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

[SJSoane]

Now for some analysis of geometry and construction. 

 

Fifteen years ago, I built the stern frame based on my observation of the first Bellona model, and the Admiralty drawings. I did not understand what I was building at the time, I just followed what I saw. And now I understand as I get into the details.

 

In the model, you can see a couple of strange things. first, the tops of counter timbers with the dovetail joints stand proud of the horizontal transom tying them together. This leaves a gap of about 2". I now realize that these gaps provide a space for the bank of sash windows to slide up into. It means that the windows can be opened at the bottom by about 9"; not a lot, but better than no opening!

 

Second, the side counter timbers, effectively the end of the hull frame, project out from the bottom of the balcony upwards. that is, the bank of windows are recessed back relative to the balcony, creating a shadow line between the upper and lower parts of the stern. There is no functional reason I can see here; it is a visual trick to emphasize the sweeping serpentine curve of the balcony from one side of the composition to the other.

That creates some complication in how this is constructed. The green line shows the aft most edge of the side counter timbers, or hull. There is one moulding along the tops of the window bank, just under the balcony, in red below. Another moulding runs along the base of the balcony, in orange below.

 

 

when the balcony swings out from the stern, the two mouldings split from each other:

 

The next item I did not fully appreciate until I started constructing the stern is how its design needs to reconcile  two geometries working against each other; the upper sweep of the sheer, and the flatter sweep up of the decks (red lines below). The conflict shows a little where the gun ports cut into the sheer in odd places.

But at the stern they really need to be resolved. One of the most important, I discovered, is that the moulding at the tops of the windows must align with the sheer of the deck at its upper edge (orange line below), since it sits flush under the deck itself; but its lower edge must conform to the sheer of the hull (dotted red line below), since it turns the corner and runs along the side of the quarter galleries that align with the hull's sheer.

 

And then one more thing about the stern. In 2014, ten years ago, I posted a question about the stern geometry (posting #173). I noticed a discrepancy in the drawings relative to the roundup of the quarterdeck. If I continued the deck aft with its normal roundup, it was a couple of inches lower than the deck as it was shown coming out into the balcony. There was a lot of discussion around posting 173 as to whether this was a mistake in the drawing, or something else. I finally found in Steel that the quarterdeck does indeed increase its roundup as it approaches the stern balcony, to give a lighter, more springy feeling to the visual lines of the stern. So the drawing is correct, and this needs to be accounted for:

 

 

So, construction next!

 

Mark

 

Edited March 2 by SJSoane

[druxey]

Late to this discussion, but it triggers memories of many years ago when I was building Polyphemus, 64 guns of 1782. I also had a time puzzling out this geometry - particularly the jog aft at quarter deck level. This is not immediately apparent in the sheer plan. It looks like "By George, he's got it!"

[shipmodel]

Mark - 

 

I don't believe that I will ever build such a large, complicated ship in the years remaining to me.  But your explanations have been fascinating and your drawings and illustrations have been brilliant.

 

Thanks for a diverting, engaging, and educational build log.

 

Dan

[SJSoane]

Hi Dan and druxey,

Your recent kind comments made me realize that I have not posted in a long time, and you inspired me to catch up a little.

 

I am still a little slow after my health issues, although things are definitely on the mend.

 

What has slowed me the most is that this stern is exceedingly complicated to construct, I have discovered. I have had to do more drawings to figure out how everyone will go together, and then fabricate a number of structural members that will eventually form the framing for the quarter galleries. This has involved cutting parts with a roundup like the decks, with a curve fore and aft, some with a slope matching the aft face of the quarter gallery and some others with a slope matching the outboard face of the quarter gallery, and then rabbets on these multi-curving surfaces for windows top and bottom. I have made many mistakes so far, keeping it all straight....

 

First, determining the location and widths of the quarter galleries relative to the hull. There are no original drawings of the Bellona stern, and only the most basic hint of the quarter gallery plan, showing the curvature of its outboard edge. So I had to reconstruct from photos of the models, and some geometry.

