William Sutherland's concept of ship hull design, 1711

[Waldemar]

 

Being fresh from reading a very interesting publication on the Restoration yachts of Charles II, I looked again at William Sutherland's work, The Ship-builders Assistant, 1711, focusing on his method of shaping ship hulls.

 

In his work, Sutherland proposes shaping ship hulls so that the submerged part of the hull takes the form of an (elongated) egg. Such a shape is called a conoid. Both the characteristic feature and result of such a form are frames with the profile of a perfect arc (connected by an additional line(s) to the keel and the posts). In addition, the radii of these arcs are different for each frame.

 

The very idea of using variable radius arcs for different frames was not entirely new at the time. Anthony Deane (1670) had already partly used this way to construct the profile of the frames (although not in the textual description, but in his drawings), the way described in the work of Georges Fournier (Hydrographie, 1643) can also be interpreted in this way, and there is a plan of the English origin in the Russian archives, in which the arc of the greatest breadth is based on the principle of variable radius (the plan is undated, but the ship has clear features specific for the turn of the 16th and 17th centuries, in particular the square-tuck stern and the considerable rakes of both posts).

 

Below I have shown my graphical interpretation of the method proposed by Sutherland. It should be stressed here that the original plan has some drawing anomalies and, moreover, does not quite match its textual description, so reverse engineering methods were also necessary to interpret it.

 

 

 

 

As can be seen, the conoid is only the body defined by the upper arcs, while the floor curves (also called hollow lines, bottom curves etc.) were constructed by Sutherland in the conventional way. Here, too, it must be added that Sutherland chose one way of drawing the floor curves in his drawing, while in the textual description he proposed a different method for use in the actual construction of the ship in true scale, but still using a simple template made of only two parts: a line and an arc. As a result, the theoretical shapes (as in the plan) and those of the actual ship had to come out slightly different.

 

 

Edited September 23, 2022 by Waldemar

[michaelpsutton2]

what was the publication on the yachts?

 

 

[Waldemar]

 

This is actually a series of publications:

 

The Mariner's Mirror Volume 107, 2021 - Issue 2
A Restoration Yacht’s Design Secrets Unveiled: An examination of a ship model with reference to the works of William Sutherland
Effie Moneypenny & David Antscherl
Pages 164-187

 

The Mariner's Mirror Volume 104, 2018 - Issue 2
A Model of the Royal Yacht Henrietta about 1679: Description and identification
Effie Moneypenny & Simon Stephens
Pages 172-191

 

The Mariner's Mirror Volume 102, 2016 - Issue 4
The Royal Yacht Isabella of 1683: Identification and principal dimensions
Kelvin Moneypenny & Dorin Paul Bucur
Pages 400-416

 

The Mariner's Mirror Volume 100, 2014 - Issue 2
The Royal Yacht Henrietta of 1679: Identification and principal dimensions
Kelvin Moneypenny & Dorin Paul Bucur
Pages 132-146

 

 

[Waldemar]

 

 

 

[druxey]

Tangent points make more sense than tangent lines for replicability. However (particularly aft) the positioning of the template or mould leaves quite a bit open to interpretation as shown above.

[Waldemar]

 

1 hour ago, druxey said:

Tangent points make more sense than tangent lines for replicability. However (particularly aft) the positioning of the template or mould leaves quite a bit open to interpretation as shown above.

 

This is perfectly correct. "Connecting tangent lines" should rather be read as "Connecting lines tangent to two arcs". This was just a mental shortcut to keep the description as compact as possible. Sorry.

 

In the diagram below a possible unintentional modification of the frame profiles in the process of actual construction is visually shown, more as a matter of showing the phenomenon itself, rather than its specific magnitude, as this effect could have been reduced by placing auxiliary (guiding) lines in better spots. Either way, the problem of constructing the floor lines was invariably treated in the early works on naval architecture at the most as secondary or not at all, and the Sutherland's description is perhaps the first to deal with it clearly enough.

 

 

 

Edited September 23, 2022 by Waldemar

[Waldemar]

 

To be sure: all the above is based on Sutherland's description. It is quite obvious that he must have agreed to such a lack of precision. 

 

Wooden floor templates were to be aligned in the following way: they should pass through the point on the keel (or post), be tangent to the frame arc, and the point connecting its straight part with circular part should lie on the auxiliary line (drawn in blue in the diagram). All of this is shown in the third diagram above. Quite tricky, yet unambiguous.

 

 

Edited September 23, 2022 by Waldemar

[mtaylor]

7 hours ago, Waldemar said:

 

To be sure: all the above is based on Sutherland's description. It is quite obvious that he must have agreed to such a lack of precision.

