La Belle by Kranck - as per the monography of Jean Baudriot

[Kranck]

As I dived into the hobby 2 years ago, I had no idea of its depth and the time it would consume. I started with HMS Victory, a common mistake I suppose. I had bought a kit for my mother that had just retired. But she told me it was too difficult for her. And, well, she had started several hobbies at the same time. So I gave it a try. I had no previous knowledge of woodworking. I'm a medical doctor but I got an engineering master earlier. So I supposed I should be able to get the best of it. But this ship was a mistake. there is no way I could master it with 3 files, some glue, two cutters and absolutely no experience. I spent almost one year on this kit though. I managed somehow to finish planking. At the same time, I joined a modeling club that happens to be quite active in my area. Interestingly, they used to manage a small ship museum in a parking building on one of the villages that border the mediterranean sea here. The museum closed when they got older, but they are still actively building.  Most of the members do scratch builds. Most are over 75 though... This is in no way comparable to the  guilds that are active around Paris, and the quality of the builds is not as high, but those folks are nice and I'm a beginner !

 

So I started building from scratch Le Cerf, from the monography of Boudriot. I learned a lot. I just finished planking. I use this build as a draft to test various procedures and learn as much as I can. I do not intend to log this build.

 

I decided to start an 'admiralty' model as I feel that I would love dealing with the intricacies of ship carpentry. I will use what I learned with Le Cerf to do what I can with La Belle, another small ship well described by Baudriot. 

I have access to lots of books and I bought a lot of tools these last months. 

Besides, I intend to design on the computer as much as I can. I 'd like to CNC as many parts as possible, especially to get precise and fair measurements for the hull. 

 

I bought or got access to the following tools:

- Byrne's saw, thickness sander and disc sander. Those are the best, enough has been said about those tools. As another modeler truthfully said: the only limitation is my current skill.

 

- Shapeoko 4 XL: this is a nice 3 axis CNC that allows pretty much anything from engraving, moulding or designing bulkheads in any kind of wood and so on. Just great. I used Fusion 360 to design the hull of Le Cerf but I feel it is too much oriented towards mechanical engineering. I use Aspire for art thingies (sculpures and so on). and I'd like to use Rhino for the hull (bought a student licence as I spent some time at the uni again for an oncogenetics diploma this year). Wasnt all for naught then. 

 

Here is an example of what can be done with it :

 

And after sanding the background, I could glue it to Le cerf:

 

 

- Metabo bandsaw: this is nice. Much better than the small modeling saw yet accurate.

 

- Proxxon Ibs/e. God bless Germany. This is also a marvel.

 

My first steps with la Belle:

 

- Scanning all blueprint and printing a copy.

- Learning what I can about the hull of those ships. I got a few interesting books to keep me awake at night.

- Learning Rhino (that will take time...)  and designing the keel and berth of the ship.

- At the same time, design the decoration with Aspire. I gave it a try for a few hours:

 

 

Aspire already provides a lot of decorative patterns that can be used. Since nobody actually how what those really looked like on the actual ship, I decided to give some leeway to my imagination. 

Like the Cerf  decoration on the stern, I will engrave it on boxwood. But finding the good bits is a pain for such tiny details. Beside, 3D CNC are inherently limited to Bas-relief 2.5D engraving. There is no way I could sculpt intricate full 3D details. 

We'll see. 

 

I'll post more when I get something nice with rhino (could be very long!).

[druxey]

Well, you are deep down the rabbit hole, aren't you. Kranck? I  wonder if it takes longer to design the files for 3D cutting than if one were to actually carve the pieces (assuming one has already acquired carving skills).

[Kranck]

Hi Druxey,

I agree that for someone with no CAD expertise, the whole process will be actually longer than carving by hand...

But I lack the carving skills ! Unfortunately there is no one around to teach me and I lack the talent for this kind of art. Besides, I do love looking at 3D models on computers. It does help me as I have some issues visualizing a ship from blueprints only.

Another pro is that you can easily start again if you fail with one part or have trouble with assembly. And I find that learning CAD is quite rewarding by itself. 

Lastly, it means that files can be shared within a community, which could be great. 

We'll see. 

Lot of work ahead.

