The Ship Model Forum

The Naval Institute might be a publisher to approach about it.

The book is usable for our purposes, and as a catalog, but it is not packed with great drawings. Peterkin documented and critiqued virtually every drawing of the USS Monitor he could find. Many of these are copies, or copies of copies, or speculative sketches or artistic fantasies. This material has great value in that Peterkin shows us what not to use, and pinpoints dimensional or conceptual errors.

The more valuable material he cataloged is the original work of John Ericsson and his assistant, McCord, and tracings made by the shipyard as an aid to construction. To me, the tracings are most CADable, since they are the closest in sequence to the "as built" version of the monitor, and because they are crisper than Ericsson's originals. Peterkin reproduced these tracings, but they are much reduced in scale to fit the page size of his book.

IMHO the best set of surviving original drawings are in the McCord collection at the Stevens Institute, and they only have a hundred odd drawings. McCord actually salvaged these and some of them are his own work. Ericsson's executor apparently destroyed most of Ericsson's drawings. This has been represented as Ericsson's own wish, but I doubt this. The executor probably thought he was protecting the estate from legal problems (questions of prior art, etc.) by destroying these papers, but it was an idiotic thing to do.

I haven't thought much about what to do with the Rhino 3D model ultimately. It is a new project with a long way yet to go.

Ah, interesting - I hadn't realized the material was so limited. Which collection do the shipyard drawings/tracings reside in? Is that the Stevens Institute in NJ?

Yes, the Stevens Institute is an engineering college here in Hoboken, NJ. It's literally only 4 blocks from where I live. I've never been able to find out how to get in and look at those drawings, though.

_________________
We like our history sanitized and theme-parked and self-congratulatory, not bloody and angry and unflattering. - Jonathan Yardley

Devin,

Stevens Institute must have a librarian or collections curator. If you contacted that person and explained your work, especially showing some of the images you have posted here, you can probably get access to the drawings. After all, college libraries exist to make knowledge available.

Your efforts are quite impressive. In my opinion you have a more legitimate claim to view those drawings than most people! You should contact them if you haven't already tried.

S. C. Williams Library at Stevens Institute of Technology:

http://www.stevens.edu/library/

There is an email contact address on their web page.

Phil

_________________
A collision at sea will ruin your entire day. Aristotle

The shipyard tracings were in the Rowland family collection. Rowland owned the Continental Iron Works.

I think the collection is now curated by the Mariners Museum in VA. The computer enhanced drawing of the Monitor I am now using was purchased from this collection.

The archivist at the Mariner's Museum is Jay Moore. He is extremely knowledgable about both the USS Monitor and the material in their collections.The powerplant is in progress. Mainly using Revolve and Extrude commands. The slides in front of the cylinder are for the crosshead. The steam chest is detailed on the drawing. I am filling in the big block components before attempting the watchworks of the double reversing valve gear.

When this cylinder and the crosshead are detailed I can mirror copy the assembly 90 degrees around crankshaft axis, offset this copy toward the bilge by the width of a con rod, and have in place the basic structure of the V-twin.

Exellent hard work.

However, I can't understand, whether the shaft was permanently connected to the tower or
it was connected when raised and disconnected when the turret was in 'travel' positiion (i.e. when the turret was resting on the brass ring)?

Can I see it on the drawings?

Or could the community provide any links?

Thank you.

In my view the drive train was never disengaged. Take a look at the first 4 or 5 illustrations in this thread. Notice that the four teeth in the crown of the main turret shaft protrude well above the deck. These four teeth cradle and engage the iron beam of the turret's floor.

The crown fitting can move up and down within the deck bearing cup. The theoretical extreme lower limit of its travel is the bottom of the bearing cup. At that level, the teeth of the crown would still engage the turret floor beam. This is theoretical in the sense that some machinery would have to disassembled or disconnected to achieve it.

The practical, working lower limit is reached by moving the wedge, or "key." This retracted the crown toward the deck by about 1.5 inches, which not enough to disengage turret.

