Patrick Matthews

That fine chain is all very fine... but if you need some chunky stud-link anchor chain, it can be printed!
 

The 3D CAD model is complete and available for free download from GrabCAD:
https://grabcad.com/library/lifeboat-uslss-1908-mclellen-type-e-36-footer-1
 
Find the 3D STEP file there, along with reference photos and the original 2D drawings I worked from.


 
 
Take the model for a spin at GrabCAD!

Designed in CAD starting in 2015. Made a few changes... since 2015, I built a whole new hull and a whole new cabin, tossing the previous versions. Accurate plans don't exist, but overall dimensions are known, and there's a decent profile from a sale brochure for Wheeler Playmates. 

Most of the model is hand built from wood of course... but certain components were printed- either in plastic, or in wax for investment casting in bronze. Such as the struts and rudder footing. The footing works with the extended wood  skeg, a common design layout in the 20's and 30's. This casting is missing from the boat at Finca Vigia, which leads some modelers to make some odd drive arrangements. If only they knew...
 
Hemingway had Wheeler make several well known modifications to Pilar, including the installation of an auxiliary 4 cylinder Lycoming for trolling. There is no evidence of this on the boat now, so this is also omitted from models.
 
. 

Cabin is constructed of ply and basswood, with all exposed wood done in cherry. The real boat appears to have been built with mahogany, but I find that cherry "works" better, and has a better scale grain appearance. The sides are built up as a 3d jigsaw puzzle, as many as 3 layers deep. They hang down straight like a curtain from the deck intersection, but angle inwards at 2° from that line up, while also following the sheer... a bit tricky.
The cabin roofs are covered with shirt cloth to simulate canvas. The cloth is secured with water based polyurethane, painted, and clear coated. The weave still shows through, like painted canvas.
The cherry is all stained with Minwax "Gunstock", and clear coated with lacquer. Warning: Most all of my cherry stained a very nice color, but occasional pieces turned orange, yuk. Always test!
.

I changed my mind after this shot and added the forward cockpit.
 
 

Roof beams
 

First piece of cloth installed
 

Thx, I think. 
p.83:  "1. 24-pdr English long gun"
 
It's already been established that contrary to an earlier opinion, she DIDN'T (for the most part) carry English guns. 
Probably doesn't matter much, they all looked about the same! 
Except for the mistaken English markings cast on the 1930 repro guns...

I've taken plenty of photos of my models, mostly in-process on the work bench. Sometimes I'll make an effort to take some beauty shots, always with so-so results. But my brother, who is a pro photographer, showed me some tricks using my PILAR model. Am I likely to pick up the same tools he has? Not too likely, but a lot of it actually isn't too much of an add, assuming you already have a decent camera.
 
My impromptu lesson took place in my garage, with a big piece of black fabric for a back drop, and a coffee table to hold up the model. 
The main "tool" was a soft box-  a big umbrella like thing with diffusers and a strobe inside. This was arranged over the model to rain soft parallel light down from above. We played with a reflector panel to get some bounce-fill, and even resorted to an old photo flood to fill in a dark spot. The process had a lot of trial and error: take a shot, examine it, adjust the lighting, it was all magic to me.
 
It was hardly a complete studio, but we made do with it to get some decent shots. See more in the scratch build gallery under PILAR. 
 
 
My brother Wil doing his thing in the garage:

 
 
Better view of the soft box and reflector:

 
 
Using an old flood to fill in:

 
 
... to get results like this:

Another update- the entire PILAR 3D model is now on GrabCAD, along with 2D drawings for the cabin structure. I may be adding more 2D detail drawings, but the underlying geometry is all there now in the 3D STEP file.
 
 
Take the 3D model for a spin at GrabCAD! 
https://grabcad.com/library/pilar-model-in-1-12-scale-1

Another CAD model posted on Grabcad for free download- my favorite tugboat, Henry Ford's "Dearborn".
https://grabcad.com/library/tugboat-dearborn-1

This is the one I tried Delftship on... but gave up and went back to Fusion360.

