A note on steam engine terminology: each cylinder is considered an "engine" - therefore if the power plant has more than one cylinder ( even if mounted on the same bedplate ) it is referred to in the plural "engines".
C O N C E P T
It should be kept in mind that from the late 1830s through the 1870s marine steam engine design was in the throes of great experimentation and innovation. The use of the screw propeller, pioneered by John Ericsson and others, relieved ocean going and coastal vessels of the need for side paddle wheels. Harnessing the power of steam to a shaft running fore-and-aft, however, required a major re-thinking of steam engine design. Improvements in metallurgy allowed higher steam pressures to be used ( 15 psi in 1830, 70 psi by 1870 ) resulting in smaller, lighter weight and more powerful engines.
Of the many interesting ( some truly bizarre ) designs of this time, the most common ( until the development of the compound engine in the late 1860s ) was the use of one or more cylinders, laid horizontally and at right angles to the propeller shaft. In one type a hollow cylinder or "trunk" was attached to the piston and passed through both cylinder heads. On the side towards the propeller shaft it housed the connecting rod - pivoted inside the piston and connected to a crank on the propeller shaft much like modern i.c. engines but having the advantage of being double acting. This trunk served to support the weight of the horizontal piston and absorb the "side thrust" from the connecting rod. John Penn Engineering in England was the major developer of this type of engine ( Penn made the engines for HMS Warrior and a "virtual tour" of her boiler and engine room is at:
http://www.stvincent.ac.uk/Heritage/Warrior/engines.html ). Maudslay and Field, also in England, developed a similar engine except the trunk piston was not used. The piston rod extended to a crosshead guide on the opposite side of the shaft. The connecting rod "came back" to connect to a crank on the shaft, hence the name "back-acting". The only surviving example of this type engine exists at the US Merchant Marine Academy and once powered the USS Ranger ( built in 1873 ). A description and drawings of these engines, as restored by the American Society of Mechanical Engineers, is available - it is a 1.11MB file in .pdf format ( you'll need Acrobat Reader ) - and may be found HERE. The advantage of both types of engines was in the low center of gravity, important in any ship. For naval vessels an additional benefit was putting the machinery below the waterline - out of the reach of shot and shell - and, despite some shortcomings, this was the primary reason the US Navy continued to use them well into the 1890s.
John Ericsson was in England during the time Penn and Maudslay were developing their engines and no doubt saw them. Initially he used the Maudslay idea of the piston rod connected to a crosshead - but Ericsson departed from the "normal" back-acting connecting rod by utilizing a system of levers which he had designed for the USS Princeton in 1843. These "vibrating levers" greatly alleviated the difference between piston position and crank position ( frequently referred to as the "problem of angularity" ) and improved engine performance. He then further improved on the Maudslay design by offsetting the cylinders and placing them on opposite sides of the keel with the crankshaft set between. By 1858 he had compacted the engine still further by placing the cylinder heads "back-to-back", thus centering the major part of the weight over the keel, the "vibrating levers" permitting this singular improvement ( Ericsson Patent No. 20782 ). The USS Monitor, however, required that the engines occupy even less space side-to-side and Ericsson, by utilizing John Penn's "trunk piston" concept, was able to do away with the crosshead and decrease the width of the engine enough to make it fit within his hull design. This became basic design for all later "monitor" engines.
The cutaway drawing above - showing the engines as seen from the after end looking forward - makes the motion fairly clear. Two pistons are set "back-to-back" in a single casting for the "two" cylinders with a dividing partition built in the middle. Each piston's trunk only extends through that piston's cylinder head ( hence "half trunk" ). The "piston rod" ( on an i.c. engine it would be called the "connecting rod" ) extends from a wrist pin located within the piston to a lever attached to a jackshaft and moves this lever back and forth ( the "vibrating lever" ). The rocking motion of this jackshaft is transmitted to a "crank lever" attached to a connecting rod which goes back to the crank and shaft ( "back-acting" ).
