Modelling help req'd - windshield
 1-11  12-28

Previous
Next
 From:  Frenchy Pilou (PILOU)
2020.12 In reply to 2020.11 
Hi Michael
Why in your model there are crips?
Could you have shell now?

Curious in my model I have a more thickness of lines?? I don't know why?
Ps Of course I have not the Plan Surface of the little part inside my previous model ;)


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:  Michael Gibson
2020.13 In reply to 2020.12 
Hi Pilou,

> Why in your model there are crips?

That's a display artifact that can happen if you have the mesh density set to lower. With a smaller number of polygons the triangle structure can be revealed.

If you want to eliminate those, go to Options / View / Meshing parameters, and check the "Add detail to inflections" option, which will create a denser mesh and should eliminate the visible triangles there - does that solve the problem for you?

Basically though that is not really any defect in the surface at all, it is just a display issue only.


> Could you have shell now?

I don't think it will shell properly - there tends to be a problem with matching up corners when attempting to shell an open surface that is made up of multiple pieces that have sharp corners in them.

Shell will tend to work best when used on single surfaces that are all smooth, or it can handle corners better when you shell a closed solid. A non-solid that has corners can be problematic, there was some discussion on this previously here:
http://moi3d.com/forum/index.php?webtag=MOI&msg=133.11

I believe this case will run into the same issue as in that link, with the surface extensions not being done properly.

For cases like this you may need to use the Offset command to construct some individual offset surfaces to build the thickened pieces manually. Or possibly a good approach would be to use Edit/Separate to break the model into individual surfaces, and then use Shell on the separate surfaces to get some pieces to work with.

Or actually now that I think about it some more probably the best option would be to shell the original sweep surface before cutting it.


> Curious in my model I have a more thickness of lines?? I don't know why?

Probably you've got some duplicated pieces there, or duplicated some of the edge curves as new separate curve objects.

When you have several curve objects stacked on top of each other they will get a bit darker like that.

- 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:  kevjon
2020.14 
Thanks Michael,

That is a really good solution to the problem. Just trying to figure out what shape the green line needs to be so my windscreen shape matches the real aircraft when the line on the side is cut is the only problem. I guess it is just a trial and error process.
~Kevin~
Image Attachments:
Size: 35.9 KB, Downloaded: 14 times, Dimensions: 671x421px
  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
2020.15 In reply to 2020.14 
Hi Kevin, yeah you may need to edit that curve to tweak the result to what you want.

You can use history to help with that - currently history will be broken after you do the trim or boolean operation but you can use it immediately after you do the sweep to help fine tune the results.

The way that it works is just after you do the sweep, select that center curve and turn on its control points by Edit/Show pts. When you move the curve's points around the sweep will update so you can adjust the shape of the curve and see right away what the effect is on the sweep. You have to just visualize what the result of the cut is though, that part does not work with history yet but I do want to improve this in the future.

I mean you still need to edit points for the NURBS construction method, it is just that the focus is a lot more on editing points of 2D curves rather than editing points of a 3D cage mesh.

- 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
2020.16 In reply to 2020.14 
Then another piece is if you want to blend off the sharp edge you can put a fillet on this edge:



I'm not quite sure if you actually wanted a totally flat plane in there or not though.

- Michael
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:  Michael Gibson
2020.17 In reply to 2020.16 
Also Pilou, if you wish to shell this it will shell more easily if you fillet that edge. Shelling of smooth pieces tends to reduce a lot of problems.

- 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:  Frenchy Pilou (PILOU)
2020.18 In reply to 2020.13 
< Artefact
I had a premonition about that :)

< does that solve the problem for you? "Add detail to inflections"...
Surely but as I am always at same default regulate, "add details" disable, it's the same for your model so pertubating :)
Because mine has not!

< Probably you've got some duplicated
He he
It was just the start generator curve of the original of Kevin's model so resolved :)

And yes with the fillet Shell is some better!