 

The key to reconstruction is that the stern windows to the upper deck are evenly spaced, and they all converge at a point some distance above the hull. In the drawing below, the 4th orange line to the right is the end of the hull itself, while the blue line is the outer edge of the quarter gallery window. Keeping the spacing of the windows the same, and on the correct angle, this then tells me the width of the quarter gallery at various levels:

 

Here are windows mocked up to determine the lengths of the stern mouldings forming the various levels of the stern:'

 

 

 

The following drawing shows some of the geometry involved in the structural parts. Nothing is true size in these, since they are bending and sloping. Everything will have to be measured on the hull itself, and parts fitted to previous parts.

 

Then I had to determine how to make the windows, and therefore how to frame for them. I mentioned before that Chuck has laser cut some beautiful prototypes for me, using my drawings. I have decided on .029" plastic, painted to match the boxwood of the hull. Using mica for the glass (.005" thick) means that I can build windows thin enough to bend to fit their curving surfaces. But to do that, I had to devise a way to trap the window frames and mica between the outer moulding and a rabbet in the structural frame behind.

 

You can see below how the structural frame behind the windows has a rabbet, and once the mica and window frame are put in, the moulding then presses the two tightly together. That is the plan, anyway! You can see in the drawing below how the structural frame (with the crossed lines) slants inward, and also of course bends from the outer edge of the quarter gallery at the stern, to hitting the hull further forward. The frames on the aft side of the quarter galleries also have to round up like the decks.

 

 

 

To make these curved and sloped parts, I made an adjustable angle table for my Foredom. I pre-cut each curve with a jeweler's saw, then sanded to the line making a sloped, curved, edge. I then refined these edges on an appropriately curved sanding stick.

 

 

I then set up a fence on the Foredom with the handle and bit turned upside down, to act as a router table with the bit coming up from underneath.

Below you can see the rabbet marked on the end of the structural piece (temporarily glued to the edge of a plywood handle, to keep my fingers away from the bit). The tops of the windows required an acute angle--made with a dovetail bit as below-- and the bottoms required an obtuse angle made with a straight bit.

 

 

 

And here is one set of frames (port and starboard) and moulding blanks for the sides of one level of the quarter gallery.

 

All of the sides are now complete. I am now working on the structural frames on the aft end of the quarter galleries.

Maybe someday I will be able to start putting some of this together!

 

Best wishes,

 

Mark

 

 

 

 

 

 

 

 

Edited June 10 by SJSoane

[druxey]

Good advances there, Mark! I made a considerable quantity of projection drawings for the stern galleries. (This was 30 years ago before the various computer graphic programs we now have were available!) I found the sane way was to expand in a single plane at a time. For instance, first correcting the tilt of the stern tier of lights, then flattening the round aft. Of course, you can do this in seconds and a few clicks now!

 

Cheers,

D

[albert]

Great beautiful work SJSoane

[Hubac's Historian]

Mark, it is amazing how complicated this whole flowing structure is.  You are doing a masterful job of drawing out all of the details and sussing through all of the convergent planes.  I’m glad to hear, also, that you are feeling more like yourself these days.

[michael mott]

Mark having just spent the better part of the morning catching up with your work on this amazing model, I fell like going and sitting in the corner and rethinking my approach to all things model related. Your tenacity to solve the methods of construction and details are enviable, I wish I had the patience.

Have you found a bit of proper English boxwood yet? if not send me a pm as I have a reasonable amount still and depending on the size of what you need I could send you some.

I have really enjoyed catching up on your build, love all the ways you use your tools and organize them as well.

 

Michael

[SJSoane]

Sorry for the long delays in posting. Lots going on in the summer outside the shop, and progress is very slow when I am in the shop!

 

Thanks so much, druxey, Albert, Marc and Michael for your kind thoughts. These keep my spirits up as I plod through the slow business of building the quarter galleries.

 

I have found it most effective to treat each level of the quarter galleries--at levels of both decks and railings--as a little platform with the structural rails temporarily glued to the outside edges. This allows me to keep the correct curve and location where the rail hits the side of the hull.

 

 

I used the mouldings at the stern as guides for the angle at which the platforms are temporarily glued in their locations, and carefully measure from the channel wale to find the correct sheer fore and aft. I check with a height gauge to ensure the platforms are symmetrical port and starboard.

