Save time and material?

I'm sure this has crossed someone's mind before but what about making reprap parts which consist of a honeycomb-like internal structure rather than solid. This could be done by using the print head to trace the outlines of the cell edges in 2d or 3d to create the part.

The you could either fill in the spaces with a different material, or leave them as is. With good part design and strut orientation, you could probably get reasonable strength characteristics with a significant saving in time and material over solid parts. Parts will also be lighter.

You can then start to begin to tune the stiffness of parts with their purpose e.g. the new base for my fridge will have a very fine honeycomb but the star on top of the christmas tree will be very coarse. (Note how often the need for strength and durability often align, so I am prepared to devote more time and material to make that stronger, finer honeycomb). This could also be done on a selective basis within a single part (similar to the way one does mesh refinement in finite element analysis)

If you need a "solid" part because it looks better, just print out a skin at the same time.

This method would also reduce the need for running reprap in an oven to merge the polymer layers and (maybe) reduce warpage.

In fact, if Vik's results are anything to go by, you might be better off running it in a fridge and allowing the structure to be built in a self-supporting fashion!

Yes, Simon posted about this sort of thing back on November 30 of last year in the main RepRap blog. The article was entitled "Faster, Cheaper, Better".

It's down at the bottom of this link...

[blog.reprap.org]

It's a great idea ... if we can get it to work. If somebody designs software that can create the cells properly within non-regular solids we'll be home free. I may well be wrong but I got the impression at the time that that the methods currently available only work on regular solids like cubes and the like.

Simon? Any comments?

I remember Simon doing a beautiful job of drawing a cube with exactly that idea, who may have also been working on it I have no idea.

One of my eventual intentions is to use material with air injected, in order to create rigid foam type material. Drywall joint compound, because it is cheap, readily available, recyclable, and can be post machined by hand tools has a great many possibilites. The cell structure of the foam and the permability of the material in general mean that the brittleness can be solved by impregnation with other materials. Not as strong as a true grid but a lot more simplistic.

Ideally I would want the cell structure to be open enough to allow impregnantion with portland cement type binders, as well as organic resin materials to give more versitility to the system.

Mike

The thoughts and ideas expressed in this post do not reflect those of my employer and are intended only as communications between individuals. Any attempts at implement are at your own risk

Edited 1 time(s). Last edit at 09/12/2007 09:23PM by ohiomike.

"One of my eventual intentions is to use material with air injected, in order to create rigid foam type material."

Some years ago I had an idea of doing something like this. I was injecting air into molten HDPE with the intention of creating a large closed cell foam. There was one little problem. While the cool air I was injecting quickly took up heat from the molten HDPE and expanded nicely, as the mass cooled the air within the cells cooled and dropped in pressure. This tended to warp the foam and caused surface collapses near the surface of the object I was trying to confect.

I was very disappointed.

Thats one of the major advantages of using a syringe type delivery system, I'm just going to create microbubbles the same way I would for an angel food cake, mix like crazy. Hopefully by changing the percent moisture and varing the level of surfactants, I will get at least a little control over the closed/open ratio and the bubble size.By staying at room tempature I can try and avoid the problem of heat based deformation.


Mike

The thoughts and ideas expressed in this post do not reflect those of my employer and are intended only as communications between individuals. Any attempts at implement are at your own risk

Edited 1 time(s). Last edit at 09/12/2007 09:24PM by ohiomike.

Do you think you could manualy select regular solid shaped areas in the interiors of objects to be honeycombed, and then leave the rest solid? Either that or cut/paste haxagonal voids to the apporiate areas.

Maybe a design time add-in honeycomb selected area. And it could just bore hexagonal holes in a pattern mating up to the enclosing area. You would have to carefully chose x, y or z axis as well as enclosing area and width of walls for appropriate strength etc. We could even do this with existing designs we have.

In my mind, this seems more like an issue which is in the realm of object design and not something which needs to be added to the software. It should be easy enough to do a subtractive combine (cut) between two layers to provide the needed output. Most design software has such capabilities built in. Though I don't know if AoI supports subtractive combines... it should.

Adding such a feature during the design phase will have added benefits such as being able to control the cell size, cell wall size, and top overlap (how do you intend the extruder to cover them). Adding such features into the software could become a nightmare, especially considering the current state of development. Personally I would consider such a software feature unnecessary and bloated.

~Seth

"Though I don't know if AoI supports subtractive combines... it should."

Yes, it certainly does. I think it might be a bit tricky to do with a process as complex as what Simon and now Michael have suggested. You could do it by hand with AoI for an object but I suspect that it would take a good long time.

