Waldemar

Thank you for your attention so far, Waldemar Gurgul

If one assumes, in line with the results of the investigation so far and with the present, general state of knowledge, that the Mediterranean methods together with its multi-arc frame construction (usually three) were only widely implemented in England at the times of Mathew Baker, i.e. in the second half of the 16th century, and that until then only the frame construction appropriate to the Northern European tradition, that is sporting just one arc (not counting reconciling or bilge sweep), but of variable radius, was employed, as shown in this presentation using the Mary Rose as an example, then this circumstance can be used to interpret perhaps more rationally and confidently other archaeological finds from this early period in a poorer state of preservation than the Mary Rose. For instance, in the first volume of the Mary Rose monograph (Peter Marsden, Sealed by Time. The Loss and Recovery of the Mary Rose, 2003), on page 141, there is an interesting juxtaposition of cross-sections of two important wrecks dating from the “pre-Mediterranean” period in England on an equal scale — the Mary Rose 1511 and the so-called Woolwich ship (possibly Sovereign 1487). The assumption of design homogeneity in this early period, specifically the single-arc frame design, allows the breadth and its height of the Sovereign to be determined quite accurately solely on the basis of the modest surviving fragment of the bottom section of the hull (subject, of course, to its minimal geometric distortion), as demonstrated illustratively in the diagram below.

Patrick ( @Baker ) came up with the idea to post copies of the publications mentioned above in this thread as well (thanks, Patrick), so they are always ready to be consulted. I think it's a good idea and that they are both in the public domain. Voilà. Cate Wagstaffe, Furring in the Light of 16th Century Ship Design, 2010: Wagstaffe Cate - Furring in the Light of 16th Century Ship Design - 2010.pdf Jens Auer, Thijs J. Maarleveld, The Gresham Ship Project. A 16th-Century Merchantman Wrecked in the Princes Channel, Thames Estuary, 2014: Auer Jens - Maarleveld Thijs J. - The Gresham Ship Project. A 16th-Century Merchantman Wrecked in the Princes Channel, Thames Estuary - 2014.pdf

It can probably be explained in some simplistic terms that initially, with the usual use of ships to carry cargo low in the holds, there was no major problem with this phenomenon, and even a certain “softness”, provided by the round sections, may have been desirable so that more violent gusts of winds would not break the masts. It was only when dedicated warships with heavy artillery began to emerge, and at an height unfavourable from the point of view of stability, that the problem hit particularly hard. The general conclusion is, although hardly surprising, that in reality there is no such thing as a universal solution for all applications, or at least it is rarely possible.

Yes, that's probably the best summary. A few dozen extra tons of difference in the original and new artillery configurations, a few dozen extra tons of battle crew with their heavy equipment. And if the repair of the ship in 1536 was mainly to reinforce her existing structure, as incidentally was done quite routinely in later times, then an additional few tens of tons of all the new reinforcing elements can be added, especially the standard knees and diagonal braces. The results of the dendrochronological studies are admittedly quite piecemeal, but at least they now suggest or at least do not contradict just such a repair scenario. All in all, at least 100 additional tons, perhaps as much as 150, and most of it at a height quite unfavourable to the ballast stability of the ship. And if still the “whole” crew was on the wrong side at the wrong time, then... Below is a graphic showing clearly which items have been examined for their age and what the results of this research are (from David Childs, The Warship Mary Rose. The Life & Times of King Henry VIII's Flagship, 2007):

Thank you for showing this . And a huge round of applause to today's marine engineers for this essentially makeshift repair! It's not the only case they have struggled with the specifics of old designs, and this despite the same applicable laws of physics now and then and all the scientific apparatus available today. Shame... I was also reminded of a case in which a professor made serious calculations from which it was irrefutably clear that the length of a ship had to be at least about three and a half times its breadth to be capable of controlled navigation. Just how to reconcile the results of these calculations with the vessel San Juan of about 1550, with a ratio of 2.8 : 1 on the draught line, which not only crossed the entire Atlantic, but still hit the exact spot it was intended to reach with a whole fleet of similar whalers.

To show graphically what Martes had in mind, here is another diagram showing the possibility of girdling/furring, hopefully in a more realistic way, nevertheless, like the previous one laden with dimensional uncertainty so it should rather be taken as an illustration. As a reminder, as the draught of a ship increases (e.g. proves to be too great after launching and fitting out, which is particularly the case for warships), broadening the hull by girdling/furring should also raise the greatest breadth of the hull accordingly, so as to maintain the desired distance, usually 2–3 feet.

