The Knowles Rotary Steam Engine

The Knowles Rotary Steam Engine

Gallery opened: 10 Apr 2014

Updated:

It is very rare for anything to be known about a rotary steam engine except bare names and dates. This one is different. I am indebted to Michael Knowles for sending me details of a rotary engine built by his great-grandfather, William Knowles. He, and both his father and his son, had an engineering business in Aberdeen in the 1800s. All three were named William Knowles. The reverse sides of the photographs are labelled "Father's engine", but no date is given. I am no expert on the history of photography, but I would have thought from the general look of the cards the date would be around 1880 - 1890.
This pair of pictures shows a professional looking bit of construction. Judging by the door behind it, the baseplate is about three feet long.

Above: Two views of the Knowles rotary engine; the original photographs

Left: The Knowles rotary engine
These black-and-white images are enhanced versions taken from the pictures above.
The engine itself is the horizontal cylinder to the left; the pipe emerging from it is the steam exhaust. In the middle is a flywheel, and immediately to the right of that what looks like an eccentric-driven pump, presumably for the boiler feed-water. To the right of that appears to be a wide pulley of the type used to drive line-shafting, which suggests that William Knowles was planning to put his engine to practical use in his engineering business.

Above: The Knowles rotary engine
In this view the actual engine is on the right. Just above it is a centrifugal governor driven by a belt from the end of the shaft, and presumbaly acting on a throttle valve in the casing below it; note the small drain cock for condensation. To the right is a pipe that is clearly the steam input as it carries a stop valve.
In terms of naming of parts, that just leaves the large cylinder at the top. This appears to be a water jacket rhat heats the feed water with the exhaust steam, as a pipe directly connects it to the feed pump.

The photographs above show only that the engine itself was in the form of a horizontal cylinder- but then that applies to almost all rotary engines. However, Michael Knowles has an old drawing of the engines's construction, and the image below is taken from that. Unfortunately the drawing appears to be undated.

Left: The Knowles rotary engine: section
This section of the engine cylinder, the result of a good deal of computer processing, shows an offset rotor with a sliding member running through it. In its basic principle of operation it looks very like:
The Woodhouse engine of 1839,
(which was almost certainly never built)
The Hyatt engine of 184?
and the Franchot Rotary Engine of 1900.
Four steam passages into the cylinder are shown, at 3 o'clock, 5 o'clock, 7 o'clock, and 9 o'clock. These are presumably two pairs of inlet and exhaust ports, though no differentiation in size appears so there does not appear to be much scope for the expansive use of steam. The two pairs are selected by the interconnected slide-valves on each side of the cylinder, interconnection being by a rocking beam pivoted on top of the cylinder. The reversing valves and lever are not present on the engine as photographed.
At the extreme right is another slide-valve which looks as though it was coupled to the governor for speed control.

BIOGRAPHICAL NOTES
William Knowles was living at 23 Cotton street, Aberdeen, in 1880-91 according to the Aberdeen street directory. Likewise, according to the 1881 census a William Knowles was at 23 Cotton St, and employing two men and a boy.
Michael Knowles tells me: "My grandfather, great grandfather, great-great grandfather and great-great-great grandfather were all called William Knowles and were all engineers from Aberdeen. The 1871 census has them at 6 St. Clement St. The 1861 census has them at 23/25 Frederick St and the 1851 census at 30 Frederick St. (probably the same place as 1861, but re-numbered)"