Finally found a photo, in the Library of Congress collection, of a Monitor's turret engines!
One cylinder is vertical, the other cylinder is horizontal, heavily obscured by the turret shaft. Find it by letting your eye follow the steam pipes down the left hand side of the page to the steam chest. This is a late Monitor design, in which the shaft was coaxial, and extended all the way up through the turret to support a (non rotating) pilothouse. If we crop in on the vertical engine cylinder and crosshead, there are a couple of things to notice. One is the wooden lagging around the cylinder. This is not in the drawings. The other is that the valve and reversing mechanism is the same, slightly exotic system used in Monadnock, in which the valve for one cylinder is propelled by a linking rod from the crosshead of the other cylinder. Note that some of the meat of the connecting rod bearing at the crosshead has been milled away to provide clearance for the motion of the valve drive link. There is a high resolution TIFF download available at the Library of Congress site here.

http://www.loc.gov/pictures/item/2002725350/

The image was originally half of a stereo card. Unfortunately the other half was lost -- it would be useful to see this in 3D.

Visible in the Camanche photo are four of the eight adjustment set screws on the Ericsson crosshead. These set and locked the clearance of the slide bearings. It must have been an art form to get these just right. A wedge was used to secure the crosshead to the piston rod. It may also have positioned the inset bearings.

Thank you,msg.

I have a pair of other questions.

1)Do you know the height of turret-to-ring spacing (when fully raised).

2) .

Do you mean that in early monitors the tower was not coaxial with the shaft or
the cup casting was not coaxial with the shaft (as one can see on your drawings).

What's the sense of any variant?

Thank you.

Re "coaxial" I had in mind a dual rotor helicopter. Maybe not the right word. In any event, the later monitors beginning with the Passaic class had a pilothouse suspended above the turret. The turret shaft was like a tree trunk ascending from the keel and supporting the pilothouse. The turret rotated around it. This is my interpretation. You might consult with Devin, who has built one, the Weehawken.This is a drive system similar to that of the Camanche, shown in the Library of Congress photo in the prior post. In both images, the central shaft is fixed. It is the off center shaft that rotates. It delivers power to the turret via a gear.

The original USS Monitor was much simpler. The main, central shaft was rotated and the turret turned along with it. The crown like casting shown in the first post of this thread was roughly analogous to a universal joint. It introduced some freedom of motion or flexibility between the vertical shaft and the turret. I think the idea was to avoid transmitting non-vertical loads to the vertical shaft, as for example when the 8-ton guns were run out or fired and recoiled.

Re spacing at the ring, I don't have a way to estimate the number you are looking for.

At the very center of the turret, the max up and down movement that could be produced by the wedge was about 1.5 inches. If the structure were perfect (inflexible) and the fully raised turret rocked all the way to the deck ring at a point on one side -- it should open up a "mouth" of 3 inches on the other side. Given the turret's masses and structure however, I doubt such a gap ever appeared.

I may have a better sense of it when I draw the ring.

The connecting rod fork.Probably the casting form was turned on a lathe with the forked end turning out, initially, as a jar on a stick. The jar was then milled away on each side to create the fork. The donuts would have been fitted and added, and the assembly used to make a mold.

This is pretty much how the piece is made in Rhino3D, using revolve, trims, and a boolean ADD to glue on the donuts. However, the back of the fork has been reshaped using Rhino's Blend Surface command, to make a fair transition from the rounded back of the fork to the flat surface of the donut piece. The con rod seems to be the most challenging piece on this engine, so far anyway.

The V-twin, sans valve gear and crankshaft. Camera is looking aft and up at the ceiling of the galley. The inboard steam pipes are the inlets, the outboard pipes are the exhausts. Below is a shot looking up and forward, toward the central bulkhead.Note that the engine frame on the right has been deepened by the width of the connecting rod, to compensate for the offset of the two cylinders. Both frames will now mount flat to the overhead deck beams.

