Deck beams and their curvature - questions (?)

[Jaager]

As I understand it.  The actual curve is an ellipse and not an arc of a circle.  I see the advantage for this.  It would have a more flat middle section and accelerating run off as it approaches the side.  I also suspect that variations in shaping the beams would probably be greater than any perceived difference between an arc and an ellipse at model scales.

My question is about how the curve is determined for each deck beam.

 

If the mid-ship beam is used to determine the curve for all beams for a particular deck.  If a parallel curve if used to shape the underside of the beams and the thickness is constant along a beam, I see the following:

All of the beams can be shaped as a batch.  
Either a block is shaped and each beam if sliced off -  much loss of shaped material.
Layers the width of each beam can be temporarily glued into a block, shaped, and each beam released by reversing the glue bond.
Or the same curve pattern can be used to shape each beam individually.

 

If a deck clamp is used to site the beams:
With constant thickness. The crown has a down camber as the width decreases.  As the width decreases the height of the crown above the line of the bottom plane decreases.  The rate of change in slope of an ellipse is lost.  But that is mitigated by the water running down hill to the mid-ship as well as to the sides.
If the spirketting is used: butting the top edge up instead of the bottom edge down, I see the result being the same.

 

If each beam has its own curve - a curve determined by the beam's actual width, the camber of the crown will parallel the slope of the deck clamp.

Each beam is a one off.  A lot more work and the possibility of error is much greater.

At model scales,  the difference in effect would probably be nearly impossible to notice.  If  one goes old school and lays  deck planking, would not there be an effect in how the planks lay?
I am wondering what the big boys did full size?

[druxey]

While all deck beams can be shaped to the same round up, as a batch it doesn't quite work. At the dead flat the beams are rectangular in section, but become increasingly parallelogram in section as one proceeds fore and aft, due to the rising sheer of the deck. Therefore the beams fore and aft need to be  a little deeper to allow for the bevel top and bottom.

[Jaager]

My mental vision missed the effect of the slope as the ends are approached.  Thank you for pointing out the additional factor to add to this equation.

Since the deck clamp follows a curve, if the beam sits directly on it as is, the beam top surface has the necessary bevel. No longer vertical are they.  If the decks are completely planked, who is to know?  I suppose the beams at the hatches at the ends would need a bevel to fake it.  That would be a lot less work.

Yet another reason to stick to modeling the actual surface and forego showing  the underlying structure.

[Mark P]

Good Morning Jaager;

 

I have never heard that the deck beams are curved elliptically; can you remember your source for this? Is it for American ships only? All I have ever seen stated is the amount of 'round-up' to be provided at the midship beam, which varied according to which deck it was. Interestingly, in the restoration Navy, the deck beams were not parallel top and bottom, but were thicker in the centre.

 

All the best,

 

Mark P

[Jaager]

Greetings Mark,

 

I believe it was in the NRJ,  maybe  in the 1970's.  Or possibly Shop notes v.1.

When I was a lab rat, I kept a card file of references to my pack of journal article photo copies.  The cards had relevant quotes about the key points.

For ship modeling, the best I did was just the basic title etc. in a database -

 

A quarter circle with a radius that is the difference in height between center and end of the beam.

The radius along the base-line is divided into 4 equal spaces and a perpendicular drawn at each of the 3 points.

The distance between the center of the beam and the end is divided into 4 equal spaces.

The three perpendiculars moved out to the three points along the beam half length.

A curve connecting these 5 points is an arc of a circle.

or

The quarter circle has a fan of three lines that each define 4 equal segments of the quarter circle arc.

The perpendicular height at each of those 3 points is instead along the same equal distant segments of the beam half length.

A curve connecting those 5 points is an ellipse ( or part of one).

 

The point being that an article showing these two geometry exercises =  a little semicircle with perpendicular lines inside and a big arc above it  and another with a fan inside the little semicircle and a different shaped arc above it,  the article in question has been found.

 

Maybe I have confused an ellipse with a parabola.  I checked HIC's Boat Building  and he shows a different way with different sort of fan.

I also seem to remember a perpendicular at the end of the half beam with a fan drawn from the mid-point. The intersection of the 3 section perpendiculars with their fan line:  connecting those 5 points is another sort of curve - but maybe that is another way to get an arc?

 

From a fast scan of HIC, I think he is saying that the curve for each beam is drawn to compensate for the difference in length.  

I also recall seeing Bob Brucksaw (I think) using a disc sander to shape the beam curve.  The beam was clamped at the end of a long stick.  The stick had a pivot at the distant end. Swinging the stick would allow the disc sander to produce a smooth and reproducible arc on the beam.  I have always been stumped. How do you get a fixed pivot point that is level with the sander table and be so far away?  The length of the stick will also need fine-tuning.

