Hi gizmo,

> Ok say I didn't want to create the main shape from a sphere or I
> needed it to be more specific shape wise?

It seems like you're still focusing on things a lot from a polygon modeling perspective? I mean here you're focused very much on the final edges and drawing those first.

It is possible to do that like Manz shows above, but really it can be a lot easier if you build a bigger shape that has the surface that you want, and then trim it back.

So if you don't want a perfect sphere as your main shape, draw a different profile and use Revolve or Rail revolve to create a different surface shape, then you can trim that using the same steps as was done with the sphere originally.

For example, if you take these 2 curves, and select the half one as the profile:



Then you can run Construct / Revolve / Rail revolve, and pick the longer one as the rail curve, and then pick the revolve axis between the 2 ends of the open curve, and you will make this kind of "customized sphere":



Also after you have completed the rail revolve, you can turn on control points for the curves and adjust them to update the revolve.

This can be an easier way to create a more customized starting surface because there are fewer curves involved to define it, and it will avoid some of the problems of trying to create a surface to a more complex boundary directly, which is possible but tends to be hard to control where points pinch together.

It takes a bit of getting used to if you are used to polygons, but with NURBS it can often be a good idea to build things larger than you need at first with more simple curve layouts, then trim or boolean them to produce your final result, rather than trying to build to the final outline directly.

- Michael

>>What I don't understand though is how you created those curves Manz?

When you have 2 standalone curves (as you have colored green) you can simply make a Construct~blend between the 2 curves (use G2 to keep curvature)

Many many thanks for all the help guys. I feel like a whole new world is opening up for me.

A quick question for anyone else still reading the thread. Can you only Boolean closed objects? I tried to union boolean the outer shell I created with the sweeped circle and all that was left behind was the sweeped circle? Should I have closed the main shell for it to work?

Right hold up I'm experiencing problems again. I've got the shell exactly how I want but how do I close it? I thought I'd simply create another curve cage but make its thickness a little smaller than the original. Ho there's numerous problems with this.
1) Copying and pasting the cage and moving it to the side to alter. How do I then get it back to exactly the same position? Object snape doesn't seem to work?
2)Because I can't get it back to the original position exactly it's out of place. When I do a Join command on the 2 shell pieces nothing happens? I'm still left with 2 pieces?

Here's an illustration of what I mean. How do I get these 2 componant to be snapped together exactly.

>>>Can you only Boolean closed objects?

Read the help file:- http://moi3d.com/1.0/docs/moi_command_reference7.htm#Construct_Boolean

Thanks Manz.

Anyone know how to snap the 2 shells exactly?? And then Join the two shells together to form one part which I can then perform a union Boolean with?
I want to go from this

to this

Hi gizmo,

> How do I get these 2 componant to be snapped together exactly.

Pick all the original wireframe and drag it from the end of that highlighted curve indicated by the arrow and there should be a corresponding point on your model to snap to as shown.


Cheers
~Danny~

Hi danny thanks for the reply. Right that seems to work in a round about kind of way. But then when I join the 2 shells together there are faults with it?
I've included 2 files. One with the 2 shells which I'd be interested in knowing how you'd snap together as well because both are shells. Rather than the shell you aligned with the curve 'cage'.


The other is the shell created using the technic you say Danny but there are faults with it when I joined them. That means that when I come to choose say fillet it for instance it doesn't work?

The union or join is probably being messed up with the fact that the 2 surfaces are "Barley touching" along the edges.

A solution would be to take say the bottom piece and make it "Bigger" so the edges stick through the top surfaces edges, then your Booleans should work.

[EDIT]Bad solution.[EDIT]

Hi gizmo,

> Can you only Boolean closed objects?

No, you can boolean open objects as well, however booleans are mostly oriented towards figuring out which pieces to keep and which to discard according to where they are located inside of a volume.

If the piece you want to cut is not a volume, it can be somewhat unpredictable which pieces will be kept and discarded.