 

I recalled reading many years ago that the fore and aft curves of the quarter galleries must be parallel to each other, although offset fore and aft by the amount of the slope of the aft windows. This is to ensure that the vertical mullions are parallel to each other and not in different planes which would create a twist in the plane of the windows themselves. Not good for flat sheets of glass! So I was very careful in maintaining the angle at the aft, outboard, intersection of the two structural rails. This ensures that the platforms above and below the window will remain parallel to each other even as the sides of the hull at different heights vary in their angles due to the tumblehome.

 

This should become clearer in later posts, when I start to lay out the mullions themselves.

 

Here are more images of the platforms slowly coming together. The first one below also shows how the structural beam at level C had to be notched for the window frame to slide up behind the stern moulding. It is actually a blind window, but I needed to do this so my windows would remain constant in their heights across the full width, for ease of construction.

 

Best wishes,

 

Mark

 

 

 

 

 

 

 

 

 

 

[druxey]

Looks good to my eye, S.J. When building these tricky structures, I often wonder who was cunning enough to invent them?

[Mark P]

Good Evening Mark;

 

I am very glad to see that you are posting again, and that you are at such an intricate, complicated stage of the work. Quarter galleries are a tremendously difficult area to do well, but they are one of the first parts to catch the eye when looking at a model; so it is vital to be absolutely happy with each stage, and to discard what does not meet your exacting standards. I don't doubt for a moment that yours will up to your usual extremely high level of excellence; a standard nearly all of us can only aspire to, and never equal.

 

I wish you all success and good health in continuing what you have begun in such an exemplary tour-de-force.

 

All the best,

 

Mark P

[Hubac's Historian]

As is always the case, Mark, your fore-thought and experimentation in figuring out these extremely complicated structures never fails to produce top results.  I am always inspired and educated when I visit, here.

[SJSoane]

Druxey, I have been pondering the same question; how did these stern structures come about? I am no nautical historian, but just looking at ships from the previous century the structures were significantly different. So someone decided they needed updating! Although we are so used to seeing the mid eighteenth century sterns that they seem natural and inevitable, the most curious part for me is the quarter gallery curving to the hull at the fore edge, leaving a thin and useless wedge of room inside as one moves forward within it. Who thought that was a good idea?🤨

 

Thank you Mark and Marc, I always appreciate your kind thoughts--they keep me going when I don't always feel like working!

 

I have been very slow in progress lately, although now working every day for a good part of each day. I ran into one of those: "ok, I will do X; wait, before I can do X I need to do Y; and before I can do Y I have to do Z.....

 

The next step in the quarter gallery construction is to glue the lower finishings in place, so I can begin to built up the structure from that foundation. But before I can glue the lower finishing in place, I need to paint the black main wale, which the lower finishing sits upon. And just when I was taping out the black wale, I realized that the gun ports at the stern cut deep into the wale, revealing a thick edge of wale around the port. What color is that, I wondered? After looking at the Bellona second model, I realized that the edges of all the ports are painted red, right out to the face of the planking. I had left the edges natural.

 

So, wanting authenticity, I had to paint the edges of all the ports red before I could proceed. And then the fun began.

 

Almost 20 years ago, I had decided to create the red color by putting Transtint dye into my poly gel natural finish. It makes a great, somewhat weathered, red. But when this is applied on wood it is difficult to keep a clean edge; the dye tends to run into the pores of the wood, under the mask, leaving unsightly streaks.

 

After some failed experiments with the red dye idea, and also with a paint brush, I decided that I would be better off spray painting the edges of the gun ports. I found an old bottle of Floquil Caboose Red that best matched my dyed wood, and then masked the entire hull leaving only the edges of the exposed planking at each port. This took me way more time than I would ever have expected--maybe five days of masking.

 

Then it was a quick business with an airbrush, giving really clean lines.

 

Now the starboard side, and then the black wales, and then the lower finishings, and then the quarter gallery structures....

 

Good thing I am not in a hurry...

 

Mark

 

 

 

 

[SJSoane]

The starboard masking for the gunport edges went much faster, only 3 days of masking.....

