Small fillets

When it comes to small fillets (<4mm radius) for a handheld product, I sometimes have a hard time figuring out what’s appropriate for the design. It might look right in CAD, but I’m never sure if it’s going to look or feel right in real life. What’s your approach for this? Is there a rule of thumb I’m not aware of? I want to CNC an aluminum “fillet reference” to keep at my desk so I can make these decisions on the fly… does this exist already?

In reality fillets always come out optically a bit larger.
If it is on a negative mold cavity, I usually do not go under .35mm. Anything over .5 is already a significant fillet making an object appear rounded rather than crisp with just the sleek light lines on the edges.
A fillet at 3mm is already substantially large that we have to apply a continuous curvature or more complex blend.

If you find the guide you describe, let us know!

Thanks Ralph. At what size do you start applying a curvature continuous fillet instead of a circular one? Since it’s so easy to do CC fillets in Solidworks now, I guess there’s not much reason to use circular unless you’re at really tiny sizes where the difference truly won’t be visible. Maybe I should make a guide with CC fillets, too.

Hi Bryan,
The standard fillets are computationally most successful so I usually only apply CC ones above 3mm, that is when you start seeing the difference.
For one-offs or 3D printed items I go for a CC only above 5-6mm. Your guide is especially useful for the tiny fillets since we have less of an idea of how those come out while they do make a large difference even at 0.10 vs 0.20.

Hi Brian,

I’ve never come across a tool like this in my work. BUT what I do to emulate the same thing is to use a circle guide (image attached) to estimate the size of the radius that I’m dealing with.

You’re right that it’s hard to pick fillet sizes, but I generally try and keep it to either very small (up to 0.5) and then larger for styling features (5-10 or more) to make the filleting more purposeful.


Ok, that’s what I figured. Yeah, tiny fillets are really what I’m most concerned with here, for exactly that reason.

Welcome to the forum, Nic.
A circle guide is helpful up to a certain point, though. That one you posted (which is very similar to what I own) only goes down to 2mm. And you also can’t use it to judge the feel of a radiused edge.

I used to use metal radius guides when I was starting out, they can be handy. There are a wide array of options and it can be nice to have something to touch.

Nice! I’ll probably grab one of those for now. Thanks for posting.

Bringing this back from the grave.
Michael, I purchased one of those not long after you posted, but it didn’t quite serve my needs. I want a full edge that I can see and feel, and the ~1mm stamped piece of metal doesn’t cut it. So I’m making something better! Here’s a sketch page showing some of the concepts I’m considering.

fun project. That would be a cool tool to have.

ditto - I’d buy a set of gauges like that. KS that sucker!

Reminds me in concept of the Montessori Trinomial Cube. These things sell for $40-50 and have wonderfully high tolerances (by kids standards).
Maybe your box can contain them in some neat way.

What I think works about the tapering tower concept is it could be a cool desktop sculpture on its own. I like the idea of getting it out of the drawer, which could be the object itself, or the packaging like Slippy is showing.

My answer was always to 3D print a variant and then decide if you wanted to go smaller or larger. Generally anything bigger than 2mm for me became a real blended surface with G2.

Before 3D Printing was easy, the old approach was a series of a dozen CNC’ed rennshape/foam tubes with various radii (D1 x D2) - those were used as HF mockups to then fine tune the approach. But these days you should be able to make a dozen of those for under $200 off shapeways.

With any handheld product, feeling it and passing it around is so critical it’s hard for me to imagine a world where 3D printing wasn’t part of the regular process. We’d always have arguments about some of the crazy shapes we made which required tons of prototyping to figure out what might feel good to me would feel awful to someone with bigger or smaller hands.

I probably printed a hundred of these. I feel especially bad to the poor bastard in Asia who had to rebuild all of these original Alias surfaces in Pro E after I left.

Thanks for the feedback everyone! Mike and Slippy, you aren’t the first to express interest, so there may be a Kickstarter in my future. I’ve wanted to do one for a while. I like the idea of something you keep out on your desk, too. I’d have it finished really nice, like a consumer electronics product.

Question–originally I was planning on 0.1mm increments, going up to 4mm. Any thoughts on that? Sounds like 4mm might be overkill.

3d printing is a big part of my process for sure, especially with ergonomics. Hopefully this tool makes it even easier to fine tune certain details. Well, one specific detail I guess.