2nd rate London 1656 – the art of the shipwright

[Waldemar]

 

Persuaded by a clever ruse by Martes, one might say - a specialist in finding both important and interesting material on the history of shipbuilding in the on-line archives, I became more interested in the plans of the 2nd rate ship London of 1656 (built by John Taylor), which were apparently made during the ship's years of service.

 

Be that as it may, the drawings of this ship deserve the utmost attention, as they come from a period from which hardly any shipbuilding plans have survived, so it is not entirely clear how exactly the conceptual methods known more or less theoretically from the textbooks were actually applied at this particular time.

 

Ship dimensions (taken from British Warships in the Age of Sail 1603-1714 by Rif Winfield):

as built: 123 1/2 (keel) x 40 x 16 1/2
as girdled: 123 1/2 x 41 x 16 1/2

 

The plans are promising, however already a preliminary graphical analysis of the plans has shown that the contours of the frames have been drawn in a very irregular way, as if the room and space value was randomly varying along the length of the hull, which clearly cannot be the case.

 

This is a fatal circumstance, as it makes the reconstruction of the conceptually extremely important longitudinal guides (rising and narrowing lines of the floor and breadth) very difficult, if not possible at all, on this basis. In other words, it is difficult to establish whether and which guides were fragments of circles, ellipses or logarithmic curves. Certainly, one can refer to written works, but the point is, after all, to read it from the drawing, not to arbitrarily select their shape from other sources. But there is perhaps no other way...

 

And here are those drawings:

 

 

 

[Jaager]

Is there a profile plan?

 

The numbering is the key to placing the stations.  Aft the stations are at two bend intervals. 

I cannot discern if the last is 22-23  or 22-24.

 

The Fore is a bit fast and loose:

0 -F  =  6 bends 

F-H =   2 bends

H-K =  2 bends  (often I was dropped - too similar to 1)

K-M = 2 bends

M-O = 2 bends

O-P = 1 bend

P-Q = 1 bend

Q-R - 1 bend

R-S

S-T

 

Now all you need is R&S.

This ship would definitely qualify for Navy Board framing. 

There is no space between the bends.  The timbers are sided wider than the prototype.

My problem with using it =

The floor is significantly longer than than its usual ~60% - the head is above the turn of the bilge. 

The area of stock required and the loss to waste makes this expensive.

F1 is also long and involves a serious arc.

It works for miniature scales.

 

I could easily reproduce a fully framed hull  just from what you have above.

For me, no reconstruct of Body/cross section is necessary.  It is all there.  The mould loft did not need anymore than this to get timber patterns for the shipwrights.  They did not need every line of every frame.  The Stations alone were sufficient.  They could fake the transitions and so can I.

 

 

I would not know any of the important details that a profile provides though:

Deck locations

Port locations

Wales

Rails

Channels

 

 

 

 

Edited January 19, 2023 by Jaager

[Martes]

That's the thing, there is no profile. There is only a midship section and ordnance list.

 

The last pair aft are 23 (left) and 24 (right), they are alternating.

 

 

 

On the link to RCT London's images are at slightly higher resolution and noticeably sharper.

 

https://www.rct.uk/collection/1047387/the-ship-london-on-her-first-voiadge

 

And direct links to images:

https://www.rct.uk/sites/default/files/collection-online/7/2/498611-1418048073.jpg

https://www.rct.uk/sites/default/files/collection-online/5/a/498612-1418048122.jpg

https://www.rct.uk/sites/default/files/collection-online/6/0/498610-1418048055.jpg

https://www.rct.uk/sites/default/files/collection-online/3/b/448394-1393596039.jpg

 

 

Edited January 19, 2023 by Martes

[allanyed]

https://www.rmg.co.uk/collections/objects/rmgc-object-136130 for a VDV portrait. 

 

Don't know that the following will help but if her keel length was 123' 6":

 

For the R&S, one could use the information in Goodwin's The Construction and Fitting of the English Man of War for a second rate ship.

Per Deane, the moulded beam is the length of keel X 3/10      123'6" X 0.3 = 37.05 feet

Depth of keel = 1/2" for each foot of moulded beam X 37.05 =  18.5"

Siding of frame = 5/7 X depth of keel     =  12.86" 

 

Assumimg an average space between frames of about 2" the R&S is 12.86+2+12.86+2 = 29 3/4"

 

Someone might want to check my math just in case🤔

Allan

 

Edited January 19, 2023 by allanyed

[Waldemar]

 

Martes, thanks for posting the rest of the graphics. Admittedly they will not be needed for the reconstruction itself, so I did not post them, but they indeed may be of interest to others.

