Manipulation with subdivided Voronoi surface produces unexpeted results

[AHTOH]

Hello there.

Here is what I get when I create Voronoi on plane exactly the way it is in Tutorials. After conversion to mesh object becomes totally unmanageable. Garbage geometry only.

Anton.

ring 3.fmz.zip

[Tech]

Hi Anton,

 

Can you post a file that contains the "before" condition so we can see what you are starting with?

[AHTOH]

Hello again.

Thank`s for prompt reply. Actually standing grid is the object before. The bent one — is after.

Anton.

[Tech]

Hi Anton,

 

Well, the Before condition is already bad:

 

 

 

In general, when creating geometry for SubDivision work, starting with a carefully constructed cage with 4 sided faces is recommended, and those randomly triangulated -- long thin triangles are not helping with your shape.  Perhaps you can start by constructing your base shape this way and get better results?

 

(While it may be possible to take "any shape" and generate a SubD object, a considerable amount of luck will be required with that approach, and working with 4 sided faces for the cage as shown in most of the tutorials will give good results much more consistently...)

[AHTOH]

Hi Paul.

Unfortunately I ca`t see how could I possibly get rid (or not to get) these triangulations. I am creating my voronoi cage as it is stated in tutorial I saw (besides I can`t find it to refresh my memory) and getting this kind of results 

This is how my initial shape looks:

And this is what I get after Sub-D:

This is my result after I bent it:

What am I doing wrong?

Anton.

[AHTOH]

Fine! Now I managed to get a glitch free model by meshing it:

And I bent it. What could possibly happen? 

It unbents and turns straight again as I click to final point.

What to say... Nice! Another 5 hours wasted.

3-3.fmz.zip

Anton.

[AHTOH]

Would appreciate to see the answer.

[¢hris £und]

If you extract the original object out of it, it will bend without issue.

 

However, looking at the extracted model,   there is a LOT of triangulation in it.   Sub-D's  prefer Quads.

So, when I go to Sub-D it again, it disappears all together.

 

I will through it into another app to see what I get (Modo, which is essentially a dedicated Sub-D modeler)

 

 

[¢hris £und]

I bet the problem is, when you bend it,  the bend tool is trying to tryangulate the base mesh, then when the Sub-d's are being applied, it is choking (as would be expected)

 

attached is an import of the model from Modo.   (bent it in modo, then applied polygonal Sub-D's to two versions)   each model is poly now, and you can use fZ's Sub-Ds on them.

 

In reality, there should be a bunch of cleanup on the model, which would make it look a lot better.

AHTON.fmz.zip

[¢hris £und]

Update,  

 

Tech,   on the file I posted back,    the two high density meshes can be Sub-D'd   where the original Density mesh (though curved) disappears when Sub-D'd.

 

The two Higher meshes aren't very clean either, They have holes that would need to be fixed

 

 

 

[Tech]

Hi Chris,

 

We downloaded your file, Separated the Joined Volumes, and Object Doctor Fixed (only) Missing Faces with less than 12 sides, and now we have 3 proper solid rings.  Generating a SubD from a base object with this many faces will take some time, but the result looks pretty clean to me:

 

 

Are you seeing something different?

 

EDIT -- Ah, ok, if you Show Edges, not all edges are showing -- thanks for your report, we will look into this...

[AHTOH]

If you extract the original object out of it, it will bend without issue.

 

However, looking at the extracted model,   there is a LOT of triangulation in it.   Sub-D's  prefer Quads.

So, when I go to Sub-D it again, it disappears all together.

 

I will through it into another app to see what I get (Modo, which is essentially a dedicated Sub-D modeler)

Hi Chris.

Thank you for your answer. Please don`t waste your time with MODO I also own it and can do that myself. 

As for triangulation: my original object was quad meshed and this triangulation was made by FormZ while converting back and forth.

My original object bent without any trouble. But all that mesh mod thing works one time of three.

Best,

Anton.

[¢hris £und]

Yes, bending the object in Z triangulates it.

 

Tech, Point, the objects that were quads, even though they were not well formed succeeded, the lowest density one and not made for f quads failed.

