A better example needed for Excercise 59 - Canoe, in training manual level 1

I know that shelling is currently included… I think that is a mistake, myself.


Dear Pascal, see it this way. I can provide the curves for normal loft as well as loose loft, but the normal loft curves will take me much more effort to get to a satisfactory end result. For the people not being naval architects it will not matter, exercise 59 provides the basic curves for the loft anyway. But showing the loose lofting will help naval architects in the right direction. But the exercise can show both methods, providing two sets of curves.

I have made another example CanoeCurvesTest4.3dm (639.5 KB), still using loose lofting, where the lofted shell goes beyond the center line. The Shell function is working on this model. Very nice.

Does anybody has a clou why the command _Shell works on this example 4 and not on my previous example 3?


I’m with Pascal on this one. I feel that it is a bad idea to model something that (some) people will expect to function as a product in a complex hydrodynamic environment in a Level 1 course. The focus here should be on the commands and workflows; not the product.
That said, I agree with Bas that, if one has a canoe tutorial, it should be correct. And I do definitely see that as worthwhile, just not as Level 1.

1 Like

Hi Pascal,

I have made a new canoe with normal loft for you CanoeCurvesTest1a.3dm (617.0 KB) with an almost flat bottom as below picture shows. I still do not like the discontinuity in the centre line. Any suggestion to get it smooth without using more tricks?

But zooming in the curvature in the bottom shows that even this bottom is not straight, which is the nature of normal loft to me.


I made a short screencast of my approach to this problem. I think it shows some important principles of modeling with Rhino. I find that you need ‘nice’ curves to start with and most often as few as possible. I also try to understand what a command does best and then I try not to move outside of that scope. Here, what I would say I’m doing is getting a basic ‘fair form’ in the simplest manner possible (which ‘loft’ can readily provide) and then modifying that form locally with methods that retain that initial ‘fairness’. It’s unfortunate that the training manual is rather linearly assembled and isn’t instead structured to jump a bit and bring together some of these more successful practices.


I also stumbled upon this item while taking the course and after spending a frustrating 1/2 hr. by myself trying to achieve the shelling of the canoe as shown by the tutorial I asked the trainer to look into it. After another 1/2 hr. he concluded that there was something wrong with my “settings” or with the exercise. I still finished the solid the hard way trimming surfaces and joining naked edges.

Now, I would like to protest the following:

If the exercise calls for a solid after the shell command, then that is what has to happen. Please don’t play down its importance because is just a level 1 exercise. I paid good money for the course and a statement like that does not sit well with me.

I know I am quoting you out of context, but here it goes. It makes no difference what the model’s ultimate purpose is. If you design a syringe tip that puts plasma into your body, but because this is just a Level 1 exercise where you might get your plasma channel plugged that is not acceptable to me. I need to know that I can build anything I want with this software and not only inoffensive rubber duckies.

1 Like

In my opinion, simply adding shelling to this particular exercise (it did not originally include shelling and was not intended to be used this way) was not a good idea. It was just fine to show how Loft works.


It is not very hard to get it smooth but its a bad idea to even try because it produces a degenerate surface. You don’t want a degenerate surface because it will guarantee that the shell command (and other commands) will fail.

Your previous technique of creating the surface so that it goes past the centerline and then trimming back to the centerline is a far superior way to make the shape if you want to continue modeling further without trouble.


You can avoid the hard way of trimming the surfaces & joining naked edges.

Execute the ‘createsolid’ command… the boat will instantly become a ‘closed polysurface’ :slight_smile:


Dear Chris,

I love your method. It is a nice combination of two methods, e.g. loft and control point manipulation.
For the training manual we need to find a method less advanced, showing the use of the loft command (with the different options like normal and loose) and mean while show naval architects that Rhino is suitable for their use.


I am yet to use this command successfully. In fact I never use it as it does not work, or I do not know how to make it work. Sometimes the object just disappears from the screen and I am being asked if I like it and want to continue. There is nothing to like.

Anyway, coming to our topic, the “createsolid” does not work with my system. Even playing with tolerances from loose to tight does not do the job.


Yes, you’re correct, I find that only the ‘normal’ loft style will result in a successful closed polysurface following the createsolid command.

There is no explanation in the help menus in regard to the ‘createsolid’ command, its only referred as an ‘undocumented command’. I think in order to be successful it requires a set of intersections that define a closed volume. In other words, a number of overlapping surfaces that enclose a space within.


From Help:

CreateSolid command

Creates a closed polysurface or closed polysurfaces from selected surfaces and polysurfaces that bound regions in space.

Note: The CreateSolid command expects overlapping and intersecting surfaces as input. If the objects are simply unconnected, the Join command is normally a better command choice.


Thanks Pascal

… this info is only available when accessing help when using ‘F1’ key. I assumed that all commands would have been described via the help ‘panel’ too. Never assume. :slightly_smiling:

Hi Sean- I see that - thanks, I never noticed. It is there in V6, FWIW…


A post was split to a new topic: What is a degenerate surface?

Hi. I am working with a canoe that I have designed from scratch. I have had many of the same problems I see in this thread and solved a few of them. I think that what you did on the “CanoeCurvesTest4” would solve the last of my issues with the hull shape towards the bow and stern. How did you arrive at, or generate the two “extended cross sections” that are situated outside the canoe fore and aft? They don’t look arbitrary but rather as if they have been generated from the previous cross sections some how.

Hi Johan,
The outer loft curves are indeed a copy of the midship curve. I first have made an interactive centerline by _Intersect and RecordHistory. Secondly I play with the control points of these outer loft curves until this centerline has the shape that I want (or that I have sketched on forehand). I see that I played also with the number of points and the pointweight (see below screendump).
Aantekening 2020-08-24 171910

Hi Bas. Thank you for taking the time to respond. I see that it’s been a while since this thread was active. This thread is the reason I joined this forum.

If I understand you correctly, you first lofted the shape from - the existing cross sections + bow and stern. You then made a copy of it by _Intersect and RecordHistory. You then took the elongated hull shape and played around with the control points for it to match the original hull shape as close as possible. You then used a vertical center plane to trim it back. Did I understand you correctly?

I don’t understand your command _Intersect and RecordHistory. This is the step that intrigues me. I’m hoping that you can explain this in more length.

From the cross sections given I think your approach works! As a canoe enthusiast though, I see a problem. Not with the generated hull shape per say. I think the problem is in the 2D design stage of the canoe. The mid cross section and the first stations out from it are just a tad to flat in the keel (maybe even has an inverted kink towards the keel?) and the mid ship rocker is also just a little bit to flat or has a kink. This is what generates the dreaded “oil canning” or inside convex bottom in the middle of the canoe.

If the curve analysis is carried out, all lines faired and adjusted to make a little rounder bottom profile, and the rocker adjusted slightly. The canoe should be technically flawless!
Then the rest is a matter of production and ones preference in canoe hull shapes.