 

Another CAD model posted on Grabcad for free download- my favorite tugboat, Henry Ford's "Dearborn".
https://grabcad.com/library/tugboat-dearborn-1

This is the one I tried Delftship on... but gave up and went back to Fusion360.

 

Very nice & complete work!
Yes, my goal is to document vessels in a manner useful for modelers. I'm not creating 3d printable files. But my documentation includes 2d drawings as well as a 3d CAD model. The cad model can be used directly, or examined to better understand my abbreviated 2d work 
I do like the DS lines drawings, and that's another area that F360 erects unnecessary roadblocks... But I'm finding ways to work around those too. See my Fire Fighter model for example:
https://grabcad.com/library/fire-fighter-fireboat-1
 
 
 

Nope on the 3D DXF. That results in a mesh not unlike an STL- points joined by straight connectors, resulting in a faceted surface. It is NOT a smooth curvy surface. 
The IGES export is smooth & curvy... I just don't know if the entire side of my hull with its compound curved surfaces comes out in one patch or not.  
 
Why do this? Because while Fusion 360 can be a pain for creating lofted hulls, it's a champ for making all the other details. And it has history, which is priceless in large complex models. So if I can make a correct hull in DS, I'd just export the shape to F360. But as I detail above, that subdivided surface from DS using a control mesh is WRONG, and  it requires a lot of tweaking to correct.  

Now, with much gnashing of teeth, I have been able to create a reasonable lofted hull in F360... so it's just a matter of which path is less headache.

My F360 generated hull:

Thanks!
My biggest question regards the IGES output, only available in the Pro version. They told me it can output many, maybe hundreds, of surface patches, which would likely look like a mosaic when used in other CAD software. But if, for example on my tug hull, it comes out as smooth single patch for the main surface, that would be fine.
They sent me a sample IGES file, but it was of a simple developable hull, so not really a test of this feature. The transom (port) was fractured, but I stitched the surface patches in Fusion360, and it all merged nicely (stbd). 
 

Here's an end view showing some of these features:
White: Original control points and the control mesh with its straight Tinkertoy connectors. The control points lay on the proper body sections from the lines drawing.
Green: Body sections cut through the DS developed hull surface... they miss by a lot. Even at the deck edges.
Pink: Nice curvy marker curves that DS draws for the second import of the original control points. These markers are the target, I need to move the green lines to match them.
 
To move any one point: Click on it, and you can enter new numbers or use the little arrow to move the point by set increments, very handy. Move each and every one this way to drag the developed sections around.

So here's the second most important thing for DS users to know:
Lines imported from a lines drawing DO NOT create a hull true to those lines!
 
First most important? It can be fixed.
 
As above: "The developed hull shape does NOT go through the control points, which are input from the original lines drawing... they just INFLUENCE the shape. Fine except it's not, because the developed form no longer matches the original. There must be a work flow that gets the generated shape closer to the desired form."
Indeed the DS developed hull cuts corners, it's always well to the inside (concave side) of input curves, except at the "corners". 
 
Imagine you have a magnetic rubbery sheet pegged by its 4 corners to the floor, exactly where you want the corners. 
Then you hang a bunch of magnets from the ceiling, and each magnet is hanging exactly where you want the sheet to stretch and rise to, defining some sort of hill (I'm picturing Richard Dreyfuss building Devil's Tower in his living room).
But while the sheet is stretchy, it's not THAT stretchy, and it hovers over the floor but doesn't quite reach the magnets. Oops. That's what we have in DS, with the developed hull not reaching the input points.
 
