Shell problem
All  1  2-13

Previous
Next
 From:  bemfarmer
4824.2 In reply to 4824.1 
Does this area matter?


Attachments:

  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged

Previous
Next
 From:  SteveMacc (STEVEH)
4824.3 In reply to 4824.2 
Probably but I'm not sure how that happened. I've got past this bit by using Rhino. If I just take the surfaces that need thickness, Rhino's OffsetSrf as a Solid can create the shell. I'm just wondering why MoI can't.
  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged

Previous
Next
 From:  bemfarmer
4824.4 In reply to 4824.3 
Erased the bottom surface, and offset the 2 sides 0.02m (20millimeters). Did not trim transom.
??


Attachments:

  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged

Previous
Next
 From:  SteveMacc (STEVEH)
4824.5 
There is something funny at the back. The original edge seems to have a piece missing. How does that happen and how do you repair it?
  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged

Previous
Next
 From:  Michael Gibson
4824.6 In reply to 4824.5 
Hi Steve,

> How does that happen and how do you repair it?

It looks like that happened because your 2 surface halves don't meet up exactly - there was a small gap between them but the gap was less than the join tolerance so it was still possible to join them together anyway.

It is possible to have gaps between the ends of edges as long as they are within a tolerance - in this case if you draw in a line you can see that the gap is something like 0.0003 units in size, so it's small enough to be in tolerance.

If you turn on surface control points you can see that the 2 surfaces have a bit of a gap between their control points.

So possibly the surfaces were not totally flush to the mirror line that was used to mirror things over - either the surface was not quite on the same line or it was angled slightly or the mirror line was angled slightly or something along those lines.

Anyway, that's what has caused that little gap - if it's an even gap and not because it's angled it can be fixed by using Edit > Separate on the pieces and then select all the trim edges on each piece and hit delete to untrim them and go back to the original full base surface, then position those so that they are totally flush with one another instead of slightly apart, and then rejoin.

Let me know if you still need help or more information.

- Michael
  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged

Previous
Next
 From:  Michael Gibson
4824.7 In reply to 4824.1 
Here I have separated and untrimmed both pieces so you can see the gap between them:





The other side seems to be touching though, so maybe something was at a slight angle when it was constructed.

One way to fix that gap would be to turn on surface control points and select the line of control points down the middle and collapse them together with flat snap using the edit frame.

But even after that it will be a difficult piece for offset or shell in MoI to handle, because of the sudden change in shape and bunched up curvature in the surface right near this tip area:



Stuff that changes shape and curvature a lot in such a small area tends to be difficult for surface offsetting to handle, it tries to make a result that tracks along the surface normal and the surface normal is changing rapidly in that small area.

To have the best chance of getting a good offset you usually don't want to have a collapsed down and pinched kind of shape like that - it's better to have a more evenly shaped surface that's been trimmed to make it have a 3-sided outline - directly surfacing a 3 sided shape that has different lengths along each side tends to produce that kind of pinching near the tip.

- Michael

  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged

Previous
Next
 From:  SteveMacc (STEVEH)
4824.8 
Thanks, Michael. I did notice the gap, so I straightened the surface points along the join. Even without the gap it would not shell in Moi. It did in Rhino though, so maybe the tolerances in Rhino are looser for this option, although the general tolerance I had set to .001, the same as MoI.

I guess trimming larger surfaces is the way to go. I still think in a Solidworks way. In SW to create this surface I would have created helper planes along the edges then used a patch with surface continuity to the helper planes. Is there a similar method I can use in MoI as I am still struggling a bit with MoI and complex surfaces? I created this surface with a sweep, with maintain tangent set, which probably accounts for severely curved edge, which is causing the problem.

This canopy is particularly difficult as the aircraft windshield and side widows are flat. The frame where the sliding canopy joins it is basically flat sided. On this model of the Tempest where they introduced the bubble canopy, the bubble part has to transition to basically straight edges at the front, so it closes with the windscreen frame. I guess that they made the mold with wood and just sandpapered it until it fitted!

EDITED: 28 Dec 2011 by STEVEH

  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged

Previous
Next
 From:  Michael Gibson
4824.9 In reply to 4824.8 
Hi Steve,

> In SW to create this surface I would have created helper
> planes along the edges then used a patch with surface continuity
> to the helper planes. Is there a similar method I can use in
> MoI as I am still struggling a bit with MoI and complex surfaces?

Since MoI does not currently have a patch with continuity or match continuity command, that particular approach doesn't really work very well in MoI - currently in MoI it tends to be better to build one big full surface instead of a half surface that's then mirrored.

It's a lot easier to get a smooth piece if you build the full thing all at once, see here for an example:
http://moi3d.com/forum/index.php?webtag=MOI&msg=1398.18
http://moi3d.com/forum/index.php?webtag=MOI&msg=1398.19

- Michael
  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged

Previous
Next
 From:  Michael Gibson
4824.10 In reply to 4824.8 
So for a case like this a shell/offset friendly version would be something like some sections like this:



Then build a loft or whatever through those to make a larger and more evenly shaped surface:



Then that can be trimmed by a side profile or by the fuselage body to cut away the excess:





Building a full surface instead of only a half piece will make it easier to get a fully smooth end result, and by building it as an extended and more evenly formed surface it will be a lot better quality for doing offsets since it won't have any tightly curved spots or areas that change curvature very rapidly in a small area like the "pole" zone of the previous 3-sided sweep will generate - trying to directly surface a 3 sided thing like that where one side is a lot longer than the others and swoops inwards in a 90 degree bend tends to make bad shaping in the tip area.

3DM model also attached.

- Michael

  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged

Previous
Next
 From:  SteveMacc (STEVEH)
4824.11 
Thats great. Thanks, Michael.
  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged

Previous
Next
 From:  SteveMacc (STEVEH)
4824.12 
If you are interested, this is how far I've got:




Attachments:

  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged

Previous
 From:  Michael Gibson
4824.13 In reply to 4824.12 
It's looking good Steve!

- Michael
  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged
 

Reply to All Reply to All

 

 
 
Show messages: All  1  2-13