Given the state of technology and tools of the time for the designer as well as the shipyard's, I would appear that "close counts".  But then, I'm not a designer or ship architect.

[Waldemar]

 

35 minutes ago, mtaylor said:

Given the state of technology and tools of the time for the designer as well as the shipyard's, I would appear that "close counts".

 

Mark, you have made me more precise on this point:    🙂

 

After reading a number of works on naval architecture, I get the impression that constructing floor curves was a routine procedure used by ship carpenters. And for this very reason the early designers and writers simply did not bother with it. In one of the source works, for example, one can literally read (from memory): "the floor curves are left to the discretion of the actual builder". 

 

Sutherland was obviously aware of these common practices as well, and this is the reason for the dualism in his work. On the one hand, it was easier to draw on paper the frames of two arcs and a line, but on the other it was more convenient to use templates in the shipyard. In other words, he knew that the lines he designed would be modified by the ways used the shipyard, so he gave the method to do it more easily and correctly. This is a fortunate circumstance, because through this we learn how these floor curves were practically determined on the mould loft.

 

 

[Waldemar]

 

And, in this state of affairs, it would not be out of place to ask today's modellers: which model would you like to make, according to the design lines or the real ones?    🙂

 

 

[mtaylor]

Considering that in England at least, the carpenters took the plans and then blew them up (measuring not photographically or any other modern way) and chalked the frames on the loft floor. As for the floors and everything else, it was again... take a stick, mark it up and use that to show where to point it.   So... IMO, the designs will get you "as designed" and the real ones would get you "as built".  Toss a coin as as either way is not exact for some value of "exact".  

[Waldemar]

 

This rather poor copy of the plan, which I estimate to date from the early 17th century, is perhaps the best visual example of these practices. The profiles of the main frame and the forward 'quarter' frame are not even designed as having floor curves. Only the shapes above the rising floor line were drawn, and the floor curves were apparently 'left to the discretion of the actual builder' by the designer of the ship (Russian archives).

 

 

 

And there are other contemporary ships' plans which, for example, show these floor curves only as a symbolic faint line drawn rather for illustrative purposes and/or completeness. Just as on this design draught of the Danish ship Tre Kroner (1604) by Scotsman David Balfour (Danish archives).

 

 

 

Edited September 24, 2022 by Waldemar

[Waldemar]

 

Nihil novi sub sole.   🙂

 

To conclude this thread, I also present a design draught of the Danish ship Printz Christian built in 1665, i.e. some 45 years before Sutherland's work was published. A quick check of the frame lines proves that this ship was designed on the conoidal solid principle, so that the cross-sections of the submerged part of the hull have the shape of a perfect arc with varying radii for different frames (Danish archives).

 

Moreover, the shape of the floor curves is somewhat 'suspicious', and I think that these bottom curves are also rather for illustrative purposes only on the plan, to be properly formed on the mould loft in the shipyard, as described in the Sutherland's work.

 

 

 

 

[druxey]

In the last example, the lower defining line of the conoid is clearly seen in the elevation/sheer plan. I cannot see where the designer defined the arc radii, though. The spread of these in body plan are only implied by your superimposed circles.

 

As you point out, the floor/hollowings are rather fanciful!

[Waldemar]

 

Trying to reverse engineer this plan in a more regular way (btw, the ship's designer is Claus Reimer). The general method is already understood, only those troublesome floor curves remain to be reconstructed...

 

 

 

[Waldemar]

 

Done. The way the designer of this ship plotted the bottom curves on the plan is not as elegant as the method described by Sutherland. However, this does not matter much in practice, as in both cases these curves were redone by the shipyard workers during the actual construction of the ship anyway. And most likely in an identical manner.

 

 

 

 

Edited September 25, 2022 by Waldemar

[druxey]

I see the centers of the  radii nicely plotted now. in a straight line.  But, Waldemar, how did you go about generating the hollowing curves?

[Martes]

13 hours ago, Waldemar said:

 

This rather poor copy of the plan, which I estimate to date from the early 17th century

 

It's a part of a very obscure set of 3 plans that are now in the collection of the Hermitage museum

 

https://www.hermitagemuseum.org/wps/portal/hermitage/digital-collection/02.+drawings/500128

https://www.hermitagemuseum.org/wps/portal/hermitage/digital-collection/02.+drawings/500129

https://www.hermitagemuseum.org/wps/portal/hermitage/digital-collection/02.+drawings/500130

 

I am not sure if their origin was ever determined, but they look very similar to Balfour's plans from the Danish archive. May be his work or copies.

