HMS Bellona 1760 by SJSoane - Scale 1:64 - English 74-gun - as designed

[AON]

Thank you Siggi,

I have been wondering how they would swing a column up if it were hinged without the foot of the column being rounded to allow it to pivot, or even pull a column out.  Hinged or not... of course they jacked the overhead beam up a smidgen to create the necessary clearance!

DUH.

[AON]

Seems the foot or bottom of the column is pocketed into the deck (hidden or dashed lines below the top surface of the deck in the image).

The head or top also is recessed into the beam from one side allowing it to slide out once the beam is lifted.

[noel_colledge]

So is the next question, What sort of Jack did they use, doubt it would be hydraulic, so some form of high lift? and Where did they store it when not needed?

[druxey]

I suspect that it was a rack and pinion device, such as pictured in Blankley's Naval Expositor (Hand Screws, page 70).

[SJSoane]

Looking again at the tops of the pillars in this image, I think the tops might be dotted lines because they are drawn the height at the centerline of the ship, but the beams are drawn at the side of the ship, covering the tops of the pillars in this view. I know in some cases in the Bellona, the pillars are at the center of the beam, so could not slide out the side. So there must be a tenon at the top as well as bottom, using the jack device to remove them. Except for the pillars that hit only partially, as seen in the far left here, or in some cases don't hit a beam at the top at all.

 

I recall a story about Frank Lloyd Wright and his masterpiece house, Fallingwater in Pennsylvania, which has large concrete cantilevered decks over a waterfall. The story goes that the workmen did not want to go under the freshly poured concrete to knock out the form supports, for fear that it would collapse upon them. The architect, confident in his design, walked under and knocked them out himself. Now whether this story is true or not, the fact is that the decks did massively sag over the years, and had to undergo significant reconstruction a few years ago. So I wonder how much the guys cranking up a deck in a 74 worried about breaking something in the process?😗

 

 

 

[rybakov]

About lifting the beams and withdrawing the pillars Dafi's Victory log might be worth a look at page 40 .

 

all the best

Zeh

 

[SJSoane]

Thanks so much, Zeh, Dafi's log is well worth looking at. His little figures really bring the deck activity to life, and the jack screw clearly illustrates how the beams were lifted in order to pull out the pillars.

 

Mark

[druxey]

A jack screw. Of course!

[Gaetan Bordeleau]

I saw of similar photo: a wooden jack screw for repair of cart and wagon wheels. (Ethnographic Museum of Elhovo, Bulgaria)

Unfortunately, the date is not written.

 

 It is clear that both french and english used something to remove pillars when needed to use the capstan. How often did they have to remove the pillars? When they needed to use the capstans as winches for anchors, sails etc.. Probably that the pillars were in place more than 90% of the time.

 

How much did they have to lift the deck beams? A fraction of inch, no problem, few inches to clear both tenons at each end of the pillar; I would be very much afraid to break many parts because  a lot of parts will follow the deck beam. A recess in the deck beam and in the floor would save a lot of "height" and would also maintain the pillars in place. I do not know how they lifted the deck beam but it had to be fast and easy,  as a simple lever by example, few feet long, something like may be a capstan arm.

 

An easy question in 1780. A difficult  answer in 2020.

[SJSoane]

Hi Gaetan, thanks so much for the photos. That really brings a drawing to life, showing how common this device must have been in a lot of different locations and for different tasks.

 

I just keep drawing away, deciding to get more things worked out on paper before cutting more wood. The cross sections are exceedingly helpful for refining a lot of the details. I should have done this years ago.... Now I have a clear record, in one place, of the sizes of things as I work my way upwards.

 

By the way, in the specifications for the Marlborough, dated 1763, there is an interesting description of the beams and clamps that I remember Gary (garyshipwright) pointing out several years ago--I just didn't understand it at the time, but now I think I do. Gary, you so far ahead of me in so many ways!

 

Taking the gun deck, for example, the lodging knees are set 1 1/2" below the top surface of the beams, to ensure more air circulation between the knee and the deck above, to avoid rot. The upper surface of the clamps touch the lower surface of the knees, and are themselves 1' 5" deep. But this means that the beams, moulded 1'-4", would cut excessively into the clamp, causing a structural weakness. So the beams are bearded up at the extreme ends (12" from the end), to cut into the top of the clamp by only 1".