 

I did a database search, perhaps of of these has the answer:

 

A JIG FOR CUTTING DECK BEAMS   WEBB,WM G   MODEL SHIPWRIGHT , 1987, 60, 52-55

DECK BEAM CONSTRUCTION   BRUCKSHAW,ROBERT V   NAUTICAL RESEARCH JOURNAL, 1977, 23,42-43

DECK BEAMS - ADDITIONAL NOTES   FLEMING,EDWARD S   NAUTICAL RESEARCH JOURNAL, 1977, 23, 99-100  

DECK BEAMS AND CAMBER   HOBBS,KEITH M   NAUTICAL RESEARCH JOURNAL, 1978, 24, 43-44

PLANKING VISE, CONSTRUCTING DECK BEAMS, TURN SQ STOCK IN 3 JAW   COLE,N R   NAUTICAL RESEARCH JOURNAL, 1978, 24, 100  

DECK BEAMS MATHEMATICS OF CAMBER   BRUCKSHAW,ROBERT V   NAUTICAL RESEARCH JOURNAL, 1978, 24 , 152-153  

DECK BEAMS  AN EARLY SURVEY OF THE WORD CAMBER   SEARLS,DELMAR E   NAUTICAL RESEARCH JOURNAL,  1979, 25 , 98-99   

 

Dean

 

 

[stuglo]

I'm sorry but I have failed to understand this thread.

Is the round up the same irrespective of beam length- if not would so it will be be greater for a shorter beam if curve is an ellipse?

For the model this may be irrelevant, but nice to know.

[Jaager]

4 hours ago, stuglo said:

Is the round up the same irrespective of beam length- if not would so it will be be greater for a shorter beam if curve is an ellipse?

For the model this may be irrelevant, but nice to know.

That is indeed the major question. i.e. is a different curve drawn for each beam? 

If a master pattern is used for a whole deck,  The beams would still be the same thickness along their length.  As each beam gets shorter, the height of the crown above the chord is less.  The ends of a beam would be up to spec, but a curve connecting the crowns would be flatter than spec.

 

As for model scale effects, at 1:48 and smaller the difference in a deck curve that is an ellipse/parabola vs the arc of a circle is beyond our ability to discern   this is my take home of the  published consensus for 50 years. 

I bought a plastic drafting tool long ago that forms an arc along the top edge.  The radius of the arc decreases as I turn a dial.  If I accept an arc as my curvature, it makes drawing the proper curve for beams of different lengths pretty easy to do.  I just turn the dial until the arc connects the two end points with the crown.

 

I have long seen that the gun-ports go from rectangle to parallelogram as the ends are approached.  I just never noticed that the deck beams also do this.

[Mark P]

Good Evening Dean;

 

Thanks very much for the well-explained clarification. Both the methods you describe have been used to draw arcs/curves for shipbuilding. I have to admit that I have no certainty as whether an arc or an ellipse results from the second method. I will set this out in CAD and see what results. I would be very dubious about the deck beams being elliptical though, for the following reason: using arcs of circles, the curvature stays the same, and as the beams shorten, the actual rise, or round-up in the centre lessens. This means that only one template is required for the curvature, which can be used for all the beams on each deck. If we start talking ellipses, then a different template would be required for each beam. Which would be both a lot more work, and lot more timber for making templates. Which I would think also answers Stuglo's question above. 

 

Although that said, the difference between an elliptical curve, and an arc, can sometimes be very small. I remember setting out an elliptical stair once, using AutoCAD's ellipse command, which resulting shape I superimposed on an 'elliptical' shape drawn to two radii given by the architect, with a short, mirrored arc forming both ends of his ellipse, and similar arcs to a larger radius forming the longer sides of the ellipse. With the lengths of the long and short axes of each shape matching the other, the difference between the outlines of the two shapes two was hardly noticeable. 

 

I will do a bit of setting out, and let you know.

 

All the best,

 

Mark P

 

 

[Jaager]

Mark,

 

I do not know the practical difference between an ellipse and a parabola when it is a small segment of the curve that we are interested in.

 

An arc has a constant slope.  The slope between any two points on the curve is the same. 

For an actual deck, I think the slope increases as the end is approached.

The middle section is sort of flat,  but the rate of water runoff out increases as the waterway is approached. 

Chapelle has the curve as a parabola.  (If you do not own Boatbuilding by Howad I. Chapelle, you may find it valuable if you did.)

His method  for drawing the curve is different from what I tried to describe above.

Maybe the difference between an ellipse for a deck curve and a parabola is that the middle stays flat(er) farther out and drops off more sharply with an ellipse.

 

By being in these weeds,  I see that no matter which curve is chosen, the same pattern cannot be used for every beam - unless you wish the difference in height between the crown and the end to decrease as the stern and bow are approached.

