Hercules Gas Engine Restoration

Peter Rooke tackles a circa-1923 Hercules Economy Model F gas engine.

Hercules engine

The circa-1923 Hercules Economy Model F.

Photo by Peter Rooke

Content Tools

This is part 1 of Peter Rooke's series on restoring a circa-1923 Hercules Economy Model F engine. Continue reading in part 2.

Hercules background

Another tired old engine has joined my collection, a circa-1923 Hercules Economy Model F, serial number 290,884. The Economy engines were first made in Sparta, Michigan, by the Holm Machine and Manufacturing Co. for Sears, Roebuck & Co. By 1912 Holm could not keep up with demand. The Hercules Gas Engine Co. was formed that year, having purchased Holm and started work building a new factory in Evansville, Indiana. The first Hercules produced engines were sold in January 1914 and were known as the Model D. The Economy engine produced for Sears differed slightly in design to the Hercules, having a more rounded lip to the water hopper, no crank guard and being painted red rather than Hercules green.

The Model E was popular, with some 220,000 engines sold. It was followed by the Model F, which was produced from 1921 to 1923. An estimated 25,000 Model Fs were produced that incorporated design modifications including the supply of Webster 1A and 2C magnetos. These magnetos did not prove popular, and many engines were later converted to use either Wico EK high-tension magnetos or fitted with the older Webster oscillating magneto. Looking at what was left of the trip finger clamp on the pushrod on my engine, it would appear it had been converted to run using a Wico EK.


From looking at photographs I received before purchasing the engine, I knew numerous parts were missing, including the muffler, igniter, magneto, rocker arm, fuel tank and fuel filler spout. There was a large toothed sprocket fitted to the crankshaft, and I was advised that this probably was used in a cement mixer. Hercules sold engines to Jaeger to use on cement mixers, so before doing anything I carefully examined the side of the engine casting to see if I could find any holes where another nameplate might have been fitted. There were none.

The entire engine was covered in rust and there was some bad pitting to the flywheels. Grease around the crankshaft bearings had protected the paint in that area and there were specks of red paint still showing. That suggested this was an Economy engine.

After using the hoist to get the engine on the workbench, the first task was to remove the sprocket and examine the crack at one end of the flywheel hub. The sprocket proved to be stubbornly in place. After applying plenty of lubrication, a hydraulic puller was used to remove it.

Once the sprocket had been removed it was possible to fully examine the damaged hub. It appeared that a piece had broken off the hub – perhaps at one time someone tried to put a wedge just in the edge of the split hub rather than the center to open it a fraction to make it easier to remove the flywheel.

After removing the clamp bolt and cleaning then oiling the crankshaft, the flywheel was removed. At first it appeared to be stuck solid, so the hydraulic puller was again used, and gradually the flywheel started to move. However, the broken part of the hub started to lift up as the flywheel was pulled and eventually broke away from the hub. Once the flywheel was off, I could see the reason why: An attempt had been made to pin the broken piece to the rest of the hub, but the drill had met a blowhole, a void caused by air trapped in the casting. To fix the problem, a hole had been drilled through the broken piece into the crankshaft and a pin inserted, fixing the broken piece to the crankshaft rather than the flywheel hub.

The piece broken off the hub did not materially affect the integrity of the hub as it was a split hub with a clamp bolt. If it had been a solid hub, I would not have attempted a repair and would have used the original as a pattern to get a new flywheel cast. It would be virtually impossible to source a replacement here in the U.K.

The first task was to thoroughly clean the hub and the piece that had broken off. This part was not too strong, with two holes in it, and there was the large blowhole in the hub itself. The first step was to grind out a “U” along the cracks in the broken piece and fill the holes. To do this, the piece was put on the brazing hearth. Once warm, a stick welder was used with cast iron rods. The weld was applied in short runs, and when finished, the piece was put back in the brazing hearth and allowed to cool slowly.

The blowhole had to be filled in the main hub, but it was not possible to preheat the flywheel. The only way to weld this was to use just enough amps to get the weld to puddle with the cast iron and weld in short bursts, then wait until the hub was cool enough to touch with the bare hand before starting again.

