Autodesk Inventor

[#1]

Hi, Larry.

For that particular sort of shape, you might try something like the file I put
in CF.

The reason your file was giving you trouble shelling, etc. is the singularity
at the nose. It's causing minute wrinkles and kinks in the surface. Put the
zebra on and zoom in very close. You can see the area and can pick out a 2 mm
edge with the measure tool where the surface is "torn". This is almost always
a problem when bringing all the U / V lines into a single point (despite the
.01 offset), except for a flat triangle surface, and is just about guaranteed
to be a problem when more than two wires converge (the three profiles bring
three wires in). While the .01 offset allows the loft to solve, it probably
contributes to the wrinkles and tear.

Have a good one,
Jeff
=====================

[#2]

Thank you Jeff! That is very cool. I will try one like that. Thank goodness
for intergalactic contacts.

Maybe you know if there is some cosmic limitation to having U & V the same
or opposite or is it something that could be allowed but isn't at present? I
keep running into that limitation and on this adventure it put me in mind of
10,000 seats of Inventor going to the Japanese ship building industry and
wondered if perhaps there may be some effort under way to address this
problem. On the other hand, I guess they could be assuming that you will be
available to consult with the Japanese on how to get Inventor to do what
they may need; ya' think?
~Larry

"Jeff Howard" wrote in message
news:4F6A26B85139BEEE0722F2DB65392DBD@in.WebX.maYIadrTaRb...
> Hi, Larry.
>
> For that particular sort of shape, you might try something like the file I
put
> in CF.
>
> The reason your file was giving you trouble shelling, etc. is the
singularity
> at the nose. It's causing minute wrinkles and kinks in the surface. Put
the
> zebra on and zoom in very close. You can see the area and can pick out a
2 mm
> edge with the measure tool where the surface is "torn". This is almost
always
> a problem when bringing all the U / V lines into a single point (despite
the
> .01 offset), except for a flat triangle surface, and is just about
guaranteed
> to be a problem when more than two wires converge (the three profiles
bring
> three wires in). While the .01 offset allows the loft to solve, it
probably
> contributes to the wrinkles and tear.
>
> Have a good one,
> Jeff
> =====================
>
>

[#3]

After taking a cruise this summer on a ship, and looking at the construction
of the various ships we saw along the way, most of the big plates that make
up the hull don't appear to have much curvature. I wonder if they design the
hull to start with as one solid or shape or if they do it as all the
individual plates of steel, each with slightly different bends, and mate
them all up.

--
Kent Keller
Member of the Autodesk Discussion Forum Moderator Program

http://www.MyMcad.com/KWiK/Mcad.htm

"Larry Caldwell" wrote in message

> keep running into that limitation and on this adventure it put me in mind
of
> 10,000 seats of Inventor going to the Japanese ship building industry and
> wondered if perhaps there may be some effort under way to address this
> problem.

[#4]

Well ... I don't know how they do the models, but having worked at Todd
Shipyards for a short while when I was a kid, I've seen how the ship
fitter's construct 'em. The hulls there, were bent (pad-eyes and jacks)
around the ribs which were welded to the beam to make manageable sections.
Crawling around in the double-bottoms flushing off all those ship fitter's
welds for the pad-eyes revealed quite a few curved surfaces as I remember.
Of course, building them and modeling them are two different things.
~Larry


"Kent Keller" wrote in message
news:1F84304E0F66EE3466CC9420A5EDBC38@in.WebX.maYIadrTaRb...
> After taking a cruise this summer on a ship, and looking at the
construction
> of the various ships we saw along the way, most of the big plates that
make
> up the hull don't appear to have much curvature. I wonder if they design
the
> hull to start with as one solid or shape or if they do it as all the
> individual plates of steel, each with slightly different bends, and mate
> them all up.
>
> --
> Kent Keller
> Member of the Autodesk Discussion Forum Moderator Program
>
> http://www.MyMcad.com/KWiK/Mcad.htm
>
> "Larry Caldwell" wrote in message
>
> > keep running into that limitation and on this adventure it put me in
mind
> of
> > 10,000 seats of Inventor going to the Japanese ship building industry
and
> > wondered if perhaps there may be some effort under way to address this
> > problem.
>
>

[#5]

Hi, Larry.

> U & V the same or opposite .......