I just checked HMS Centurion's draught.  The lines defining the difference in height for the crown and at the deck clamp are parallel all the way aft and in the fore up to the last station. The lines converge from there to the FP (rabbet at  the stem.

[druxey]

It makes no difference surely, as a vessel is not static like a building. There I can see that an ellipse might be used: roofs - generally - do not taper! A ship rolls, so why bother with an elliptical round-up?

[CDR_Ret]

Please permit me to put my oar in, since this topic has recently been of some concern for my research project.

 

I have finally found some contemporary and near-contemporary wooden ship construction rules for my late-19th century brigantine merchant.

 

William Crothers, in his book American-Built Freighters and Packets of the 1850s, makes the following statements:

• "The true form of the camber curve is the arc of a circle of great radius, which is difficult to draw due to the space required." (p. 55)
• "Three methods of developing a proper curve for deck camber—one mathematical, one geometrical, and one natural—are illustrated [in Figure 3.6]. In all cases, the maximum moulded breadth of the deck is the database." (pp 55, 57) [Frankly, the first two methods he shows appear to be two different geometrical methods. The third is springing a long uniform wooden batten between marks at the ends of the reference beam on the lofting floor and marking the curve at the required roundup.] Brackets are my additions.
• "This [the deck camber] was arbitrary and was commonly 6 to 8 inches in forty feet." (p. 57)
• "The established camber curve, or round up, is constant throughout the length of the vessel." (p. 57)

The following passage was completely new to me when I first read it. It refers to "springing" the deck beams.

• "An attempt to spring a [straight] beam, which might be moulded eight to ten inches, to the maximum height of camber curve, would exceed the elastic limit of the substance of the beam. The maximum allowable spring in a beam is about one-eighth of an inch per foot of length, but this figure is reduced to one-tenth of an inch in actual practice. The solution was to cut a portion of the camber into the beam itself on the upper side and, if necessary, on the under side. In any case, the required moulded depth of the beam had to be retained. (p. 57) [Farther on, he explains that the beam is placed in the ship and fastened to the clamps, then jacked up in the middle to attain the required camber. A permanent stanchion under the beam is then installed, which transfers some of the weight of the beam to the keel below, relieving the load in the ship's sides.]

I discovered in the Record of American and Foreign Shipping-1890 that deck beam scantlings were determined by vessel tonnage. Beams adjacent to hatches or mast partners were 10 percent larger than standard beams. The underside was to be straight, and cambered only on the upper surface. All beams in the vessel (for a given deck) were to be maintained throughout the vessel. An exception was made that beams toward the ends of the vessel could be reduced by one-eighth throughout the length of the beam. A later standard identified the "ends" of the ship to be one-fifth the length from the stem and stern post.

 

One question I haven't quite answered yet is whether the bottom surface of a beam is parallel to the moulded sheer of the deck as the top surface would be, or is it parallel to the waterline? Crothers suggests that the latter is the case. He writes, "At the beam ends, in way of the clamp, the lower surface of the beam was snaped to rest flat, or horizontal, on the clamp. If necessary, the clamp itself was trimmed to accommodate the end of the beam." (p. 207) This makes sense only if the lower surface of the beam was parallel to the waterline.

 

Any thoughts on this information. None of the foregoing applies to wooden warships, sadly.

 

Terry

[Jaager]

2 hours ago, CDR_Ret said:

[Farther on, he explains that the beam is placed in the ship and fastened to the clamps, then jacked up in the middle to attain the required camber. A permanent stanchion under the beam is then installed, which transfers some of the weight of the beam to the keel below, relieving the load in the ship's sides.]

Hi Terry,

I have always looked at stanchions as being a push fit or wedge fit element.  Not as something already resisting really serious downward force.

In a model, I tend to see them as optional because with my style, they would never be seen. 

 

Random thoughts:

 

This would probably never work in a model.  It would be like planking the hull without pre-bending the curve in the plank.  But the forces involved would be much much more.

 

In a multi-deck warship, doing beams thus way would make the decks a tightly tensioned spring.  The effects of a projectile eliminating  a stanchion: ...  The beam alone would exert downward force, wanting to go horizontal.  Adding on the weight of the guns,  I do not think I have seen illustrations of stanchions large enough for this.

 

Jacking up the middle would pull ends of a beam in.  The joinery during the build would be interesting. If it is a tight fit before being jacked, placing wedges at the ends would fill the gap and help the stanchion, but there would be serious outward force on the frame it is mated with.

 

If I am reading this correctly,  some beams  had bottom edges that were chords of the round up and not parallel to it.  a lot easier to model.   No jacking up with this.  Smaller vessels only.  For a large vessel, the timber thickness would possibly require a Redwood tree sized balk.  I would think economy built merchantmen only.

 

The trimming to sit on the clamp -  it would need to be from the inner corner up.  The outer corner up would affect the height of the crown. 

