Interesting feedback from Fillet command - Cannot chamfer between these two entities.

[robert.1.hall72202]

Command: FH
._FILLET
Current settings: Mode = TRIM, Radius = 0.5000
Select first object or [uNdo/Polyline/Radius/Trim/mUltiple]:
Select second object or shift-select to apply corner:
Cannot chamfer between these two entities.


Why does this say chamfer?
This is the fillet command.

[david_peterson]

They may not be co-plainer? I suggest you look at the start and end Z coords for each line. From what I remember the fillet command uses the champer command as a base then applies the radius between the start and end points. Just a thought.

[Steve Johnson]

Why does this say chamfer?
This is the fillet command.
I would say it's a micro-bug, and the two commands use code in common and they forgot that bit when they were setting up the alternative messages. However, as it doesn't happen for me in 2007 when I would expect it to (e.g. fillet a pline and arc), I won't say that. :?

[robert.1.hall72202]

I did check the Z cordinates, all are at 0.

I also notice that the fillet command will return a
"radius is too large" error. Makes a bit more sense than
"cannot chamfer".

Merely an observation.

[sinc]

I've been having a lot of problems with FILLET in 2007. There is definitely something wrong, but it's been pretty hard to pin down. Probably because I'm growing more and more convinced that there are a number of bugs here, all of which are only triggered in special circumstances. Makes it hard to find a solid cause for any particular problem.

One of the issues is that FILLET will often only work to one end of a polyline, but not the other. Reversing the polyline, or exploding it, causes FILLET to work. Another issue is that, when used on lines, FILLET often creates arcs that are not snapped to the endpoints of lines. This means that PEDIT JOIN may not work, or you may need to use the "M" option so you can enter a "fuzz distance" (although this doesn't always work, either).

There is another issue caused by the Dimension Input portion of Dynamic Input. This can cause random problems with lots of commands. I run into it most often with the ARC command. For example, I will start the ARC command, enter the start point, type "E" and enter the endpoint, then type "R" and try to enter the radius. Depending on where I leave the cursor on the screen, the command will either work or return a weird error like "Start and end angles must differ". Turning off Dimension Input makes this problem go away. The problem you are seeing with FILLET may be part of the same mess.

[robert.1.hall72202]

I have noticed arcs that do not snap to the endpoints of lines.
I thought it was related to some other issue.
Now I know that I am not seeing things, no screw lose here, lol.

[david_peterson]

I did check the Z cordinates, all are at 0.

I also notice that the fillet command will return a
"radius is too large" error. Makes a bit more sense than
"cannot chamfer".

Merely an observation.When you checked them, what did you have your units set to? even if they are off by a little bit (1/256") they will not be co-plainer.

[Steve Johnson]

When you checked them, what did you have your units set to? even if they are off by a little bit (1/256") they will not be co-plainer.
If they are not coplanar, AutoCAD reports "Lines are non coplanar." This appears to be a different issue. Robert, can you post a small example drawing to show us what you do to make AutoCAD generate this message?

[sinc]

If they are not coplanar, AutoCAD reports "Lines are non coplanar." This appears to be a different issue. Robert, can you post a small example drawing to show us what you do to make AutoCAD generate this message?

It's tricky to get. I suspect it has something to do with one of two things:

1) FILLET does not always work to the beginning of the polyline, but will work to the end of the polyline. When this happens, exploding or reversing the polyline causes FILLET to work.

2) FILLET seems to have some issue related to floating-point precision. I think this issue was solved a long time ago, but was reintroduced in the 3D revamp. This can cause two entities not to FILLET, no matter what. Moving either of the lines a microscopic distance will place the two entities in relative positions that no longer trigger the floating-point precision error, and Autocad will then be able to accomplish the FILLET.

As far as it saying that CHAMFER cannot work between those two entities, I think that is a seperate error. Autodesk programmers seem to have accidently used the same error message for both the FILLET and CHAMFER commands (which are very similar in functionality, and undoubtedly share code).

[robert.1.hall72202]

If I can remember what file was producing the dialogue, then I will post it.
I have to look back through my design log to see what I was working on last week.

[steve.perkio]

I'm having a similar issue in AutoCAD MEP 2009.

I'm working in basic autoCAD with construction lines and extrusions and have a vertical line situation that refuses to fillet.

I have attached an example to show what is going on.

Why won't the vertical lines fillet to the arc? It works in plan view.
I tried changing my ucs to left and front with the same results.

Perk

[david_peterson]

Sorry, but I don't have 2009 yet. I'm stll back in the stone age.
I'd check to make sure that all your lines are in the same plane and actually intersect. If that still doesn't work, try the trim command instead from a 2d wireframe plan view.
Just a suggestion.

[steve.perkio]

Yes, I thought of that. I was able to fillet one side, and then the other side (I had to extend the horizontal line some first) but they were both planer (one at a time).

Perk



I'd check to make sure that all your lines are in the same plane and actually intersect. If that still doesn't work, try the trim command instead from a 2d wireframe plan view.
Just a suggestion.

[jaberwok]

Odd one.
If you set UCS onto the lines it still won't make the second fillet but you can draw a circle using the TTR option and it will all trim to make what you want. It look like a bug in the Fillet command. :-( Especially as the Chamfer command does work (with a suitably small distance value) even without setting the new UCS.