Once the blowhole and broken piece had been repaired, they were cleaned up to marry together and the crack line grooved. A piece of steel plate was put in the split in the hub as a guide and the broken piece clamped in position so that it could be tacked in place. Again, welding was in short bursts to first tack the piece in place, then it was built up with a strong weld around it. Great care was taken to not overheat the flywheel, and to allow it to cool right down before applying more weld. When finished and cool, the weld and surrounding area were cleaned up with small grinding wheels and a file.

The big decision when finishing an engine is whether or not to repaint it. As mentioned, there was a lot of rust pitting, but there were also specks of red paint, uncovered by carefully cleaning the engine using a paint brush and kerosene, allowing the kerosene to soften the accumulated grime and hard grease. No hard abrasive was used, the toughest cleaner being a plastic pan scourer.

In places of solid rust, where there was no chance of paint remaining, a metal scraper was used to remove the worst of the rust, but care was taken not to expose bright clean metal. With the generous use of kerosene the engine was covered with softened dirt and a pool of liquid had accumulated in the well under the connecting rod. I save the sawdust from my power wood working tools, and this is ideal to rub down the engine casing; I find it easier than using rags. It will soak up any puddles of oil or kerosene and is also useful for clearing up whenever I have spilled oil or other dirty liquids in the workshop.


The old pushrod has seen far better days, and appeared to have been welded at some stage to compensate for wear or damage. It was also bent and twisted, leaving no alternative but to make a replacement if the engine was going to run at its best. The hardened roller was still in good condition, and after tidying up the hardened steel of the catch plate with a diamond file it could be used again.

The first step was to put some 0.75- by 0.375-inch long steel in the milling chuck and, after centering it, drill a 0.331-inch hole ready to thread 3/8-inch by 16-TPI threads for the roller retaining screw. Next, copying the measurements from the original, a slot was milled in the rod 0.15-inch deep for the catch plate. This was milled to be a close fit for the catch plate so that it did not rock. To finish this part, a 0.177-inch hole was drilled in the center of the slot and threaded #12 coarse for the catch plate retaining screw.

To finish the pushrod, the last 6 inches of the end had to be rounded to fit in the guide hole in the cylinder head. The quick way to mount the bar and get a good rather than perfect fit in the 4-jaw chuck was to set the jaws to hold a round piece of 0.750-inch diameter bar. The bar can then be released by slightly loosening two adjacent screws. They can then be retightened in the reverse order they were released so that the rectangular bar is then held in the chuck, using two pieces of 3/16-inch steel to pad out either side of the thinner section.

Rocker arm

The rocker arm was missing, but a friend provided me with a couple of photographs and some measurements, so it proved relatively easy to fabricate one. Before starting on the rocker arm, a new pivot pin was made to replace the original as it was badly worn, with numerous ridges in it.

The first step was to drill a 0.375-inch hole through some 0.875-inch diameter steel for the pivot. For the exhaust valve actuating arm, a piece of 3/8-inch by 1-inch steel was roughly shaped using the milling machine to form an inverted “T” section before being roughly shaped to mirror the photograph I had. This piece was then profiled to a close fit against the pivot piece before the contact edges were ground to give a channel for the weld to fill. The two pieces were then welded together, then cleaned up with grinding wheels to a rough shape and finish filed.

The arm actuated by the pushrod was not as simple, as it curves down to the pushrod. Fortunately, I throw nothing away and soon found a suitable scrap piece of steel I sawed to basic size and rough milled to shape. This was tacked in place with a couple of spots of weld to check alignment with the pushrod. It was necessary to adjust the arm a couple of times before fully welding it in place. The hole for the adjustment screw was then drilled in the new pushrod arm for a 3/8-inch UNC thread, followed by extensive use of grinding wheels and files to get the correct shape to the arm. A 0.1875-inch diameter oil hole was drilled near the top of the pivot.

The adjustment screw was first made out of 0.750-inch diameter steel, but the head was too small as the pushrod could push past the rocker arm at bottom dead center. Another was made, the head being just under 1-inch diameter, to prevent this from happening.