 

In the vessels with jacked up beams,  the cross sectional curve is an arc.  But each beam would have an arc for a different circle.

 

3 hours ago, CDR_Ret said:

"The established camber curve, or round up, is constant throughout the length of the vessel." (p. 57)

OK.  Someone here ( I think Bob Cleek) pointed out that camber is the curve of the deck seen in profile - i.e. from bow to stern.  It is not the curve along the top of a beam as seen in a Body plan ( cross section ).  To do this, each beam must have its own individual curve as seen in cross section. ?

 

Dean

[druxey]

Terry: The lower surface of the beam ends might be trimmed horizontally, but normally the underside would arch up to maximize the (limited) headroom below.

[Charles Green]

I recalled an article in Wooden Boat magazine that addressed this topic.  It has taken me this long to find it - "Deckbeam Moldes - Ye Olde Mythe of Boatbuilding" by Andy Davis, Wooden Boat, #165, April, 2002, pp., 40 - 45. 

 

The answer is, in order for the height of each beam's crown - at center-line - to match its height - as seen on the shear draught - each beam must be made to its own arc.  The "constant camber" approach will only work on vessels shaped like barges, with parallel sides and no shear.  Davis outlines several "tried and true" methods for establishing a uniform arc for all of a vessel's beams and shows how they all fail to produce a fair deck.  Davis describes how a fair deck will only result from these methods after shaving and shimming to make up for each methods inaccuracies.  Davis also describes a sliding-batten method for establishing each beam's curve, that will result in a fair deck. 

 

On a vessel with deck shear and tapering ends, the inaccuracies of a constant camber show up as a dip in the stern and bow areas of the deck.  At model scale, you might be able to live with it.  It all depends on the model's scale, the degree of bow and stern taper and the amount of deck shear.  

Edited July 10, 2021 by Charles Green
terminology

[CDR_Ret]

Reviewing the responses here, I am having some difficulty reconciling these views with what I have read in resources contemporary with the age of sail. Granted, I am not an expert on this topic, so please bear with me.

 

First, dealing with camber. My understanding is that "camber" refers to the round up of the beam athwartships, not longitudinally. Camber is visible in the body plan, not the profile view. The curve of the moulded deck in profile is the centerline deck sheer. This definition of camber is confirmed in both Crothers' books and in various dictionaries. The term "camber" itself is not used in the Record of American and Foreign Shipping 1890, but the context of the word "crowned" referring to the shape of the deck beams is unambiguous. My understanding is that the American Shipmasters' Association (ASA, and its successor, the American Board of Shipping, ABS) were essentially American construction standards setters for underwriters similar to Lloyds of London.

 

I'm not sure what Mr. Davis's point was in the Woodenboat article, but if the deck beams with a "constant camber" (meaning to me that all beams use the same camber template created at the deadflat body plan) are installed in the ship to create a fair sheer curve along the ship's centerline, then all deck planks laid parallel to the centerline will also have the same deck sheer. It seems to me that the sheer of the outboard rail is contingent on the intersection of the deck beams with the moulded hull surface, assuming a constant rail height above the deck is to be established. A sweet and fair rail sheer in profile depends on the deck camber, the moulded beam of the ship, and the shape of the hull. If each beam had its own round up based on its length, then the deck planks would not provide a uniformly fair surface. Perhaps I have misunderstood what was intended in the above comments.

 

In attempting to reconstruct my grandfather's ship's main deck using the principles Crothers stated and the ASA's construction standards, the following image shows how the moulded deck's surface lies true and fair. All beams have the same camber. The gaps resulted from variations in beam spacing. The deeper beams are in way of hatches and masts per the ASA Record...-1890.

Brigantine Galilee's reconstructed main deck beams, constant camber (2-1/2 inch roundup). Model in DELFTship.

 

Dean's concerns about introducing the "spring" in deck beams in a model are correct—up to a point. There is no reason to include this feature in a model because the stresses relating to an actual ship aren't present. In any case, the forces involved springing a beam compared to the other stresses in the hull were minimal. A 40-foot beam would be sprung only 4 inches (0.83%), and that would be included in the total design camber for that beam. Beams near the ends of the ship would have little or no spring. Recall that the purpose of the beam spring and stanchions is to reduce the weight of beams on the sides of the ship by transferring that weight to the keel.

 

Regarding the shape of the beam bottom surface, it is pretty clear from both Crothers and the ASA Record... that the bottom of the beams (at least merchant ship beams) in the latter 1800s were not curved. From the Record of American and Foreign Shipping-1890, it states "The beams may be reduced in moulding one-fourth at their ends; the underside straight, and upper side crowned." (p. 38, emphasis added).