EDITED: 2 Oct 2008 by PILOU

  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:  kevjon
2020.19 In reply to 2020.15 
>You have to just visualize what the result of the cut is though
This is the problem working with nurbs when you are trying to model stuff that looks like the real thing and you have to define the overall shape first and then do booleans rather than doing it the way I was on the first post of this thread which is define the shape of the object with curves (get that part looking right) and then create the surfaces.

For example the fuselage of the F9F panther (see pic) is quite simple (perfect circle) but trying to define the shape of the cross section curves so that when you boolean out the cockpit opening so it is the right shape and size is a bit of geometric puzzle due to the subtle curvature extending from the fuselage up to underside of the canopy (particularily around the rear of the canopy).

>I'm not quite sure if you actually wanted a totally flat plane in there or not though.
It very slight (see photo) but thanks for showing the idea of how to address this area.
Image Attachments:
Size: 112.2 KB, Downloaded: 39 times, Dimensions: 600x450px
  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:  BurrMan
2020.20 In reply to 2020.19 
For the canopy area as you mention, you could build the shape first, then cut out your inside section.

EDITED: 19 Jun 2012 by BURRMAN

  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:  kevjon
2020.21 In reply to 2020.20 
Thanks BurrMan

The blend of the fuselage into the cockpit is quite a soft and subtle which needs to be handled by a variable fillet radius. As MoI does'nt have the option you would need to create it manually via a swept surface. There is also a subtle chine along the spine of the fuselage which adds to the complexity of shape.
I'll post up my attempt at it over the next couple of days when I have time to work on it. I always try and minimise surface joins as much as possible as they often result in messy mesh joints that need to be cleaned up in the polymodeller.
  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:  BurrMan
2020.22 In reply to 2020.21 
Look forward to the outcome. Details, details! :)
  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:  BurrMan
2020.23 In reply to 2020.22 
Thats kindof what I was getting at. Mine is kindof cheesy but if you know the direct shape you want , it should be fairly simple.

EDITED: 19 Jun 2012 by BURRMAN

  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:  BurrMan
2020.24 In reply to 2020.23 
>>I always try and minimise surface joins as much as possible as they often result in messy mesh joints that need to be cleaned up in the polymodeller.

MoI's surface mesher is second to none! Looking forward to a great model!
  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:  kevjon
2020.25 In reply to 2020.23 
BurrMann

Here is my first attempt tackling the shape which is the logical way to model such a thing. But I get pinching at the back of the cockpit due to the hard joint in my curves.



As Michael has stated with nurbs my method above is flawed because with nurbs it is best to do the overall shape then cut out the cockpit opening. However the problem with that approach is what shape do all my cross sections need to be so that when the cockpit is cutout it looks right in side and plan view ?

So my next attempt will be the way you show which results in more surface joins and more messy mesh joins but looks like it might be the only way. However creating the blends from the fuselage to the canopy will probably still result in the pinching shown above but I'll give it a go anyway.
~Kevin~
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:  Michael Gibson
2020.26 In reply to 2020.19 
Hi Kevin, thanks for the image - it is a bit small though and only from one angle.

Since you're trying to very exactly match the profile from different directions, do you have images zoomed in on just the canopy from different directions (like front view, side view, etc..) ?

That would definitely help otherwise it will probably be many more rounds of just trying different shapes that then you'll say are not quite exactly right...

Maybe NURBS is just not a good fit for your project or the way that you like to work.

Trying to do contoured surfaces tends to be a more difficult area since they are not as totally clearly defined by side profiles as a purely mechanical shape. So it is pretty natural that things like aircraft and automobiles are some of the more difficult shapes to work on.

- 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
2020.27 In reply to 2020.25 
Hi Kevin,

> Here is my first attempt tackling the shape

Could you please post the .3dm model file and also those blueprint background images? That should help...

- 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
 From:  kevjon
2020.28 In reply to 2020.27 
>Could you please post the .3dm model file and also those blueprint background images?

I'm at work now but will do later tonight when I have access to those files. As far as close up photos of the canopy area goes, no I don't have any. I do however have many slightly distant photos of the aircraft from different angles and will post some of them up as well. It will be enough to get the fuselage canopy blends close enough.
~Kevin~
  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:  1-11  12-28