 

Allan, thanks for the calculations. I am still wondering if and how such data can be effectively used in this case, and in particular room & space. The problem is that evenly distributing the ribs as drawn on the original plan along the hull, according to any regular room & space, would result in a wavy hull.

 

One can ignore this circumstance and start building the model according to the original contours, but such an attempt will only end up with a hopelessly wavy as well as generally spoiled hull, because the original plans are also deformed in other ways. Anyway, this thread is not so much about woodworking, but it is rather on the underlaying concept employed by the creator of these 17th century plans.

 

* * *

 

Relatively the easiest part of the job is recreating the ship's master frame. It's hard to imagine a more classic shape than this one. As it happens, the proportions of the design grid perfectly match the ship's known dimensions.

 

 

 

[trippwj]

Are those radii set by trial & error or based on one of the extant sources?

[Waldemar]

 

Right, good question. If you mean the radii of the main frame, they were read by me from the original drawing by fitting circles of different diameters to the original curves.

 

In contrast, the early modern shipbuilder chose these radii arbitrarily, within the limits imposed by past experience, the geometrical constraints of the designed ship's hull, the properties of the available material and the result he wanted to achieve (e.g. greater capacity or better seaworthiness of the ship, etc.).

 

 

Edited January 20, 2023 by Waldemar

[trippwj]

Your approach would be what I called trual & error (Adams refered to it as "by construction").

 

I doubt the shipbuilder of the 17th C chose the radii "arbitrarily". Yes, it was based on the institutional knowledge of the builder (this era was the beginning of the split between design and construction) - the conceptualization of form was seperately applied to the construction via preplanned hull shapes rather than application of a master frame.

 

For our subject, the test of methodology is how well the master mould fits frames fore & aft. This specific set of drawings is awkward as they are not plans but, rather, appear to be generated from off sets at less than standard intervals. If geometric radii were used they should fit over other stations accomodating narrowing (can be found from these drawings by measurement) and rising (not quite so easily found but diable). To draw those on a profile equal intervals longitudinally are nice to have but not essential - use of a spline or similar flexible batten would allow fitting a smoothed line to the points available.

 

Edited January 20, 2023 by trippwj

[Waldemar]

 

4 hours ago, trippwj said:

Are those radii set by trial & error or based on one of the extant sources?

 

Oh...! Almost forgot... This should be perhaps clear to everyone, but to lead you out of this strange misconception I will also add that these very plans of London 1656 are extant sources themselves, and one might even say – extant sources par excellence. Together, of course, with the values of the specific radii they contain.

 

I am also pleased to inform you that the graphics in period works of technical nature are as important as the text they contain, exactly as they are in modern technical documentation.

 

Wayne, is there anything else I can do to help you?

 

 

[trippwj]

Thank you for the update. I fully understand the value of the drawings, however my curiosity concerns how well the radii you calculated correlate to any of the existing treatises from the period. Would be beneficial in determining whether the treatises actually represented common practice or were more specific to the author.

 

Hopefully at some point in time the full wreck site will be excavated and documented. Currently only a few floors seem to be exposed.

 

Many thanks.

[Martes]

The basic intention, to my understanding, is to extract as much data from these plans as possible. It's not that we have many alternatives. The fact that using the base dimension we get quite round and feasible values for radii of the curves may, possibly, indicate the process might be going the right way.

 

Hopefully, @Waldemar would be able to figure out what is wrong with the sequence of the stations.

 

There were some concerns about the condition and deterioration of the wreck, I vaguely remember.

Edited January 20, 2023 by Martes

[Mark P]

Good Evening Martes;

 

You are quite correct in your recall; the wreck is located close to a busy shipping lane, and the prop-wash from passing ships, and tidal scour of the exposed remains, are causing rapid deterioration. A further unhelpful factor is the refusal of Historic England to allow any items to be removed from the wreck site. This means that artefacts and ship's timbers can be observed when they are exposed, but cannot be removed to preserve them. Important pieces of the ship's structure and artefacts have been noted, which have now disappeared without any chance to make an official record or attempt to preserve them. Which is very frustrating for all those who want to know more about this time capsule's contents. 