This is correct, because we do not know precisely this level after the ship's refit, and in any case it was variable depending on the actual weight of the cargo. Thus, of necessity, these particular diagrams simply show the general principle or idea of carrying out this process that could have been applied, without exact dimensional reconstruction. There is no need for this anyway, as it has not been put into practice. As for the illustration showing the Gresham ship, it may not show the geometry very accurately. It is rather a more or less loose graphic reconstruction, also out of necessity. Please see how fragmentary the preserved parts of this wreck are (illustration also from the report mentioned above). Actually, the greatest breadth of the hull need not or should not be far away from the draught level. Thus, if the ship's draught was not planned to increase, there was no need to raise it.

Did the Mary Rose have to capsize? Apart from a better, or more prudent, vertical distribution of weights on the ship, a very effective and often used way to improve the lateral stability of ships sporting tumblehome was to increase their breadth. However, this does not mean increasing the breadth of the hull per se or by any means, but ideally this had to be done in such a way as to “maximise” the breadth above the draught line (typically 2–3 feet) while keeping it as intact as possible at the water level itself. This is shown in the diagram below (dashed lines). In this way, a dramatic improvement in so-called shape stability (as opposed to ballast stability) can be achieved. This procedure was called furring or girdling, depending on the structural way it was performed (for more on this see, for example, Cate Wagstaffe, Furring in the Light of 16th Century Ship Design, 2010). In an archaeological context, such a case is exemplified by the so-called Gresham ship of the 16th century (for more on this see Jens Auer, Thijs J. Maarleveld, The Gresham Ship Project. A 16th-Century Merchantman Wrecked in the Princes Channel, Thames Estuary, 2014) and a graphic from this report specifically illustrates the essence of this commonly used solution for unstable ships on a concrete extant shipwreck. It is no coincidence that the Gresham ship also features a round hull section having a very narrow bottom, precisely as the Mary Rose 1511.

Verification of the results obtained with the (published) archaeological evidence is naturally a mandatory component of the whole exercise. Here this will be demonstrated by using the cross-sections of the hull of the wreck published in the monograph by the Mary Rose Trust as an example, but the analogous cross-sections in the ship's monograph by Douglas McElvogue are equally good. In contrast, many of the other illustrations in both monographs are inadequate for this purpose, due to their over-interpretation by the responsible authors. The placement of the frame stations in this reconstruction does not always coincide with the stations adopted in the monograph (this is particularly true of the bow section) and in such situations there cannot, of course, be full correspondence of the lines, but then one can judge by the parallelism of the contours being compared. The outer contours of the frame timbers, or in other words, the inner contours of the outer planking, should be taken as reference lines.

Remaining conceptual frames (bottom & reconciling curves) In the final stage of forming the conceptual frames, the botttom curves are joined to the lower breadth sweeps by reconciling sweeps (red), tangentially at both ends, with the reconciling sweeps starting from points on the line of the floor. Where needed, the straight bottom lines are completed by arcs, here all with a radius of 10 feet, the same as for the arcs of the bottom of both quarter frames. This, with the exception of the first frame, where this arc is also tangent to the vertical line of “keel”, and its resulting radius is about 6 feet 11 inches. The diagram also shows the geometric construction used to draw the last frame, which requires special treatment due to its position close to the specific sternpost/fashion frame assembly. * * * Reasons for the disaster of the Mary Rose Diverse, sometimes quite conspiracy-oriented reasons have been put forward as to what may have led to the ship's disaster in 1545. It is difficult to argue with theses for which there is no hard evidence in fact, however, it can be said with certainty that the very shape of the Mary Rose's hull is already quite unfavourable from the point of view of lateral stability. Without going into complicated explanations of a theoretical nature, this shape can be compared to a circular in cross-section beam floating in water, which, when set in motion, easily rotates around its axis. In terms of Mary Rose's specific round cross-section, as long as the ship's centre of gravity was relatively low and the line of greatest breadth was sufficiently high above the water (seemingly 3 feet by design, which is quite a standard value), there was little danger, and the ship could even have excellent seaworthiness. However, with the reconfiguration of the artillery armament to be much heavier than the original, and in addition the embarkation of a battle crew of several hundred, together with heavy combat equipment, there must inevitably have been a significant raising of the centre of gravity of the entire ship, and worse, a simultaneous lowering of the greatest breadth of the hull to water level, which already ultimately devastated the lateral stability of the ship. Actually, it probably no longer matters how the gun ports on the Mary Rose were closed, directly by the gun crews or by someone else, for example on the upper deck. Vasa 1628 had the gun ports operated directly by gun crews, and still did not avoid disaster for the same reason. Over time, the relationship between stability and the cross shape of the hull was better realised, and in 1643 George Fournier was able to state in his Hydrographie: Although the practice [of employing round hull sections], which I described in the previous chapter, has long been and is still widely followed, quite a number of brave workmen, whether French, English or Dutch, depart from it for two reasons. The first, that such ships, being almost round, heel too much in the water. Secondly, because they usually have too narrow a bottom [which is precisely the particularity of Mary Rose – WG]. Below, a round hull cross-section from Fournier's Hydrographie according to the old manner on the left and on the right the shape according to the new manner sporting a wide bottom, laterally more stable.