[mtaylor]

Kranck,

I moved this topic to the appropriate Scratch Build area.

[Kranck]

ah yes, my bad !

Thank you

[mtaylor]

9 hours ago, Kranck said:

ah yes, my bad !

Thank you

No worries.  Happens more often then one thinks.

[Kranck]

So,

First try with rhino lofting. 

I fight with the geometrical errors of the monography as well as my own incompetence. 

Yet, there are several interesting resources on the forum to learn about lofting thoses surfaces. 

The aim would be to intersect the inner and outer surfaces of the frames to get a perfect shape before planking. 

But well, this is a remote goal. 

For the moment, time to beautify this loft. I will follow a step by step process according to the webinar one of the guys here gave a few years ago. 

Darn rabbet. 

 

 

[Kranck]

 

Getting there. 

The tormented whirlwind (funnier than mathematical singularity) in the front disappeared due to the creation of a new section (new rhino command for me!) at the appropriate distance from the bow. 

I prolongated the surface at the rear (thxs to wise advice from Waldemar). 

But let's face it: It is a pachidermous monstruosity. The curvature diagram is a mess. 

Let's ponder and meditate about fairing tomorrow. 

[mtaylor]

If carving is an issue... have  look here https://modelshipworld.com/forum/84-build-logs-for-the-carving-group-project/.  Basically, it's nice learning how to carve using simple tools.

 

I think that with the tools and plans you have, the Rhino won't add much as the parts can be cut from the plans using the tools you own.  

[Kranck]

Hi,

I agree that this is not essential to the building of the ship. I could do it by hand.

But I used to be an engineer and I'm having some fun designing the ship on the computer too. It is clear that for the moment, it will be extremely time consuming. 

But in future builds, it should be faster and more useful. Let's say it is a proof of principle experiment I'd really like to develop and CNC mill the planks. 

Besides, I have no experience with naval architecture so depicting the parts on the computer helps me visualize the overall construction of the ship.

[Kranck]

So,

A happy and fruitful year to you all. May our ships remain fair and waterproof.

 

I worked some more on the surfaces. 

Unfortunately, more issues than successes. 

 

* Working with the outer surface of the frames and connecting it to a lofted 'inner surface' is actually very difficult:  The tangency and paralellism of those two surfaces is somewhat inaccurate and hard to control, especially at the bow and stern of the ship. As a consequence, some frames have a thickness that does not change properly along the Z axis

* I chose to ad thcikness to the sole outer surface instead. It is safer but less accurate. I should add some thickness later to the base of the frame, modifying manually the surfaces of each frame one by one. 

 

* I found that I lack proficiency in the reading and understanding of the blueprints. I'll ask some friends about that. 

 

* A lot of work to solve (partially) the issue of the continuity problem in the bow. It turns out that the 'wirecut' tool is a lifesaver for it.

 

 

[Kranck]

Bow (allonges d'écubier ? Futtocks?) with a first try at choosing a material. 

I had no idea these were designed this way in older naval architecture. It is quite satisfying to see that the monography is quite accurate for this part of the ship. 

However, I'm begining to have second thoughts about how to actually make these parts. No way 3D CNC is sufficient. 

 

[Kranck]

So,

Things are getting along. I'm doing better with the software and the overall strategy. 

I was about getting started cutting frames in 3 when the harsh truth struck me: I got an issue with the overall thickness of the frames, mainly at the center of the ship. 

As you can see on the second picture, I only added thickness on the outer surface of the hull. But this thickness is not the same everywhere. I deemed it negligible at first, but there is up to a 3 mm difference between the blueprints and the frames I drew. 

I found out just today how to edit the inner surface of the loft. But it is a serious setback, as I must come back to an earlier version of the work. 

There is also the alternate solution to modify each frame individually. I doubt it would be rewarding. 

Thinking.

[Kranck]

Some update:

- I redrew and lofted again the outer and inner surfaces. Much better. Some futtocks are still a little too thick, especially towards the bow, but this can be dealt with later with smaller cuts with an uptdated version of the inner surface loft if needed. 

- The stern is quite satisfying. Lots of projection of curves on surfaces etc, and a final touch directly dealing with solid edition but it is quite accurate. 

- Currently cutting the futtocks in smaller parts as shown on the blueprints. 