One interim measure, however, might to do what it appears that Vik has done with the cross-hatched infill in printing which is to make the infill less dense, viz, spread the extrusion tracks out to where they aren't welding on either side but not so far that the next layer sags into the interstices. This makes an openwork rick of polymer that can greatly reduce the amount of material in a solid. Of course, the strength of the object is not as great, but then that may not be all that important for any number of things that we would like to print.

Hmmm...

I'm speculating, mind you, not having used Aol.

How about making the solid, then making a shell of the solid, subtracting a template of a honeycomb from the original solid, and then adding the shell back in?

I don't know if you can do this. I don't know if the result would be useful. As someone mentioned, it wouldn't consider needed strength, so the fourth step would probably be hand designing extra trusses where strength was necessary, using the shell as a template, and letting these overlap with the honeycomb pattern.

Oh you could do it. I think, however, that you've missed the bit where the honeycomb template has to fit within your object.

From a practical standpoint I suspect that the boolean op to do the removal would require an awesome amount of cpu time and memory. I also shudder to think about what the .stl file coming out of the process would look like, never mind the complexity of the .aoi file.

Perhaps so. As I said, I've not worked with Aol. To me, subtracting zero from zero, ten inches from the item, should make no difference at all.
If Aol insists on remembering that there is a void in the middle of nowhere, ten inches from the solid, that would definitely present a problem. With memory, if nothing else.

Can we "borrow" a GNU meshing tool from a FEA application? Then transform the mesh generated into a web of polymer.

It may need decent cpu power but I know that Solidworks has a simple FEA add on that doesn't take up too much resources or time.

If only we had some mechanical engineers to comment on the feasibility of removing more material as appropriate from pieces with bulk. Maybe someone familiar with the current design, and having experience with creating and modifying AOI/STL models. Say, who was it that made those cute little RepRap logo drip holes in the parts. Wonder how much plastic and therefore extrusion time could be squeezed out to improve the take off curve.

If an FEA mesh tool is used to generate the honeycomb internal structure, you are halfway to doing a full-blown stress analysis already....who needs mechanical engineers!

Of course, you will need to understand how the part is used and what sort of forces it will need to withstand.

There is a quick and dirty way of removing material from an object that is relatively easy to implement. You simply increase the distance between individual tracks of your diagonal infill. I saw Vik doing something of that sort months ago and have adopted the dodge largely to reduce corner curling. I'm running at about 50% infill as a result. This isn't a big problem for most parts in that you need nothing like the mechanical properties of solid HDPE in them.

I have seen your and Vik's blogs on that technique and i have been thinking:
wouldn't the obtained pieces be just great as a gridwork to be filled up with a casting resin, effectively achieving a composite material, with a resilient thermoplastic and a tough/hard resin?

i don't know if PE would work as it's just not good at gluing together with anything unless it's heated.. but then maybe we could use a thermoplastic with a lower melting point than PE to fill the piece out. It would lightly melt the surface of the PE threads and weld them in, but not completely so, keeping the piece's grid structure....

If you use CAPA or PET or ABS as thermoplastics, you could use any resin as they would glue to the threads quite well.

Just an idea...

I was thinking the exact same thing, similar to my experiments with drywall compound where the extruded material is only suppling a form to contain the casting resin. Maybe design gates in the infill so that the casting resin can join up with inself throughout the finished part and the thermoplastic would only act as structural reinforcement.

Of course that leads to all sorts of interesting ideas, like make the support material water soluable, infill primarily with support material, then use a water based resin to dissolve the support material as it takes its place. Or use two part rigid foam resins as infill to provide strength without adding weight or cost to the finished product.

Or use a tempature stable casting resin, like say a solubilized polyurathane perhaps even with a light filler load to lower cost. Then bake to reclaim the thermoplastic, and you have a finished product without losing the starting materials. Use the thermoplastic to create cavities in this finished product, and it could be used as a pouring mold for aluminum casting (althouh it would smoke like crazy).

Mike

The thoughts and ideas expressed in this post do not reflect those of my employer and are intended only as communications between individuals. Any attempts at implement are at your own risk

Edited 1 time(s). Last edit at 09/12/2007 09:19PM by ohiomike.

When I get some time, I was thinking of modeling it all in Inventor, since that what I use. Then I could do FEA on changes..... Inventor is not real fond of STL files... it will export them, but not import them.

AOI (with approprite addins) supports Wavefront in and out so you should be able to use that as transfer medium between Autodesk Inventor (with approprite addins) and AOI.