At this point the spacer plate is a drawing aid -- it helps to have a uniform datum surface. However we may find that Ericsson mortised one frame into the deck beams to accomodate the cylinder axis offset -- and bolted the other frame directly to the beams. The solution he used may make itself clear when the V-twin is mounted to the crankshaft, since this should establish the height of the frames and mounting points.

This is the basic slide valve and reversing link for one cylinder. Sliding the link from one end of the slot to the other reverses rotation of the engine. It was possible to accomplish this shift, via a long and elaborate linkage, using a handcrank inside the turret.

Three pillow blocks are mounted on a plate closing the "V" of the two engine blocks. These provide bearings for the valve control mechanisms. The two side bearings are for the two T-shaped reversing links for the two cylinders. One is shown. The central bearing carried the control lever for stopping and reversing the turret engines.

The lever and its long, intricate linkage to the turret was probably the most vulnerable apparatus on the ship. It had to be dead reliable and well protected. For a confederate saboteur, a logical target perhaps. The single main control lever was apparently sandwiched between these frames and the underside of the deck, underneath (and thus somewhat shielded by) the turret. On later monitors, the whole system, including the V-twin turret engine and its critical control linkage, was moved to a safer place -- mounted deep in the hull.

Here is the reversing gear for the V-twin steam engine.This is what you would see looking up at the ceiling of the galley, with your back to the boilers. The two long rods extending away from the mechanism are the valve rods. Each links to a slide valve inside of the steam chest of its respective cylinder. The slotted, T-shaped links were a commonplace of the steam era. They can be seen on many steam locomotives. Reversing is accomplished by sliding the valve drive rod from one end of the slot to the other.

Because this is a V-twin, two reversing T-links are required. They are not quite mirror images. They are slightly different in shape and vertical placement -- basically to keep them from colliding.

On the central shaft are mounted two control levers. The levers are locked together -- probably as a single casting -- with an included angle of about 100 degrees. Via short links, these levers pull and push the valve rods along the slots.

When the valve rod is In the middle of the slot the turret does not move. At one extreme of the slot the turret rotates at full speed in one direction. At the other extreme it rotates at full speed in the other direction.This is the same machinery seen from above -- though you could only capture this view while the ship was under construction. The master control lever for the control of the turret is mounted on the middle shaft, and protrudes forward, toward the turret bulkhead. This lever is the beginning of a long linkage leading, ultimately, to a control crank mounted inside the turret.

Missing from this picture are two rods. These were connected from the end of each T-link to the crosshead of the complementary cylinder. From a CAD standpoint, it seems best to install these final rods after both cylinders have been positioned relative to the valve gear. Here for visual context is the valve gear shown with a single cylinder.

Very nice!

_________________
A collision at sea will ruin your entire day. Aristotle

Thank you Phillip.

Here is the completed and aligned valve and reversing system for both cylinders of the V-twin. All of the other components of the turret engine have been turned off.

Two straight, simple rectangular section rods connect the crossheads to the tips of the T-links. These two reciprocating rods completely replace and supplant the elaborate cam driven valve gear more typical of 19th century steam engines. This excellent simplification worked because one cylinder of a 90-degree V-twin is always a quarter of a revolution ahead of the other cylinder.

Here is the complete engine, ex-crankshaft.

We know the USS Monitor used a standard, off-the-shelf V-twin "donkey" engine supplied by the Clute brothers of Schenectady. I think it was a rolling mill engine -- already developed and proved in iron works and mills. The features of V-twin rolling mill engines include reliable and sudden start/stop operation, quick reversing, high starting torque, the ability to operate without a flywheel, and stall-proof starting. They were also called "reversing engines." There is a print of one here: http://www.prints-4-all.com/cgi-bin/cat ... 011875916#

The specific engine I have modeled in Rhino3D, the turret engine for the USS Monadnock launched in 1864, was supplied by the Atlantic Iron Works in Boston. How closely does it resemble the original Monitor's turret engine?

