Gap Appears When Filleting

Rhino Rhino for Windows
,
1

testfile.3dm (2.0 MB)

Inside the dashed circle, you’ll see there’s a gap where I tried to fillet. The rest of the fillet around the part worked fine. This was a closed polysurface before the filleting.Why is this happening.

Also, my Rhino has adopted the habit of continually modifying the cplane. This is not only in one window. I’m having to set back to the appropriate world view each time it happens. I’ve been using Rhino for years and this is new and consistent.

thanks

rex

2

Rhino has a new thing called autocplane. I’ve never used it myself, but I presume it’s a configurable option and you have somehow activated it. The help should explain it.

3

If this is happening in Rhino its because you are using FilletEdge to make the fillets.
This is very common occurrence. FilletEdge is based on a defective algorithm that tries to make one fillet for every edge the user selects. No other CAD program tries to do that and yet Rhino has doggedly tried to make that work for more than 25 years.
In your model there are places where the number of edges does not match the number of fillets required and that leads to FilletEdge failing.

You can make the fillets manually using FilletSrf. However, to make clean fillets you should make an effort to make the base surfaces tangent at the joined edges where the fillets are going to cross them. In your file, there is at least one place where the curves used to extrude the main shape have poor continuity.

In this file I fixed the the non-tangent break in the construction curve and then made fillets with FilletSrf

testfilex.3dm (440.5 KB)

2 Likes
4

Jim, thanks so much.
We’re treading in an area I’m unfamiliar with. How did you discover the non-tangent edge (I’m not sure what you’re seeing) and how did you fix it?

thanks again!

rex

5

Use GCon between the ends of two curves. Use MatchCrv to align curves G0, G1 or G2.

1 Like
6

I discovered the non-tangent edges when I tried making the fillets to replace those that were incorrect. My plan to make the 5 missing fillets that were needed and do a little trimming and that would be it, but the fillets did not connect end to end and that indicated the base surfaces were not tangent.

At that point I could have just use MatchSrf (for position) to make the fillets connect, but that just adds to the number of surfaces that are supposed to be tangent but aren’t. For me, its better to make surfaces tangent that are supposed to be tangent . Modifying the one errant curve that was not tangent to begin with meant creating a new surface with a slightly different shape and then also fixing all the adjoining surfaces.

7

Jim, if you could indulge me a little further. I think I understand what you did, but not clear how you did it. A couple questions…

  • Which surface did you recreate and did you have explode the poly surface to do it?

  • How many actions did you perform to fix this? Is it possible to list the steps?

  • I really appreciate you fixing the file for me and I can work with that. At the same time, I’d like to get a complete understanding of the process for future reference.

thanks much,
rex

8

Jim, it occurs to me that the surfaces you recreated are colored cyan. I see what you did. Looks like a long process. I get it; thanks!

rex

9

First of all, I’m assuming you want the arc based cylinder that is the top of the part to be concentric with the 7/16" thru hole. To make that work given the existing geometry the radius of that cylinder has to be a cylinder with a radius of 0.34927". The previous file I posted I used a radius of 3.5.
Actually as it turns out the original model has the correct cylinder, but it is trimmed and joined with its neighbors at point where the surfaces are not tangent. On the side where you want the fillets it misses the tangent point by almost .01"

Anyway, the easiest way to get to the correct geometry is to use the curve fillet tool and make the red curve shown in the pic below. You can then extrude the red curve to make the green surfaces and extract the original badly trimmed surface and replace them.

All the purple surfaces have to be adjusted for length so that they match up to the new correct geometry. The purple fillets are not native Rhino fillets. Extending them is going to be precarious.
It may be easier to just replace them with Rhino fillets that are easier to work with. The light purple surfaces need to be extended because they got trimmed back incorrectly by the failed filletedge.
The green torus shaped fillet can be made as a revolve surface or if you have Rhino wip you can make the torus with FilletSrf

When the base surfaces are all correct then you can make the 3 orange fillets with FilletSrf.
Then use Mirror to complete the other side.

image
image1496×923 276 KB

Here is the file: testfilex2.3dm (456.1 KB)

The cyan surfaces are all the surfaces extracted from the original. Only a few were replaced as described above. Also some of the cyan surfaces are involved in fixing the hole that filletEdge made in the bottom of the part. The file I posted above shows an alternate way of dealing with that.

10

Thanks Jim. The crux of this issue is that this file was provided to me by a client and then I needed to add the filets before machining. This is why I like to do my own CAD design. :blush:

thanks,
rex