 

Next step will be setting the decks. 

[Kranck]

I designed the decoration of the poop with Aspire. 

It is not exactly as Baudriaut drew it, but I had the opportunity to use a bunch of components already available. It should be ok to machine it with a very small engraving bit on boxwood. I worked quite well with the decoration of Le Cerf. 

[Kranck]

Dire times. 

I designed some more elements in the hold and at the bow. 

I tried to project the contour of the inner planking of the hold, but the results is not satifying. 

Obviously, it is wobbly and irregular because of some bumps on the frames due to an inner surface that is not fair. The outer surface if much better. 

Projecting the planks and adding thickness on those frames highlights the defects of the surface. 

It should be better with actual wood, since there will be some rigidity with the material and some sanding that will allow for a compensation of the irregularity. But up to what point? I'm not sure it will be sufficient to smoothen the biggest bumps. 

 

I don't know what to do. I'm not sure I can get a fairer hold. 

Any slight offset or UVN change in the lower part of the hull surface results in big changes in the actual XY position of the projected contours of the planks, which is quite natural since the surface is almost flat at that location. 

 

Is there another way to model those planks? 

Should I loft again the surfaces? 

Should I stop designing here and get to woodworking with the hope of compensating the defects?  

[Richard Dunn]

In my opinion it is worth the effort to have your frames fair and correct in 3d before you start, no matter how long it takes.
If you get the outside fair by lofting every 5th frame ,more at the bow and stern you should..... should have a fair surface.
Mark the lines on that like ribbands that correspond to the futtock joints and then do an extrude normal to surface inwards to the depth the frames at that point, use the top of the floor line on centreline of course and then once done section those at the SAME frames you loft on exterior, this forces the loft to behave in the same way as the outer surface. and then run waterlines and buttocks etc to evaluate the surface.
Looking at the images above the frames don't look that bad, but obviously they are if that's the result.
did you trace the Monographie from the actual frame sections or the body plan?
In mine I used the body plan but also set up every 5th frame as a reference to make sure it was within tolerance.

I will give you a copy of my rhino file later so you can see how its done, I will remove some stuff though for copyright reasons.
I should note to that that pink surface looks really nasty in regards to the isoparms, what type of surface is it? once you get a surface for the inner just define the edge of one mating face offset that on surface by width and cut and then do an offset surface with make solid on so complete the part, if you get joins you need to rebuild the curve first but pay attention to the tolerance of any change and keep it within .25mm..well I do anyway.

Could you send me the rhino file with just the lofted inner and outer surface and the curves for me to look at?.

Edited January 24, 2023 by Richard Dunn

[Richard Dunn]

Also I just noticed this image,

the isoparms on this are clearly showing a non smooth surface, I use the word smooth not fair because these ships are seldom fair in the way 3d software mans fair, so its all about looking at the lumps and removing wandering isoparms.
as a rule isoparms should always be a good indication if this are right.

To me one or more of those 2nd third or fourth lofts are not right, also try to establish a fair sheerline, it will help.

the other option is to turn on gaussian curvature and manually move points with Move UVN  and curvature graph to remove bumps but I still think a heavy handed result like that should not be needed based on my experience with these books.
but this surface is for sure not right.....

 

[Kranck]

Thank you so much for your interest in thiws work. 

This is really important work for me as I intend to do this properly for each ship I build. I think that this is not bad for a piece of software I didn't know just 2 months ago but still...

 

I fully agree with your statement: It is worth getting a correct loft before proceeding with the rest. 

I started lofting again with without using 'normal', it is much better. But the accuracy cannot be as great as intended and I get some curves quite far from the monography. 

I lofted from the stations drawings. 

 

The issue I run into when using your method is that when I extrude from the futtocks joints with the thickness at that point, the resulting frame (whose section I could use to loft the inner surface) is too thick close to the deck and too thin close to the keelson. Unlesse I misunderstood our method ! but I agree it should force the loft too behave in the same way as the outer surface, which is what we want. 

 

The pink surface is obtained according to the following:

- Draw the contours of the hold planking in the XY plane (monography)

- Project those contours on the inner surface 

- Create a surface joining 4 sides of each plank

- Solid Offset this surface to get the plank. 