 

The bottom surface of the beams being flat seems to be evident in a photo taken of the Galilee's hold in the 1960s, 70 years after she was built. I'm not sure that any "spring" in the beams would be evident from this vantage point. Also, considering that the keel is basically gone at this time in the vessel's history, eaten by ship worms, the spring, if present, would have been released.

 

Derelict Galilee's hold, showing main deck beams, hold ceiling, beam clamp, and knees (c. 1965).

 

A question I have is this: Though the beam bottom surfaces were straight (not crowned), were they horizontal? Or were the bottoms angled fore and aft to follow the deck sheer? This would be significant only at the ends of the ship.

Reconstructed forward main deck beam in Galilee with a horizontal bottom surface (perspective looking starboard toward centerline). Only the port half of the beam is visible in this image.

 

Same deck beam but with the bottom surface following the local moulded deck sheer line.

 

Based on the quotation from Crothers in my previous post, I would say the intent would be that the lower surfaces of deck beams were horizontal. This would also reduce the work in fabricating beams and simplify installing the beam stanchions to bear against a perpendicular surface.

 

Thoughts?

 

Terry

Edited July 10, 2021 by CDR_Ret
Edits for clarification.

[druxey]

I suspect that you are making thing unnecessarily difficult for yourself. Let us assume constant arciform round up. Think of the sheer plan view of the vessel. If you place the beams with the upper surface of their outer ends on a smooth sheer curve (that at the ship's side), the line of the beams at the centerline will mirror the curve at the sides for a short distance fore and aft. As you approach bow or stern, the sheer line at the center will converge with the sheer at the side in a smooth manner, ending together at the bow. 

[stuglo]

Fascinating. The "straight underside" is interesting. Saving cost (work hours) vs extra weight .Thanks everyone. Perhaps there was a variety depending on age, size and use.

[CDR_Ret]

David,

 

Yes, that likely is the actual way the deck sheer line is created.

 

The method I described reflects the process  I  had to resort to because Galilee's original plan's rail sheer wasn't fair to begin with. I suspect that a draftsman before the advent of 3D software may have had to resort to an iterative method to arrive at both a fair rail sheer and a fair deck sheer.

 

Thanks for injecting a bit of reality into my comments. 😬

 

Terry

[Jaager]

Terry,

 

I suspect that when discussing 1650-1860 warship construction, there are details and practices that are very different from 1900 +/- 50 merchant.   The strength and physics of wood would be the same, so necessary scantlings in proportion to vessel size would be reasonably close.  I also have ASA 1870, 1885, 1903 and I use them for 19th century vessels when Meade is mute on the part in question. (My interest ends at 1860.)  The camber vs sheer vs round-up is a point of confusion.   I have slowly come around to realizing that academically generated scientists and engineers with professional discipline,  standards, and vocabulary is a late 19th century creature at best.  The earlier data sources are generally the work of enthusiastic amateurs or individually trained professionals.  There was no board of grey-beards and silver-backs who edited their work for errors or inconsistencies in nomenclature.   I feel  that being dogmatic, rigid, and self-sure about much of this is neither wise or supported by reality.   I go by, does it match what existing plans and information is available? For the gaps, does it match known contemporary  practice within reason and does it seem it fit its time? 

 

As far as deck beams,  in order to fit what I see on plans,  I will shape the curvature of each beam as an individual.  To use the same curve for all and still match the plans as far as the crown would require that I shim the land at the clamp.  By doing this, I would not be following the plan for where-is the underside of the deck at the side.  My plans are 1732 not ~1900.  I am going to have to soak a while about how I will deal with there being a bevel on the top where the slope of the profile line becomes significant.

Batch production of deck beams is an idea that I will file in the specious folder.

 

Dean

[Jaager]

22 minutes ago, CDR_Ret said:

The method I described reflects the process  I  had to resort to because Galilee's original plan's rail sheer wasn't fair to begin with. I suspect that a draftsman before the advent of 3D software may have had to resort to an iterative method to arrive at both a fair rail sheer and a fair deck sheer.

Terry,

Being sort of tongue in cheek,  given the situation involving your subject,  I would probably pretend that my customer was rich and valued form over substance.

I would do the ideal version of the vessel - how it would have looked if the designer and builder had possessed superior skills.

[CDR_Ret]

Hey Dean.

 

My customer is this guy named CDR_Ret. He's neither rich nor values form over substance. He is really particular about substance.

 

Not having ever built an actual model ship from scratch, I'm probably overthinking many of these details that would never see the light of day in a finished model.

 

Since my nuclear engineering background was founded in following rules and specifications, I find it natural to review the basis for every detail that I don't have personal knowledge of.

 

And since it is more than likely that I will never actually build a model of the ship I am researching, I hope to at least create the most accurate set of plans possible based on all available resources, so someone else can have that privilege.

 

Sorry to have hijacked your thread. Did you find the answers you were looking for?

 

Best wishes.