In cases like that, switch to use the Edit/Trim command instead - it allows you to cut objects with one another similar to booleans, but you manually control which pieces you want to discard by clicking on them.

So in the case that you are describing use the Trim command instead.


> I've got the shell exactly how I want but how do I close it?

This would likely be easier with the method that I described earlier about building a larger surface first and them trimming it back.

When you build a surface like you have there with the corners of the 4-sided surface sheet totally smooth to one another, that will create a degenerate surface normal in that area and prevent some commands such as shelling or probably filleting from working in that spot.

That's another reason why it can be better to build a larger more simple structured surface first and then trim it, rather than trying to build a surface directly to your final outline which is more of what you would do if you were building with polygons - remember you are not building with polygons here, with NURBS it is generally better to use trimming and cutting type operations to make a final outline if that final outline is different than a 4-sided sheet.


> Here's an illustration of what I mean. How do I get these 2
> componant to be snapped together exactly.

The usual method is to find a key point like an endpoint and then grab the object close to that key point to snap on to it and drag it over to the matching point on the other object, Danny showed this above but also just to be clear here is a screencap:



- Michael

Hi gizmo,

re: shell with fault - if you zoom in you can see that the pieces did not get really aligned properly with each other:








The pieces could join because some parts of them were close enough, but other parts were not close enough so did not get joined into a single shared edge.

You'll need to get a better alignment than that to get a clean join.

But after that I still would not be surprised if you could not fillet it, because you've got those degenerate edges in the surface patch - that's where your 4-edged surface has a tangent edges on its surface corner instead of a distinct corner, that causes problems with offset operations and part of filleting involves an offset calculation.

Again if you would use the technique described earlier on building a larger surface then trimming it back you are more likely to avoid many of the issues you are running across, you are attempting to do things in a more difficult and low level way right off the bat and that is increasing the difficulty level for you by quite a bit.

- Michael

Hi gizmo,

Michael wrote: you are attempting to do things in a more difficult and low level way right off the bat and that is increasing the difficulty level for you by quite a bit.

This is so true, your head will explode ;) trying to do surfacing type stuff like that without exploring and seeing what MoI can and can't do or having a Cad/nurbs background.
But don't get too discouraged or frustrated, the more you find out about MoI and techniques shown on the forum the more easier it will become :)

I've attached your file, after having a play with it, and as Michael mentions there is a bit of low level work to be done before you can use the Boolean and filleting tools successfully so it is possible.
One hint is that if you look at the model you will notice it's in 4 pieces joined together, so I just worked on a 1/4 of it, got that right then mirrored and joined them together to make a solid then the holes and fillets were easy after that.

.

Don't know if this is of interest/value.

My starting point was a sphere with thickness.

Does the pdf in the zip explain the MoI moves well enough?

I have enclosed the .car file for those who could use it.
I have fun!
Brian

Thanks again everyone for your input. The willingness of everyone here to help me out and explain things is greatly appreciated.

It is really frustrating getting my head around Nurbs concepts but I really want to crack it after several failed attempts over the years. Moi's user friendliness is proving to be my saviour in this aspect.

Anyway back to the case in point. I understand totally what people are saying but what I can't get my head around is how would I be able to create the shell (exactly how I want it) without doing it the way I have? Starting from a sphere I can't see how I would be able to precisely replicate the shell dimensions I want? Perhaps someone could try and do this and prove me wrong!? Use mine as a template but create it entirely differently?

As for how my model is currently. Danny, how did you achieve the result you had!? That's 'almost' exactly what I want! I'd just like the thickness of the shell to be less. As you said I've broken mine into a smaller part but how then am I to create the underside of it like you've done?

Thanks again in advance!

Hi gizmo,

What version of MoI are you using, demo or V2 beta ?


Cheers

The demo. V2 is locked off to current owners isn't it?

Hi gizmo,

>The demo. V2 is locked off to current owners isn't it?

Yeah it is , just wanted to know to see which version to use to try some things out , V2 has some new features which help reduce the steps.