 

I did not understand the logic of this years ago, but now I do. The clamp is providing great longitudinal structure, and cutting large notches into its top surface would reduce the effective structure down to whatever was left below the notches. This needs to be minimized. At the same time, the structural capacity of the beams is mostly determined at the center of its span. Notches at the extreme ends do little to compromise the structure; although, it does provide a place for splits to start. The pillars at the centers would provide a lot of additional support at the center, reducing the load that has to be carried at the end over the clamp. An interesting balance of avoiding rot, while maximizing structure. They were clever guys, those shipwrights!

 

I already installed my clamps several years ago, and placed them without this refined detail, so they are lower by a few inches. I am not inclined to move them at this point. So my drawings will reflect the true detail, but the actual model will be simplified. Don't tell anyone!

 

Mark

 

 

 

 

 

 

 

 

 

 

 

[druxey]

Another rot-reducing measure were holes bored into the beam ends for air circulation. A longitudinal hole was drilled into the end for about a foot or so, then another smaller hole bored up from under the beam to join it. The inner surfaces of the holes were then sealed using red-hot irons to char them.

 

Be sure to include this detail!

Edited October 26, 2020 by druxey

[AON]

the things I learn here....

[SJSoane]

druxey, I just need to figure out how to hold the miniature red hot iron...😉

[druxey]

Heat the needle holding it with tweezers!

 

[SJSoane]

druxey, I was hoping for a 1:64 electric soldering iron...🙃

 

Continuing to draw more details, now the stern balcony and screen bulkhead.

 

I learned a lot about some very refined design characteristics. Because the roundhouse beams round up higher than the quarterdeck beams, the transoms below the windows take a sweet curve split between the two. All of the intermediate window bars are equally spaced between the two. The doors in the last bays at the outboard ends of the screen are disguised to look like the rest of the windows, with very thin rails framing the door window hiding behind the columns on the fore side of the bulkhead. And best of all, pocket doors to the quarter galleries from the balcony itself, which slide aft into just enough room in the panelling at the side of the balcony. You can see in the upper left sketch below how the door just fits between the hull frame and the finish panelling. this has to keep very thin so it does not intrude on inboard side of the fair damsel in the stern carving, sketch to the upper right. And the bottom sketch shows just how close the sliding door (slid aft position in dotted lines) comes to the end of the panelling.

 

Oh, and the sash window to the wardroom one deck below has a little pocket between the transom and the aft most mouldings, so it can slide up 8" for some fresh air.

 

There was some very sophisticated design work going into this finish joinery at the stern. There is something kind of thrilling about this detective work, reconstructing the design of someone working 270 years ago. It is almost like sitting there with him at the drawing board, as he thinks through how to make all of these parts interact nicely with each other.

 

Mark

 

 

 

 

 

Edited November 1, 2020 by SJSoane

[dvm27]

Although I've no doubt that you will tweak the design 12.75 times before settling on a a final version I very much enjoy seeing how an architect's mind works. Most of us are happy to achieve something that looks right. You are only happy when the design is not only aesthetically pleasing but actually works from a human perspective.

[SJSoane]

Hi Greg, when I started looking at the finish joinery and bulkheads, it began looking like more familiar territory for me. Besides the human factor, I realized that much of the detail was inspired by Classical architectural design, in the columns and pilasters on the bulkheads particularly. Some contemporary models and drawings I looked at were pretty fanciful in their interpretations of the traditional Classical orders (like Tuscan, Corinthian, Doric, Ionic), while others were pretty accurate. The rules in architecture are very precise in proportions and shapes of mouldings. I suppose the officers came from an upper class used to living in Classical houses, and wanted something like that in their own quarters.

 

So finally moving on from drawing, I fabricated more beams while the garage is not too cold to work on the router table. I refined my jig, and it worked like a charm to crank out all of the upper deck, forecastle and quarterdeck beams. 

 

The jig has to do several things: 1) cut a smooth arc on the upper surface; 2) cut a smooth arc on the lower surface concentric to the upper surface; and 3) cut these surfaces with the correct angle at every beam location, which gets increasingly steeper towards the bow and stern due to the sheer.