I think it means that some of the flow lines (U/V's) are folding or looping
back onto themselves. I notice it most when a profile weight value is too
high. I put an IGES in CF that, I think, illustrates what's happening. It
can be imported into MDT and viewed. The middle surfaces were "blended"
between two planar surfaces with a "maintain" value too large to allow the
transition to solve correctly and causes some of the flow lines to loop back
onto themselves. There are also "good" surfaces and wires shown for
comparison. The different entities are segregated by layers with descriptive
names.

> 10,000 seats of Inventor going to .....

Yea, the CSSC deal. Probably a lot of politics, smoke and mirrors involved
there. "The collaborative capabilities of Autodesk's software will help CSSC
promote digital information exchange among ship design enterprises and ship
owners." Web noodles and balancing international trade deficits... I guess
there's enough of the design that's like any plant installation / mechanical
design (interior structures, plumbing, machinery) that IV would be a valuable
part of the effort and it's possibly an improvement over what they were using
(drafting boards and lofting floor?) if it's the primary design environment.
It would be interesting to get some follow up on stuff like this. Seems like
we always get these news flashes (in the pan?) and never hear any more about
it.

> ...having worked at Todd Shipyards for a short while when I was a kid....

Always thought it would be fascinating to visit a ship yard and watch the
construction.

I don't know the first thing about it, but I believe Kent was correct, in a
sense. I think that large hulls are designed so they can be (primarily?)
constructed from "flat wrap" panels (correct me if you know better) vs. panels
with compound curves. Rhino and, I suppose, (m)any of the specialized hull
design software packages have loft functions that will aid in the creation of
the surfaces where that is the desired result (Rhino calls them "developable"
surfaces and flat patterns can be generated from them). How it would be done
if you are lofting manually... I have no idea, but I guess people have been
doing it for hundreds of years. It's always so humbling to consider things
that have been done for eons that I haven't the slightest comprehension of.
Where's the phrase "I am only an egg." from? 8~)

===============================

[#6]

Jeff:

If you look at a ship in dry-dock, you'll see that they are not
streamlined in the same sense as a racing yacht. Bear in mind that
their primary function is to carry freight. A yacht's shape sacrifices
a lot of stowage volume for the sake of incremental increases in speed.

You can approach the design philosophy from either end. Start with a
yacht shape and trade off the losses in hydrodynamic efficiency (higher
fuel costs) for the additional payload (higher revenue). Conversely,
start with the highest capacity, a rectangular prism, and trade off a
bit capacity as you increase the hydrodynamic efficiency.

As they say, 'Form follows function'.

Richard

[#7]

Well, that is strange. First time I opened the iges in Inventor the two
surfaces were side-by-side and I could clearly see the flaw. I tried to
import it to MDT but guess I don't know how to do that and couldn't find
anything in help except how to create one. I saved it as a sat but that
didn't work; I think maybe surfaces can be brought into MDT as sat files.
Anyway, I extracted the file if CF again and opened it in Inventor again,
and now the two surfaces are stacked on top of one another instead of
side-by-side, but it's probably still in memory and I will have to start
Inventor again. Can you say how to get an iges surface into MDT? Bet it's a
command-line thingy which I try to avoid whenever possible.

"...profile weight value is too high."
I don't think I know exactly what that means. The profile shape could be
adjusted to make it work, maybe?

"...a :maintain value too large"
This either. No idea what a "maintain" value is.

Todd Shipyard was a long time back and I mostly worked with the sections
that had already been put together, (flushing off welds, burning in decks,
burning out holes, untangling hose, etc.) but I think, for the most part,
the skin started out as plane ol' flat plate. The welders would weld it in
place on one end and weld some pad-eyes on beside the location of a rib.
Then the ship fitter would put a jack on the inside of the rib, use the jack
to bend the plate around the rib and the welder would tack it in place. Then
they would repeat that procedure until that piece of skin was flush to the
ribs.
~Larry


