flatten quilt

[nrollins]

Hello, Happy Friday...

I am trying to flatten a quilt. I have used the "Flatten Quilt" tool
before. It sure is a great tool when it works in "auto" mode.

I think my problem today comes from the fact that the quilt is an extrude
with a variable draft:



Any hints on how to get this to unwrap?

[dgschaefer]

Are you sure the geometry will flatten? Thinking about this and looking
at your picture, I don't think so. The variable draft angle makes it
'twist' in a way that's no longer a single ruled surface, and therefore
not flatenable (is that a word?).

Doug Schaefer

[nkaminakis]

why dont you try to make a sheetmetal part kai try to unbend it???
i dont't know if it works .. but you can try...

-nk/.

On 9/21/07, Doug Schaefer <> wrote:
>
>  Are you sure the geometry will flatten?  Thinking about this and looking
> at your picture, I don't think so.  The variable draft angle makes it
> 'twist' in a way that's no longer a single ruled surface, and therefore not
> flatenable (is that a word?).
>
> *Doug Schaefer*
>
>
>

[nrollins]

Well,

That was suggested. I did convert it to a SM part, but unbend is just as
un-cooperative. I appreciate Doug's comment, and had a suspicion that it
would be an issue before I started. BTW, Doug, thanks for the terminology
"single ruled surface".

Now the question is, How does one get a flat pattern of a "non-single ruled
surface"?

I know that this is geometrically impossible - like trying to flatten an
orange peel without tearing the edges. But How about an approximation?
This thing deviates from a straight lined edge at the draft transition by
only a couple of thousandths at most (I did not measure - just a guess...)

I am thinking I could maybe create a VSS with a transition from 5° to 3.5°
and keep a single ruled surf...? Is that impossible? Or force some sort of
approximate flattened pattern.

This is for a label die line. How do those guys do it?

Thanks again...

_____

[ptc-3962496]

Personally, I'm not convinced that this cannot exist as flat (maybe I'm dense, it IS Friday after all). If all walls are straight and it's only a draft angle change, I'd think it would flatten. We're not talking sphere to like the orange. Think about it this way ...
If all walls of this part are linear top to bottom, set this part on the floor and roll it. The path it rolls over will represent the flat pattern. If there is always a linear contact point touching the floor an nothing is slipping, it should flatten just fine.

[brianlynn]

I would guess that it is because a label starts flat and then is wrapped
around the object, instead of starting off curved and then flattened.
Probably the best way to handle this is a four-sided datum curve that you
can wrap around the part. I think you could use a blended surface to
represent the label then. I have never tried any of this, it is all
conjecture and speculation so YMMV especially with how skewed my mind is.

Brian S. Lynn
Technical Coordinator, Product Engineering

[dgschaefer]

The more I think about it, I'm not 100% convinced it won't flatten either. Just not convinced it will.

Try a 2 trajectory VSS using the upper and lower curves as trajectories. You'll need to create your curves somehow, of course. That should result in a straight section at any given point around the part. Of course, a draft feature should as well ...

Doug Schaefer

[ptc-680841]

As an aside .



I recently developed an approximation of the flat pattern for a non-ruled
surface using ProE Analysis features and BMX to "map" the surface and then a
variable section sweep driven by datum graphs to model the flat. Graphs
used splines with the X, Y points read in from the Anaysis feature's point
mapping.




article somewhere to describe the process.



Regards,



Andy "Flat Guy" Burke







_____

[nrollins]

Well, At least I can be assured that I have the world's best minds on it.

Doug, I actually have your very nice explanation on VSS terminology from
just yesterday in my box to review when I get a chance. Today might be the
day. Thanks a lot for that, too!

I really like Karl's vision of rolling the thing on the floor, but I fear
that it may rock slightly along the extrude's axis, as well as roll along
over the drafts. As is my custom when trying to trouble shoot things, I
like to go to the extreme. This is what happens when you change the
transition to 10 x 1 degree:


I think it is NOT a single ruled surface. Pro/e wouldn't lie...
This even shows that the bottom surface becomes nontangent.
This sucks more than I thought it did.

Thanks again for all your help.

-Nate

_____

[Dmi3U]

I remember doing something similar. Here is a cheat: I needed to extend
one of the edges to have flat tangent surface. If you have
a sketch for your original extrude, just add a little tangent line there
to extend beyond the "rip".

[nrollins]

Hello,

The solution is in fact possible, but not using draft. I had a couple of
people try to flatten my variable drafted version without success. If the
sides need to be "single ruled surfaces" then they need to be constructed as
such using a VSS. Variable draft is a great tool, but I guess the
mathematics it uses do not consider such a constraint. A simple sweep of a
straight line using the top rim as the origin and the bottom foot perim. as
the x-direction gave me exactly what I needed.

Thanks to all who tried and replied, but most of all to Mr. Bertram who dug
me out of this hole for real.

-Nathan


_____

[bbrejcha]

To flatten the quilt ... the most common mistake is you did not supply the
surface a datum point. The PNT needs to be on one corner of the surface.
call me if you have troubles and I'll walk you thru it over the phone.
Monday of course.

Bart Brejcha
*DESIGN-ENGINE|**EDUCATION*

[jon.hadley1]

Depending on the the surface and its edges... the point may or may not
need to be on the corner, it may just be along an edge somewhere. -Jon