 

Getting the right curvature for the beams on each deck level starts with finding the rise at the center of the beams at their widest point on the given deck. This comes from the drawings or specifications. I then drew this curve on blocks of wood. I was fortunate to obtain an old set of drafting curves, which makes this simple; but arcs can also be constructed by traditional means.

 

 

I then cut the curves on the bandsaw, and rubber cemented sandpaper to one side. I used this to smooth the opposite side, then put sandpaper on the smooth block to clean up the first block. When this is complete, you have a set of convex and concave blocks at the correct radius, which are very accurate and smooth arcs because rubbing one lengthwise against the other cleans up any irregularities and forms a perfect arc of a circle. I use these all the time for sanding edges of planking, leveling decks, etc.

 

 

For the deck beam jig, I used these to form a convex and concave edge on a piece of plywood, I did this by attaching them to the plywood with double sided tape, and then running them against a piloted router bit on a router table. I then installed spacers, first one with a straight edge for cutting the initial convex edge, and then a curved one for cutting the concave edge (because the beam now has a convex edge after the first cut, which is placed against the concave stop for the second cut). You see below that I also rubber cemented sandpaper onto the jig, to provide more holding power so the blanks don't slide around when they are being cut.

 

 

I then hinged this template to another piece of plywood, also cut with the concave and convex edges, but recessed well enough away from the template above so it would not run into the nut on the router. Its job is primarily to provide a smooth and stable surface for sliding on the router table.

 

 

 

 

Two nylon screws allow me to adjust the angle of the template relative to the router table, thereby creating the angle needed for each beam due to the sheer. I used to use an angle gauge held against the drawings and then against the jig, but now I can get the angle off the CAD drawings, and use a digital angle gauge to measure as I turn the nylon screws.

 

 

And then finally, a third piece of plywood is screwed down with the black plastic knobs, to provide a clamp to the workpiece (a scrap strip of wood the same thickness as the beams is put in the clamp at the opposite side to keep the upper plywood piece parallel to the template). We see it set up here with the first cut ready to go.

 

 

And then the template is run against a piloted router bit in the router table.

 

 

It took a bit of work to get all of the jig parts to work, but once set up, it allows a very rapid and precise cutting of beams, each one tailored to its unique location in the sheer, and with parallel surfaces of the correct radius. Here are (from top to bottom) the beams for the forecastle, the quarterdeck, and the upper deck, including the half beams at the aft end of the upper deck.

 

 

I took extra care in marking out on each beam which edge should be convex, and also the direction of the angle on the end. The danger of a jig is that you automatically cut, remove, add next, cut, remove, etc. and it is so easy to get things turned around if you are not scrupulous about keeping everything in the same order. Having a visual clue to the correct orientation when putting in a blank helps keep things straight.

 

Once each set of beams for a given deck was complete, I took the jig apart, recut the concave and convex edges with the radiuses required for the next deck, and then put it all back together again.

 

Those with a mathematical outlook will realize that my upper and lower surfaces are not exactly concentric, since they are the same radius but the moulded distance of the beam apart. I decided at this scale that it would not be significant.

 

 

So my only remaining challenge is how to store all of these beams so they don't warp. How does one sticker so many small pieces?🙂

 

Mark

 

 

 

 

 

 

 

 

 

 

 

 

Edited November 7, 2020 by SJSoane

[druxey]

Wood will move (or not), no matter what you do! Hopefully it is well seasoned and will not move.

 

I like your built-in bevel refinement to the jig. When I cut my beams, I cu the convex edge first, then rout the concave side using a width stop, which keeps the curves concentric to each other. However, it does not produce a bevel unless one has a tilting table. I simply cut the fore and aft beams a hair thicker to allow for bevelling later. Your method is more sophisticated, Mark!

[SJSoane]

It has been a long time since my last posting. I started work on the gundeck spirketting and quickwork. Here are a few images in progress. Apologies for the dust I see in the photos; time for a little housecleaning!

