In this video I fix all your problem. You have a lot of naked edges, to find them I used script that is in the end of this post.
You chose very bad approach to make your object, I think first you had to make perfect cylinder or whatever you wanted and then trim it and modify. Intsed of this as I can see you made it from diffrent pieces.






script: /* Show naked edges */ var gd = moi.geometryDatabase; gd.deselectAll(); var breps = gd.getObjects().getBreps(); for ( var i = 0; i < breps.length; ++i ) breps.item(i).getNakedEdges().setProperty( 'selected', true );

And your project is here: http://we.tl/lvGFDTLFCL

Hi Sebastian, you can get rid of a bunch of the naked edges by just selecting your object and using Edit > Separate followed by an Edit > Join, that gets rid of about half of the unjoined areas, but there are still 11 edges primarily on the outer shell that are not joined (use the script shown above in Andrei's post and also described here to see them: http://moi3d.com/forum/index.php?webtag=MOI&msg=6051.2 )

When 2 pieces do not join together, it means their edges are more than 0.005 units apart from one another, so that's kind of an indication that the surfaces are not all that accurately hugging one another.

If that accuracy problem does not bother you, then one quick workaround is to scale your model down (which also reduces the distance between edges too), then join, then scale back up. If you do that in this case you will be left with only 4 naked edges on one surface so you can focus on rebuilding that one surface.

So that scaling rejoin sequence goes like this - open your file, select the object, run Edit > Separate to break it into individual surfaces. Run Transform > Scale, and at the first prompt that says "Pick origin point" type in 0 and push enter to specify 0,0,0 as the scale origin. At the next prompt that says "Scale factor", type in 0.1 and push Enter, your object is now scaled down by 1/10 in size. Now do an Edit > Join, and now repeat the scale this type using a scale factor of 10 to scale it back to its previous size. Now use the select naked edges script to see the 4 remaining naked edges that bound one surface - the actual problem there seems to be that there's a duplicated surface in that spot so it's not able to get a clean join in that location. Delete one of the duplicated surfaces from there and do one more Edit > Separate and Edit > Join and that will get you a solid with your existing pieces.

But like Andrei writes above, it is generally better for things like this to be formed by a single taller simple cylinder piece that is then cut, rather than trying to directly loft a surface between the bottom edge that is curved in the z direction with the top, that kind of direct-to-irregular-edge surfacing has left some ripples and bumps in your surface that you may not really want:




But anyway that sequence above can be used to get a solid out of your current object just as long as you don't mind that some of the surfaces are not touching each other to a very high degree of accuracy.

- Michael

Michael,
>>> When 2 pieces do not join together, it means their edges are more than 0.005 units apart from one another, so that's kind of an indication that the surfaces are not all that accurately hugging one another.

Never know about this, now I clear undertand why you use to be to advise to small down object to make join.
A lot of things are now clear for me... Thank you very much for explonation!

And one thing I'd like to know, in his model was an issue when I made sweep between 4 edges, but can not join becouse of one of edge has one little segment in edge in the corner, so I run command rebuild edge and repaeir it. My question is what happen with this small segment when we small down object and join it and then scale it up. This segment will just deasapere?

Hi Andrei,

> Never know about this, now I clear undertand why you use to be to advise to small down object to make join.
> A lot of things are now clear for me... Thank you very much for explonation!

Things are slightly more complex than what I wrote - the join operation does use a "relative tolerance" so if the bounding box around the things being joined is very large or very small it uses a scaled value. But it does target 0.005 units for the max join if the object is between 1 and 50 units in size (measured by bounding box diagonal).


> My question is what happen with this small segment when we small down object and join it and
> then scale it up. This segment will just deasapere?

It probably won't disappear, it is possible for that to happen but an edge has to be squeezed down to a very small size for it to actually go away through a join.

What exactly will happen kind of depends on how the edges relate to one another - but it is possible for the ends of the small edge to divide the larger one up into pieces and then the pieces get joined.

- Michael

Also now I understand that size affect modeling process.
What other things are depend on size? I remember something about fillet and size.
In what scale is better to work in MOI?

Hi Andrei,

> What other things are depend on size? I remember something about fillet and size.

Some operations used a fixed tolerance of 0.001 but I'm generally in the process of switching things over to the relative scale method. There is kind of a mix of things currently, but a lot of things do use the relative tolerance method such as booleans, sweeps, lofting, offsets, ...


> In what scale is better to work in MOI?

It tends to be best to work with things around 50 to 0.1 units in size, that way things that still target only a fixed 0.001 tolerance won't be either too accurate or not accurate enough.

- Michael

Got it, thanx)

>It tends to be best to work with things around 50 to 0.1 units in size, that way things that still target only a fixed 0.001 tolerance won't be either too accurate or not accurate enough.