"Jeff Howard" wrote in message
news:5AFE6D472B7C7E7DC30F62186F95C686@in.WebX.maYIadrTaRb...
> Hi, Larry.
>
> > U & V the same or opposite .......
>
> I think it means that some of the flow lines (U/V's) are folding or
looping
> back onto themselves. I notice it most when a profile weight value is too
> high. I put an IGES in CF that, I think, illustrates what's happening.
It
> can be imported into MDT and viewed. The middle surfaces were "blended"
> between two planar surfaces with a "maintain" value too large to allow the
> transition to solve correctly and causes some of the flow lines to loop
back
> onto themselves. There are also "good" surfaces and wires shown for
> comparison. The different entities are segregated by layers with
descriptive
> names.
>
> > 10,000 seats of Inventor going to .....
>
> Yea, the CSSC deal. Probably a lot of politics, smoke and mirrors
involved
> there. "The collaborative capabilities of Autodesk's software will help
CSSC
> promote digital information exchange among ship design enterprises and
ship
> owners." Web noodles and balancing international trade deficits... I
guess
> there's enough of the design that's like any plant installation /
mechanical
> design (interior structures, plumbing, machinery) that IV would be a
valuable
> part of the effort and it's possibly an improvement over what they were
using
> (drafting boards and lofting floor?) if it's the primary design
environment.
> It would be interesting to get some follow up on stuff like this. Seems
like
> we always get these news flashes (in the pan?) and never hear any more
about
> it.
>
> > ...having worked at Todd Shipyards for a short while when I was a
kid....
>
> Always thought it would be fascinating to visit a ship yard and watch the
> construction.
>
> I don't know the first thing about it, but I believe Kent was correct, in
a
> sense. I think that large hulls are designed so they can be (primarily?)
> constructed from "flat wrap" panels (correct me if you know better) vs.
panels
> with compound curves. Rhino and, I suppose, (m)any of the specialized
hull
> design software packages have loft functions that will aid in the creation
of
> the surfaces where that is the desired result (Rhino calls them
"developable"
> surfaces and flat patterns can be generated from them). How it would be
done
> if you are lofting manually... I have no idea, but I guess people have
been
> doing it for hundreds of years. It's always so humbling to consider
things
> that have been done for eons that I haven't the slightest comprehension
of.
> Where's the phrase "I am only an egg." from? 8~)
>
> ===============================
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>

[#8]

Relief will be on it's way soon...... I'm presenting a class on translations at
AU this year.... the papers will be available after AU on Autodesk's site, or
may be obtained by emailing me after December 9th....

Dennis

Larry Caldwell wrote:

> Well, that is strange. First time I opened the iges in Inventor the two
> surfaces were side-by-side and I could clearly see the flaw. I tried to
> import it to MDT but guess I don't know how to do that and couldn't find
> anything in help except how to create one. I saved it as a sat but that
> didn't work; I think maybe surfaces can be brought into MDT as sat files.
> Anyway, I extracted the file if CF again and opened it in Inventor again,
> and now the two surfaces are stacked on top of one another instead of
> side-by-side, but it's probably still in memory and I will have to start
> Inventor again. Can you say how to get an iges surface into MDT? Bet it's a
> command-line thingy which I try to avoid whenever possible.

--
Dennis Jeffrey
CAD Associates - Fort Wayne
Autodesk ASC - Autodesk University Instructor
(260-432-9695 x 221

[#9]

Hi, Larry.

Sorry, I left out the "how to get into MDT" part. The command is IGESIN.
There are good and bad surfaces superimposed, so turn off or freeze the layers
you don't want visible. If that doesn't work and you are interested, I can
post or email some jpg's.

Profile weight.... referring to the loft dialog (in IV6) Conditions -
Weight. It's value must not be the reason you see the error. I guess it can
be generated when lofting between mulitple profiles, but an example doesn't
come to mind. I may be misinterpreting what the error message is telling us,
too (assuming it's not picked by a random number generator ). Any how, if
you'll post one when you come across it, I'll see if I can offer an
explanation for it.

Maintain value... same thing as the profile weight.

How thick were the plates?

Jeff
============================

[#10]

I had read this a while back in the Fabricator magazine. It doesn't get
into to much detail, but it is interesting non the less. It sounds to me
they panels are fully formed now days before being welded (if I remember
right....didn't read the whole article)

Anne could probably shed some light on this for us 8^)

http://www.thefabricator.com/xp/Fabricator/Articles/Fabricating/Fab02/02web4
99.xml

--
Kent Keller
Member of the Autodesk Discussion Forum Moderator Program

http://www.MyMcad.com/KWiK/Mcad.htm

"Larry Caldwell" wrote in message


but I think, for the most part,
> the skin started out as plane ol' flat plate. The welders would weld it in
> place on one end and weld some pad-eyes on beside the location of a rib.