Rhino Mini Tutorials

[Willja67]

The resurrection of the Z place prompted me to return there and look around. While doing so I came across the attempt I made at starting a tutorial and I felt somewhat guilty for abandoning the attempt, but I also realized 2 things;

1. There are plenty of tutorials for people to learn how to use Rhino(at least the basics).

2. My attention span is too limited to complete such an undertaking.

Perhaps this is a conceit but I like to think that I come up with creative ways of doing things and that my approach to designing paper models only follows conventional paths when i think that is the best way, but I try and find other ways than the conventional of doing things.

Although the title says Rhino, I will presenting ideas that I think can be used with other software and if anything is more about a mindset than software. After all the fanciest software won't help you if you can't wrap(pun intended) your mind around the shape of what you are creating. As such there will be little regarding commands used and more about thought processes.

To that end I will be presenting (perhaps infrequently) small tutorials about how I designed certain tricky portions of my models. If any of you have seen something on one of my models that you wondered how I accomplished in the computer ask here and I will try and show you.

I hope that I don't exhaust the topic too quickly but the fact of the matter is that there are surprising few principles involved in the design of paper models, there are just infinite ways to apply said principles. Hopefully this will help all those fledgling designers (and maybe some veterans as well)out there learn to think a little differently about what can be done.

Edit:

I don't want to appear to be an arrogant know it all. If anyone has any tips they wish to share feel free. I study every paper kit I can get my hands on to see how other designers tackled a problem. The intention of this thread is to increase the knowledge base.

[Willja67]

To start off this series I borrowed something I had already posted from my Super Corsair thread.

One thing to keep in mind when designing is that you have to be smarter than the software you are using. You have to realize that the software deals in absolutes but you don't necessarily have to. Rhino may be a extremely well trained beast (it knows more commands than you ever will) but you have to string those commands together in the right way to make something happen.

The point I wish to make here and is one of the few core principles mentioned above that has infinite applications is this:

We often think that we only have to break up the shape we are modeling into individual pieces. You can break up what is going to be one part into many pieces in the design phase then put them back together for the finished product.

There are many variables but depending on that shape you are modeling you can have curves on more than 2 sides and still get it to lay flat as long as you hide a straight line in the middle somewhere.

The first pic is of the cooling vents on the aft portion of the cowling with the red box indicating the portion of the model shown in the screen capture(the carb intake). The numbers 1-3 are next to curved lines and honestly it took me a while to see the simple and elegant way to get a fairly accurate and pleasing shape with only one piece of paper with no slits in it. Line #4 is the straight line that breaks the compound curvature. Generally Rhino won't unroll 2 surfaces joined together like I have these but sometimes it will. On those occasions when it won't the edges don't always mate precisely but I figure that a gap half the width of a piece of paper can be fudged and no one will be the wiser.

[Alcides]

Will this thread is a good idea. Thanks for you sharing your tips

As an aspiring designer find very useful this type of post.

[Zathros]

I will probably be copying this thread with word to make an offline archive is that is alright with you. I have been a longtime admirer of your work. This is really great.

[Willja67]

Alcides,

I'm glad you are finding this informative. Feel free to ask questions, or post a tricky shape you are working on. I won't be offended if others share their methods either. I frequently chew on a problem for a while before I come up with a solution I like, and nothing says mine are the best.

Zathros,

Sure you can make an offline version. Hopefully others will contribute as well.

[Zathros]

Thanks. I have been redoing my ESKA 2 Ekranoplan and decided after seeing how you do the wing root to go to that method as the wing drops around 100 degrees from the root to the tip according to the drawings Billiy Lelveld has given me (excellent drawing I must admit). This would allow the builder to get the right angle instead of gluing a completed wing assembly to the side and hoping for the best. The main problem is that the aircraft was made of fiberglass so where do you put lines. I look forward. I still can't afford Rhino 4.0 same am trudging along with Rhino 2.0.

[Willja67]

Contrary to what I posted at the beggining of the this thread I will concentrate on a specific command in this edition. Zathros and his A90 brought to mind this technique.

I must admit I learned the basic idea from Gil and his C5 thread. He designed his 3d Rhino model with no develope-able surfaces. In other words they were all compound curves. His method was to create an accurate 3d model then split it apart as necessary to make it into a paper model. Up till that point (and mostly till now) I have done the opposite. I design the model from the outset to be flattened out and consequently sacrifice some accuracy.

I am going to show how to create a compound curved canopy using 3 curves.

First pic is of the cad model (my own original airplane design that i might one day build 1:1 scale) sans canopy. As you can see I have the bottom and aft edges already defined and the profile drawn in on the model.

Second pic shows the lines copied away from the model. I am only going to model half of it (the whole thing could be done if desired). The curves are numbered in the order that they should be selected for the sweep1 rail command.

In the sweep1 rail command it first asks for the rail curve (you can only select one) then it asks for cross section curves (select as many as you like). I don't understand the math well enough to kow what exactly the program is doing but if you play with it enough you can get an idea of what it will do. Sometimes though the results are somewhat unexpected.

That brings us to the third screencapture. You will note the strange swirl shape at the bottom right of the surface.

The rectangle in the second screencapture shows the area zoomed into on the 4th screencapture. As you can see the ends of the curves are not perfectly joined and Rhino didn't like that at all so it went all loopy when it created the surface.

The last screencapture shows the final result after the 2 curves had been dealt with.

At this point you can use the split command to break up the surface into segments, duplicate the borders and loft the cross sections to get segments for your part, or if the isocurves (the lines that make up the meshes) don't appear in the places you want them you can project lines onto the surface and then loft between them.

The sweep2rails command operates on a similar principle and I reccommend anyone that plays with this technique also try substituting it in place of sweep1rail.

Any questions?

[Zathros]

I have found sweeping 2 Rails as my best friend. I also found that after I get the first surface, if I like it, I use the surface edge as the first first rail, you get a tighter fit.. Also sometimes, try drawing the rail curve from the other end if the first end does not give the wanted results. I also use the "Split Isoparm" command a lot and make a lot of shapes to use as cutting tools. When doing the rear end of the Ekranoplan (fuselage end) I found that if I split the outline shape of the tail end with an extruded line as the "cutting tool", I was able to run the surface edge of the last fuselage former as the rail curve and use the flat curve segments ( that formed the side view of the fuselage end)to "Run 2 Rails", it produced a perfect tail piece. I believe this would work for the bow of a ship also. If anyone has or is using Rhino, I think this thread could turn into an incredible resource. The information in these few posts would allow many models to be created. I work slow as I have physical disabilities but someone who does not have these limitations could indeed produce high quality models and a lot of them. Taking these methods that Willja, Gil and others contribute can save a lot of time and effort.

[Gil]

I kind of like some of the weird shapes that unclosed curve sets produce. Sometimes it takes several minutes to compute a tangle.

+Gil

[Golden Bear]

I apologize for coming in late on this so do not know the starting premise, but...

Why aren't we using "surface from network of curves?"

In the attached image, the back pair shows something like the canopy used earlier in this thread. I made it simply by intersecting three ellipses. For something simple like this, with only the neat edges, I grabbed all three curves and clicked "surface from network of curves" and you can see the result.

The near pair of images shows that I imposed another line, simply a straight one, in order to show that other cross section curves can be used with this command. However, when the shape gets trickier, the command can have problems and thus I only did the half canopy.

I use this command extensively when forming hulls for my ships. I like to get a complete half hull from bow to stern and then break it up on my own later when it comes time to smash the segments flat.


Carl