Tray1

Tray1

The STL file and a link to instructions on how to make the waffle bottom are here: Thingiverse and here: Pinshape

May 19, 2017

Birk Binnard Birk Binnard
Revered
616 posts

6 replies


Nice waffle bottom :-)
I find that my belts are grateful if I avoid sharp corners in the XY-plane so I would give each one a fillet. Reduces noise and ringing too :-)
I use at least 1mm fillet radius if possible, but sometimes go down to anything that is bigger than the radius of the nozzle if things are too cramped.
With my weapon of choice (CATIA V5) I would probably make one single pyramid stump, give it fillets, then repeat it along X and Y. I would define the size of the individual "primitives" so that there is always a whole number of them in each direction. Would that also be feasible in Rhino/Grasshopper (which I don't know, but like the name of)?

May 19, 2017

Kerrberos Kerrberos
Beginner
28 posts

The 2 links I gave include a link to a blog page where I describe how I made the waffle bottom - I used a totally different approach from the one you suggested. If you are interested you can see the blog page here: waffle bottom

Your discussion about filleted corners is interesting. My printer, an Atom2 delta, has never had any problems related to drive belts - so square corners were an obvious choice for me.

CATIA, of course, is a far more robust program than Grasshopper + Rhino, so it is probably easier (and more reliable) to make filleted corners with it. However, my sense is the filleted corners would take much longer to print due to the tiny straight line segments the final toolpath would require. The method you describe could be done with my software - I'll give it a try and see how it compares to my current method. But I'll omit the fillets because Rhino has problems exporting a filleted surface as an STL file.

One of the things I did with my printer to smooth out it's operation was to tweak the Acceleration setting in the printer's firmware. Doing this on your printer might help the drive belt problems you mentioned. There is a setting for Acceleration on my printer's display screen. I reduced it from something like 10,000 (or maybe it was 1,000) down to 600. This greatly improved the printer's motion control and made for much quieter printing. It does slow down the printing of complex shapes (ones with lots of short movements), but most of my parts are large so this usually isn't a problem. It does slow down printing the waffle bottoms however.

*Update*

Well I tried your method of making an array of single bumps. It didn't work because of the bug in Rhino having to do with Boolean Difference. Basically this Rhino function is known to "not work in all cases", and I have verified in the past that this is a correct statement. It should probably work in Catia though. Unfortunately that one is a bit out of my price range (even in the capitalistic US (<== for Bart.)

May 20, 2017

Birk Binnard Birk Binnard
Revered
616 posts

I read your blog page (I regularly do!) before I answered and it triggered my how-would-I-do-this neurons!
I have a medium-old CB+ with bartified hotend and so the famous M1203-option, which will usually find itself set to something between 1200 and 1500mm/s2.
After making such a suggestion I have just done a test design to see how it would turn out - the fillets would of course not meet completely but the tiny bridging that would cause should be no problem. CATIA is also known to have some strange behaviours with booleans, particularly when attempting to subtract identical areas/volumes from each other, so that the result would be excatly 0, but in this case (creating a simple 2-dimensional array) it works fine...

May 21, 2017

Kerrberos Kerrberos
Beginner
28 posts

Well I'm happy to hear my blog is of some use to you. What's interesting to me about it is that every time I think I've run out of things to post about (like right now), something new or different crops up and I end up making a page about it. It will be interesting to see what comes next.

Your info about Catia is worth noting - if Catia has problems with 3D Boolean operations then I expect there is not going to be a fix for that anytime soon. From what I can tell, Catia is the top of the heap in the 3D design world, and what you said about it mimics the posting the author of Grasshopper said sometime ago.

Back then I started a discussion about these kinds of problems and what he said was that the solution to a Boolean subtraction boils down to a potentially very complex mathematical problem, and for some of them (depending in the geometry involved) there is no way to solve the problem. He went on to say this is not an end-user issue, but a deficiency with the code in the 3D software. His contention was that software should work the way the end user needs it to.

If only that were always the case.

May 22, 2017

Birk Binnard Birk Binnard
Revered
616 posts

Among the vastly humourous and predominantly male German engineering fraternity Catia is sometimes known as "the little French girl" or just "she" - due to a certain delicacy and desire to be pampered.
Don't read too much into what I said about Catia's treatment of booleans - it handles almost everything you throw at it flawlessly. Sometimes the things I give it are just unfair (example: two design iterations of a 1500-part assembly where I want to know which parts have had a design update => combine both assys into a single body and subtract the two from each other, then reverse the order of subtraction. Anything left will belong to a part that has been changed between the two models. No wonder it sometimes gives up...).
There are conflicting schools of thought regarding Catia (probably not limited to Catia, but it's all I get to see at work) booleans. At least the following can be observed among the people I work with or have had trainings with:
1. Some are of the opinion that a model of a part manufactured by conventional machining should be made up exclusively of boolean subtractions to ensure that no material is added where a tool has already passed (part would otherwise be impossible to make).
2. Some think that use of booleans demonstrates a sophisticated design approach (like grouping all drill holes in a positive body and subtracting this from the main body in one fell swoop).
3. Some avoid them "like the devil avoids holy water" and complain that they are unable to make their design look as they would like it to.
4. Some will accidentally on purpose misunderstand design guidelines and do the entire design in a sub-body, finally adding this and only this to the until-then-empty main body.
5. Some use them where they should be used. Everyone thinks they themselves belong to this category and are largely surrounded by morons who have not quite understood the most basic basics. Huge fun/mind numbing boredom can be had listening to conversations about "good design philosophy".
These differences in personal approach would not be possible if Catia were not in fact able to handle pretty much anything in some way or another.
--------------------------
On a side note, I have read up on booleans a bit in the Meshmixer forums - they also (used to) have boolean problems, particularly when original and to-be-subtracted surfaces are almost coincident - due to the huge amount of tiny "confetti" that produces...

May 27, 2017

Kerrberos Kerrberos
Beginner
28 posts

You make some very interesting points about how people use Catia; they show how there is a very significant difference between designing things and actually making them. In the Grasshopper forum, which is mostly frequented by architects, there are many postings that boil down to "here's my design, why can't I get it to work?" In my case, "get it to work" almost always means "get it to slice with no obvious errors."

Overall I have probably spent as much time doing re-design in order to get successful slicing as I did coming up with the original design. Hopefully the software we use will continue to improve and thus minimize this kind of requirement.

May 28, 2017

Birk Binnard Birk Binnard
Revered
616 posts
To start a discussion or reply to a post please Login or Create an account.