In the Scientific American for December 12, 1863, on page 372, there is a long article entitled "The Monitors". This volume of the Scientific American (V9) can be downloaded as a pdf from Google Books. The article is dense with specifications and dimensions for the original USS Monitor and for the nine Passaic class monitors that followed it. For all these monitors, according to the article, the turret engines had a 12 inch bore and a 16 inch stroke. The USS Monadnock also used a 12-inch bore. By CAD measurement inside the cylinder I get a maximum possible stroke of about 16.5 inches. The crankshaft throw is 7.993 inches, suggesting a 16-inch stroke exactly. From the USS Camanche photos posted above, I would say the Monadnock and the Camanche also had essentially the same turret engines.

Finally, here is a CAD analysis of an overhead view of the USS Monitor's crank and crankcase. The original image from which these lines were lifted is in the Stevens McCord collection. Net of it, the USS Monitor turret engine was designed with a 16-inch stroke. The bore I have not yet pinned down.
As a practical matter, I intend to use the USS Monadnock engine modeled here to finish the CAD model of the USS Monitor turret drive. It is an approximation of the original, but it may be the best approximation we find until the actual USS Monitor turret engines are recovered from underwater.

My guess is that this newer, Boston-built Monadnock engine varies in some details from the Clute Bros original of 1861 -- but that the size (i.e. bore & stroke), power, valve gear and controls of the turret engines were essentially unchanged from the USS Monitor to the USS Monadnock. Subjectively it seems to me that the double reversing gear, the camless valve drive, and the single master control lever all ring true.

Michael

Michael,

That's some really remarkable work. Do you plan to continue modeling the Monitor once the turret drive is complete? Based on what you have accomplished so far I have a feeling such a model would be extremely impressive

Thanks also for the pointer to the Scientific American article: http://books.google.com/books?id=gDRJAAAAYAAJ
It looks like section twenty five has the article in question. Rather neat to see how artistic and detailed Scientific American issues were in that era.

Many thanks to you for the detailed National Archives drawings of the USS Monadnock turret engines. The whole model has crystallized around these machines. For example, from Capt. Peterkin we have drawings of the control linkage inside of and descending from the turret. It uses something rather like a helicopter swash plate to communicate motion of the control crank inside the (rotating) turret. However, until you posted the USS Monadnock drawings, it was unclear where the linkage ended. We can see now that its "target" is the single master control lever in the center of the V of the V-twin.

The model will include 6 to 8 frames of the mid-section of the monitor, that is, 18 to 24 feet sliced out of the middle of the ship. Much larger and you begin to obscure the view of the machinery with other stuff, e.g., the boilers. It is essentially a model of the USS Monitor under construction, showing the state of the build with mostly open framework in late November, 1861.

The extent of the model is also limited by the memory on my computer, which is already beginning to creak and complain a bit.

It is surprising that the Scientific American would print all those technical specs, for anyone to see, in the middle of the civil war.

Michael

I'm really thrilled that it turned out to be useful - I copied it more or less on a whim at the time.



Hmm. I'm not intimately familiar with the details of Rhino, but as far as obscuring parts go there are probably ways to make them translucent/transparent, or turn on/off various sub-components. In BRL-CAD this can be done, certainly - an example is here: http://brlcad.org/gallery/s/renderings/ ... 2.jpg.html Just a thought of course - a "during construction" model is also quite a legitimate approach.



Heh. Yeah, what usually is done in those situations is model components individually and then "assemble" them into place. BRL-CAD isn't a good place to create NURBS models at the moment, but it is quite good at visualizing large models and can import pieces from elsewhere - see for example http://brlcad.org/gallery/s/renderings/ ... t.jpg.html for an example of a very complex model. It might be possible to model the individual pieces in Rhino and then assemble them into a complete Monitor in BRL-CAD.

