A 'One of a Kind' Waterwheel

Article Tools

Image Gallery

Print

E-mail

Comments

RSS

DIFFERENT STROKES. . .

Now that the wheel and framing were completed, I felt I'd really accomplished something . . . but I still had to figure out -and construct-a pump that'd work with my setup. To keep things simple, I set my sights on a single-acting pump (which creates suction and pressure on one side of the piston only) so I wouldn't have to provide a packing gland for the piston rod. At the same time, I figured I'd mount its cylinder on a swivel, thus eliminating the need for a connecting rod and guides.

But the question of how much displacement I could have still remained. I knew the length of the pump mechanism would be limited to 42" (the distance from the wheel's axle center to the upper frame member). So I finally settled on a 9-inch throw for my crank, which meant that my piston would travel 18 inches . . . and that the two parts together would take up 27" overall. This would leave 15" for the pump fittings and swivel bracket, a working space I could easily live with.

Choosing the cylinder bore proved to be a compromise between what I wanted and what was available in standard hardware. A quick wheel-force diagram showed me that the 15-pound force acting on my paddles at 35 inches from the centerline of the shaft and then transferred through a 9-inch crank -would amount to 35 divided by 9 (or 3.88), multiplied by 15 . . . a healthy 58.3 pounds of force.

Dividing this number by the area of the piston (I decided on a 2" Schedule 40 PVC pipe cylinder, so its cross-sectional area using good of -r2-was 3.14 square inches) gave me the available pressure of the pump (18.6 pounds per square inch). From there, since I knew that it took 0.43 PSI to raise water one foot in height, I could calculate the total head the pump would deliver (43.25 feet . . . plenty for my purposes) by simple division.

Furthermore, with all my "unknowns" known, it was easy for me to calculate that, guessing my wheel speed at 4 RPM, the 2" X 18" pipe pump could deliver roughly 13,565 cubic inches, or 58 gallons, per hour . . . which would amount to almost 1,400 gallons in a day!

Acquiring the pump's components was simply a matter of going to the local hardware and plumbing supply store and picking out the parts I needed. My cylinder was just an 18" length of 2" Schedule 40 PVC pipe, equipped at one end with a 2" slip-to-pipe female adapter and at the other with the same type of fitting terminated in a male thread. I screwed the latter end into the common of a 2" galvanized "T", then bushed the T's other two ends down to 1/2 inch. On the intake end, I used a short nipple, an elbow, and a 3/4" plastic hose adapter . . . and on the other, I ran a short nipple to a 1/2" swing check valve. Then I concentrated on creating the piston parts.

The piston rod, a 5/8" X 26" stainless steel shaft, required a bit of machine work, since it had to be threaded at each end. One tip just needed a 5/8-11 thread 1-3/4" long, but the other had to be cut down to a 1/2" diameter for a distance of 1 inch. A 5/8 "long 1/2-13 thread was added to this stub . . . which left a 3/8" length of smooth surface between the threads and the step to the rod's 5/8" diameter.

With this done, I drilled a 21/32" hole directly in the center of a 2" galvanized pipe plug, then slipped the piston rod through that opening so its 1/2" end was on the same side as the pipe threads. A couple of body washerswith 1/2" center bores and 1-5/8" outside diameters-made excellent buttresses after I curved their edges slightly with a file and sandwiched two 2"-diameter leather piston cups between them. A 1/2" nut and a cotter pin held this homemade piston in place on the end of the shaft, and to alleviate the effects of water on the leather, I soaked the parts in neat's-foot oil before slipping the piston in place and threading the pipe plug to the cylinder's plastic adapter. (The leather plunger will slide home much more easily if you take the time to taper the end of the cylinder's bore with a file.)

The connector at the opposite end of the piston rod was just a 1 " X 1 " X 3" block of steel drilled and tapped 1 inch deep at one end to match the rod, and cross-drilled at its other end to a 17/32" diameter. The wheel crank-a 3/8" X 2" X 11" piece of flat steel-was drilled out on a 9" center to join the 1-3/8" axle and the connector, and I welded a 1-9/16" length of 1-1/4" pipe to the large-holed end, and a 3/4" piece of 1 "-diameter cold-rolled rod (tapped for 1/2-13 thread) to the other, to serve as crank collars.

Once I'd fabricated those metal parts, I constructed the swivel bracket that would hold the pump in place. That involved trimming out a 6" X 8" piece of 1/8" steel plate, welding two 1/4" X 3" X 3" X 3" angle iron sections to one endof this platform so that the lips were aligned with the plate's sides, and drilling a 1/2" hole 1 inch from the end of the plate at the butt joint of the two angles.

To mount the pump to the bracket, I drilled a 7/8" hole that was centered and 2 inches from the baseplate through each of the upright angle iron lips and let the 1/2" pipe nipples on the T serve as fastening pins. I then pinioned the swivel bracket to the 8" X 8" mounting plate on the wheel carriage, using a 1/2" X 1-1/2" bolt, two nuts, and a cotter pin.

Having come this far, I breathed easy: The rest was just a matter of fitting and fine tuning. To check the wheel's lateral play, I cut a section of 1-1/4" Schedule 40 pipe into three lengths (3", 4-11/32", and 4-19/32"), and used two as spacers between the wheel hub and the bearing blocks and the third as a keeper collar on the end of the shaft opposite the pump. Also, I placed 2-1I4" outside diameter flat thrust washers between the pipes and the wooden parts, tightened the hub's 1/2" setscrew, and locked the collars to the main shaft with 3/8" X 2" machine bolts.

Then I fit the crank to its end of the shaft (securing it by-once more-using the technique of tightening a bolt run through a hole drilled in the axle) and fastened its other end to the piston-rod connector with a 1/2" X 3" bolt and a lock nut. (I'll tell you now, if you try a similar project yourself, be sure to employ some form of locking mechanism-whether it's cotter pins or a thread sealant-on all your fasteners, or you'll be retightening them on a regular basis!)

Before I set the wheel into the water, I had to finish up the plumbing. On the suction side, I just ran a 6' length of heavywalled plastic hose that terminated at an adapter and a 3/4" foot valve submerged in the water. To protect the pump from foreign matter, I wrapped the intake with aluminum screening . . . and later made up a gravity-operated sand trap-from some 2" Schedule 40 PVC pipe, a T, a 90 ° elbow, and a bit of woven filter packing-which I placed in line between the foot valve and the pump to catch and settle out the debris.

On the pressure side (the one with the swing check valve), I installed a 112" nipple and T to the check and threaded a small pressure gauge into one of the T's ports. I then ran my supply hose to the garden area and terminated it in a garden-hose fitting and another T . . . which I plan to put to use later on when I can get around to installing a pressurized holding tank and a small, decorative fountain.

Finally, I filled the wheel hub with grease to keep it from rusting to the shaft, ran plastic oil lines to the oaken bearings, and let the machine down into the water.

RSS | Comments | E-mail | Print