Cylinder head

Looking at the cylinder head, it was clear it had endured some hard use and it would require more than just a little clean up.

Once the cylinder head had been removed it was evident that the valve stems and the guides would need some work. There was a lot of sideways movement and the springs also needed replacement. Both the elbows for the exhaust and the mixer were firmly in place, and although these could have been removed with a large wrench it wasn’t necessary for the work needed.

After carefully cleaning the head to retain as much original paint as possible, the valves were examined. When they were removed, a letter punch was used to mark them “E” for exhaust and “I” for inlet so they would be replaced in their original position.

Both valves were well worn around the stems and one was slightly bent, but the heads were in good condition. The valves had been made by peening the heads over the stem, so it would not be difficult to replace just the stems.

To clean out the valve guides, a 0.3125-inch reamer was coated in cutting fluid and tried, but it rattled around in the guide. Going up a size to 21/64-inch, resistance was felt as the reamer bit into the cast iron. After running the reamer through, clean metal was exposed.

This meant that oversize stems had to be fitted to the valve heads. The old valves were held in a lathe collet chuck and the peened over section of the stems were carefully cut off so that the valve heads could be removed from stems with the aid of a hammer and punch.

New valve stems were made by reducing some 3/8-inch drill rod to the new diameter of 21/64-inch. These were made to be the same length of the originals, with the ends reduced to 19/64-inch so that they would be a push fit into the heads. Before fitting the heads to the stems, a 0.125-inch hole for the spring retaining pins was drilled through each stem.

The original method of joining the two parts could have been relied on again, but to be completely certain the joint was coated with flux and the two pieces brazed together. There was still 1/8-inch of stem sticking through the heads, and once cold from brazing the stems were peened over to act as a safety check.

The valves then had to be ground into the seats in the cylinder head. Using grinding paste and rubber-cupped lapping sticks, the valves were worked until the metal on the seat and valve appeared clean and a good fit. A thin coating of engineer’s blue was used on the valve head to double check the fit. This threw up small differences and a little more grinding with fine paste was needed to get a perfect seal.

Once the valves had been ground in, all traces of grinding grit were cleaned away and the valve stems oiled. Fitting new springs and split pins completed this part of the restoration.


The governor assembly was stripped for cleaning and examination. Wear to the weight pivot pins was clearly evident and it was a simple task to cut some 0.125-inch steel to length, drill 1/16-inch holes in each end and fit new split pins.

The speed-change arm had been replaced at some stage, the new one having a blob of weld at the end to push against the swinging arm. A new one was made from 1/8-inch bar, slightly thicker than the original, with weld used to build up the end before shaping with a file.

The pivot pin for the detent blade holder arm was also replaced, the top formed over to fit the recess in the top of the arm. The original spring was broken and was replaced with a salvage one, the length being adjusted later when the engine was up and running.

Fuel mixer

There is no easy way to check out the fuel needle/valve in the mixer. The iron needle itself had rust pitting to one side of the tapered part, but aside from this was in good condition. It was straight, so after cleaning the threads and the knurling on the head it was refurbished.

The stem of the needle was pushed through the lathe collet chuck and clamped tight. The chuck was turned so that the good part of the taper faced the front so that a dial gauge could be pressed against it. Using this dial gauge setup, a compound slide was adjusted to follow the taper. It was then a simple task to take fine cuts and clean the taper.

To clean up the seat for the fuel valve a taper reamer with the same profile as the needle tip was used to lightly cut the seat. Once done, the fuel inlet was blown through while turning the adjustment needle to assess whether or not there was a gradual adjustment to the flow.

The check valve was made from a 1.125-inch length of 0.635-inch wide hexagonal brass. Male and female threads of 1/4-inch NPT were cut in the ends and a 1/8-inch hole drilled through. This was enlarged at the female end with a 1/4-inch drill to leave a clean seat for the 1/4-inch check ball.

Contact Peter Rooke at Hardigate House, Hardigate Rd., Cropwell Butler, Nottingham, NG12 3AH, England • peter@enginepeter.co.ukPeter Rooke's Engine Page