 

I said earlier that in order to keep a clean line between the stained spirketting and the natural wood waterway, I would not be able to make individual strakes of this inner planking, level them down, and then stain in place. I did what Rob Napier discovered in his refurbishment of the 18th century model of the Princess Royal; he named this typical 18th century model builder technique SWOPEM- "Situation Where One Piece Equals many". I scored the hooked scarp joints onto a single strake for the spirketting. The butt joints between the pieces will be covered by the standards.

 

Laying out the sprirketting and joints:

 

 

The SWOPEM scored joints in the spirketting and quickwork:

 

The aft most piece of spirketting took such a severe twist and bend that I decided to carve it rather than steam it. My steaming always required clamping into final position, and there was no good way to clamp this:

 

and now to see the guns against the red:

 

 

 

[druxey]

Oh, you crafty creature, Mark! Certainly SWOPEM is a good way to go at times. And I agree that aftermost piece of spirketting is easier carved than bent.  However, it's covered by the knee anyway! She's looking very good indeed.

[garyshipwright]

Outstanding Mark. A real joy to wake up to seeing your masterpiece. Thank you for sharing her. Gary

[giampieroricci]

Wonderful model!

I really like the choice of colors of the various types of wood!

[SJSoane]

Thanks so much, druxey and Gary. You keep me going through some tedious steps. I am working on quickwork now, which was not so quick when I started; but it gets faster with repetition until my brain wanders and I cut too short...

 

Thanks also, Giampieroricci. The wood is all South American boxwood, and the red is a stain mixed in with the polyurethane finish. I can't stain up to a line because the stain travels through cells in the grain past the edge of the stain. That is why I have to stain first and install later.

 

Mark

[dvm27]

Great Swopem Mark! One of the models in the Kreigstein book has all the external planking scored onto the solid hull. The skill to execute that boggles the mind.

[Hubac's Historian]

Your scarfed spirketting looks like individually carved plank -  Beautifully done!

[SJSoane]

Thanks, Greg and Marc. Once I discovered the 18th century model builders did SWOPEM, it was an open door for me! But as you point out Greg, sometimes it was harder to do it in one piece than in many. The paradoxes of the world.

 

I have finished cutting all of the quickwork, and realized that now I have each piece off the hull for staining, I could drill for the hardware for the gun tackles and breaching ropes. So I started laying out the locations for the eyebolts and rings, and discovered interference with the knees and sometimes the standards. Here is a sample.

 

The rules for location I found in Allan Yedlinsky's great book, Scantlings of Royal Navy Ships; a note on page 151 locates the ironwork relative to the timber frame member adjacent to the port opening. The eyebolt for the port tackle is centered up and down on the port, 1/3 of the siding of the hull frame away from the port; and the eyebolt and ring for the breeching rope is a quarter of the height of the port above the sill, and 2/3 of the siding away from the port. This would ensure that the bolts run through a solid frame, and are offset in the same piece of wood to avoid splitting.

 

But following this rule puts the ironwork sometimes right on top of knees or standards. I can see putting in compass timber knees where necessary to avoid the conflict, as I have done in a number of places below, but on the standard #6 in the second image, the conflict is not fixable. The standard is located by the beam below. So, would the eyebolts in this case have been shifted to the other side of the standard, or bolted through the face of the standard itself?

 

Just when you think you have everything figured out.......

 

Mark

 

 

 

 

 

 

 

 

[druxey]

I suspect through the standard in such a case. The bolt needs to be ithrough solid wood, not just planking!

[SJSoane]

A little more drawing this morning, and I was able to clear the eyebolts and rings in all places but one, by casting the hanging knee a little further fore or aft of the port. The one remaining question is where standard #6 cannot move because it is related to the beam below. The ironwork either has to go on top of the standard, or moved fore of the standard as shown in the dotted line. Maybe the latter position is more consistent with everything else?

 

Mark

 

 

[SJSoane]

Thanks, druxey, I I must have been posting my comments when your's came through. Yes, the only way the dotted line position makes sense if if there is a hull frame in that location. I will have to look at that more closely. 

 

Mark

[druxey]

Are you sure that the hanging knee off 15 is to the aft side of the port? A little fettling of the standard's tupper arm , then the shorter route would be for it to locate between the standard and hanging knee 14A. If I were master shipwright I might be tempted to place 14A forward of the beam as well....