[Waldemar]

 

1 hour ago, druxey said:

Waldemar, how did you go about generating the hollowing curves?

 

Firstly, I determined through fitting that all green arcs have the same radius, and all blue arcs have different radii. Then I determined more or less arbitrarily the triangle for the blue curves.

 

The red lines were drawn first, then (inside the triangle) the blue arcs as tangents to the red arcs and the horizontal lines, and finally the green arcs tangent to the previously drawn sections of the profiles. 

 

It would even be similar to the conventional moulding method if it weren't for all those variable radii. For greater clarity, I have included another diagram below.

 

 

 

[Waldemar]

 

2 hours ago, Martes said:

I am not sure if their origin was ever determined, but they look very similar to Balfour's plans from the Danish archive. May be his work or copies.

 

Many thanks Martes for these links, much appreciated!

 

And you may be perfectly right about the creator of these draughts now kept in the Russian archives. Similarities to the other Balfour's draughts are striking indeed. Once, when examining just a copy of this plan in detail, I also noticed many similarities to the specific proportions, shapes and methods as described in an English manuscript from around 1620. If I had copies of the originals at the time, rather than just this poor redrawing...

 

Anyway, the original drawing, in contrast to its poor copy, show the frame profiles as having the floor curves, so, as a replacement another draught, better suited to my narrative (from the Danish archives). It should replace the first one inserted in post #12:

 

 

 

[druxey]

Thank you, Waldemar, for demonstrating the rest of the (re)construction. It is similar to the 'propositions' in the Newton manuscript.

[Waldemar]

 

8 hours ago, druxey said:

Thank you, Waldemar, for demonstrating the rest of the (re)construction. It is similar to the 'propositions' in the Newton manuscript.

 

Well, I try very hard to keep to the spirit of the age...  🙂

 

Many thanks to you as well, because through this very exercise I too have become more familiar with the ins and outs of period ship design.

 

🙂

 

 

[druxey]

It is certainly a fascinating study!

[Waldemar]

 

Thank you very much Druxey.

 

To finish the job one more diagram comparing the design profiles of the Danish ship with the possible profile shapes that could have actually been traced in the shipyard (according to the Sutherland way). I have made no attempt to get these two sets as similar as possible.

 

Now I ask myself: which set would I choose for my model or reconstruction of this ship if such an attempt were made?

 

 

 

 

[druxey]

Pretty obvious which might be a more fair set of lines. It would be interesting to run diagonals or waterlines to see which had a better 'flow' without tweaking!

[Martes]

19 hours ago, Waldemar said:

Similarities to the other Balfour's draughts are striking indeed. Once, when examining just a copy of this plan in detail, I also noticed many similarities to the specific proportions, shapes and methods as described in an English manuscript from around 1620.

 

I once posted them here:

 

https://modelshipworld.com/topic/24626-plans-found-in-hermitage-collection-st-petersburg/

 

and there was a comment that it's early 17 century English style, but I did not know about Balfour at the time.

 

Also, did you see these plans:

 

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

Edited September 25, 2022 by Martes

[Waldemar]

 

Martes, thanks a lot again for pointing me to previously unknown material from 1657.

 

As for this ship plan from the Russian archives (Hermitage), a poor copy of it was first published in the Russian journal Судостроение 1971/08 (Shipbuilding 1971/08) in an article by A. I. Dubravin, Shipbuilding in the age of the Northern War (in Russian). This plan is described in this article as an example of a design drawing from that time, i.e. circa 1700. This is, of course, nonsense. Personally, I think they were simply brought to Russia by Tsar Peter I from his travels in western Europe.

 

I reverse-engineered this plan a few years ago and found many of the proportions consistent with the 1620 shipbuilding manuscript. As in the example illustration below.

 

 

 

In the process, I have also made a simple 3D model as a feasibility test. 

 

 

[Waldemar]

 

Waterlines and diagonals of the reconstructed hull section of the Danish ship. I have not made any corrections.

 

 

 

 

[druxey]

So, some corrections were needed to make the hull fair. No big surprise! Thank you for the demonstration, Waldemar.

[Martes]

@Waldemar, concerning the Danish design you reconstructed - are you sure that the red curve was not intentional and would not be copied at the shipyard? Why go to such lengths to create a relatively complex shape for all of it to be simplified and shaven off during construction?

 

Could it be possible that the shape as designed was in line with even more extravagant experiments, like the later Lord Danby's Maggot:

 

 

https://www.rmg.co.uk/collections/objects/rmgc-object-385580

 

(oh, how I'd love to see the planking plan for that one!)

 

and had some hydrodynamic meaning?

Edited September 26, 2022 by Martes