The engines built for USS Monitor were part of the "hundred day wonder" ( the time Ericsson was alloted by the US Navy to build her ). There were many compromises to make them fit into the available hull space and some of the engine geometry was definitely "weird" - but the design was good and the whole vessel such a success that, within a few days of her engagement with the CSS Virginia, the US Navy ordered 10 "monitors" of "improved design" called the "Passaic" class ( then, as now, the class name was taken from the first vessel built ). In the fall of 1862 the "Canonicus" class of 9 monitors was ordered from Ericsson. This class was much improved over the "Passaic" class and it is the design of the engines of this class that forms the basis for Artemis' engines.
An aside: we in the United States do not fully appreciate the genius of Ericsson and the impact of his Monitor. On receiving news of the Monitor - Virginia battle, the London Times remarked ( with good cause ):
"Whereas we had available for immediate purposes 149 first - class warships, we have now two, these two being the WARRIOR and her sister IRONSIDE ( sic, Black Prince ). There is not now a ship in the English Navy apart from these two, that it would not be madness to trust to an engagement with that little MONITOR."
The other major world naval powers were equally affected!
D E S I G N
As noted on the "DESIGN" page: an 18" diameter by 24" pitch propeller is the "best starting point"; a step-up ratio, engine to propeller shaft, of 1 : 3 was selected so that the normal engine operating speed would be 125 rpm and the moving parts could be "seen"; 5 design HP is required. The "Monitor" type engines were of 40" bore, 22" stroke and 13" diameter trunk. These dimensions, when drawn out, give good geometry and events throughout a full revolution and I decided simply to scale them down to 1/8 size.
Having determined the horsepower required, available space for the engine was the next consideration. In the case of Artemis the engine will be located in line with and close to the boiler. As the Ericsson VL ( vibrating lever ) engine is of very low profile and relatively short fore-and-aft length, the width becomes the primarily limiting factor. As the boiler will be 20" wide the engine should not be much wider - for reasons of both simplicity in mounting and safety ( due to open, moving parts ).
The boiler safety valve will be set for 100 psi, allowing for a normal operating pressure of 85 psi. A vacuum of 20" Hg ( because of the use of a condenser ) is not unreasonable and a pressure drop in the main steam line of 5 psi should be expected - thus 105 psia is available ( psia = pounds per square inch absolute - measured from a complete vacuum; i.e., 15 psi greater than the "steam gauge" pressure ). After much "fiddling around" with the P.L.A.N. formula, and taking into account the difference in piston areas due to the "half trunk", a bore of 5" and stroke of 2.75" with a trunk diameter of 1.625" was selected. This agrees with my scale of 1/8 and is easy to work with. The engines will produce the 3 Shaft HP required with a MEP ( mean effective pressure ) of 78 psi - approximately 50% cutoff - and consume 100 pounds of steam per hour at a valve chest pressure of 97 psia. Based on experience from the previous Artemis' boiler, the proposed boiler will be adequate.
Now to design this for a 20" wide space. I want to use "off-the-shelf" components for bearings, shafts, fasteners, packing, fabrication materials, etc. to minimize custom castings and machining. And it has to be possible - without "special" tools - to effect most repairs with the engines "in situ". The components need to be sufficiently rugged to handle unexpected stresses, therefore proportioning components to the size of "originals" is not reasonable. So - compromises must be made. The overall width of the engine taken from the extreme positions of the vibrating levers will be 20.5". This is acceptable and, because the bulk of the engine will be near the deck level and have "guards" around it, neither I nor any passengers will be endangered or inconvenienced. As soon as I have completed a computer acceptable, dimensioned drawing I will post it here.
The valve gear will be of the single eccentric type. As, due to the use of the "Kitchen Rudder", not only is a reversing mechanism not needed but, as the vessel's speed is also controlled by that device, the engine may be allowed to run at its most efficient speed and the valve events "fixed" accordingly. The drawing below ( showing the engines as viewed from aft on Artemis ) is a good approximation of how the valves will be activated.
C O N S T R U C T I O N
Once work starts details will be found here.
P E R F O R M A N C E
When it's built and installed, pertinent data will be available.
Latest update, July 6, 2004
We are not a library or research service, but if you are interested in more information - or just want to leave a comment - you may access our COMMENTS form by selecting the "whistle" to the left. Don't ask for plans. If they aren't here, they aren't available!