Unanticipated Nightmares

Have you ever tried to model something you thought would be relatively simple and straightforward, only for it to turn out to be the ultimate brainbending shitfest from hell?

For me it’s a steel cable, or wire rope, or whatever we’re calling it these days. How about you?

I enjoy working out the details so never think of it in a bad way. I’ve modeled rope and cable before but this is the first time I’m modeled a thimble.

Think of it as a learning experience.


It’s only a learning experience after you’ve learned how to do it. :grin:

1 Like

So, has nobody had any nightmare models, or is it just that no one wants to admit it? :grin:

Here’s an early attempt at steel cable.

It is six strands of 0.5 inch diameter wire (“6x19”) approximately 90 inches long, the max length I could Follow Me it before SU choked and gave up. These six strands have to be further reduced to approx. 20 smaller individual strands, but the file is already more than 5MB, and my laptop has become a slide projector. (I deliberately giganticized the circles – 60 segments – because earlier attempts with only 24 looked like total garbage.) Also, just to keep things interesting, the cable has to follow a changing curve. Seems as if I’m gonna have to explore those component thingees @DaveR and others are always harping on about!

yup. a “nightmare model” is a puzzle. I like puzzles.

I assume you’re using a tool that rotates in addition to the follow-me (there is one, can’t think of its name right now)
it looks like you modelled each strand individually. don’t.
model the whole cable as one outside skin. you’ll loose the individuality of each strand but gain a lot in geometry.

On the left, it’s 674 entities. on the right it’s 386. from the outside, they look the same

zoom. out. if you keep working on tiny details with your eye stuck on them, you’ll always model too many things.

this spring on the side of the lamp is a 4-sided spring.

looking at it, it’s not pretty.
but it’s a 12 cm spring, cable less than a mm thick, on the side of a lamp, on the side of the desk.


Yes, please tell us some! :grinning:

That cable was modeled by manually creating a spiral coil of the appropriate shape and length (although SU truncated it) and then a circular face of the correct diameter was FMed. Then that strand was rotated 60 degrees five times until the cable was complete.

The problem with low res models and components is they render like ■■■■, and I have to generate very high quality renders of everything (not just the overall model) due to the specific requirements of this project. Clearly this 60 segment strand is not the right way to do it, though. :laughing:

Repeat after me “Sketchup is fun, Sketchup is fun” :wink: posting a model from my early days using SU. Keep in mind what Dave said it really is just a learning opportunity. :sunglasses:


Regarding the general topic of modeling objects that turned out to be more complicated than expected, noting comes to mind really. The closest situation for me perhaps is occasionally when forming fillets, if I decide to make extra work for myself. I use a mix of @Fredo6 's extensions and native tools, depending on situation (and if the extension produces an undesirable result). I have wasted/spent a fair amount of time on some fillets when I wanted to achieve perfect symmetry (e.g., in a 90 degree corner to have the start and end fillet curves be true 90 degree circular arcs that are perpendicular to the face being filleted, and to have all the intermediate fillet curves that sweep through the corner also be true 90 degree circular arcs). Here is an example; all these fillets were done with native tools, and if you select any of the arcs SketchUp will show them as arc entities:

I have modeled a few simple small-diameter stranded wire cables, here are examples:

Here is a cable by itself with the softened edges made visible. The cross-section is a six-lobed shape similar to what @ateliernab suggested. I used an extension named something like FollowRotate (I forget exactly) to sweep the cross-section along the path. Softening most but not all edges was done quite tediously manually (so that the edges where two “strands” meet are not softened). There are about 4400 edges and faces in this little component. The goal was representational rather than high-fidelity.


What was the difficult part?

Have the folk at JPL seen this? Because they ought to, and I bet they’d love it.

JPL had very little to do specifically with the Viking lander, for what that’s worth. Martin Marietta (now part of Lockheed Martin) designed and built the landers. JPL’s major contributions to the Viking project were to design and build the orbiter (based on their earlier Mariner spacecraft) and Flight operations (both major contributions for sure!). The Viking Project Office (where the overall project was managed) was located at the NASA Langley Research Center in Virginia for most of the project’s nearly 15-year lifespan, except for the final couple of years when funding was minimal at the Project Office was transferred to JPL.

I have shown the 3D Viking lander model to some Martin Marietta folks (in person even), and they have expressed pleasure at seeing it. That was gratifying. :slight_smile:


Nightmares? Absolutely, but when I descend to Hell, I call it Art and Move on. I’ve got some really crazy beautiful unbuildable forms. :wink: After that I check in with my SU Wizard and he solves it. You’re already in touch with him. Cheers