That's right. To put it yet another way, these are ‘entire’ single circle arcs . * * * Remaining conceptual frames (curves of the bottom) Now, to make the surface of the bottom/“flat” fair, auxiliary lines need to be drawn on the body plan (dashed lines on the first diagram), extending from point A and tangent to the corresponding bottom arcs of the quarter frames. The intersections of these auxiliary lines with the vertical line of the ‘keel’ yield vertical co-ordinates (points B), which should then be transferred to the quarter frames stations on the side projection (second diagram). Longitudinal (other) auxiliary lines can now be drawn, the function of which is essentially to sharpen the hull in zones close to the keel. The intersections of these longitudinal auxiliary lines with the vertical station lines give co-ordinates, which are then returned to the body plan, so that bottom curves for all the remaining conceptual frames can be drawn (as straight lines at this point, except for the last station, which should in principle be an arc from the start – to be discussed separately).

Remaining conceptual frames (breadth sweeps) The procedure for tracing the lower breadth sweeps is fairly straightforward, and involves first creating a vertical ‘scale’ (separately for the two halves of the hull), which horizontal tiers (lines) match the corresponding frame stations. If the distances between the frame stations are equal, the result will be that the distances between the horizontal tiers of this ‘scale’ will also be equal, and vice versa, as in case of the last, additional height of tuck “station”. The initial reference segment for creating this ‘scale’ is the vertical distance between the respective levels of the lower breadth sweep quadrant points of the previously defined master and quarter frames. All that is left is to draw the arcs themselves for all the remaining conceptual frames, in such a way that they are tangent to the horizontal lines thus determined (see diagram). The upper breadth sweeps for the aft half of the hull sport fixed radius, equal to the radius of the lower breadth sweep of the master frame, while for the forward half of the hull these radii are variable, equal to the corresponding lower breadth sweeps.

Double master frame & quarter frames Shaping the hull surface starts with defining the contours of the double master and quarter frames, and this step can already easily be carried out straight away on the mould loft, without first drawing these contours on paper. As far as the master frames are concerned, the lines of the bottom/„flat” were first traced according to the corresponding coordinates taken from the line of the floor for the master frame position, followed by the lower breadth sweeps drawn tangentially to the deadrise level for the master frames position, and finally the two sets were joined by reconciling arcs (dashed red), tangentially at both ends (see diagram). Hopefully for greater clarity, the centre and radius of the lower breadth sweep of the master frames can still be described in Martes' words: „take the difference between the deadrise level and max. breadth level, then take this distance from max. breadth horizontally, and that's the center”. In the quarter frames, on the other hand, the bottom curves are arcs (as opposed to straight lines). The radii of the two sweeps that make up the frame contour, i.e. bottom and lower breadth sweeps, were chosen to obtain sharp hull shapes, suitable for warships/privateers, and even necessary for short hulls as well, and at the same time to make the subsequent reconciling curve (dashed red) close to a straight line while maintaining tangency at both ends. For both quarter frames of the Mary Rose, the respective radii appear identical, and are 10 and 12 feet respectively (see diagram). It was important that the surfaces of both quarter frames below the draught line should be similar to maintain the correct longitudinal balance of the vessel, a point clearly made in French shipbuilding manuals of the mid-18th century. This particular method of design, that is, in particular with this specific use of quarter frames (including the subsequent stages described further on), was still being used and described (albeit rather vaguely) as late as 1697 by the Dutch ship carpenter van Yk, and also in the construction of the French ships of Louis XIV's fleet.