 

Terry

 

 

[Charles Green]

CDR_Ret:

 

This is a case of "a picture being worth a thousand words", but the picture in question is copyrighted and I don't have the skills to electronically transfer an illustration from one place to another anyway.  So bear with me. 

 

Imagine laying deck beams, all with identical camber, in a hull with fore and aft taper but designed with no deck shear.  In other words, the mid-ship deck beam shown curved (cambered) on the body plan, with the centers of all the others lined up on the water-line plan, creating a flat/straight deck on the shear plan.  And, consider the geometrical theory stating circles/arcs are composed of an infinite number of straight lines progressing around the arc's center, all tangent to the arc at its radius.  In other words, circles/arcs, are composed of straight lines.

 

Delving into matters of infinity isn't necessary for this discussion, but in this example, as the beams approach the tapering bow and stern of this hull, they become shorter, and since they are all made to the same radius, their curvatures begin to approximate a straight line.  This is to say, as the beams shorten, the high-point of their camber will begin to approach the same height as their beam ends. 

 

On the shear draught, this vessel has no deck shear.  But as-built, with all beams made to the same camber, it will have deck shear, and a strange one at that.  The high-point of the mid-ship, deck-beam will be the high-point of this deck's shear.   From that point on, 'though the beam ends are all mounted at the same height above the keel,  the height of their centers will create a down-hill taper, 'till at the bow and stern, the height of the beams' mid-points will approximate the height of their beam ends. 

 

On a vessel with a tapering bow and stern, designed with deck shear and with all beams made to the same camber, the tapering distortion described above will show, as-built, as a dip or depression in the designed deck shear as the deck approaches the stern and bow.  The only way to get around this distortion on a deck built with all beams of the same camber, is to sequentially elevate the clamps to keep the beams' centers on the designed shear line.  As far as I know, this is never done.  And it would create another type of distortion on the deck near the beam ends.

 

Davis states no conscientious builder would produce a deck with a swale on each end.  And indeed none do.  But the use of  strong-backs, shims, the shoring of low beams and planing off the high beams at the transition is work that must follow to be able to make these beams lay a fair deck. 

 

A sliding batten, as described and illustrated by Davis is a way to lay out a distinct camber for each beam that will lay a fair deck. 

 

  

[Jaager]

Terry,

 

No hijack has occurred.  Seminars will drift a bit and as long as it stays reasonably close to the subject, it is what is supposed to happen.   A wide search light is my preference.  I doubt that I am the only one who could use help with this subject.  

Yes.  I have gotten the answer that I was seeking.  I just excavated my Acu-Arc.  The Acu-Arc is essentially a sliding batten.  The plastic is still plastic and the springs still pull.  So doing each beam as a unique individual will be less painful than having to draw a lot of geometric constructs and connecting the dots for each beam.

I have the 14 inch one and I am certain that I did not pay anything like $60 for it.  But I bought it 40 years ago.

 

You might reconsider building a physical model instead of a virtual one.  It is a different sort of reward. 

 

Dean

 

[Dr PR]

I have read through this thread several times and most of it makes no sense at all. There are too many undefined terms, ambiguous wording and simply irrational statements.

 

It seems several people have said you cannot build a "fair" deck if all the deck beams have the same camber. I assume "fair" means smooth and not wavy, as in a smooth curve or surface. However, virtually all modern ships are built this way.

 

This image shows the curvature (camber) of the transverse (side to side) deck beams (red). They all have the same curvature or camber. The green line from bow to stern is the centerline along the deck surface. The center point of each red transverse deck frame is positioned on the green centerline. The black line is the curve of the deck edge from bow to stern, as shown on profile views. This drawing shows quite well that anyone saying you cannot build a "fair" deck using the same camber at all stations (or deck beams) doesn't know what he is talking about.

 

Actually, the deck surface is just a section of a cylinder, like a pipe, where the diameter (curvature) is the same all along the length of the cylinder.

 

This is an image of the actual boat. It is a 40 foot personnel boat used by the US Navy for the last half of the 20th century.

 

 

Note that the curvature of the hull where it joins the deck surface is "fair" - not wavy. The entire deck surface and hull surfaces are "fair." The deck does not curve down into "swales" at the bow and stern.

 

Another example of a model of a 19th century wooden ship constructed with the same deck curvature (camber) at every frame can be found in this post:

 

https://modelshipworld.com/topic/19611-albatros-by-dr-pr-mantua-scale-148-revenue-cutter-kitbash-about-1815/?do=findComment&comment=603771

 

 

In this case the deck has sheer (curvature along the longitudinal direction). The deck surface approximates a hyperbolic section, although in this instance it isn't an actual mathematical hyperbolic surface. It resembles a saddle surface, or the surface of a doughnut hole, where the deck curves downward from the centerline in the transverse (side to side) direction (camber), and curves upward from midships in the longitudinal (lengthwise) direction (sheer). The line where the curved deck surface intersects the molded surface of the hull is not wavy, but is "fair," and the deck surface does not dive down to a lower level at bow or stern.