 

It is possible that the circumstances have now changed, and recovery of at-risk artefacts etc is now allowed; my information is not up-to-date; but I am not overly hopeful of anything involving a complicated bureaucracy's minions. Especially as many of them are now, post Covid, probably 'working' from home. 

 

There is a charity devoted to preserving/raising the remains, 'Save the London', managed by the Nautical Archaeology Society. 

 

All the best,

 

Mark P

[Waldemar]

 

From the initial chaos, something is beginning to emerge, and although these plans are not a demon of precision by today's standards, it is already clear that they have been drawn according to the rules of the art by a professional.

 

One of the main goal of this reconstruction is to find the correct types of curves used by this shipwright, or in other words, to find the specific geometric or mathematical methods employed to profile these crucial to the conception longitudinal guides. This is considered far more both important and interesting than the comparatively secondary, case-to-case variable proportions.

 

Below just reverse-engineered line of the breadth aft is shown. The found method is identical to that presented, for example, by Dassié in his L'Architecture Navale published in 1695. The application of this particular transformation means that the hull had a slight transverse kink at the main frame. The scattered points close to the blue arc in the top diagram correspond to the original contours of the uncorrected frames.

 

 

 

[Martes]

11 minutes ago, Waldemar said:

slight transverse kink at the main frame.

 

Is it normal and would it be visible, or the outside planking would cover this?

[Waldemar]

 

The bend is not very dramatic, you can see it in my post here:

 

https://modelshipworld.com/topic/32948-bow-shape-of-le-françois-1683-and-la-néréïde-1722/#comment-935696

... and this was most likely smoothed out during actual construction.

 

 

Edited January 21, 2023 by Waldemar

[Waldemar]

 

Typically, designers used proprietary methods in a fairly uniform way, so the shape of the breadth line fore was already somewhat easier to reproduce. The logarithmic curve was a perfect fit here, however, a curve composed of two circles of different radii, which approximates the logarithmic curve quite well and commonly used by English designers up to this time, could also be chosen with almost equally good results.

 

I have also updated the master frame diagram. Here the only difference is the correction of the lower breadth sweep from 13' 10" to a round 14 feet.

 

 

 

 

Edited January 23, 2023 by Waldemar

[druxey]

Interesting extrapolation of the half-breadth in the body plan producing a good fit!

Edited January 21, 2023 by druxey

[Jaager]

Going by the instructions in Deane   so at least up until 1670:

The Body plan shape creation was only done three times during the design.

The three were placed on the WL plan and battens used to define the run.

The forward or aft cross sections could be moved to get the sort of entry and exit needed to theoretically favor speed or stability or capacity.

Doing a mental extraction from the above premise:

The complete Body plan was derived by plotting the points from the WL plan. 

The Diagonal plan was important to confirm that there were no problems in the run in the swimming body.

 

I am suspecting that those who were writing an advocacy of math formulas over shipwright intuition did not gain control until about 1700.  This being limited to English practice.

Edited January 21, 2023 by Jaager

[druxey]

I don't think 'proof diagonals' were used until well into the 18th century.

[Jaager]

@druxey  not on paper if you are correct.    But, if you look at the Navy Board framing models, the top edge and bottom edge of the overlap band look to me like they would serve the same purpose.

[Waldemar]

 

Jaager, what you have written is utterly misleading. You are confusing Deane with van Yk. You are confusing the English with the Dutch, and their practices. At least read Deane's work, it is in English. Apart from that, you can also read the numerous English works on shipbuilding that were written before Deane, both printed and unprinted.

 

But if you intend to create an alternative history of shipbuilding, no problem, you can always do that, best in your separate thread.

 

 

Edited January 23, 2023 by Waldemar

[Waldemar]

 

Consistently following the designer's paradigm, the rising line of the floor aft has been reconstructed (green line). At this point, it is now possible to apply a futtock sweep of a fixed radius tangentially to both arcs – the lower breadth sweep and the floor sweep (thick red lines).

 

 

 

And this is the reconstructed shape of this guide line as seen in the sheer projection:

 

 

 

[Waldemar]

 

This stage proved to be quite a killer, probably the most difficult in the whole reconstruction. The surprise is total, because the floor sweeps turned out to be arcs of variable radius. What a tremendous advance has been made in just a few sparse decades over the methods known from the English early modern written works of about 1600–1625 on naval architecture!

