Hi Jim, I haven't taken a close look at your model yet but one big thing to note is that it looks like there is not much space available in those areas that you are trying to fillet.

A fillet will not be successful if you ask for a radius that is too big to fit within the available space.

Sometimes it can be a bit surprising how much space fillets take up, especially since they can kind of eat away at pieces from both sides.

So because of that one thing to try in general when you run into a fillet problem is to reduce the radius that you're asking for and see what that does. For example try asking for a radius of 1/2 of what you did initially and then maybe even 1/10 as much.

- Michael

Hi Jim, so a fillet of 0.04 appears to work but not any larger than that. But that's not due to running out of space, the filleter seems to be getting confused about something.

One thing that I've noticed before is that the fillet engine that MoI uses does not seem to like it when you have planes fragmented into multiple smaller plane pieces that are joined together instead of larger simple planes. You've got a few areas like that in this model (co-planar fragments adjacent to one another), so I'll see if replacing those with a more simple larger plane helps any.

- Michael

Hi Jim, so here's the kind of coplanar fragments that I'm talking about (the ones selected here):



Those pieces are all on the same plane as each other. If those are combined together to be just one large planar surface like this:



Then the filleting in this case should work a lot better. It just kind of happens to be an area that the fillet engine that MoI uses is kind of sensitive to.

The way those areas can be combined into one large plane is by deleting them to leave a large hole there, then select the object and run the Construct > Planar command which will build a face on any planar holes and join it in to seal up the hole.

I'll see if I can make a tuned up version of this object for you.

- Michael

Hi Jim, so yeah it will work a lot better with those planar pieces combined into just single larger plane pieces.

I've attached an updated model file with those changes in it.

With this version you can just select the whole model now instead of picking edges individually (when a whole object is selected and then Fillet is run it will fillet all sharp edges in the object) and you can go up to a radius of around 0.15 before running out of room:



- Michael

Hi Michael...

Today many user have problem with filleting. Is there any plan to improve the moi engine or maybe replace with another engine for example Acis. ViaCAD is sample of affordable CAD system that work with Acis. I dont know another "affordable" price geometric engine. From Deelip.com I heard Spatial will sell geometric engine called CGM. But maybe the price is higher compare to Acis. Looks like there is no improvement from Solid++. So are you not considering to replace it in the future ?

Your very good skill in programming and designing CAD system must be supported by a good geometric engine :)

Just my opinion,,,

Thanks...

- Anis

seems also that is also often users who have some error in their objects before filletering ;)

Michael,

Thanks so much for looking into this! I'll try this and keep this in mind for the future. By the way, is there a status message window, or log, something you (the user) can look at when operations fail?

Cheers

Jim

>>seems also that is also often users who have some error in their objects before filleting ;)

Oh boy, ain't that the truth! Check the model twice before hitting the "fillet" button.

C.

Hi Anis,

> Is there any plan to improve the moi engine or
> maybe replace with another engine for example Acis.

Currently I have no specific plan for replacing the geometry engine, but it is something that I might look into in the future, probably not as a full replacement but maybe replacing just some particular operations like filleting.

The Solids++ engine does continue to be developed and improved as well though.

- Michael

Hi Jim,

> By the way, is there a status message window, or log,
> something you (the user) can look at when operations fail?

Not really... There is a debug log but it won't really contain the kind of information that you're talking about here.

In the future I would like to have some kind of console area that I could use to spit out various messages.

- Michael

Hi Michael...

>The Solids++ engine does continue to be developed and improved as well though.
Sound good, but looks like here in the MoI forum many user already complaint about this problem ( fillet ).
I am not sure the power of Solid++ in the surface continuity. Hope there will be more improvement in this area.

Keep on good work Michael... ;)

Thanks

Notice though that 98% of the time it is a "poor geometry issue", and not a bug if the filleter.. I have tried a couple other packages that just "succeeed" in all areas, though they completley change the topology in areas, and the output can have open gaps, bad trims and poorley joined geometry.... Stuff that didnt go into my CAM package and work well...

It never hurts to learn how to model better.

Hi Anis,

> I am not sure the power of Solid++ in the surface
> continuity. Hope there will be more improvement in
> this area.

Solids++ does have some low level tools to help out with this area, I do expect to be working on making some new tools for MoI for surface continuity in MoI v3.

- Michael

Hi Burr, yeah definitely having a good clean high quality model will help filleting a lot.

If you have some kinds of geometry anomalies like surfaces that bunch up and overlap back on top of themselves, that will cause problems in pretty much any fillet engine.

One area where the more sophisticated fillet engines can have a big advantage though is in handling more kinds of corner configurations where multiple fillets are converging on a single point. The Solids++ guys are adding more corner cases to be handled by their fillet engine as time goes on as well though.

- Michael