A 'One of a Kind' Waterwheel

Enlarge Image Image Gallery [1] This unassuming creek supplies the power for the wheel . . . and water for the McCray garden. [2] Once the ideal size and shape of the paddles were established, Dick used a sheet-aluminum tem plate to assure con sistency. [3] The paddies were mounted to the iron wheel with U-bolts and small aluminum bridges. [4] The PVC-pipe pump mounted to its swivel bracket. The intake hose with its foot valve is on the left, and the T's right arm delivers pres surized water.

Article Tools

Image Gallery

Print

E-mail

Comments

RSS

I got my first taste of waterwheels at the age of five, when my great-grandfather McDowell made me a toy paddle turbine to run under the spring-fed spigot at the back of the house. Today, some 65 years later, I've applied the principles of that early lesson to building a full-sized undershot wheel that provides me with every drop of water I need to supply my thirsty garden throughout the entire growing season.

My vegetable plot, you see, is quite a distance from the house and its plumbing. True, the garden is located not far from a small perennial mountain stream that forms the southern boundary of our property . . . but that "crick" runs a good 8 feet or so below my patch!

Now I certainly don't have an aversion to honest work, but the drudgery of using a hand pump to fill a washtub, lugging the sloshing vessel around, and repeating this operation at least six times every time I wanted to water my garden forced me to look for a less labor-intensive means of getting the job done. Naturally, that first waterwheel in my life came to mind, so I set about researching the design and operation of functional "paddle pumps" in hopes of building one at my site that'd handle my watering chores with a minimum of maintenance.

EARLY PLANNING PAID OFF

Because there was only about 4 inches of fall in the part of the stream bordering our land, an overshot wheel was out of the question. Unfortunately, I had little luck digging up specific information on undershot wheels, so I had to use common sense-and a by-guess-and-by-golly approach-to make my project a success.

Early in the game I decided that an all wood wheel would be too expensive and time consuming to assemble. So, considering the fact that an undershot design uses paddles rather than the intricate buckets of an overshot apparatus, I figured I'd search for a metal-spoked wheel about 4' or 5' in diameter and simply fasten some plywood paddles with 1'-square blades to it.

I started by roughly calculating what I had to work with in the way of water. The creek usually runs about 2" deep at an 8' width. To create a weir that would direct the flow toward the center of the stream (thus enhancing its depth and velocity when at normal levels) but still withstand the punishment of occasional deluges, I piled rocks in the creek until the center channel was 16" wide and the water—normally—was about 8" deep . . . dimensions that I thought would be about right for the size of wheel I had in mind.

Then, to get an idea of the water's force ;within the channel, I held a foot-wide board Win that slice way and used a small scale to determine the stream's "push" on the piece of wood . . . a force of about 7-1/2 pounds.

e Next, I calculated the velocity of the wa ter. I laid two long poles across the creek, 10 feet apart, and floated several sticks on the surface while I timed their progress. The average I came up with-four seconds to cover the 10-foot distance-was then multiplied by 15 to arrive at a figure of 150 feet per minute.

Now comes the fancy part: I'd found a 4-1/2'-diameter, 16-spoke metal wheel from an old hay-rake, and figured out a way to mount a 12" X 12" paddle to the rim at each spoke (I'll tell you just how I did that later in this article). To establish the estimated rotational speed of the completed wheel, then, I just added the 27" radius of its steel rim to the distance from that rim to the 8" centerline of the water's force on the paddles, and came up with a lever arm of 35".

The old formula C21rr gave me an overall circumference of 220 inches, or 18.3 feet. Using the flow rate of 150 feet per minute through the weir channel, I calculated by dividing 150 by 18.3-that the whole shebang would theoretically turn at about 8 revolutions per minute (RPM). Of course, knowing from experience that theory tends to be more optimistic than reality, I covered my bet by cutting that estimate in half . . . to take into account slippage and friction after the wheel and pump were installed. (As it turned out, I could have given the wheel a little more credit: Its actual speed averages just over 5 RPM!)

At this point I knew approximately how fast the wheel would rotate, but I still had to deduce how much torque it would provide in order to size my pump properly. A scaleddown sketch of the side view showed that with one paddle dipping 8 inches into the water, the adjacent paddles would each be submerged 4 inches, which would be equivalent to having two paddles fully in the water at any given time, each furnishing a 7-1/2-pound force from the current. By multiplying the 35" lever arm by the 15-pound force from the two paddles, I arrived at 525 inch-pounds-or about 44-foot-pounds-of torque available to power the pump.

Armed with the numbers I'd worked out, I could now sit down and think about just how I was going to mount the wheel so it would function under various flow conditions . .. and work out a pump design that'd be compatible with the arrangement I would finally settle on.

RSS | Comments | E-mail | Print