Looks like nick hit the pattern, It's exactly the pattern in the picture.

Nick did it. Nick, how did you do the holes between the arms radiating out from the first hexagon group? I know that you can make the first group and then array it around at 60deg but how did you make that first group? The "group" are the black filled holes in my illustration above.

Michael, I like the script idea since it answers the question about deselecting edges that would not be chamfered. Also, it will take my skill level up a notch; I have not tried using Moi3D scripts. Thanks.

BurrMan, one thing about your hole pattern illustration, the first ring out from the center would have six holes. Having twelve results in one less outer ring and too many holes in the "group". Thanks for helping out.

As it is said, "The proof is in the pudding." I will put this to use and post what I come up with.

Thanks,

Leonard

>Nick did it. Nick, how did you do the holes between the arms
>radiating out from the first hexagon group? I know that you
>can make the first group and then array it around at 60deg
> but how did you make that first group? The "group" are
>the black filled holes in my illustration above.

I'm not sure I follow you... but I think, that you might be trying to hard (I let MoI do the thinking :) ). The only thing I did is what you see in the screen shots. I just used Array>Circular for everything and added 6 to the "Item Count" field each time. Thats it! I never concentrated on the white or blacks, just the # in each tier and where the base line was. :)

You are so right, "trying to hard". And, it can make my head ache. :) I now understand you; the key: "...just the # in each tier..." So simple!

Brilliant, Nick. And brilliant, Michael!! Really!

Thanks,

Leonard

Edit: it took about two minutes to complete the hole layout! A note for others struggling with this in the future: in the # field of the circular array info, the first tier number would be 6 (holes) and each successive tier, n+6, where n is the previous tier's number of holes.

Thanks again all.

Leonard

Here is the kitchen utensil.

Michael, I tried the surface and hole selection but the fillet did not work. It remained in calculation for a long time and then ceased without applying a fillet. I tried several times.

Leonard

Hi Leonard, it is looking really great so far!

re: Failed fillet - it looks like the object has a pretty small thickness in that area.

That will limit the fillet radius or chamfer distance to need to be a really pretty small value in order to fit within that small area, like maybe something around the range of 0.01 or so?

If you can post the .3dm model file here, or send it to me at moi@moi3d.com (if you wish to keep it private), I can take a closer look at it.

- Michael

Something to grab "Malasada's" out of the fryer.

The trick would be to pick just one hole and get the value right, then do the surface slection and wait it out.

Try something like 1/3 of the thickness of the hole (can a hole have a thickness? there you go thinking too much again :) ).

Thanks for the cheer, Michael.

Yep, I had the value set at 0.01. So, I will go with 0.007 and smaller.

BurrMan, "Malasada's". Like our doughnuts? Yum!

Good point: do one hole and then surface selection. I am still newish to this compact but elaborate toolset that makes the workflow so easy.

Hey Nick. Hole thickness? I will think about it.

Thanks,

Leonard

Hi Leonard, I also sent you this info through e-mail but I also wanted to mention it here.

When you start to approach radius values around 0.001 units, that will start to run into different problems there, since the fitting tolerance for the fillet operation is 0.001 units in size.

So currently when you want to go much under about 0.01 you may need instead to scale up your object (to do this select it, go to Transform/Scale and type 0 <enter> to place the origin at 0,0,0 and type 100 <enter> to scale up by 100 times).

I'm going to take a look today to see if I can tune this up for the next v2 beta, to make sure that the fitting tolerance is adjusted to be tighter if your fillet radius is approaching too close to it.

Then the other problem that you will probably run into is adding in a few hundred fillets is going to increase the complexity (in data size and memory consumption) of the model by quite a lot, you'll need to set MoI up in "large model mode" for that to avoid consuming a lot of memory. To do this means to adjust the display mesh density to be coarser - that's under Options / View / Meshing parameters . Switch the "Mesh angle" to 20 degrees (from the default of 10), and uncheck the "Add detail to inflections" button.

With those settings some of your objects will have a less smooth appearance, but it will help to greatly reduce the amount of memory consumption, since otherwise there will be a zillion little polygons made on all those little fillets.

You'll probably want to adjust these display mesh settings before doing the fillet.


- Michael

Hi Michael,

Scaling up: my metal thickness is 1/32", or 0.031". Are you suggesting I scale up 100 times? If I scale that up 100 times my metal thickness will be 3.125" thick. Am I right on that?

Got you on the "large model mode". I did experience slowdown. Do I change those settngs back prior to exporting?

Leonard

Hi Leonard,

> Scaling up: my metal thickness is 1/32", or 0.031". Are you suggesting
> I scale up 100 times? If I scale that up 100 times my metal thickness
> will be 3.125" thick. Am I right on that?

Yup, but the scale up is just a temporary measure to avoid approaching too close to the 0.001 tolerance value for the fillet calculation.

Once the fillet has been calculated on the larger scaled object, you can then scale that filleted version back down to its original size again.


> Do I change those settngs back prior to exporting?

Nope, that is not necessary - those settings only have an effect on the display of the model in the viewport during normal modeling.

When you export that calculates a separate mesh from that, which is instead controlled by the "Meshing options" dialog that is displayed during the export. The export mesher does quite a bit more work to try and avoid making an unnecessarily dense mesh.

If you switch back at some point later on to working on a more simple model instead of a denser one, then you may want to switch the settings back at that point in time to get the smoother display which looks nicer on the screen.

- Michael

That all sounds great, Michael.

Moi3D is not only a great and fun modeling app but it is a stimulating lateral thinking exercise, too.

Thanks,

Leonard

Hi Leonard, I have tuned Fillet up now for the next v2 beta, so that it can handle smaller radius values better.

When the radius gets small enough to approach fitting tolerance, it will now automatically adjust the tolerance to be tighter relative to the radius, so fillets with small values (for example 0.005, 0.001, etc...) will work better in the next beta without needing the "scale objects up" manual workaround anymore.

But putting in a whole lot of little tiny fillets will still increase the model complexity (in file size and memory consumption) by quite a lot though, so that is still something to take into account.

If the fillet is going to be very very small on the screen on the final renders, you might want to put a chamfer in there instead of a fillet - it will look about the same if it is really tiny in the render but will have quite a bit less increase in complexity and in final polygon count too.

- Michael

Hi Michael,

Thanks for the news. App updates in real time - new to me. :)

I agree with chamfer ILO fillets or no treatment. With no budget or schedule, it is for when it does matter. The app lures one to questionable and undisciplined behavior. (Fault transfer. ;) )

Thanks for the good works,

Leonard