(Just an aside - if you decide at some point you have an interest in posting the CAD files themselves online, you might want to glance at the Creative Commons licensing options and see if any of them sound interesting (http://creativecommons.org/licenses/) - they cover a variety of situations (non-commercial use only, derivative works allow/not allowed, etc.) and seem to be fairly widely used - Wikipedia for example uses one of them.)

I'm wondering if there might be some way to actually simulation motion with that geometry using something like PAL (http://www.adrianboeing.com/pal/index.html)...

Michael,

I am curious. How large are your files? Are you running into memory limits or CPU speed limits (excessive drawing time)?

My CAD model of the USS Oklahoma City is something like 500-600 Mbytes so far. It takes about 90 minutes to 2 hours to render the whole ship the first time. After that each new image takes about 1/10 as much time to generate.

Note: I say "something like" and "about" because I haven't assembled the entire thing in about two years. I have added a huge amount of details in that time. One of these days I'll bite the bullet and put it all together again. Patience is a virtue!

Phil

_________________
A collision at sea will ruin your entire day. Aristotle

The largest 3dm file to date is about 75 megabytes but, curiously, I have more trouble with one of about 56 meg. It is a stack of 10 identical frames, created as an array. It is not that difficult to work with these files, but they don't want to be rotated in 3-space. Rotation has hitches and glitches. This computer has 4 gigs, but I am not sure it knows how to access all the memory that it has. In general I make an initial sketch of a part in position. Then Export it to a file of its own -- finish it -- and then re-import it to the main file. There are several "main files": one for the main bulkhead, one for the frames, one for the engine, etc. I hope to begin pulling it all together shortly.

The BRLCAD site is very interesting. The animation idea would be really trick. In Rhino I can turn layers on and off, and vary transparency for any object on a scale between 0 and 100. There is also an X-ray like visualization. But what I was thinking about is a bench model of the USS Monitor under construction. The valve linkage, for example, seems to me a plausible candidate for photoetched brass. The main frames of the ship also have some flat, repetitive patterns that might be photo etched.

Take a look at the cylinders depicted here, in a pioneering CAD drawing from a paper presented at The Third International Conference on the Technical Aspects of the Preservation of Historic Vessels convened in San Francisco, California on April 20-23, 1997:

Subjectively, it seems the cylinders are oddly small in diameter and, therefore, in cylinder bore. The image is intended as a schematic representation, not as a scale drawing. it is a helpful sketch and it shows well the basic layout of the components of the turret drive. But for a scale modeler it is important to note that the cylinder size is not to scale.

How did these tiny cylinders happen? I think it was a CAD issue. It is a question of how you read and interpret into CAD the sense of the original drawings.

For example, here is a reproduction of a small element of a fine USS Monitor drawing that hangs in the Swedish American Museum in Philadelphia.
The drawing has been screened for printing in Peterkin's book, hence the bothersome dot pattern.

The original artist -- possibly John Ericsson -- heavily shaded the cylinder along the edge of a horizontal shelf that demarks the slide valve's steam chest. In this overhead view, the steam chest and its horizontal shelf mask from view almost half of the cylinder.

Given the poor reproduction quality of the drawing, it is easy to confound this heavily shaded line with the edge of the cylinder. If you create a hypothetical cylinder in CAD using this subtly deceptive but absolutely false line, it looks like this:One clue that this stunted cylinder is wrong is that it is offset to one side of the frame. The centerline of the piston rod and crosshead (green line) does not correspond to the centerline of the cylinder.

Another problem is absolute size. The oak deck beam crossing the picture diagonally is 10 inches wide. If we allow for cylinder wall thickness, the bore of this hypothetical cylinder must be down in the range of 5 inches. Compared to the 12-inch turret engine cylinder diameters of the later Ericsson monitors, a 5-inch bore doesn't seem consistent.

If we ignore the shaded line marking the edge of the steam chest, and instead revolve in CAD the upper, outer edge around the centerline, we get a much more robust and normal looking cylinder:In this one the centerline is correct. The cylinder's outside diameter corresponds well with that shown in another drawing in the Stevens McCord collection in Hoboken.