 

I am extremely familiar with the construction of the Cleveland class cruisers of World War II, having studied thousands of blueprints for 14 years to build a detailed CAD model of one of these ships, and I can assure you that these ships were constructed with the same deck camber at every frame on the main deck and most higher weather decks:

 

https://modelshipworld.com/topic/19321-uss-oklahoma-city-clg-5-1971-3d-cad-model/?do=findComment&comment=590447

 

 

I have given examples of ships that were built with the same transverse curvature (camber) at each deck beam/frame, in cases where the vessel had no sheer and where it did have sheer. In each case the lines of the deck and hull were "fair." And in no case does the deck curve down (longitudinally) at bow or stern.

 

Now can anyone explain why a ship built in ANY period could not have fair lines if all deck beams (in the same deck) had the same camber?

 

NOTE: If you do, please define every term you use so we have a chance of understanding what you say. Otherwise it will just be more incomprehensible gibberish.

Edited July 11, 2021 by Dr PR

[Jaager]

Phil,

My apology for the confusion in definition of "camber".   The original definition as I understood it turns out to be the correct one - the transverse or cross section curve from side to side.

I have always understood sheer as applying to a profile line from bow to stern.  But from HIC's plans where he names a continuous line at the side of the ship going from bow to stern as "sheer".

The lines that define center of a deck from bow to stern and underside of deck at side from bow to stern - may not have a generally accepted name.  They are usually present on most of the plans that I work with.  I am at present hacking away at HMS Centurion 1732.  These lines are present for all of the decks.  They are also parallel on every deck, except on the two lower full decks where beginning maybe 10 feet from the rabbet of the stem, they start to converge.  I have recently seen plans where close to the hawse holes, some ships had this deck with a sharp dip down.  I see the utility of isolating water coming in with hemp anchor lines in a forward well  for runoff there.

 

If you work the geometry, using the same curve for every beam cannot produce this result.  Up to 1860,  most every ship seems to follow this pattern.   When iron and steel became dominant,  what you describe maybe became the norm.   I see that world undergoing at major change around 1860.  That is why I stop at 1860. 

Perhaps wood and steel have different requirements for economical deck fabrication?   The deck camber may also have been parabolic before 1860 and an arc after.  I think that the sort of work done on deck may have been very different on either side of the tech inflection date.

 

I think this is an a situation of comparing apples and oranges.

 

As an aside, I propose that except for small individual yards, the discipline and methods of wooden shipbuilding gained over several hundred years was lost as the older generation aged out and did not pass it on.   The books that reflect the methods used during the brief resurgence in large wooden hulls around 1900 - 1914-1918 -  appear to me to be a translation of steel methods to wood.  

 

[archnav]

1 hour ago, Dr PR said:

Now can anyone explain why a ship built in ANY period could not have fair lines if all deck beams (in the same deck) had the same camber?

Hi Phil,

unfortunately, there are very few explanations of this that can be used for the 17th, 18th and 19th centuries. David Steel 1805, Klawitter 1835 and other authors have given descriptions and drawings how to construct these deck cambers.

Imagine, that the shipbuilders of that times had to "draw" this camber on every single deck beam. Only in midships where the length of the beams were nearly the same they could use the same pattern for all the beams in that area.

 

I have made a construction template of how these lines were taken out on the beams sides. It is shown by Steel and Klawitter.

If you study the way of construction, there is no doubt that the arc of every beam that is longer or shorter than at midships (to the aft and bow section), will differ in shape ! 😉 Make a try of this way of construction on CAD and you will see that the curvature of the deck, looking from fore to aft, will differ slightly. You will get NO fair lines along the whole deck.

Of course, these discrepancies were hardly noticeable, but they were noticeable if you looked closely.
The line of the camber definitely showed more of an ellipse but in no case a linear curve.
I think one cannot apply the old construction methods to the 20th century, that would be a mistake.

 

I hope my post helps you all.

[druxey]

Thank you, Phil, for graphically showing what I tried to describe in post #16.

 

As for definitions, they have changed over time. In the period I've studied most, the 18th century, round up referred to the curve of the beams athwartships and camber referred to the downward sheer of a deck; usually at the bow towards the hawse holes.

[CDR_Ret]

Phil,

Your perspective diagram is just what is needed to explain the concept of "constant camber" as it applies to deck form. I was considering using the term "saddle" in an earlier post in order to explain the essential shape of a ship deck where all beams had identical cambers along a curved sheer, but I was concerned that would just add to the confusion. Well done!