 

So now I am ready for the black Wales at last. Just one more round of final sanding for the sides (some masking tape ghosts to clean up).

 

As I was preparing to paint this last round of red, my airbrush compressor suddenly stopped pushing air. I looked around for the cause and noticed that the hose connecting the compressor to the filter had completely severed at the fitting pressed onto its end, either rotted away or twist stressed. I did my best to cut a clean edge on the hose and reattach it to the screw fitting, only to discover that there were slow leaks at several other cramped points in the hose. I taped those up, completed the painting, and then the other end of the hose severed from its pressed on fitting. I had to order a new hose, and the black wales will have to wait for delivery. The compressor was probably 15 years old, maybe the shelf life for hoses....

 

Mark

 

 

 

 

 

 

[dvm27]

Your meticulous research and execution really shows up in the above photo. Not a waver can be seen on any of the runs of planking or moldings. Such a pleasure to watch this model come together!

[druxey]

Some lovely color accents now, Mark! A helluva masking job, though. Do treat yourself to a new hose.

 

The tapered wedge shape forward of the quarter gallery allowed the occupant a view forward along the ship's side while ensconced....

[jdbondy]

"While ensconced"...love it!

[SJSoane]

 

Thanks, Greg, for the kind comment on the project. And druxey, thanks for clarifying why the quarter galleries shape to a point at their fore end. While "ensconced", the Captain could keep an eye on the trim of the sails from a different perspective!

 

After a great deal of fussing with airbrush and compressor, I finally got the starboard wale painted. It is currently flat black, but will have a coat of the transparent poly gel over it, the same finish as for the rest of the hull. it will give a little sheen and lighten the color, according to my test pieces.

It looks nice with those mean black cannon poking through the ports...

 

 

Another fun masking process before the paint. I had to turn the hull on its side so I could spray the under edge of the wale, avoiding using an airbrush pointed upwards....

 

 

Next, the port wale, and then back to constructing the quarter galleries!

 

Mark

 

 

[jdbondy]

Your polyurethane topcoat, is it oil based or water based? I just used some water based polyurethane on a non-ship model project, and the question arose in my mind.

[Chuck]

absolutely stunning.  Really nice progress.

[Stuntflyer]

42 minutes ago, SJSoane said:

It is currently flat black, but will have a coat of the transparent poly gel over it, the same finish as for the rest of the hull.

One gorgeous ship, Mark! Can you tell me what name of the poly is?

 

Thanks,

 

Mike

[SJSoane]

Hi JD and Mike, I use General Finishes Gel Topcoat, Wipe On Urethane in Satin (cleans up with mineral spirits). It is a rub in, rub off finish. On the raw boxwood, it darkens it up and gives the slightest sheen. You can see it on the exposed futtocks below the wale, compared to the raw wood above the wale. It is a little challenging to apply to all of the ledges and carlings on the exposed deck frames; I haven't quite figured out an efficient way to apply it between the spaces, and then rub off well into the corners. 

 

I used Vallejo Model Air acrylic for the black.

 

Thanks, Chuck, for the kind comment. I am trying some Vallejo colors for spraying the plastic window frames to look like boxwood. I'll let you know what works best!

 

Mark

 

 

[SJSoane]

And in just one more day, the port side is painted. Much faster second time through.

 

I got more efficient with the masking:

The port side. I am glad I took the time to paint the edges of the planking at the ports with red, giving a very clean junction where they cut into the wale at the stern.

 

And very satisfying to see both black wales from the bow:

And now back to the quarter galleries!

 

Mark

 

 

 

 

[yvesvidal]

Absolutely superb.

 

Yves

[scrubbyj427]

Wow! That is just beautiful! Excellent work!

[mcpwilk]

A work of art.

 

Mike

[SJSoane]

Thanks so much, Yves, Scrubby and Mike, for your kind comments.

 

I have not posted in a while, because I was having to work out a few new challenges as I  begin to work my way up from the lower finishing.