Cheers

Hi gizmo,

> Starting from a sphere I can't see how I would be able to precisely
> replicate the shell dimensions I want?

Actually the key thing is not that you have to start from a sphere, but rather that you start from a piece that is extended larger than what the final result will be.

The sphere was just one example of such an extended shape - you can use other tools such as revolve or sweep to build a customized extended shape instead of a sphere.

Here are some steps for how to do this kind of an approach with your shape.

To start with I used these 3 curves. Especially note here that each of these 3 curves are simple planar curves, that is one of the biggest benefits of this approach, it is a lot easier to control and shape curves that are just 2D. The model file of the curves is attached here as pad_extended_trim.3dm



Select the top curve:



Run Construct/Sweep and select the central one as the sweep rail. This will create the extended shape:



One thing to watch out for there is you don't want to have a large profile shape that goes through a tight bend in the rail, which can force the swept surface to bunch up and kind of overlap through itself, you don't want that.

Now select the shape, run Edit/Trim, and select that 3rd profile curve as the cutting object:



That will cut that shape up and then the Trim command will allow you to pick which pieces you want to keep or discard, select this piece to remove:



Now with that trim complete, it will give you your basic form:



Note how that was constructed without having to build curves that swoop around in 3D?

Also another big benefit of this kind of construction is that shapes that are built to an extended simple form and then trimmed back will tend to have a very evenly curved shape throughout them. When you try to surface something to a fancier outline directly it will tend to cause different amounts of curvature in the surface as it approaches corners and pieces where your curve framework comes together.

Generally surfaces constructed in this extended manner will be higher quality in their curvature characteristics.


Now you can use Construct / Offset / Shell to turn this surface into a solid by giving it thickness:



This will cause a piece to kind of jut out past the original trim outline - you can select the shelled solid and use Construct / Boolean / Difference to slice it with that same trimming curve, which will cut the shell into 2 pieces and then you can delete this piece:



That leaves you with this shape:



Now run Construct/Fillet to round off the edges:






So to use this method, you have to construct your initial surface to be somewhat larger than your final shape will be, then you use trimming or boolean tools to cut it back to your final outline, rather than trying to construct your shape initially to the actual final outline like would tend to do with a poly modeler.

It's a much different approach and it may take a little while to get used to it.

Hope this helps show you how this approach can work though!

Once you get used to it, you can create shapes using just a very small number of 2D curves though, which tends to be really quick.

- Michael

And another slightly different approach.
Brian

Michael, thank you so much for taking the time to explain that to me. I would not have thought to model like that had you not.

However, one step I do not understand though is where you Construct / Offset / Shell and then use Construct / Boolean / Difference to give the model volume. I don't get how you've done this? The offest seems brilliant but it doesn't 'close' the model or give you the option to? Also using Boolean Difference I don't see how you've managed to close the object?

I hope you can spare me some more time to fill in these last parts of the puzzle for me!

EDIT! Hold everything, just seen the 'Shell' submenu in Offset! What an idiot.

Anyway thats sorted now. I've got my shell looking great with thickness and a fillet. Now I've sweeped a circle along the shells outside curve and then attempted to Boolean/Union the 2 objects. Unfortunately it doesn't work? I've applied ajoin to both object but it hasn't made any difference. Furthermore I can't seem to 'jion' the sweeped circle? It only stays in 2 halves?ve attached the current file.

Hi gizmo,

> Furthermore I can't seem to 'jion' the sweeped circle? It only stays in 2 halves?ve

This is how I see it, the curve that is extracted from the model is quite complex as in, if you turn on the control points for this curve you will see there are a lot of points, so I think the auto solver for the sweep is having a hard time calculating a nice sweep, and if you zoom up into the joins, circled in red, you will see that the ends have gaps and are not square or plumb.

To get this sweeping more accurately, I again only worked on a quarter of the curve and placed the circle profile on each end of the curve 90 deg to each other, then swept it with no caps, then mirrored and joined.



I've attached your file with a completed sweep and one set up ready for sweeping on the model so you can see what I mean.

~Danny~