Longitudinal design lines (risings & narrowings) Geometrically, these lines are exclusively single or combined circle arcs. At a later stage, in order to achieve a generally similar run of these design lines, the combined circle arcs will quite commonly be replaced by more advanced curves, such as ellipses or logarithmic curves. The extra rise of the line of the floor at the very end of the stern is actually desired to prevent concave hull surfaces in this region (apart from lifting the flat transom panel out of the water for hydrodynamic reasons), and this is not, or need not be, related to the presumed change in stern configuration during the refit of the vessel. Somewhat anticipating, it can already be said that, conceptually, the Mary Rose exhibits extreme geometric consistency without any anomalies of this kind (assuming, of course, the correctness of the results of this investigation), which practically excludes any local alteration to the ship being made during its life, and especially the lengthening of its keel (according to the hypotheses put forward so far). It is also of note that these hypotheses are probably based only on the circumstance that some of the wreck's components, dated dendrochronologically, revealed newer wood, that is, fitted after the date of the ship's construction in 1511. Yet, the replacement of degraded elements in wooden ships is in itself not unusual at all, especially as the Mary Rose was already a quarter of a century old at the time of her great repair in 1536, which still exceeds the normal lifespan of wooden vessels. In conclusion, there is nothing in the shape of the ship's design lines which seems to suggest any changes to its design during its service.

Read or assumed dimensions of Mary Rose 1511: Breadth outside planking – 40 feet Breadth inside planking – 39 feet 4 inches Length between posts (at 3rd deck level) – 3.5 x breadth outside planking = 140 feet Length between rabbets (at 3rd deck level) – 3. 5 x breadth inside planking = 137 feet 8 inches Draught at midship (without keel) – 1/10 x length between posts = 14 feet Height of maximum breadth above waterline at midship – 3 feet Height of 1st deck at midship – 10 feet Height of 2nd deck at midship – 7 feet Height of 3rd deck at midship – 8 feet Rising of 3rd deck aft – 6 feet Lengthwise division (number of equal length segments between posts) – 13 Forward rake – 2/13 x length between posts = 21 feet 6½ inches (21. 538 feet) Aft rake – 1/13 x length between posts = 10 feet 9 inches (10.77 feet) Keel length – 10/13 length between posts = 107 feet 8 inches (107.69 feet) Breadth of the bottom at midship – 1/4 x hull breadth = 10 feet Deadrise at midship – 6 inches Mainmast position – half the length between posts Keel assembly, lengthwise division The base dimension of the vessel – the breadth, being decided, all other dimensions could already have been calculated on the basis of the general proportions commonly used at the time, possibly adjusted by the specific function of the vessel as well as the experience and inclination of the designer. The method of carrying out and the result of such calculations for Mary Rose 1511 is shown above, bearing in mind of course that the values presented here are more or less hypothetical due to the extraordinary distortion of the wreck. The issue of dividing the vessel into 13 equal parts is quite important, as it not only results in specific values for forward and aft rakes, but, even more importantly, sets the stations for all conceptual frames, including the double master frames and both quarter frames (see diagram). The waterline is shown horizontally, however, according to the original design it may have had a trim within 2–3 feet. The mainmast position falls halfway along the entire hull.

Thanks. The reconstructed design lines may indeed suggest some modifications to the stern. I'll show what's involved a bit later, in the correct order, as it's probably still a bit early for such details now, and the presentation graphics are not yet ready.

The results of the analysis of the design method applied to the construction of the iconic ship Mary Rose presented here, although perhaps no longer entirely unexpected at this latest stage of investigation undertaken, quite decisively complement and correct previous understandings of the history of shipbuilding in this early period. This case emphatically demonstrates that shipbuilding methods that can be called Northern European (as opposed to Mediterranean) are not, as hitherto thought, confined to the early modern Netherlands, from where they were supposedly spread over time to the other regions of the continent. On the contrary, the increasing number of examples being studied show that this is in fact a building tradition that is omnipresent throughout all northern Europe. Suffice it to say that long after the Mary Rose 1511, an exactly identical design method, with all its specific paradigms, was still applied at least 200 years later, for example, for the construction of the very successful Flemish predatory privateer ships, such as the highly regarded Neptunus of about 1690, and described by Chapman as an „extraordinary sailor”, or vessels of Louis XIV's navy of various sizes, such as the light frigate l'Aurore of 1697 (detailed presentation forthcoming). Even if one is not interested in issues such as the historical context of ship design, familiarity with these methods and the ability to apply them in practice may prove useful to today's authors of reconstructions in order to obtain reliable shapes. It is also worth adding that the ancient design methods, correctly applied, virtually guarantee the immediate fair shapes, without the later, punitive synchronisation of cross-sections, waterlines and buttocks, which was in fact not practised in this early period at all. This particular study is based on the published documentation of the ship in two excellent monographs of the Mary Rose — Mary Rose. Your Noblest Shippe, ed. Peter Marsden, 2009 and Tudor Warship Mary Rose, author Douglas McElvogue, 2015. In contrast, one need only caution against the disastrous in content and effect chapter ‘Hull Design of the Mary Rose’ in the first of these monographs, both in a general sense and for me personally. The attempt there to reconstruct the Mary Rose's design method seems to have led to an even more distortion of the lines of the shipwreck than nature has done in 500 years, which consequently led me astray earlier and a lot of time was wasted to finally sort things out. As always, I could also count on the invaluable help of my friend Martes. So much for the introduction, and before the actual detailed explanations, here are a few welcome renders showing a graphical overview of the results obtained:

A very attractive effect. To be honest, I personally even prefer such a more or less monochrome convention. It allows a better focus on shapes and structure, modelled, after all, with such care.

👍 Apart from the occasional printing errors (usually unintentional rescaling of graphics, one or two dimensional) and subsequent possible distortion of the paper, inaccuracies and errors in the original drawings themselves are quite common, especially when drawn by hand, but can also occur in computer CAD drawings if the software operator is not disciplined enough. Unfortunately, this has to be taken into account even when dealing with first class plans, otherwise problems will arise later if not checked beforehand.

Many thanks, Denis, for the demonstration. I was particularly curious just about this stage of model creation. I think it is enough to appreciate the essence of your methods, indeed so closely linked to the specifics of the software. Once again — great results .

Thanks a lot for the explanation, Denis, but to be honest, I wasn't referring to the mesh density of the 3D objects already made (for display, rendering, printing, etc.). Rather, I was referring to the prior stage of modelling the shapes of the ship's components sporting, after all, extremely complex geometry, and yet in such a way that all those thousands of components fit together perfectly (at least that's how I see it in your renders), taking into account all those carpenter's mortises and tenons etc., while still maintaining the rigour of the required shapes, individually and as a whole. I know that even in specialised CAD software, modelling such complex structures is very difficult, and I was curious to see what it looks like in Blender, just for the sake of comparison. And, it's absolutely clear that you are fully comfortable with Blender, as can be seen by the results you get .

Normally I pay attention to that too, but in this case it's all very carefully done indeed, especially compared to some wood models where the grain can run completely perpendicular (and quite absurdly it has to be said ) to the contours of the frames or other proportionally elongated elements. I'm rather puzzled by something else — is a mesh based 3D modelling program (as opposed to NURBS geometry) really optimal for such magnificently detailed structural models and yet of such extremely complex geometry for most components?

Rightly so. Resistance to side loads and minimal runout are critical features of these rotary tools, as indeed with all devices of this type, so that drill bits would not make cones in the air, making it nearly impossible to start the hole in the right spot (can be checked if buying personally in a local shop, best at highest speeds). Not to be overlooked are the numerous accessories that Proxxon offers for its rotary tools (e.g. drills presses of different types, router bases, etc., also of good quality straight away from the box or only after a small/easy adjustment). However, there is a detail, probably usually unnoticed or underestimated, for which I additionally appreciate Proxxon's rotary tools, namely the metal (i.e. rigid) neck in the shape of a perfect cylinder with a diameter of 20 mm, which is standard on their entire range of miniature drills. This makes it very easy to make various holders for Proxxon rotary tools yourself, while keeping decent geometry of the whole setup (parallelism/perpendicularity). Below is one of my self-made holders, which in the attached photo is in turn mounted in a lathe toolholder, but can also be mounted in vices, adapted to various accessories (including by other manufacturers), etc.

My experience with Dremel has not been good and my assessment is that the product is highly overrated. The main objections (but there are others) — huge vibration and deafening noise, especially at higher speeds. This is a result of the poor precision of the tool and mainly cheap plastic components. But perhaps the newer models of Dremel rotary tools are a little better, check with your dealer how it actually behaves if you can. For comfortable and precise work, I don't think there's an alternative to Proxxon, at least in the popular sector. Its 'strategic' components are very precise and made of metal. Quiet and vibration-free, even at the highest speeds. That said, the highest possible speeds are not really necessary, medium and low speeds are much more useful, provided the torque is sufficient. The lower the speed a rotating tool can achieve, the better, as it makes the tool truly more universal.