Naseby 1655 - reverse engineering the ship model

[-Waldemar-]

.​

The constantly appearing round values no longer leave any doubt that the model was built according to a re-scaled plan for a 120ft keel ship.

The perspective diagrams show those elements that would have been drawn on paper, i.e. without the contours of the frames (only the shape of the master frame and also quite possibly two more end or quarter frames.).

.​

 

[Hubac’s Historian]

.

Marc, if you see your needs as more important than my project and what I would like to show, and cannot wait with them, I have good news for you. Dr. Kroum Batchvarov has just opened a sort of information centre for modellers. He is here and it is hard to imagine a better reference. Why don't you take advantage?

I'm also glad you found a good place to write about the Knicks, whatever that is.

.​

The only need I have is to understand what is being discussed, simply because I am interested in what you are doing. My needs certainly don’t superscede the pace of what you are doing. With utmost humility, I can assure you that I am merely trying to keep up.

I will check out Kroum’s page, as well, because having met him recently I know he is a valuable source of information.

 

[Don Case]

All this time I thought ship design was art, turns out it's math and geometry

 

[Allegheny]

@Waldemar, The level and degree of analysis sophistication is simply breath taking. I have worked in the engineering field for many years and have seen many types of analysis.

Yours by far is incredible given what you had to work with - using a 3D scan of a wooden museum model from the 1600's which had been broken and reassembled. The new and old information gleaned from this model's analysis is beyond description.

If I may suggest, that you assemble a document documenting your analysis from beginning to end and submit it for publication in a journal or to the museums archives. I'm certain they would be very interested in your findings.

It is quite obvious that you are well versed on this subject matter - at least how to conduct an analysis of this nature.
Out of curiosity, where did you learn this analysis technique - is this something you do for living?

 

[Allegheny]

All this time I thought ship design was art, turns out it's math and geometry

Don,
In reality, the design of anything like that of a ship, a building, an engine, a transmission, etc. is both an art and science.

It takes years of practice to reach this state in which the designer or simulation analysist can achieve the ability to take merger bits of information and some forehand knowledge of the subject to be able to present an analysis that has less than a 1% margin of error.

Simulation tools if used effectively and with great care can glean incredible amounts of information in the right hands.

 

[-Waldemar-]

.

Thank you very much Allegheny. Recognition from a professional engineer really ennobles, thanks.

Your idea of publishing papers e.g. in thematic periodicals is very good and also tempting, yet online publications such as here on the forum also have very serious advantages and the choice is not at all that easy. Quick publication without the often cumbersome publishing process and wasting time on formalistic so-called scientific apparatus (references, footnotes, indexes, etc.), usually a much larger potential audience and ongoing interaction with other users. All this is difficult to ignore. I have already mentioned Martes, I might say my (somewhat hidden) partner in these endeavours, several times, but I will do so again, because our skills complement each other perfectly, and it is only this synergy that has borne fruit which can be seen. I don't know if such an intensive collaboration would ever have been possible outside the forum. Paradoxically, even the quarrels that sometimes occur, in most cases ultimately produce positive results in the shape of new findings or conclusions.

I have indeed carried out similar projects on this theme on behalf of museums and for remuneration, even quite recently, but basically the skills used were acquired in a hobbyist way. Both in terms of using CAD programmes and gaining knowledge of ancient shipbuilding.

.

 

[Martes]

All this time I thought ship design was art, turns out it's math and geometry

The ships we are discussing come from the English shipbuilding school, which, in that time, was indeed all about math and geometry, and it was, as it seems, possible to "compress" a ship design to a relatively small amount of key numbers and then reconstruct the plan from them.

It became rather different (and complicated) in the 18th century.

 

[Allegheny]

@Martes and @Waldemar

Thank you for your response to my suggestion.
Given the complications you've outlined concerning publishing a formal paper, could you both more or less create some sort of summary of your findings here on this site?

I believe it would be very helpful for folks like myself to read through the process of how you approached the subject and findings of your detailed analysis.
Specifically, I'd like to see how the model's dimensions align to that of the actual wreck.

You've both have done a lot of work and I believe you should put it all together in someway for folks to read.
It would be a shame for this work to disappear into oblivion once this thread ceases to be read.

Now that the work has been completed, I think it would be interesting to hear from some our members who felt that this model could not have been related to Naseby in anyway.

Has the analysis changed your mind in anyway or shed new light on the subject?

 

[Martes]

@Allegheny,

You rush things a little. The process is by no means complete - what has been found is a tentative system, to which Waldemar is now reconstructing the master plan in a cleaner way, and we are probably going to explore the whole thing for many iterations to come. Please be patient.

If we get it right, we'll gain a glimpse into mid-17th century English ships that otherwise was denied, since no plans and only a single, relatively crude model in Greenwich museum survives from the period, apart from very tempting, but themselves unfortunately useless for 3D reconstruction, drawings by Van de Velde.

 

[-Waldemar-]

.​

Upper works. And you would never guess – further round values were found, shown in the diagrams.