[¢hris £und]

I can only extract the objects you provided, but I see the first object is made of Triangles and Quads,  the second is made of on n-gons.  it appears that there was a mesh operand placed onto it at one point.

 

Even in Modo, this would take significant cleanup.

[Gyro]

I'm also attempting to create a Moroni object but one from a closed surface, a sphere for example, am I right in thinking Formz will only create these from a flat plane.

This video <

shows a form being created from a  sphere, is this possible in Formz?

Ultimately I want to create the forms in the attached link:

http://forums.formz.com/index.php?/topic/3077-generative-design-webinar/?p=14197

 

I'm unsure of the best approach. Compounded by the fact I want a rough surface.

 

Any help appreciated.

 

Thanks Andrew

www.gyro.co.nz

 

[Tech]

Hi Andrew,

 

You should be able to create forms like this with the Scatter Points tool and then Voronoi tool with the proper parameters.  In the attached screenshot, the sphere has a 7' radius (obviously this might be larger than what you need, but you can scale down once you get this shape, or you can scale down all of the measurements before creating your object), 200 points scattered, Separate By: 1'

 

Then select the Voronoi tool and set the Tool Options to Lattice and a radius of 2".  Does this help?

[Gyro]

Thanks for your reply.

The diatron I'm building will be be 3D printed.

I'm attempting to overlay a star, over a soccer ball that has been modified as per the Generative Design Webinar.

Some seemly insurmountable (for me) have arisen.

 

I wish to streamline the process by overlaying a star and the soccer ball, but the star even though set to soccer ball has a different number of facets 6 as opposed to the soccer balls 5, so they can't be lined up.

I have tried using a cone but it can't be formed from the meeting point of the soccer ball faces, always wanting to snap to a face, I've tried also placing a sphere inside and pulling a cone from there but almost impossible to line it up through the meeting point of the soccer ball faces.

 

Is there a way to roughen the surfaces of the spikes prior to blending then to the soccer ball?

 

Am I proceeding down the wrong path with construction of this model, if so could you please give me some tips.

Thanks in advance

 

Andrew Turney

 

 

 

 

 

Diatron Test.fmz.zip

[Tech]

Hi Andrew, 

 

It seems like you are on the correct path with the "soccer ball" portion of the model... but I think the cone idea would yield better results rather than using the star tool.  Unless you knows how many "meeting points" are on the soccer ball and the star has the same amount of points, but trying to rotate it to fit properly might be more cumbersome than orienting one cone at a time.  To do this properly, place an X at the center point of the soccer ball and then try setting snaps only to Intersection and then ghost each cone that is complete once they are lined up with the meeting points.  

 

To “Roughen the surface,” try using the disturb tool or maybe a bump map that will give your object a more custom, subtle texture.

 

Does that help?

[Gyro]

Again thanks for your help, here's where I have got up to, but I wish to blend the spikes into the ball, then ideally roughen the surface. 

 

I've tried Converting the object type, subdividing, using the Derive tools to no avail.

Is it due to the object types differing?

 

I have attached a file if you have the time to check it please.

 

Thanks Andrew T.

Diatron v3.fmz.zip

[Tech]

Hi Gyro,

 

We are not sure exactly what you are trying to do here, but you could Union these objects and Round their intersections if you wanted:

 

 

Is this just something you are creating "for fun" or is this a real project of some sort?

 

(Unioning all these spikes and Rounding all these segment sequences will be quite computationally intensive and create a very complicated object.)

 

Alternately, if you want to create the whole thing as a SubD object, you should start with a very simple facetted cage object composed of primarily 3 or 4 sided faces -- but that will be quite complex as well...

[Gyro]

Thanks for your reply, I'm creating this as a movie prop that requires 3 printing. There is an image  in my previous  post. Roughening  the surface will produce a larger file still I suspect.

I have attached a different approach that occurred to me about 1am this morning, would this be better, it has 1 hexagonal hole but round all faces doesn't seem possible, especially when applying that to the whole model.

 

Cheers Andrew

Diatron WIP.fmz.zip