How to fix it? Just keep raising the magnets until the sheet rises to the proper place.
In DS, I did this, though there may be more efficient ways:
1. Import the control points once to define the control mesh and its resultant malformed hull.
2. Import the control points a second time as "marker curves". Markers are static curves, immutable, and they serve as references of where the original control points and desired hull sections are. 
3. In an end view, compare the developed body sections (which are wrong) to the desired shape shown by the marker curves. Now start moving control points sideways or vertically to slowly pull the body sections towards their markers. Yes, this modifies the original control points and mesh! 
4. Note that this is iterative... moving one point affects the adjacent surface faces, so you'll need to sweep back and forth over all the points to get everything close.
5. Also note that the four edges of your rubber sheet between the corner pegs were also pulled off the floor when the magnets started lifting the center. Same thing in DS- the edges at deck edge and keel also have moved out of place. So the adjusting exercise needs to address those too.
 
I did this for my tugboat hull, and while it still needs some adjustment, it's much closer to the desired shape now (below).
 
More: While you can get close by doing this, you can also turn on curvature displays to better smooth (fair) your new shapes. I haven't done this yet, may not need to for my present purposes.
More #2: The Pro version of DS has an "automatic fairing" tool. I'm not sure how this works, but it could be of help.

Another CAD model posted on Grabcad for free download- my favorite tugboat, Henry Ford's "Dearborn".
https://grabcad.com/library/tugboat-dearborn-1

This is the one I tried Delftship on... but gave up and went back to Fusion360.

 

Very nice & complete work!
Yes, my goal is to document vessels in a manner useful for modelers. I'm not creating 3d printable files. But my documentation includes 2d drawings as well as a 3d CAD model. The cad model can be used directly, or examined to better understand my abbreviated 2d work 
I do like the DS lines drawings, and that's another area that F360 erects unnecessary roadblocks... But I'm finding ways to work around those too. See my Fire Fighter model for example:
https://grabcad.com/library/fire-fighter-fireboat-1
 
 
 

Nope on the 3D DXF. That results in a mesh not unlike an STL- points joined by straight connectors, resulting in a faceted surface. It is NOT a smooth curvy surface. 
The IGES export is smooth & curvy... I just don't know if the entire side of my hull with its compound curved surfaces comes out in one patch or not.  
 
Why do this? Because while Fusion 360 can be a pain for creating lofted hulls, it's a champ for making all the other details. And it has history, which is priceless in large complex models. So if I can make a correct hull in DS, I'd just export the shape to F360. But as I detail above, that subdivided surface from DS using a control mesh is WRONG, and  it requires a lot of tweaking to correct.  

Now, with much gnashing of teeth, I have been able to create a reasonable lofted hull in F360... so it's just a matter of which path is less headache.

My F360 generated hull:

Thanks!
My biggest question regards the IGES output, only available in the Pro version. They told me it can output many, maybe hundreds, of surface patches, which would likely look like a mosaic when used in other CAD software. But if, for example on my tug hull, it comes out as smooth single patch for the main surface, that would be fine.
They sent me a sample IGES file, but it was of a simple developable hull, so not really a test of this feature. The transom (port) was fractured, but I stitched the surface patches in Fusion360, and it all merged nicely (stbd). 
 

Very nice & complete work!
Yes, my goal is to document vessels in a manner useful for modelers. I'm not creating 3d printable files. But my documentation includes 2d drawings as well as a 3d CAD model. The cad model can be used directly, or examined to better understand my abbreviated 2d work 
I do like the DS lines drawings, and that's another area that F360 erects unnecessary roadblocks... But I'm finding ways to work around those too. See my Fire Fighter model for example:
https://grabcad.com/library/fire-fighter-fireboat-1
 
 
 

Here's an end view showing some of these features:
White: Original control points and the control mesh with its straight Tinkertoy connectors. The control points lay on the proper body sections from the lines drawing.
Green: Body sections cut through the DS developed hull surface... they miss by a lot. Even at the deck edges.
Pink: Nice curvy marker curves that DS draws for the second import of the original control points. These markers are the target, I need to move the green lines to match them.
 
To move any one point: Click on it, and you can enter new numbers or use the little arrow to move the point by set increments, very handy. Move each and every one this way to drag the developed sections around.

