied to hold it in place before a final check. Leaving the wheel on the axle made it easy to apply the final weld. To finish each wheel the rims were pressed in place and after measuring and adjustment to run true the rims were welded on. All that remained was to grind off surplus weld to end up with a completed wheel.
The front axle was the most complicated to make. This was left until the end as it was then possible to double-check that the final dimension of the wheels and axles were the same as on the plan. Again, a scrap piece of board was marked up with the outline of the pivoting axle bracket. The original was made in a “U” shape, so some 0.625-inch thick bar was used, with the center part or channel removed by milling.
The first step was to cut the bar to size, cutting at angles as necessary to get the shape needed. After a trial fit to the template to check the accuracy of the cutting, the central channel was milled away, leaving a 0.275-inch wall on each side. The two holes for the axle rod were cut out once it was easy to hold the flat section in the vise. A 0.5-inch hole was also drilled for the axle pivot pin.
The five pieces used for the axle bracket were aligned to the template and securely clamped to the bench after fitting the axle rod, having already prepared the edges for welding. Once more a grinder was used to tidy up weld and remove sharp corners. The final part needed was the front cross member to support the axle and a pivot pin. The cross member was made from some 0.375-inch thick steel and a further piece of steel 1.5 inches in diameter was welded to both it and the 0.5-inch pivot pin that goes through it.
The handle for the cart was made from some leftover 0.5-inch steel. This was not long enough to make the handle from one piece, and this was probably just as well as it is far easier to make a single rather than multiple bends. A temporary jig was made and used to form the two rings in the side arms where they would fit around the axle rod. Once the rings had been formed the arms were trimmed shorter than the length of the handle. The handle section was then bent 90 degrees at both ends, followed by grinding a taper to the ends of the handle and the side arms. It was necessary to carefully set up these parts for welding, one side at a time. Clamps and scrap steel bar were used to keep pieces in perfect alignment.
Regular bolts were used to assemble the cart in order to check fit and alignment, but it really needed square-headed fixings, and the easiest way to get these was to make them. Some round steel was threaded 7/16-inch UNC and square bar was drilled and tapped to the same thread. Sections of this square bar were then sawn off to provide heads for the threaded rods and also for the nuts. One end of the threaded rod was then tapered before fitting in a head and welding together, the taper providing an extra space for the weld. The surplus weld was then ground flat.
After drilling holes to mount the engine and 0.1875-inch holes in the axles for the split pins, the cart was assembled, the final task being to apply a spot of weld to each axle to hold it in place on its support.
Over the years, Economy engines were fitted with slightly different styles of filler spout, ranging from a simple pipe with a cap to flip-top lids to a sliding lid under spring tension. From the 1920s, a thin flip-top lid was fitted, but access to one of the earlier models with a sliding lid meant this style was copied.
A short length of steel 1.75 inches in diameter left from an abandoned project was in the scrap bin. It was already partially tapered and had other machining marks on it, but these would be removed when shaping it, tapering the body of the filler.
The inside was bored out with a taper starting at 1.125 inches wide at the top, reducing to approximately 0.5-inch. The outside was turned, first to a diameter of 1.625 inches. Next, the top was turned down to 1.375 inches to leave a lip, 0.125-inch wide and deep. This was then tapered to approximately 1-inch at the bottom, the overall length of this piece being 2.35 inches. For a connecting stub, a short length of steel was drilled with a 0.578-inch hole ready for tapping 3/8-inch NPT for the pipe nipple. This was turned to a diameter of 1.05-inch before reducing to 0.5-inch to create the connecting stub. The body of the filler was then held in the 4-jaw chuck so a 0.5-inch hole could be bored to seat the short stub. Boring was used as a drill might wander down the tapered side of the filler. The end of the stub was then shaped with a small grinder to give a smooth channel for the fuel to flow through the filler. The two pieces were then welded together before final shaping.
To make the projection at the rear of the spout for the pivot pin for the sliding top, a 0.8-inch length of 0.5-inch steel was drilled and then tapped with a 10 x 24 thread. This was then welded to the filler spout and filed to shape. The sliding lid was made from steel 0.125-inch thick. A bead of weld was applied to the underside and then filed to create an indexing mark to locate in the recess cut in the top of the filler spout. All that remained was to make a threaded pin to hold the spring that presses against the lid and fit it all together.
Having obtained some photographs and measurements for a fuel tank, a piece of 22-gauge sheet metal was cut to size. This would be formed into the main body of the tank, measuring 14 inches long, 5.5 inches wide and 3.5 inches deep. The sheet was cut with tin snips, and then the cut edges were clamped between two lengths of straight steel bar, the surplus beyond the cut line sticking out so that it could be trimmed by running a file along the steel bars. Once trimmed the sheet was marked out to show the bending points. Layout blue was first applied to the general area of each bend and then a rule and scribe used to make the line.
The straight edge bars were also used to hold the sheet when it was folded over along the bend line, a small polishing hammer being used to create a clean corner after initial bending by hand. A smoother finish could have been obtained if a large head leather or wooded mallet had been used. The final bend was a short tab of 0.5-inch for the soldered joint. Before bending, the two contact surfaces were cleaned back to clean metal, flux was brushed on and both surfaces, heated before applying a thin coat of solder to tin them. After bending, flux was brushed into the joint before it was clamped together and heat applied to melt the solder.
The position of the inlet pipe, vent tube and outlet pipe were marked and the holes for the threaded fittings were drilled and reamed to size. The threaded fittings were made from brass to accept 1/4-inch and 3/8-inch NPT pipe nipples. It is possible to drain this tank from the ‘T’ fitted to the outlet, but at this stage the decision had not been made whether or not to paint the engine and/or cart. Given the effect of modern fuel on enamel paint it was felt safer to also fit a drain hole in the underside of the tank so there would be no chance of splashing the paint-work if this were used.
The end caps for the tank were made from 22-gauge steel. They were first cut to overall size, marking out the bends. The four corners were cut away ready for bending inside. The corners were cut with a saw to make sure they were not overcut, easily done with tin snips. Fortunately, the end caps would fit inside my large vise, so this was used to bend and then hold the end cap for the second bends that wrap around the tank. These were made using the hammer. If a large vise is not available blocks of steel can be used as formers.
Once the end caps had the second row of bends completed they were trial fitted and then the metal was again cleaned. The tabs were tinned where they would touch both the inside and outside of the tank. The walls of the tank that would touch the end-cap tabs were also cleaned and tinned. An end cap was fitted in position, after flux had been applied, then heat was applied to solder it to the inside tank wall. Once cool, the rest of the tab was bent over, fluxed then soldered. A short length of brass tube was soldered to the top of the tank as a vent tube before the tank was cleaned of surplus solder and flux and given a coat of galvanizing paint. The nipples and other fittings were then assembled to test their fit.
On this engine there were four projecting lugs cast inside the engine base. By wrapping thin wire around each lug, the tank could be held in place, this wire coming down the side of the tank, then being wrapped around a pair of 0.25-inch thick metal rods, one toward the front, the other toward the rear. Both ends of these rods were given a sharp bend to stop the wire possibly slipping off.