 

This is too complicated, I should use your method to cut the inner surface.  

 

In your second post, you mention a picture of my first loft. I discarded it for those same reasons: isoparms were too nasty

 

I will prepare today a file with my new loft. I tried a new method to get the bow in a separate way. 

 

Many thanks

 

PS : Just bought some wild cherrytree planks. Eager to machine !!

[scrubbyj427]

You could also take sections through your butt lines and then offset each one to the thickness you want the frame (in profile view)  and then proceed to build the new frames around those lines as a reference. Also as @Richard Dunn mentioned, offsetting by the outer surface of the frames and creating a solid from that would be  beneficial. The issue is that the inner hull and outer hull are shaped differently.

Have you tried to just offset your hull surface by itself and see how that results?

For my model I just used offsets from the hull surface at the thickness I wanted for the frames and extruded solids so the tops of the frames are as fair as the outside. These were always taken from the higher side of the hull, for example, in the bow I would take the section from the fwd facing side and extrude aft and the opposite in the stern. This allows for proper sanding. In rhino you would do the same but run you spline that you would use to create planks, from the highest side of the frame to the next highest one, slicing right through the part that would be sanded… if that makes any sense.

 

Edited January 25, 2023 by scrubbyj427
Added photo.

[Kranck]

The link to the rdm files follows, if the rules of the board allow it. 

 

https://drive.google.com/file/d/11s1BxOKmqo73I2yY2q8fYigun9q_qBle/view?usp=sharing

I removed everything that was not necessary. I kept some image files within.

The external surf is not linked to the curves because I tweaked the UVN motion, which breack history. 

However, for the internal surf, you can still use curve motion to control the loft. 

 

This is the best I could do so far. 

The frames I get are not that bad. 

I chose not to draw the bow this time (which might be mistake).

[Kranck]

43 minutes ago, scrubbyj427 said:

You could also take sections through your butt lines and then offset each one to the thickness you want the frame (in profile view)  and then proceed to build the new frames around those lines as a reference. Also as @Richard Dunn mentioned, offsetting by the outer surface of the frames and creating a solid from that would be  beneficial. The issue is that the inner hull and outer hull are shaped differently.

Have you tried to just offset your hull surface by itself and see how that results?

For my model I just used offsets from the hull surface at the thickness I wanted for the frames and extruded solids so the tops of the frames are as fair as the outside. These were always taken from the higher side of the hull, for example, in the bow I would take the section from the fwd facing side and extrude aft and the opposite in the stern. This allows for proper sanding. In rhino you would do the same but run you spline that you would use to create planks, from the highest side of the frame to the next highest one, slicing right through the part that would be sanded… if that makes any sense.

 

 

Hi,

 

I did try to just offset the hull surface. There are two issues:

- If the surface is offset to a defined thickness, the top of the frames is too thick and the lower part is too thin.

- I the offset is variable, it is very difficult to get something nice in the stern area with the tools I found in rhino. It gets very bumpy and not accurate since the offset variation is linear. 

 

I will think about your method, I'm not sure I understan. 

 

Best regards,

 

Nicolas

 

 

[ccoyle]

3 hours ago, Kranck said:

The link to the rdm files follows, if the rules of the board allow it. 

 

Sharing the files for discussion is fine. The question becomes a bit trickier should it ever seem that you are actually distributing the files -- that would bring up the question of whether your work is sufficiently derivative to not be considered a copyright violation (a determination that ultimately can only be decided by a court). One way to avoid that might be to make the files read-only.

[Kranck]

Thanks for the clarification

Alas, my work is not proficient enough to even resemble a copyright violation...

 

Cheers,

 

Nicolas

[Kranck]

Reworked the inner surface once again 

The outer one is fair.

 

In color, the old frames.

In grey, the new ones. Much fairer. I managed to get something much smoother and removed the bumps here and there.

Still a bot of work at the stern, where I  will keep the sternpost (the design is fine, no need to do that again).

I'll have to deal with the bow too. But how? Many possibilities. 

 

Sorry for the lack of consistency in this log. Ideas come and go, and I write what I think is right at the time...

[Kranck]

work in progress. 

[Kranck]

Finally the main structure of the ship is finished. 