 

Since some of the issue seems to center on the historical use of the term "camber," it's always helpful to go to the etymology of a word to understand where it came from.  The etymologyonline.com source provides the following entry for camber:

 

camber (n.)
"convexity on an upper surface," 1610s, nautical term, from Old French cambre, chambre "bent," from Latin camurum (nominative camur) "crooked, arched;" related to camera. As a verb, "become slightly arched," from 1620s. Related: Cambered; cambering.

 

I consider this website to be fairly authoritative. They provide a link to their list of principle sources, which is quite extensive.

 

It is possible and even likely during the past 400 years that writers and draftspersons could have misappropriated the word, not fully understanding its origin, and applied it to situations not intended by its initial usage. We even see the same thing happening in more modern times. A cambered road surface is crowned to facilitate water runoff (✔️). But the word also now applies to banked race tracks (❓) as well as the design feature of vehicle steering mechanisms that facilitates restoring turned wheels to the straight-ahead orientation (⁉️).

 

Tom, I must respectfully disagree with your conclusions. If you create a deck beam template with a particular round up or camber at the deadflat, then use that pattern for every beam forward and aft of deadflat arranged along the moulded centerline deck sheer, you will obtain a fair, curved surface. (See Post #15 in this topic, which illustrates this approach I used.) This is a mathematical necessity, as Phil noted. The length of the beams is determined by where they intersect the ceiling timbers (at least that's the way it worked for 19th century ships) but this in no way affects the moulded deck surface. It is true that the beams appear flatter near the ends of the hull, but their surfaces are still parallel to adjacent beams with the same camber, so the surface they form will be a fair one. I will defer to those who have studied ship construction in earlier periods where builders may have reduced the camber toward the ends of the ship, since the camber was less effective to shed water on shorter beams. But this wasn't a geometrical necessity, and eventually was superseded by the more standard practice of constant camber as stated in the references I have cited above. 

 

Terry

 

[Charles Green]

Ladies and Gentlemen: Remember, I am only the messenger!  The message comes from Andy Davis's article in Wooden Boat #165.

 

The illustration in post #26 is present in Davis's article as well.  It's even on page 112 of Chapelle's Boat Building.    Following, is what Davis has to say about that method of laying out camber..."In reality, this method produces a curve that is not even a particularly good approximation of an arc in that it does not have uniform curvature and is about 5% inaccurate relative to a true arc.  Nonetheless, it is generally in boatbuilding books as being more accurate and technical - it is neither."

 

Now, on to my own observations of the CAD drawing in post #24:  Near midship, where deck beams begin to be represented and looking to the bow; the first Red line representing a deck beam makes a perfect intersection with the Black shear line and with the vertical Red line descending from the Green line.  The vertical Red lines are important.  They represent the distance the Black shear line is from the Green line; the Green line being a straight line drawn from the point where the deck meets the bow to where the center of the deck meets the transom.   The curve of the Black shear line is a function of its intersections with the vertical Red lines.   And the Black line is a smooth, fair curve. 

 

To create a fair deck that follows the curve of the Black shear line, the intersections of the Red beam lines with the vertical Red lines must agree with the intersections of the Black shear line with the vertical Red lines.  And as stated above, looking forward, the first Red beam line and the Black shear line intersect the vertical Red line at the same location.  At this location, the Red beam line and Black shear line are in agreement.  

 

Now, go to the second beam forward.  A tiny discrepancy can be seen in the intersection of the Red beam line with its relationship to the Black shear line.  It appears low; maybe too close to call. 

 

Go the the third beam forward.  The discrepancy between the Red beam line and the Black shear line is greater.  The Red beam line definitely lies below the Black shear line. 

 

By the seventh beam forward, on my computer screen, the Red beam line's intersection with the vertical Red line lies about 1/8 inch below the Black shear line's intersection with the same vertical Red line. 

 

With the next Red beam line, very near the bow, the discrepancy begins to correct itself.

 

This is exactly the problem created by deck beams all made to the same camber as described by Davis, and this drawing proves his point.  The same type of distortion occurs in the aft section of this drawing.  With deck beams all cut to the same camber, the intersections of the Red beam lines with the vertical Red lines cannot match the Black shear line's intersections with the same vertical Red lines.

 

On a deck made of beams all cut to the same camber, the amount of distortion, as built (the Red beam lines), compared to the shear, as designed (the Black shear line), is variable, depending on the amount of camber, the rate of taper of the bow and stern and the amount of shear as shown on the shear plan.   

 

 

[Charles Green]

On modern steel vessels known to have been built with all deck beams of the same camber, especially warships:  For the sake of economy and fast construction, a uniform camber makes sense.  Is there any evidence of the use of shims or some type of adjustment on the beams at the points of deck fasteners to fair the deck? 

 

 During repair and refits of historical vessels, when deck planks have been removed, does anyone know of finding evidence of shims or wood removal to make the original deck fair?