Very handy tip, I always thought it was between 1 and 1000. I'll have to work smaller in future to minimise errors in joining things together.

Hello,

You can see your naked edges here. I see that you have a draft to that part( 2°? ), so it might be good to extrude out your cylinder with that taper and then start building out a tapered helical cutting curve or surfaces. Building out then carving away from your part should leave you with some good surfaces.

Hi Kevin,

> Very handy tip, I always thought it was between 1 and 1000.

That's generally close enough that it should be fine. There isn't any exact cutoff point, it's just that if your objects are too large in relation to the fitting tolerance the results can be too heavy and take too long to calculate (because they're trying to reach some very tight accuracy level in relation to your object), and if your object size is too small and approaches too closely to the fitting tolerance then fitted geometry is not accurate enough.

Usually it is worse for your object to be too small in relation to the tolerance though.

But the reason why I'm generally moving to a relative tolerance system that is based on some fraction of the object size is so that you won't have to worry about this as much.

- Michael

Thanks for all your input guys. The scale method worked perfectly.

As to my design approach, unfortunately this was the only way to do it. The bottom (contoured edge is a split-line for a cavity while the top edge (dia 20.5mm) is prescribed by the client. I had to work within those limitations. Cannot reveal more due to NDA.

Thanks again :)

Some thoughts about subject:
http://moi3d.com/forum/index.php?webtag=MOI&msg=6499.12

Hello Andrei,

I guess that set of recommendations are fine for some situations but the original part that was posted looks like it's being used in a molding operation. Finding the simplest way to model something doesn't always leave you with an optimized or usable model for a particular application. Having a plan is a good idea, but you have to work within the limitations of what the model is being used for.

>>Some thoughts about subject:
http://moi3d.com/forum/index.php?webtag=MOI&msg=6499.12<<

I really don't know how this is supposed to help...

OSTexo, bigseb,
Becouse your base shape is simple but made in complex way. It have complex geometry and you could not even join it... So even if client give you sourses in that bad state, you could rebuilt them to get clear geometry. It is not so hard to do.
I can see only one justification for this muddy geometry it is if clien need all that muddy edges and stuff... But I really can not see any reason for client to ask for this? May be I do not know something, tell me :)

My Portfolio: www.samardac.tumblr.com
A lot of my Tutorials!
Subscribe to my youtube channel: http://www.youtube.com/samardac
Russian community of MOI 3D: www.vk.com/moi3d

What you don't is how this all fits together. I can't reveal the rest of the project due to NDA. The geometry of the part I uploaded is not open change. It is not always an option to do things the simple way.

bigseb,
One thing I can say for sure - everything can be fitted or connected much more easer to a good geometry than to a maddy one. Peace :-)))!
_________________________________________________________________________________
My Portfolio: www.samardac.tumblr.com
A lot of my Tutorials!
Subscribe to my youtube channel: http://www.youtube.com/samardac
Russian community of MOI 3D: www.vk.com/moi3d

Hello Andrei,

That file is not simple, you're misinterpreting the model and it's intended use. It's pretty obvious from his sig that the person who posted the file is in the mold design business. It also stands to reason that someone who is adding draft to parts and naming files thermoform in another thread is using the file for some sort of mold operation as well. Formulate a plan based upon the intended use of the model, trying to do simple all of the time is not possible and in some cases will end up causing more problems later in the process.

Hello OSTexo,

Look pleas what I made, I rebuild its model from his edges, and you say that there is no way to make it geometry clear like this? Some molding process not allow it? You can trim it any time you like to get those pices but still keep clear geometry.

I made it more clear not only in geometry but also in surfaces, look I mark it in red. He made it with mistakes and got those jagged faces. And it is not becouse some molding process or client wish it is just becouse of absence of modelling skills.









He got all that troubles and horrible geometry becouse he do not follow rulles that I wrote in link I gave.

First, he break first rule, START FROM BIG AND GO TO SMALL. He has to be started from Cylinder and then make all other details but he tried to make Cylinder in the END and got all that troubles! Looks like he started from that bottom spiral detail and after that tried to make Cylinder.
And if even that bottom spiral detail was given by client, hi has to make sylinder absolutley another way... May be like I made it.

Then he break second rule:ALWAYS TRY TO FINDE THE SIMPLEST WAY TO MAKE THINGS.
He started to make Cylinder by joining all that sigments and got a lot of problems. It was not the simplest way but also wrong, becouse it produced jagged faces.

So guys all your arguments looks like you are trying to protect BAD GEOMETRY and absence of modelling skills arguing it by some foggy circumstances.
I can not see any arguments to keep that BAD geometry!


____________________________________________________________________
My Portfolio: www.samardac.tumblr.com
A lot of my Tutorials!
Subscribe to my youtube channel: http://www.youtube.com/samardac
Russian community of MOI 3D: www.vk.com/moi3d