So here's the second most important thing for DS users to know:
Lines imported from a lines drawing DO NOT create a hull true to those lines!
 
First most important? It can be fixed.
 
As above: "The developed hull shape does NOT go through the control points, which are input from the original lines drawing... they just INFLUENCE the shape. Fine except it's not, because the developed form no longer matches the original. There must be a work flow that gets the generated shape closer to the desired form."
Indeed the DS developed hull cuts corners, it's always well to the inside (concave side) of input curves, except at the "corners". 
 
Imagine you have a magnetic rubbery sheet pegged by its 4 corners to the floor, exactly where you want the corners. 
Then you hang a bunch of magnets from the ceiling, and each magnet is hanging exactly where you want the sheet to stretch and rise to, defining some sort of hill (I'm picturing Richard Dreyfuss building Devil's Tower in his living room).
But while the sheet is stretchy, it's not THAT stretchy, and it hovers over the floor but doesn't quite reach the magnets. Oops. That's what we have in DS, with the developed hull not reaching the input points.
 
How to fix it? Just keep raising the magnets until the sheet rises to the proper place.
In DS, I did this, though there may be more efficient ways:
1. Import the control points once to define the control mesh and its resultant malformed hull.
2. Import the control points a second time as "marker curves". Markers are static curves, immutable, and they serve as references of where the original control points and desired hull sections are. 
3. In an end view, compare the developed body sections (which are wrong) to the desired shape shown by the marker curves. Now start moving control points sideways or vertically to slowly pull the body sections towards their markers. Yes, this modifies the original control points and mesh! 
4. Note that this is iterative... moving one point affects the adjacent surface faces, so you'll need to sweep back and forth over all the points to get everything close.
5. Also note that the four edges of your rubber sheet between the corner pegs were also pulled off the floor when the magnets started lifting the center. Same thing in DS- the edges at deck edge and keel also have moved out of place. So the adjusting exercise needs to address those too.
 
I did this for my tugboat hull, and while it still needs some adjustment, it's much closer to the desired shape now (below).
 
More: While you can get close by doing this, you can also turn on curvature displays to better smooth (fair) your new shapes. I haven't done this yet, may not need to for my present purposes.
More #2: The Pro version of DS has an "automatic fairing" tool. I'm not sure how this works, but it could be of help.

I did receive replies from DS HQ:
1. The forum is not broken, new posts are moderated. But still, not so much recent activity, which is curious.
2. Hull surfaces exported as IGES are "smooth parametric surfaces, but please keep in mind that the exported hull can consist of many (sometimes hundreds) of such surfaces." So it's not clear yet whether that would look like a broken mosaic when imported into my CAD.

Thanks and yes, but a lot of material is a bit dated. I don't see much recent activity, makes me wonder if this is still a popular tool?
I sent an email to DS HQ asking why their forum is broken... no answer yet.
 
But I'm wrestling with topics like:
- The developed hull shape does NOT go through the control points, which are input from the original lines drawing... they just INFLUENCE the shape. Fine except it's not, because the developed form no longer matches the original. There must be a work flow that gets the generated shape closer to the desired form;
- Deck camber- the "Extrude Curved" tool is supposed to generate a nicely cambered deck, except it doesn't... it's generating garbage for me;
- You can pop 160 euro for the Pro version which will output the surfaces in IGES format. But what format? A faceted mesh, or a proper smooth surface?  The available exports in the free version just give the faceted mesh, which looks terrible when imported into, say, Fusion360 (as seen below). Unusable. 

The whole reason for me exploring Delftship is its better ability to generate hull shapes, compared to the buggy tools in F360, whadda headache (that area around the sternpost/stern counter is a BEAR to loft). Maybe I should pop for Rhino, $995 ouch.

 

I'd like to get help at the official Delftship forum:
https://forum.delftship.net/
but it seems broken and/or unused... I can't successfully post there, and most topics and responses are at least weeks old...