 

The first dimensional values are beginning to appear (still to be verified):

 

stempost radius: 40 feet (equal to the ship's width)
height of the rising line of the floor at the stem: 12 feet
distance between ribs: 3 feet

 

Actually, the underwater shape of all the frames is ready, it only remains to apply hollowing curves to them.

 

 

 

Edited January 23, 2023 by Waldemar

[Waldemar]

 

First, the parameters of the rising line of the floor have been modified slightly, by reducing its height fore from 12 to 10.5 feet. This is precisely half the height of this line aft (a typical value for the period).

 

The hollowing curves were applied to the fore body. From the frame '0' to the frame 'M' they take the form of straight sections tangent to the floor sweeps. The next ones are templates consisting of a straight section and an arc with a radius of 30 feet, which is the same as the radius of the futtock sweep. I have already described how to apply the hollowing templates here:

 

 

 

Looking at the changes in the position of the futtock sweeps in the individual ribs (thick red lines) it becomes clear why the name 'hauling down/pulling up futtock' has been attached to this method.

 

 

 

Edited January 25, 2023 by Waldemar

[Waldemar]

 

The application of the hollowing curves in the aft section of the hull is the same as for the fore half. Up to the frame '15' these are straight sections, and later templates consisting of a straight section and a 14ft radius curve.

 

 

 

[Martes]

What especially astonishes me here is level of data compression.

With these methods, how much numbers you have to memorize to correctly unpack the ship plan afterwards?

Edited January 24, 2023 by Martes

[Waldemar]

 

Well, don't really know how to answer this question, but if I understand it correctly, not very much. Over time, once you become more proficient, it's more a matter of getting a feel for the shapes/curves and where they can be applied, and the range of possible proportions. Other than that, it's very much a matter of using templates and geometric (graphical) transformations, where you don't really need to use many numbers, hardly at all. Simply put, it's the geometry rather than the numbers that count.

 

It may be added that it was not only possible to reproduce the way these plans were drawn, but it is now also possible to try to reproduce the actual shape of the London 1656 hull in a feasible way, and maybe some other ships from this period too. This would only require slight modifications to the curves at the very ends of the hull. For example, using exactly the same methods/transformations discovered here.

 

Overall, this method is very typical of English 17th century practices, only details were, and are still somewhat blurred. It was very advanced for its time, but by about this time, elsewhere in Europe and among leading designers, other concepts were already being born. But that's another story...

 

Below, the finished frame contours for the fore part.

 

Thank you for watching so far,
Waldemar Gurgul

 

 

 

[Martes]

17 minutes ago, Waldemar said:

Well, don't really know how to answer this question, but if I understand it correctly, not very much. Over time, once you become more proficient, it's more a matter of getting a feel for the shapes/curves and where they can be applied, and the range of possible proportions. Other than that, it's very much a matter of using templates and geometric (graphical) transformations, where you don't really need to use many numbers, hardly at all. Simply put, it's the geometry rather than the numbers that count.

 

I thought of peculiar nature of industrial espionage in this case. Usually one would think that to steal a design of a ship, a whole lot of documentation must be obtained and transported. But in this case the required information would fit on a scrap of paper or could be just memorized.  You just have to have somebody who knows what to do with them

 

There are 2 portraits of the ship by Van de Velde, one mentioned above by @allanyed:

 

https://www.rmg.co.uk/collections/objects/rmgc-object-136130

 

And another in a Dutch collection (visit the site for full size, the cat faces on the gunport lids are quite adorable):

 

http://collectie.atlasvanstolk.nl/data_nl.asp?q0=85301&startc=1&subj=27&bron=theslijst

 

 

The stern decorations on the second drawing are somewhat different those on the plan, and they are probably Commonwealth era.

Edited January 24, 2023 by Martes

[Waldemar]

 

Thanks Martes for additional pictures. With all this data, an attempt can already be made to reconstruct London 1656. The hull lines, according to the surviving plans, look as in the diagram below (no corrections have been made). It is quite certain that a few extra ribs (in the shape of the main frame) need to be added in the midship area. This is a fairly typical feature in ship design, both today and then.

 

 

 

[Martes]

7 hours ago, Waldemar said:

Below, the finished frame contours for the fore part.

I may be mistaken, but - purely visually - doesn't the transition around M appear to slightly distort the surface?