Is it me or there’s still a problem with draft features that can’t be arrayed or mirrored?
How come such a basic capability is still flawed in such a way?
Maybe there’s something I don’t know but I’d like to! Almost every day I have to re-organize the feature three and apply the drafting after the arraying/mirroring.
I’m not sure I understand what you are refering to as ‘draft features’ Mr Gibeault. Care to clarify..?
If it’s an applied feature (fillet, draft, chamfer, etc.) one isn’t able to mirror/array it.
1 Like
When the three looks like that:
The Draft will only appear on the original geometry, not on the arrayed one (even though the preview shows it with draft)
When the three is like that everything is working:
But I have to apply the Draft features to all the arrayed/mirrored items individually. Very time-consuming and prone to errors.
I know that there’s the “Geometry pattern” option but that works 1 time out of 20.
Block with drafted cut:
Patterned:
Mirrored:
I rarely use draft, so this may be an oversimplified example, but it does seem to work for cuts and extrudes. Just be sure to select the original feature and the draft feature for the pattern/mirror.
The logic behind the draft feature really doesn’t allow for patterning unless you have things really buttoned down. You’re much better off patterning bodies or even faces. You really have to look at the draft feature and CAD geometry in general differently to get patterns of draft to work.
Geometry Pattern is another one that you can’t get to work if you think about it incorrectly.
The only real way to get real help is to post your part.
4 Likes
I understand the limitation, etc.
The reference plane for example may not be relevant for drafting to work for all instances.
But most of the times I want to use this, it fails for no good reason.
Here, the preview shows what I want:
But here’s the result:
In Jim Sculley’s example patterning or mirroring the feature works because the situation is identical for each one.
My lay understanding of mirroring or patterning feature is that it tries to apply the exact same inputs to other locations, unless it’s a simple block as in the example it is likely that the exact same feature inputs may cause instances of the features to fail. It may be over simplified, but I’ve considered a pattern to be like a macro that runs the patterned feature(s) over and over in new locations.
In general drafts and fillets should be at the end of the tree for a given body. In my experience, mirroring and patterning bodies tend to be much more robust method.
Maybe a work-around.
Make the ‘pin’ an individual body, draft it, pattern the resulting body and then merge
3 Likes
In your example, the pin appears to be moving away from the far face of the base extrusion. So much that the vertical face on the seed pin would shrink and then not exist on the last few. My guess this has something to do with it. SW is saying, “If I blindly apply the exact same steps to all of these some of them don’t work.”
1 Like
Please upload part file as Matt suggested or add screen shot of the draft feature property manager, with enough of the model view to see the selections.
Another fun surprise:
I mirrored some features and amongst them, a 60deg chamfer.
On the mirrored side the angle was the opposite way!
That’s the kind of fun to catch after the file is sent to tooling…
1 Like
@mgibeault What version of SolidWorks on you on?
I extruded it both ways, added some fillets, and used instant 3d to move it around, and half off the base, it still didn’t fail.
Interesting. I don’t quite see, the array is the last feature?
I am on 2024 SP4
It works for me on 2024 SP5.
A while back I attended some “Power Hour” webinars by our VAR. I want to say John Setzer put it together. His presentations are usually good, he seems more interested in how Solidworks actually functions than marketing bells and whistles. Anyway, I think the one was called “Using Surfaces to Fix Your Solids”. It was not about surfacing as much as how the kernel is solving the features and really shed light and made sense this type of behavior that looks strange, but really it’s expected if we have a bit of insight of what’s going on internally.
For me at least this type of training helps make sense of what users feel are unexpected behaviors, so much so that they become clues that the software is giving back that the user needs to tweak feature parameters or a different method to try.
If I recall correctly, Matt touches on these things a bit in one of his books. It may have been the Solid Edge Bible, but the concept of thinking like the Parasolid kernel can mostly apply to both SW and SE.
4 Likes