There are two sheer guides. The first (in violet), the design one, is needed to delineate the contours of the frames, and starts from the full height of the design grid, i.e. 37.5 feet above the keel. The other, in red, limits only the physical height of the toptimbers. At midship, it has a height equal to the depth of the design grid under the breadth line, as suggested by Newton's manuscript c. 1600.

.​

 

[-Waldemar-]

.​

The toptimber lines for the aft part of the hull are plotted in the conventional way, i.e. using the coordinates of the sheer guide line defined on the paper plan and/or mathematically calculated. However, for the fore part of the hull they are plotted in a completely different way, shown in the diagram. In this case, the lower and upper variable arcs are connected by a third set of tangent sweeps with a fixed radius of 30 feet (thicker lines).

Basically, the reason for using variable arcs for the upper parts of the hull was quite mundane – it was to ensure that the lines did not overlap or cross each other as they were drawn. In addition to other reasons, this was also one of the factors for the part of the hull below the line of greatest breadth.

Lastly, of note, there has also been an update, involving a slight modification to the reduction device for the upper breadth sweep radii (shown in the diagram below).

.​

 

[-Waldemar-]

.​

I have made the difficult and fateful decision to reduce the hull width from 37.5 to 37 feet (without planking). The length of the wing transom will also be reduced, from 25 to 24 feet. As the method is proportional, this requires changing quite a few other values and redrawing the whole thing. This will take some time.

The good news is that the geometric structure itself remains unchanged, and essentially it is now 'just' a matter of changing the parameters (values).

.​

 

[-Waldemar-]

.

Not yet ready to show the final result, so just graphics showing the work in progress. Basically only the hollowing curves are left to complete.

.​

 

[-Waldemar-]

.​

Nevertheless, I still made a few important modifications, about which more later. Now checking the resulting lines by comparing them visually with the model, among other things.

.​

 

[-Waldemar-]

.​

A few more illustrations from inside the model. Despite various corrective manipulations, the 3D model is still not completely symmetrical.

.​

 

[-Waldemar-]

.​

The diagram below shows more or less what would have been drawn on the mould loft in a shipyard (not on paper in this period, at least in the proper production context! As to this, see the so-called Newton manuscript ca. 1600, anon. manuscript ca. 1620, Bushnell's Vademecum 1664, as well as quite a number of working design drawings of the period, notably by David Balfour).

A keen eye will pick up a few changes from previous iterations of the reconstruction:

– the radii of the two most important master frame sweeps have been reduced: of the floor sweep from 10 to 9 feet, and of the futtock sweep from 30 to 24 feet. This has in effect imposed
– correction of the shape and run of the conceptually most important rising/narrowing line of the floor (green line),
– the hollowing curves have also been refined: they now all have a radius of 10 feet, except for a number of bow frames, where a two-radii hollowing template has been applied (shown in the diagram below together with its guide line on the actual body plan marked in thicker line).

.​

 

[-Waldemar-]

.​

This is not yet the final iteration of the reconstruction. Martes has now directly rejoined the analysis and kindly carried out an evaluation of the resulting shapes. It turned out that the reconstructed surface of the hull, although already very good, could still be improved in the midship area at the height of the line of greatest breadth. An obvious candidate for such a modification is naturally the device for reducing the radii of the lower breadth sweeps.

It is not just a matter of obtaining the ideal shapes expected from a professional like Sheldon. Also, it must be said, the actual shapes of the cross-sections of the Ö 3 model begs for such a correction as well.

Below are sample screenshots of the 3D model, quickly and efficiently created by Martes, the subject of our discussions and suggestions in private correspondence.

.​

 

[DonatasBruzas]

@Waldemar
I wonder if you have attempted or not (and why?) to fit any type of curves based on arithmetic functions as described in the anonymous treatise on shipbuilding, c.1620 in the Admiralty Library?

 

[-Waldemar-]

.​

Yes, I was ready to check these equations as well, but in the first instance I found characteristic kinks in both the floor and breadth lines. These kinks do not match the one-element smooth curves as described in this 1620 manuscript, but they correspond perfectly, also in shape, to the two-element curves (precisely of two arcs of circle with different radii) as described by Bushnell 1664.

.​

 

[-Waldemar-]

.​

In terms of the smoothness of the resulting surfaces, but perhaps even more so the nature or specificity of their curvatures, I think it would be difficult to achieve better results. Although, it must be said, Martes is not quite happy with the very small concavity (or maybe better: lesser convexity) still left in the midship area, which must nevertheless be retained, being a feature of the original design.

However, starting the analysis with only the left half of the model, while discarding its right half, ultimately proved to be a mistake that now requires further iterations of fits. It is perfectly clear what needs to be done – generally speaking, the volume of the hull needs to be increased in certain places. In order to achieve this, first the run of two of the three main design lines (floor and sheer) needs to be properly corrected, then the whole set of frames needs to be redrawn, and only having accomplished this whole procedure, the resulting shape can be compared with the scanned model. Perhaps even several times...

.​