 

First challenge: how to frame and plank the quarter gallery surface above the lower finishing? This is the blue band between the lower stool rail and the lower gallery rim, below:

 

It curves, twists, and fays to the side of the hull with a shape I could not determine in advance. Also, how would I plank the inner surface once the framing was in place? I remember seeing a model of the Superb, with its stern in pieces. With the quarter gallery lying on its side, it is possible to see that this piece appeared to be cut out of a single block. So I decided to do that. This way, I could fair each edge to a known surface, and thereby discover the actual form. Then I could scribe the planking on the outer surface later.

 

First, I drew the top and bottom profiles as best I could determine from my drawings. I pasted one on top and bottom of a blank, then shaped the outer surface to match each edge. After refining while checking against the actual model, I came up with the shape and how it fit against the hull:

 

Here you can see the wicked twist in the shape.

And here you can see it pinned temporarily in place (this is a trial blank I made out of basswood, to quickly see if this would work):

 

Next, I had to work out the shapes of the mouldings on its lower edge, which is the knuckle between the lower and upper counters. It took a long time to realize that the moulding here is the not the same profile along the counter as it is along the side of the quarter gallery. These have to miter cleanly at the corner, but they are very different profiles.

 

Looking carefully at the sections below (cut through the quarter gallery on the left, and through the knuckle of the lower and upper counters on the right), you can see that the face of the quarter gallery is almost vertical, whereas the face of the upper counter slopes back quite substantially. I finally realized that in order for these to miter together, the important thing is the thickness of the moulding in both locations has to be the same at the top and the bottom.

 

Also, the top edge of the moulding on the quarter gallery has to align with the roundup of the moulding along the counter, whereas the top edge of the moulding on the counter has to align with the sheer of the external hull planking.

 

A little geometry worked out the profile for each, so they would fair to their own surfaces while mitering cleanly at the corner.

 

 

 

 

Next challenge: how to cut the profiles with a scraper? All of my earlier mouldings were just straight blanks cut on a flat surface and later gently bent to the sheer of the hull. But these mouldings are on sharply curved surfaces. I decided for reasons of accuracy in the overall construction to shape the blanks of the mouldings to their final curves, rather than making flat mouldings and bending them to the shape later. I saw too many places for things to get out of alignment if I did not make accurate blanks to start with.

 

So how to scrape a shape on a curved blank, which also has top and bottom edges not at right angles to the face of the moulding? I decided to make formers, to which I could temporarily glue the blank, and use the former to guide a fence on my cutter holder.

 

 

Here you can see the cutter for the moulding on the quarter gallery. The face of the fence is at an angle to the cutter, corresponding to the angle of the top of the moulding. You can see the necessity of the angle, because if the fence were at right angles, the cutter would tear up the shape as it progressively works its way down.

 

Here is the scraper at work:

And here is the moulding temporarily spot glued in place, to check its fit and final size. This is also the final faux planked piece, cut out of boxwood after the basswood trial worked:

 

The flat piece faintly labeled STAR BD eventually goes away. This was a temporary former for getting all of the railings in their proper location and shape. The rabbeted rail is the base of the window frames still to come.

I have frankly struggled with making scraper profiles, until Chuck suggested I just make them out of brass rather than fussing with tool steel, softening to file and hardening to use. Each profile here is only going to be used for one or two parts, and then can be thrown away. So I worked out profiles based on the two semi-circles, and accurately drilled these in the brass blank on a mill:

Then I carefully filed the remainder of the profile using a needle barrette file and a jewelers clamp to ensure right angles to each side. This is important for the cutter to work the same in both directions.

 

At last, I got accurate profiles with no heat treating challenges. Thanks, Chuck!

 

Now on the port side, and the big moulding along the knuckle here.

 

Best wishes,

 

Mark

 

 

 

 

 

 

 

[druxey]

Nice and methodical. My variation on your approach was to carve the piece as you have, but thinner, then plank the outer surface with very thin stock.

[TJM]

11 hours ago, SJSoane said:

Thanks so much, Yves, Scrubby and Mike, for your kind comments.

 

I have not posted in a while, because I was having to work out a few new challenges as I  begin to work my way up from the lower finishing.

