Hi
V5
This shape is uber simple.
DO NOT EXPECT MASSIVE FAILURE. but its Rhino,
expected a gully and got this.
initially I unioned the cylinder with the horseshoe, to get a D shape. and again , so two D;'s
and the cutters became the cut object.
so try just the C cutter, still it gets amalgamated in .
why ?
dir seems ok.
Steve
Hi,
Dear McNeel
please add this bug to the fix it list, this to me seems primary school failure.
and shoudnt exist in V8 let alone V9, should have been dealt with decades ago.
@bobmcneel
@Japhy
@ anyone that cares.!
I could say the cause is to me irrational !
Steve
Hi,
that such a simple shape has taken 4 days of my life, is not good,that Rhino cannot even boolean such is beyond belief.
That V9 still cannot is UNFORGIVEABLE.
I run RebuildUV, select U and 16 point (wouldnt have known that in a million years)
and it loses the front end. no longer is it starting from tangent.
so now what to do ?
buy Catia !
I would love to see a table of how things are in leading CAD progs, what fails in what etc.
Boolean difference sees cutter become the cut object.3dm (2.3 MB)
Steve
check the normals of the “egg”
_dir
flip them so they point outwards.
NO
avoid self-intersecting surfaces
edit - my workflow:
boolean union the cutters first
fix direction _dir of egg
boolean difference
(rhino 8 on mac intel)
keep stuff linked - see the other post for the same object:
happy modelling - kind regards - tom
and now my sweep1 method,
purple magenta purple with rail green
repeat on 4 areas and join.
and boolean difference works.
sweep1 method sees Boolean Difference work.3dm (3.7 MB)
I wonder best way to get the gulley swinging down into the hole, try for filletEdge might win, else redo the shape .but with the arc ended rectangles that could be fun.
ah and now Tom .
The actual item has a seam at where the intersecting shapes were, I saw the shape and thought ooh thats as per the real thing.
I couldnt get the one to union, dont know why.
But I need to get it to swing inwards, a radius of 4.5mm swing in.
Probably will have to craft it to do so.
tried for a filletEdge 4.5mm but got a V channel.
sweep1 method sees Boolean Difference work filletEdge 4.5mm not good.3dm (3.9 MB)
Steve
Hi Steve, Reading through this i can’t tell if you are having drawing errors, workflow issues or actual bugs. If you need me to file a bug or a workflow enhancement i need steps to repeat the issue. Thanks
He provided a file, an explanation of what he did and a picture.
If you cannot repeat what he did that’s very sad.
Once again you have demonstrated that your addiction to booleans is a huge waste of time.
Trimming and joining will get you the result you expect with just a few clicks of the mouse.
I’m looking at the thread as a whole and not seeing steps to repeat a particular issue.
When reporting a bug please provide clear steps to repeat and a minimal file. All the rest is just noise and makes getting into the issue more difficult.
Hi,
When I posted this problem someone replied saying it was a bug and still exists in V9
I think the reason given was it was a Rational ..something. cant remember the second word.
I am sure it was Rational something. prompting my play on words reply irrational !
Such prompted my ask to remove this bug and my disbelief of such a simple shape having a bug.
Now its as if I dreamt that, as the post has vanished.
I see David Cockey has removed a post..was it Davids ?
@davidcockey
So is it not a bug, is it due to as Tom has spotted,..normals internal, or overlapping cutter ?
Flipping Normals solves the issue.
All I know is I didnt dream it. mention was made of there being a bug with Boolean Difference and Rational something.
is that so, is there such a bug still existing as far as V9 as was said.?
My Sweep1 shape doesnt see an issue.
But can you not replicate this failure by taking my file (uses the RailRevolve method) and my cutters and try for a BooleanDifference. it fails, as normals are flipped.
Steve
I had incorrectly identified the cause as the surface being rational. When the correct cause, inverted normals, was identified I removed the post with the incorrect cause.
i updated the other topic with a subD-Modelling approach…
Hi,
Tom.
I have trimmed the overshoot (see separate post on why it happens) then mirror copy and join.
With that I also hand crafted the 4.5mm rad sweep inwards, moved profile to where that starts, then where it ends, essential to move profile, in fact my photo was showing where it starts anyway. gulleys 11.5mm apart.
trimmed cylinder to fit with it. feared that would be a struggle yet was easy.
now its as it should be, and no oddity at tip.
Now for the second one.
Then hopefully FilletEdge and give the rounded off edges.
Steve
and now both sides success,
Porcelain insulator 4.5mm arc inwards success both sides.3dm (3.2 MB)
So pleasing when things actually go right !
Now for filletEdge, or have I spoke too soon.
Steve
Manual work
Hi,
FilletEdge on the one gulley produces a fillet but doesnt cut away the material to reveal it.
What is wrong ?
Steve
This is wrong:
Forgot about this.
I opted for a flat area in the cutout, and minimise with a fillet at the front. I found it surprisingly awkward, but I think that for me, the key was maintaining degree 3 on the initial egg shape, and maintaining tangency.
One of the things is certainly the choice between design intent and result. The hole does not have the slight elongation in the fillet that the evidence shows.
Though, I must say, I’d think if it needs to be a near-perfect replication, I’m not sure why you wouldn’t just 3D scan it, then remesh if needed.
Side view I started from.
Using Bobi’s Alias-like view
Render, with minor displacement.
By the time rendering arrives, it seems the material and displacement override the need for perfect surface accuracy anyway.
OMG, it looks real.
Its this rendering that I havent a clue on and wish to get into without having to learn on a load of subjects that dont relate to my interests or subject matter.
Scanning would cost me money, and as its shiny it might be tricky, and as its valuable the insurance of postage would also add to cost.
It should be easy enough to model in Rhino I thought. I am almost there.
If I knew how to make porcelain like that, no doubt there is a porcelain material, whilst there might not be a blued aluminium material, or a slightly rusting steel sheet material.
If you could do a step by step of how you did that rendering, that would sure be of interest.
Also you got the radiused edges, a Bobi method, ?
I ran filletEdge and it failed to trim.
This I presume is due to the squiffy end ?
Tom, why if I trim on datum the ‘overshoot’ then mirror copy and join, would I still get a squiffy V gulley ?
I see by the way that I will have to extend the shape, as after gulley making it falls short of the actual shape. So that means redoing the cutters.
Steve
Even with 3d Scan spray?
