Designing a motor (PartDesign/Part Next and creative process)

[DeepSOIC]

Hi!
I am making a small project for fun and learning. At the same time I'm testing, how good is my general-purpose body container that I proposed some time ago.

The project is somewhat similar to my motor demo project I made for showing off Part-o-magic. Yes, it's a motor !

That demo one wasn't intended to be accurate in dimensions, complete, or made for real. And, it didn't involve much of creative process - I had a really good idea of all the features to be modeled. This time, everything is different. I want to build a real motor using 3d printer. That means it has to be precise, all pieces must be 3d-printable, preferably without need for supports. And I only have a general idea of the construction, so I need to figure out lots of details during the modeling time. That is, lots of changing around, not just straightforward modeling to achieve a known model.

Vitamins.
* 2mm diameter shaft
* collector (salvaged from another brushed motor)
* inductor core (toroidal common-mode choke from ATX power supply)
* laminated wire for windings,
* neodymium magnets (small box-shaped magnets; I won't bother ordering custom-curved ones)
* some random metal things for brushes and contacts

More to come...

[DeepSOIC]

So far, the model looks like this:

motor v07.3.png (70.56 KiB) Viewed 10062 times

I'm not attaching the project file, because no one will be able to open it properly (general-purpose body container is not supported in master FreeCAD). Maybe I will convert the bodies into PartDesign ones for showing off the project...

Those three wings on the rotor, that interfere with magnets, are helper structure to make windings. The plan is to wind the coils around them (flush to the core), wick coils in superglue, and break these wings off.

Now I'm busy printing parts...

[DeepSOIC]

Impressions about modeling process.

My old projects, modeled with Part + Sketcher + Lattice + random stuff, were just an absolute mess. As projects got moderately big, I began to have serious problems locating the shapes I needed in model tree, as well as finding all the final shapes to unhide and export.

Now, with Part, Body, and Module containers, that has massively improved. Plus, Part-o-magic automates hiding-unhiding process, so the need to find and unhide the final shapes after completing another modeling step was removed. It feels I can comfortably work on much larger projects now!

My dependency graphs have changed, too. They used to look like this:

iosp v02.1 depgraph.png (496.55 KiB) Viewed 10049 times

Now:

hole png was huge; I had to shrink it

motor v07.3 depgraph_shr.png (267.3 KiB) Viewed 10049 times

The new ones are very polluted with origin features. But different parts tend to be grouped together.

Still long way to go, but I very much like the new container functionality of FC v0.17 (i.e. Parts and Bodies).

--------------------------

Sketcher constraints display is quite significantly broken. When a sketch is in Part or Body, constraint symbols like Point-on-object and Horizontal are rendered right on top of the curves/points they are applied to. It is not a show-stopper, but it is quite annoying

--------------------------

General-purpose bodies (Modules) lack UI to change Tip. I have put in some simple automation of Tip setting into Part-o-magic, but when it makes it wrong (often!), I have to go to Py console to fix the Tip.

[triplus]

I guess back in the old days dependency graph like that would raise an eyebrow or two.

P.S. Looking forward to see printed product.

[DeepSOIC]

Progress (actually, I made all this a few days ago)

WP_20160605_20_36_03_Pro_.jpg (342.34 KiB) Viewed 9874 times

I have not printed magnet holder yet, so didn't test it. But the list of things to fix is quite long already.
* The thickness of windings should be increased. Current is 0.45 mm. This is too low for good reproducibility.
* The gaps between wings should be decreased, so that I can make the windings almost touch each other. Now they are too far apart. Maybe I should wind the coils in completely different manner, with some overlap. Thinking required.
* The gap between stator and rotor is too small.
* Holes in ABS are not good bearings. Friction is significant, and overheating will probably seize the shaft. I plan machining some bearings off PTFE.

[bejant]

I enjoy reading about what you're working on; it's always something unique and interesting!

[yorik]

I enjoy reading about what you're working on; it's always something unique and interesting!

Me too...

Offtopic: We could try, for ex. on twitter/facebook/google+, every now and then, to post a couple of links to the most interesting/active discussions of the forum at that moment... That could motivate people to come discover the forum.

[A_3]

Nice work!

My intention is not to preach, this is a constructive criticism.
First thing I spotted in your project was a bearing problem, PTFE is good solution. But friction is not single problem, there is a clearance between shaft and bearing, rotor imbalance. Idont see any housing and covers (where you can put bearings e.g), the rotor should be kept in the axis by bearings in covers. Еasy maintenance is important too, design of bearings and how to replace them. What keeps the magnets in position, clearance or glue?

Yes, I know, the motor is small and for home use, but...

[DeepSOIC]

Thanks you all for feedback! Makes me happy

What keeps the magnets in position, clearance or glue?

How can clearance keep magnets in position? Glue, of course!
I printed the magnet holder. The clearance between rotor and magnets is... negative! So the rotor simply can't spin at all.

The magnet holder will probably face a major redesign, as the current one turned out to be very poor on precision. It also doesn't allow to wrap a ring of metal around the magnets, needed to complete the magnetic circuit and boost magnetic field.

And I think I found the culprit of all my clearance fails. Filament I use is 1.81 in diameter, while I thought it is 1.75 (and fed this to the slicer). So I had slight over-extrusion everywhere.

[A_3]

Мea culpa, the hurry = the error. Оverlapping, (as you say "negative clearance" ) with a little help of a force for mounting and glue to be secure.Maybe the grooves in magnet holder are too shallow? It could make a problem, loose magnets... Grooves go to the end of a holder or not? You can make a hole at the end of a groove to stuck a magnet in place. Sorry, I was working on electric motors, balancing rotors and mechanical repair only, and your project reminds me of good old times.