There are several defects, especially in the aft area as depicted in the monography. 

But those are acceptable. 

I planked the deck. I will try to plank the hull later and perhaps deveelop the planks for CNC cutting. 

 

Speaking of CNC, I decided to cut the first parts, starting with the keel. 

I wanted to start designing a process and see wether it had any chance at all to be successful. 

Results are mediocre so far but I should improve. the process is the following:

- Designing the part with Rhino

- RhinoCam is too expensive (but new licensing option are available for 2023 that may be interesting). So I export the part as a stl file

- Importing the file in Vectric Aspire. It is a nice piece of software available at my club. 2-sided machining seemed quite straightforward. 

-  Preparing 2-sided machining on Vectric. Ok, it was not so straighforward. Tip: using 3 asymetric dowels to reposition the stock when machining the other side. Use tabs to avoid motion of the part.

- I bought nice planks of wild cherry. But they are too thick: 25 mm, wich is too much for 1mm bits. 1 mm bits are sufficiently precise to capture details and approximate rectangular pockets for instance with the corners looking too 'round-shaped'. The issue is that 1 mm bits are not that long, otherwise they would be too fragile. So a first step roughing the stock with a 1/4'' endmill is necessary. Roughing area should include lateral margins around 1 cm to account for the width of the router collet when the 1mm finishing bit plunges far below the surface of the stock.

- Cut and hope for the best.

 

I will not summarize here the pain of this trial and error endeavour. It was fun in the end though. 

The first part is not that nice. Many errors. But the next will be better. I shall use the mill to improve the accuracy of some corners but the surface is real smooth .

[Kranck]

Here is the wood. Next one will be better hopefully

[Mic_Nao]

Your screenshots show your email address. It's risky...

 

For your work, it's good !

[Kranck]

So, after a few months on and off the hobby, I just wantd to update on the project.

 

 * CAD drawing was satisfyingly accurate. I stopped at some point to start getting some real parts for this ship.

 * Getting accurate 2D or 2.5 parts is rather easy with one sided machining on the shapeoko 4. It took just a few days to get a .2 mm accuracy, which is sufficient for our purpose. 

 * Getting 3D parts with two sided machining is an entirely different matter:

1. Consider for instance a frame or a part of a queel with a rabbet on each side. The two main issues with two sided machining are getting an accurate thickness when turning the part, and be able to accurately zero X and Y. 
    
2. But the nightmare starts way before that: first, you have to choose a CAM soft to be able to transfer your part into a series of G-code instruction. To whom it may concern: most pieces of software available for hobby are NOT accurate enough. Only engineering-grade soft are good enough. For instance, Aspire, from VECTRIC, does not handle properly 3D parts (an issue with voxel size). Other just can't handle 2 sided machining properly. 
    
3. I ended up using fusion 360. But first, it is therefore necessary to transform the rhino files into fusion 360 readable files. You don't want to know, but it is not an easy tasks (many restriction for the free version of fusion 360).
    
4. Since even in fusion 360 there is not 'intuitive' way to prepare your files for 2 sided machining, It does take a lot of time foreach file to do so. And an error in one single step will wreck your piece on the machining table. 
    
5. Then you have to deal with all the diffficulties of preparing appropriate dowels, double sided tape, and be extra careful for each manipulation of  the part while machining. Of Note: the more you machine your stock, the less wood is left for rigidity. It WILL start to bend. 

 

Well, as you can infer, I finf this step extra difficult. I increased my knowledge of the process and I got better and faster. I do enjoy the accuracy of the curves, but with such tools, it is wrong to think that all parts will magically fit perfectly. Sanding is indeed required. 

 

I just wanted to show the queel and the first half-frame.

 

I do wonder how people can get parts to so nicely and tightly fit. Is it necessary to finish machining with a manual milling machine? 

I also wondered how do hobbyists highlight the edge of the parts with a black line? I think caulking shouldn't be visible at this scale, but the effect is nice. 

 

Anyway, my main difficulty is that I have no bakcgroung in manual machining of wood. I hoped to compensate with CAD and CAM but it's hard. 

 

Any good books on the woodworking tehcninque itself as requied by the hobby?

 

Best regards, 

 

Nicolas

Edited September 2, 2023 by Kranck