 

First challenge: how to frame and plank the quarter gallery surface above the lower finishing? This is the blue band between the lower stool rail and the lower gallery rim, below:

 

It curves, twists, and fays to the side of the hull with a shape I could not determine in advance. Also, how would I plank the inner surface once the framing was in place? I remember seeing a model of the Superb, with its stern in pieces. With the quarter gallery lying on its side, it is possible to see that this piece appeared to be cut out of a single block. So I decided to do that. This way, I could fair each edge to a known surface, and thereby discover the actual form. Then I could scribe the planking on the outer surface later.

 

First, I drew the top and bottom profiles as best I could determine from my drawings. I pasted one on top and bottom of a blank, then shaped the outer surface to match each edge. After refining while checking against the actual model, I came up with the shape and how it fit against the hull:

 

Here you can see the wicked twist in the shape.

And here you can see it pinned temporarily in place (this is a trial blank I made out of basswood, to quickly see if this would work):

 

Next, I had to work out the shapes of the mouldings on its lower edge, which is the knuckle between the lower and upper counters. It took a long time to realize that the moulding here is the not the same profile along the counter as it is along the side of the quarter gallery. These have to miter cleanly at the corner, but they are very different profiles.

 

Looking carefully at the sections below (cut through the quarter gallery on the left, and through the knuckle of the lower and upper counters on the right), you can see that the face of the quarter gallery is almost vertical, whereas the face of the upper counter slopes back quite substantially. I finally realized that in order for these to miter together, the important thing is the thickness of the moulding in both locations has to be the same at the top and the bottom.

 

Also, the top edge of the moulding on the quarter gallery has to align with the roundup of the moulding along the counter, whereas the top edge of the moulding on the counter has to align with the sheer of the external hull planking.

 

A little geometry worked out the profile for each, so they would fair to their own surfaces while mitering cleanly at the corner.

 

 

 

 

Next challenge: how to cut the profiles with a scraper? All of my earlier mouldings were just straight blanks cut on a flat surface and later gently bent to the sheer of the hull. But these mouldings are on sharply curved surfaces. I decided for reasons of accuracy in the overall construction to shape the blanks of the mouldings to their final curves, rather than making flat mouldings and bending them to the shape later. I saw too many places for things to get out of alignment if I did not make accurate blanks to start with.

 

So how to scrape a shape on a curved blank, which also has top and bottom edges not at right angles to the face of the moulding? I decided to make formers, to which I could temporarily glue the blank, and use the former to guide a fence on my cutter holder.

 

 

Here you can see the cutter for the moulding on the quarter gallery. The face of the fence is at an angle to the cutter, corresponding to the angle of the top of the moulding. You can see the necessity of the angle, because if the fence were at right angles, the cutter would tear up the shape as it progressively works its way down.

 

Here is the scraper at work:

And here is the moulding temporarily spot glued in place, to check its fit and final size. This is also the final faux planked piece, cut out of boxwood after the basswood trial worked:

 

The flat piece faintly labeled STAR BD eventually goes away. This was a temporary former for getting all of the railings in their proper location and shape. The rabbeted rail is the base of the window frames still to come.

I have frankly struggled with making scraper profiles, until Chuck suggested I just make them out of brass rather than fussing with tool steel, softening to file and hardening to use. Each profile here is only going to be used for one or two parts, and then can be thrown away. So I worked out profiles based on the two semi-circles, and accurately drilled these in the brass blank on a mill:

Then I carefully filed the remainder of the profile using a needle barrette file and a jewelers clamp to ensure right angles to each side. This is important for the cutter to work the same in both directions.

 

At last, I got accurate profiles with no heat treating challenges. Thanks, Chuck!

 

Now on the port side, and the big moulding along the knuckle here.

 

Best wishes,

 

Mark

 

 

 

 

 

 

 

This is fantastic! Thanks a lot for going into details on this!

[Hubac's Historian]

Mark - beautiful progress, as always.  I like your approach to constructing the quarters.  You are so good at getting the complex geometry right.

 

As you work your way up the quarters, you might find some useful ideas in Siggi’s Tiger build, beginning with post #1003, page 34.  Like you, Siggi is brilliant with the smallest details: