plastic (composite) frame for reprap printers?

I must admit that I am frustrated at the direction reprap is going. I can understand that for maximum rigidity and easy building going for metal frames is a common sense solution. But that is not reprap anymore. Unless we will be able to print in metal those structures. Reprap spirit is fading and everything is becoming more and more rep strap actually. Most of the much lauded 70%,80% or even 90% reprap 3D printers out there have all but some non-critical parts made out of metal. How can a reprap 3D printer claim to be 90+ percent rep-rappable when all that is actually 3d printed are connectors? There are some attempts at fully reprap like the Snappy (without motors/electronics) and there is Dollo that are doing a great job to keep reprap alive but I am sure we are not tapping even the tip of the iceberg of what advanced engineering structures can do for our parts.

I am sure that reprap can be achieved should all these bright minds get away from the confort of taking it the easy route. There are 3D structures that are suitable for rigid plastic frames. I am but annother reprap enthusiast without much expertise in trusses, micro-lattices, corrugated hollow-core panels etc but I can see that the plastic composites industry is going aerospacial more than metal. Of course we can't expect PLA/ABS frames to perform at those exceptional levels but there is room for adequate performance out of what we haveat hand.

I can see that most printers that use plastic for frames use acrylic panels. Is that the most rigid plastic out there yet? Can trus/lattice panels be printed in acrylic? What are some reprap options in this direction?

I'd like to hear your thoughts about this.

---

RepRap Lander concept on Concept Forge
RepRap Lander concept on RepRap Forums
My Things, mostly experimental stuff

I hear your concerns about the "pollution" of the RepRap ideal, but for me, when I started my adventure in 2011, my primary objective was to buy or build an affordable, upgradeable 3D printer. "Buy" was way out of my budget (how things have changed in 4 short years!), so "build a RepRap" was the only viable option. I only encountered the RepRap project because it was the only affordable way into 3D printing for me - the ideals of self-replication were (and are) admirable, but for me they were secondary to the end goal.

My RepRap started life as a 1st-generation Prusa Mendel, which was pretty much the bee's knees in affordable DIY 3D printing back in August 2011. I bought a kit of RepRapped plastic bits and "vitamins " on eBay, and self-sourced all of the other components (stepper motors, electronics, etc). The buy-and-build process was itself instructional (and fun!), and about two months after construction commenced, I had a workable 3D printer in operation. Since then, it has been upgraded in several respects (notably by replacing some of the original plastic components with metal alternatives, such as axis slide bearings, Z-Rod Couplers, print-head, etc). It still LOOKS like a classic Prusa Mendel, but in many ways, it owes as much to later generation machines as it does to its predecessors. It prints faster and with much better fidelity and reliability than it ever did as a "pure" Prusa Mendel, and it is still user-maintainable and upgradeable - and that to me is much more important than the fact that its RepRapped plastic content has dropped by about 20%.

The Prusa Mendel and its predecessors truly embodied the core philosophies of the RepRap Project: These days, if you start a similar adventure, you are more likely to try building a 3rd-generation Prusa i3 or similar,which embody the accumulated learnings of both the originators of the RepRap Project, and thousands of RepRappers. The kits are cheaper than ever, but use a variety of materials and forms of construction, with less of an emphasis on using 3D printed materials for all components, for reduced cost and increased reliability / performance. (Steel or aluminium are just better at some jobs than thermo-plastic!)

Some may argue that this somehow "dilutes" the essence of RepRap, but for me, it makes the goal of ownership of an affordable and reliable 3D printer far more attainable, while still allowing freedom for experimentation and development.

Edited 1 time(s). Last edit at 11/01/2015 08:55PM by julianh72.

---

Follow my Mendel Prusa build here: [julianh72.blogspot.com]

The whole "printers printing printers" thing was a nice idea, but it's pretty hard to make a decent printer that way. If your goal is to say your machine is 99% printed, then by all means keep at it. But if your goal is to produce quality 3D prints there aren't many places in a machine where 3D printed parts perform as well as metal parts. When you consider the amount of work that goes into making a printer, wouldn't you want the best result you can get for all your effort? Even if you build a printer with no printed part content, you can always use it to print parts for other machines.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

Of course I want a reliable printer as much as the next guy but I read a lot about high-tech materials that are in chemical composition (composites, nano-fibers infused plastics, etc) and in shape (like I said: micro-lattices, organic hollow core based on the expected loads and forces, etc) and I wonder where would we be if those wouldn't be left for aerospace/military use only. If we set our minds on the comfortable side of the industry innovation will stop. Granted this approach can be as expensive if not more expensive than building with metal mass-manufactured parts, but for those we are just starting to payback the costs of pollution from its manufacturing, costs that were hidden from the final price as the environment took a huge chunck out of that price.

Let's try and take a e step back and discuss possibilities, come up with things to experiment with. Of course first plastic parts used for Prusas and other repraps were not intelligent. The 3D printing has changed a lot in the last years but if we go metal we will never even see the true potential of the 3D printing intelligent plastic components and all the reprap will happen very slowly if at all. And we can always rep-strap it at the begining because some materials are way more effective to be mass-manufactured at this point.

I don't want to talk philosophy only in this thread. Let's try some brainstorming and let's copy from the high-tech industries.

For example we can easily print honeycomb structures. Also truss beams could be 3d printed.

The strongest panels out there are made of a quite rigid honeycomb structure glued to a non-strechable sheet of some sorts for the faces. It is usually aluminum (as in singcore's patented product). What materials can we use for these skins? I don't expect any reprap 3d printed out there to be able to 3d print in carbon fiber like the MarkForged but embedding a criss-cross of carbon fiber filaments in the plastic used for the skin would make it quite non-stretching. What about rep-strapping it by printing the honeycomb panel with a outer skin made of the same material and then gluing or applying some sort of non-stretching tape ontop (aramid maybe)?

---

RepRap Lander concept on Concept Forge
RepRap Lander concept on RepRap Forums
My Things, mostly experimental stuff

Hi guys,

Of course some, like realthor want to use a maximum of plastic on a perfect design.
I don't understand the distinction between Repstrap and Repraps.
To me, metal frames are just threaded rods frames made better.
Repraps were Repstraps since the start.

To me, there's good and bad about plastic parts.
We know acrylic frames are the worst material ever.
It needs to be realy thick (and expensive, obvioulsy) to get the same stiffness than metal.
Otherhand, they surely won't be as flat and as precise than a laser cut metal frame anyways.

For sure we can print tetrahedron comb structures or any complex shape.
But it will probably be thick to get good properties, even with carbon fibre filament.
Probably it will be more expensive and energy consuming than a part out of metal or wood.
I don't believe in a full (or mostly) printed Reprap.
Unless we reinvent physics, plastic cannot be a reasonable choice for a lot of parts,
even some we print today (gears, pulleys, bushings...).
Otherhand, we still have a lot to experiment today, like polymer bearings, and so on.

Actualy, any CNC lathe is much a self replicating machine than any mendel,
because it can make any of it's components, from frame to the last screw.

Quote
Adrian Bowyer
RepRap is about making self-replicating machines, and making them freely available for the benefit of everyone. We are using 3D printing to do this, but if you have other technologies that can copy themselves and that can be made freely available to all, then this is the place for you too.

So hopefuly Repraps are not only about plastic.
IMHO, we should enlarge our vision of the self-replicating process away from materials.
Mostly Repraps are additive 3D printers, but it must envolves or we will reach a dead end.
Stay entrapped into a single technology (FDM) and the Reprap project will die
as soon as there will be any new challenging technology.

We need DLP Repraps for high detail additive parts.
We need Metal Deposition Repraps for special strong parts.
We need Mill/Lathes Repraps for soustractive parts out of unextrudable materials, like wood or steel.
We need Laser/Waterjet Repraps to cut sheet materials
We need Pick&Place Repraps for our electronics.

We should think about Repraps not as a single FDM machine,
but as a machine farm with specialized machines.
Because there's no meaning about spending a lot of energy printing a frame for hours,
when a proper mill or laser can make it in few minutes for a frament of that cost and energy spent.

Don't get me wrong, I understand what realthor wanted to discuss about :
new FDM applications applied to Repraps components.
That's a respectable topic, and "which parts should be printed or not" is another topic.

As some of our parts can be improved, I have some Ideas, but that needs multiple extrusion, for example :

To reduce wear and noise on printed gear, what about composite wheels,
with Igildur/nylon for tooth and any mecanical/hardware contact and PLA/PETC for the body ?

To reduce weight, cost, and allow to make guides out of construction square tubes,
what about polymer bushings with Igildur/PTFE skates, like initiated on the Foldarap ?

Another think. We usualy want a working part just out of the printer.
In our designs, we are far away from what industry already do today.
We forget that assembling different parts produce often better results.
For example a foldable flat part out of the glass, make a nice box with shiny surfaces on any face when assembled.

Mix materials needs to be explored. I had interesting result on printing on a piece of fabric.
The FDM shield and/or add strenght or shape to the fabric which offers some interesting hinge or filter.
For logotype purpose, I printed on a piece of veneer too. I want to try to print over glass or carbon fiber.
We could do much more if we could print from a 5 axis printer around any volume part/material.
A 5 axis printer could reduce the need of using support too. Just like a CNC FDM pen.

Endless discussion, I guess...

++JM

---

^ Things I said. My thingiverse here. My website there (in french).

Good points JM and I don't want you or any others to get me wrong, I don't want to reach a dead end either .

My point is, like you have briefly mentioned, that new materials come out every day, each made for different purposes. I want to apply what the high-tech industries already do with composites to our reprap philosophy. They build lattice panels out of carbon reinforced plastic for some time already. There are nano-cellulose infused plastics coming our way to overcome the heavy toll on the environment of the petrol-based carbon fiber so there will be options. But if nobody prepares for that, even with less good materials, then the trend will be set and everybody will go that way.

Regarding the embodies energy of metal plus the energy spent to machine it I don't think one can even compare with plastics, especially the non-petroleum based ones.

Also I am not a fan of a myriad of machines doing very specialized jobs because you scale that up and you end up at the situation today, the mass production. I believe that we need multi-purpose machines that can do as many tasks as possible to get some raw materials to their final form, all in one pass if possible. Of course that current 3d pritners can't do CNC that will always remain in the realm of heavy duty machining and few people will have CNCs at home. CNC and Laser Sintering and Water Jet Cutting those are for shops/ mini factories. While 5 axis 3D printing, light duty milling, laser cutting and DLP, and pick-and-place for electronics/metal parts/etc will be done by a single machine at home, hopefully one no bigger than a cupboard.

For the purpose of this thread, I hope I made myself understood and that we can continue going forward with the kind of ideas you expressed in the latter part of your post.

Multi-material one-pass gear is a great idea from both noise reduction and strength pov. Also one could replace the worn out crown instead of the whole gear, maybe while the gear remains attached to the part, no need to unscrew everything. Can this be done with dual extruders and with the technology we have today on repraps? Or maybe we can put an already 3D printed part in a DLP machine and have the laser build the perimeter structure around the part with very fine detail and maybe strong (expensive) resins, using much less of them in the process.

I'd like to see electroplating for strength not only for design. 1-5mm of electroplating mass of a hard metal over the contact surfaces would greatly increase the lifespan although maybe it will be more difficult to recycle at the end of the parts life. I think this too is possible without too expensive equipment. It is already done by Stratasys.

I would really like to see how can we print organic shaped "honecomb core" / "corrugated" / "micro-lattice" shells for the exterior surface of the parts (ex: http://wiki.theprovingground.org/april-14th-2010-space-truss). You can't mass produce such things and that's why I believe we need to dive in the details of designing such parts. We can use the process to build the frames of the repraps but the technology is not limited at that. We can't dismiss it as a valid process to build repraps frames just because we can't see other uses for it. What about replacing aluminum extrusions with isotruss structures made of some fiber-reinforced plastic? I don't know if it's possible or cost-effective right now but it will never be if we don't push thinking in that direction.

Edited 2 time(s). Last edit at 11/03/2015 03:56AM by realthor.

---

RepRap Lander concept on Concept Forge
RepRap Lander concept on RepRap Forums
My Things, mostly experimental stuff

Funbot

OK: Challenge Accepted!

I have a concept for a 3D-Printable structural member, suitable for use in making the frame of a RepRap 3D printer (or other application). The typical member would be a 20 mm x 20 mm square hollow section (SHS), which is the same overall dimensions as the HFS5-2020 aluminium extrusion as used in the MendelMax RepRap and similar 3D printers.

Individual members could be printed as long as your 3D Printer will allow, but a Joiner piece could be used to join multiple segments into longer members. You should be able to print a member > 200 mm long by arranging diagonally on a typical 200 mm x 200 mm print bed. Joints could be screw fastened and / or glued. (See attached assembly sketch, which would use 4 x 3 mm bolts at each Joiner.)

The members would be designed to be printed in either horizontal or vertical orientation without support. Some experimentation with both orientations would be needed to see which gives the strongest and most rigid component.(I'm inclined to think printing horizontal would be preferable, as it would give you continuous "fibres" along the long axis of the member, without having weak inter-laminar layers which could separate when the member is stressed in bending etc.)

The joints between segments would probably be the weakest link. Assuming that you can achieve good tight joints, the flexural rigidity (EI) of the printed members (in PLA or ABS) should be approximately half that of an HFS5-2020 aluminium extrusion - but that is still about 5 to 6 times greater than 8 mm steel threaded rods (as used in my 1st-Generation Prusa Mendel). Assuming that you can achieve good tight joints, the flexural rigidity (EI) of the printed members (in PLA or ABS) should be approximately 1/20 that of an HFS5-2020 aluminium extrusion, or about 65% of an 8 mm steel threaded rod (as used in my 1st-Generation Prusa Mendel).
[Edited to correct basic arithmetic error.]

I envisage that 3D-printable corner joints would be used as the nodes for the machine frame.

I imagine someone must have looked at something like this before, but I haven't found anything on the RepRap Wiki or Forums.

Is anyone interested in pursuing this to see if it can work? (Or is this already tried and proven to be an "impossible dream"?)

Edited 2 time(s). Last edit at 11/03/2015 07:04AM by julianh72.

---

Follow my Mendel Prusa build here: [julianh72.blogspot.com]

Have you seen the GUS simpson? no precision linear guides. no belts, everything but the print bed and head is reprapped plastic, and It's precision puts many cartesian bots to shame.

[www.youtube.com]

The nice thing is its precision is based on geometry, linkages to control the motion paths and tensile strength of the fishing line, you are not limited by the quality of linear rail or leadscrew you can get.

EDIT: silly me, i should have recognized your name from the lander project. of course you are familiar with GUS

Edited 1 time(s). Last edit at 11/03/2015 05:19AM by John Meacham.

1) is the square shape the best for hollow members? ALU extrusions used in repraps use the [T|V]-slot profile and I assume that is for two reasons: one is to increase the number of surfaces that can't be bent due to their close proximity to corners, which will increase the flexural rigidity and second is for easy attachements via T-nuts along its length. I am sure that if you can print a square you would be able to as well print a T-slot profile. Note: t-slot has been printed before not sure of the results though (example: [www.thingiverse.com], here's an openSCAD library: [www.thingiverse.com]).

2)

Quote
julianh72
Assuming that you can achieve good tight joints, the flexural rigidity (EI) of the printed members (in PLA or ABS) should be approximately half that of an HFS5-2020 aluminium extrusion - but that is still about 5 to 6 times greater than 8 mm steel threaded rods (as used in my 1st-Generation Prusa Mendel).

How did you get to these numbers? It would be interesting to actually come with such numbers for different profiles (t-slot, round) and compare.

3) Why did you choose singular members for a part that is plastic-made instead of printing a larger part that also includes at least some part of the faces? I mean if you would have a cube where all the edges are aluminum then it would have enough rigidity but with plastic I assume you would have to change the strategy a bit to achieve comparable rigidity and that is by also printing the faces or at least some wings can be printed in the same part. This would change the design from a columnar member design to a corner piece that would have more area for bolting and would also offer a way to attach the face between. Going one-to-one with the metal profile we want to emulate will not work as expected.

My take on individual vertical members would be more of a mixed approach: i would print a truss-negative part, run some pre-creeped Spectra fishing line along the truss' negative channels, tension it with a jig at the bottom end and then put a sleeve on and pour some resin to embed the spectra. The connectors will be also the insert points for the horizontal members, which, depending on the load on them, would have a more busy truss embedded. It can be split in half and printed horizontally if layer failure seems like an issue. The Spectra can be run in the channels after all pieces are connected together to form the required length and the sleeve mustn't be the same length with the final piece, it can be used to pour resin in sections. I made a quick sketch:



I hope we get some input from more experienced (structural?) engineers around here.

EDIT: we can also look at Project Locus for inspiration.

Edited 5 time(s). Last edit at 11/03/2015 06:35AM by realthor.

---

RepRap Lander concept on Concept Forge
RepRap Lander concept on RepRap Forums
My Things, mostly experimental stuff

Well, firstly, I AM an experienced Structural Engineer, so there is in fact some method to my madness!

(However, I am a lowly hobbyist RepRapper, with only one build to my name - A 1st-gen Prusa Mendel - so my method may be very misdirected! )

Secondly, I need to correct a mistake in my previous post: The 20 mm PLA square tube will have a flexural rigidity of about 65% of an 8 mm steel rod, not 5 to 6 times the stiffness. (Decimal place error - Doh! )

While I admire the ingenuity of your concept of a spiral-wound tube, I am a bit puzzled, as it seems to me that it doesn't really get you much closer to your goal of an all-printed machine frame.

Anyway ...

Why a square hollow section? I wanted a profile which was fully printable, in multiple orientations, on pretty well any 3D printer. I didn't NEED T-slot sides for the assembly details I had in mind, and in my experience, sections with thin free-standing walls and edges tend to be a bit tricky to print successfully without stringing and blobbing, etc. (And the projections and overhangs could make it virtually unprintable in any orientation other than vertical - and I suspect that printing horizontally may give a stronger section for bending, to avoid laminations across the member.)

The profile I am looking at is an internally stiffened square hollow tube (very easy to print; quite tricky to extrude in aluminium!), with solid ends for the joints. The internal stiffening should stabilise the walls and allow the member to approach its full theoretical bending strength and stiffness.

Which brings me to:

How to predict the flexural rigidity (EI)?

E is the Elastic Modulus, and is a property of the material. It is typically about 200 GPa for steel, 70 GPa for common aluminium alloys, 3D-printed ABS can approach around 2 GPa, and around 3 GPa or more for PLA, provided your 3D printer can produce good quality prints. (E.g. see [www.appropedia.org] )

I is the "Second Moment of Inertia" of the cross-section, and can be calculated by simple formulae for common shapes, or slightly more complex methods for complex shapes.

For an 8 mm steel rod: E = 200 GPa; I = 201 mm⁴; so the Flexural Rigidity is given by EI = 40 N.m²
For an HSF5-2020 aluminium extrusion: E = 70 GPa; I = 7420 mm⁴; EI = 519 N.m² (12 times the rigidity of an 8 mm steel rod)
For my 20 mnm x 20 mm PLA SHS: E = 3.3 GPa; I = 7870 mm⁴; EI = 26 N.m² (65% of the 8 mm steel rod)

OK - we might need a somewhat bigger cross-section to get the required frame rigidity, but maybe we're on the right page?

My profile is NOT a "one-to-one" replacement for the aluminium extrusion, because the forming process is different (3D-printed vs extruded), and the end-use and jointing methods etc are different. The main function of the internal stiffeners is to stabilise the side walls, which I think answers your question about how I arrived at my section design. On rechecking my maths, I realise maybe we should go a bit bigger than 20 mm SHS if we are seeking to emulate the rigidity of a MendelMax, but because I goes up to the 4th power of the section dimensions, a relatively modest increase in tube size can yield a big increase in rigidity.

Anyway, I'm printing a couple of samples now, and I'll see how they go together (although I don't have the time or resources to make an entire 3D printer).

If you want to look at my native design files, they are a shared "work in progress" on my OnShape account at [www.onshape.com] - just search for "RepRappable Structural System", and you can make a copy, edit the design parameters, etc. I will post the files on Thingiverse when the design is a bit more "fleshed out".

---

Follow my Mendel Prusa build here: [julianh72.blogspot.com]

Hi guys,

To me the Locus project is the perfect example you don't want to follow.
I've printed a small piece of 2060 extrusion to check other parts fitting, I'll post a picture later.
For many reasons, the part was definitely not good enough, except for testing purpose.
It took about 20h of printing time for a single 150mm piece.

We should stay pragmatic.
As a meter (36") of 2020 extrusion cost less than 3.20€ ($3.40) for a perfect thing.
Spending time, money and energy to print extrusions
you need to joint with glue and a threaded rod,
and feel that with PU foam is a pure nosense.
Definitely it takes longer and cost more.

Let's be honnest, trying to reproduce existing working materials like sheets or extrusions is just stupid.
If there's a reality about printing some structural stuff, it's obviously anything different.
His structure should be advantageously replaced by PVC tubing and any central tensioning system.
IMHO we don't want to reinvent the wheel or the aluminium profile, it will be obviously worst.
Let's think about things that already working with plastics : clip-on stuff or things like that.
Small parts, that prints quick and easy, that clips together to make a reliable structure.
Something based on connectable triangles mabe.

++JM

Edited 1 time(s). Last edit at 11/03/2015 07:36AM by J-Max.

Since were getting all silly about plastic frames here, has anyone considered or used common PVC water pipe for a 3D printer frame? 2" schedule 40 pipe is pretty rigid in short pieces... It's available everywhere, cheap, easy to cut/drill/mill, and fittings like tees, crosses, and elbows are as available as the pipe itself. You can build a frame then 3D print fittings to attach guidde rails, motors, etc.

For that matter, large diameter pipe could be used as the frame/enclosure with the printer being built inside a piece of it. PVC pipe comes in 24" diameter- big enough to make a pretty good sized printer- a delta could be made by attaching linear guides to the inside walls, and a door could be cut into the side for accessing the print. The trick would be figuring out a way to align the rails parallel to each other and set 120 degrees apart.

For that matter, any kind of large diameter tube, such as a concrete form, could be used the same way. A 16" diameter x 4' long sonotube costs about $20.

I have two cats, so I also have a bunch of plastic buckets from litter box sand that are about the right size and pretty rigid- such a bucket could be used as the enclosure/frame for a 3D printer. It would be at least as rigid as the best i3 printer I've ever seen.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

I like the large tube idea.

PVC pipe/connectors are already there in the wild: http://www.thingiverse.com/thing:932732.
Also: http://reprap.org/wiki/Easy_build_delta_printer.
Also original reprap Morgan is PVC pipes based. One example: http://letsmakerobots.com/blog/hoff70/morgan-3d-printer-adventure
Also reprap Lisa by Nicholas Seward has been consdtructed with thick PVC pipe: https://www.youtube.com/watch?v=N7XVxtgHHhY

Smartfriendz and others used to use wood panels for the frame but that was affected in time by humidity and also was too resonating and amplified the noises inside the box.

I don't think the simple shape of the pipe confers it enough rigidity for a quite high-vibration, high-jerk application like 3D printers. Unless you go overkill with the thickness. Also not reprap (yes, I am annoying with that). Still a vitamin by reprap standards.

Regarding concrete pipes/tubes, I've tought about that some time ago but I don't think somebody will put such tube in an apartment.

Edited 1 time(s). Last edit at 11/03/2015 08:44AM by realthor.

---

RepRap Lander concept on Concept Forge
RepRap Lander concept on RepRap Forums
My Things, mostly experimental stuff

I've wondered about combining structural elements printed in different orientations to maximise strength. I could imagine that nested square tubes might work quite well, perhaps with an outer printed vertically and two inners printed horizontally and rotated 90 degrees to one another. To get the maximum benefit you would probably need to bond them together - perhaps printed in abs and bonded with ABS juice would work?

Hi guys,

I agree, PVC tubes are just vitamins.
I dunno how it is in USA or abroad, but here in EU,
It's not realy cheaper than small aluminium extrusions.

The stack of 4 tubes as PVC frame make me smile.
As I said earlyer, you can have a pretty rigid structure
just by adding a tensioning thing inside the tube.
When any tube is in compression it makes a pretty stiff ensemble.
So you can get good results with small tubes.
It works with light aluminium profiles too.

I though again about small clippable triangles, and found that :

It's called the Polydron. With that kind of parts, you should be able to build any frame or shape.
If the clip is strong enough and the shape stiff enough, it should be suitable for Reprap building.
It just need a hole system at the centre to fix some hardware.

++JM

Edited 4 time(s). Last edit at 11/03/2015 09:52AM by J-Max.

Quote
J-Max
I though again about small clippable triangles, and found that :

Polydron_09506.jpg

It's called the Polydron. With that kind of parts, you should be able to build any frame or shape.
If the clip is strong enough and the shape stiff enough, it should be suitable for Reprap building.
It just need a hole system at the centre to fix some hardware.
++JM

I like this idea but just don't know what stiffness can be achieved with so many parts. There will be play in all those hinges. Maybe we should aim for a cylindrical final shape that can be fastened via a mechanism like the hose clamp.

I have just found this Harward study for roofing systems that uses lego-like modules, it may be of some help: http://www.3ders.org/articles/20141111-harvard-graduates-replace-corrugated-iron-roofs-with-3d-printed-plastic.html

About everything else I can find online talks either about cellular structures (aka honeycomb core, micro-lattice, corrugated) or some form of truss structure. The usual injection molding type of making plastic containers and large parts sturdy is by using ribs but we shouldn't try to copy that otherwise there would be no added benefit of the complexity we can achieve with 3D printing. Injection molding is best at what it can do.

I believe we should really take advantage of the celullar structures, especially the micro-lattice variety Injection molding can't do closed cellular chambers while 3D printing can. 3D printing can also vary the sizes or the concentration of the cellular chambers based on the expected stress (for example the Voronoi lattice). I think that slicers should include such algorythms as to transform the surface of the object in two separate surfaces linked by a cellular/lattice network of spikes which would maximize their strength.

Here's an image of such a lattice cylinder:


Edited 5 time(s). Last edit at 11/03/2015 03:24PM by realthor.

---

RepRap Lander concept on Concept Forge
RepRap Lander concept on RepRap Forums
My Things, mostly experimental stuff

or create the latice in the 3D software

http://www.autodesk.com/products/within/overview

Edited 1 time(s). Last edit at 11/03/2015 08:56PM by MechaBits.

All those structures are for reducing weight and raw materials. For a printer of modest size I don't think weight or material costs are the primary issue.

For most typical components, most of the strength and stiffness is contributed by the outer "skin", and the interior will function perfectly well as an open honeycomb or similar. It's main function is to provide stability to the outer load-bearing shell, preventing it from buckling under load. In my design concept for a RepRappable structural strut, I used 20% honeycomb fill for the two "solid" ends (where concentrated loads at the bolts etc will be highest), and used a much lower density thin-wall "diaphragm fill" for the main shaft. I haven't done any real testing yet to see whether my internal diaphragms are "optimised", but they use a lot less material than using a honeycomb fill for the whole part, and it prints a lot faster as well.

Where you are using honeycomb filled parts, you can easily increase the strength, stiffness and durability by increasing the thickness of the outer "shell" (increase the number of perimeters and top / bottom layers), and I find the default Slic3r "honeycomb" fill at 20% works fine for the parts I am making.

This is not to say that we should not pursue more optimal ways of filling the cores of our designs (such as optimised variable-density honeycombs, etc), but for me, the default 20% honeycomb fill works fine. Don't forget that optimising the fill would require an iterative design / analyse / modify / repeat cycle - and I suspect that the vast majority of RepRapped parts never see even a basic stress analysis, so what would be the basis of optimisation for these cases?

---

Follow my Mendel Prusa build here: [julianh72.blogspot.com]

I Can't wait to get going with honeycomb infills, but wish there was a way to shift them on a layer by layer basis, rather than keeping continuous, I have to learn how to make sure those filaments get smooshed into the perimeters to make sure they are connected.
But an optimized voronii infill would also be cool.

Quote
MechaBits
I Can't wait to get going with honeycomb infills, but wish there was a way to shift them on a layer by layer basis, rather than keeping continuous,

Why? Having a vertically-stacked honeycomb infill provides really good mechanical strength. What would be the advantage of random (or leaning) honeycomb infill layers?

If you want something that seems more "random", Slic3r also offers other infill patterns such as Hilbert Curves and Archimedean Chords.

---

Follow my Mendel Prusa build here: [julianh72.blogspot.com]

I dont want sloping, or random, either shifted by half the hexagon to lines all line up still, or half scale on alternating layers, depends if using it for decorative or flexible material uses, maybe not on a layer by layer, maybe every few mm, I have a lot of experimenting to come.
Why, because I hoped it might solve some of the issues from poorly chosen infills, when people are pushed to save time on machine,
Some of the parts I had done, I wished had been solid, only by having my own machine can i decide to try for longer print times.

Edited 2 time(s). Last edit at 11/04/2015 02:33AM by MechaBits.

Hi guys,

We can already make tubings and/or block structures to make lego alike assemblies. Fastened or glued.
Lattice structures are what we already do in our prints, whatever the infill we use.
Micro lattice depends only on the resolution, and we will be far away of what can achieve a laser printer.

Triangular structures are known to be the strongest structures in architecture.
I believe we can make hinges with no play except on the axis of the hinge.
And that's the more important, because when you joint different triangles together,
the next triangle will constraint the last hinge.
If your structure is a loop of triangles with no holes, the shape is perfectly constraint.

We could make some triangulated beam structure adapted to a single printer. We can already do that.


It should be at last better than any chinese acrylic frame.
But will it be realy innovative facing the existing Holliger and Tantillus Repraps ?
Just awnser some questions :
Why those Repraps did not succeed ?
And why the Prusa I3 and the Kossel did ?

I believe the key is to find a polyvalent shape, quick and easy to be printed,
which can be assembled in any direction (or preferabily in selected fixed angles),
and which can hold anything you need to fix on it. Kinda "open shape".

The more simple the shape, the more versatile it will be, allowing to build any printer's architecture,
from single arm to gantry cartesian, to delta, to CoreXY.
But anyways, even if the shape can be printed, it will be cheaper in injection as soon as a mold is built.

++JM

Edited 3 time(s). Last edit at 11/04/2015 04:26AM by J-Max.

While I have studied triangle truss in the past, I don't think it is suitable for plastic. Not even the fiber reinforced one. The specifics of the polymer structures are better suited to large surfaces that's why you will mostly see Polycarbonate and PolyPropylene honeycomb core or corrugated panels throughout the construction and architecture industry. The skins are more problematic at those because plastic skin is not that good, they use some form of metal film or aramid/kevlar/carbon fiber sheeting.

For me another important aspect of cellular parts is their thermal conductivity, which is suited to insulate the printing chamber from the exterior. An enclosed printer with a heated bed would have an inside temp while running below 70deg so having an enclosure that is cellular in the walls is a plus. That's why I, like MechaBits, would like the honeycomb infill to have some flexibility, in my case I would like the capability to create closed hexagonal chambers (at least by creating another skin every X mm in height but I'd like those to be random for each hexagonal tube, not all at the same height) rather than continuous tubes.

Something like this:



Regarding the whole structure pf Polydron-style hinged panels, it has to be squeezed after being put together, to eliminate all play, I am thinking at a combination of a band-clamp style for the large spans on the sides and a quick-release plastic clamp at the corners or some sort of lugguage-style clamp that would tension the whole assembly together:



I believe that we should take advantage of the whole size of the bed when printing the individual pieces and this way we would end up with fewer pieces that can envelop a large space.
I have found a square polydron design already on thingiverse: http://www.thingiverse.com/thing:186440

Regarding the injection molding of a standardized way of doing reprap enclosures I aggree it wil be much more economic and fast but until such a time we have to come up with a solution for our small runs, individual needs.

One more thing: isn't the polydron patented like the Lego?

Edited 1 time(s). Last edit at 11/04/2015 06:32AM by realthor.

---

RepRap Lander concept on Concept Forge
RepRap Lander concept on RepRap Forums
My Things, mostly experimental stuff

Ok, found some more inspiration ideas:

* The 3-Way joint: http://www.thingiverse.com/thing:1085447



Maybe we can apply this to whole square 3d-printed panels, will have to think about it but now I am in brainstorming spree, searching ideas on the web like a mad man.
Even introducing a snap feature at the 3-way joint? Having the panels snap-in the next from different sides, alternating from the inside, then from the outside, then have panels that snap-in from below or above...

* WenTAI Snap Boxes: [www.5thpostulate.com] . These are better suited for injection-molding though. I would print closed such boxes with the cellular honeycomb fill on the inside.



* Togetherfarm Blocks are basically Lego: http://togetherfarm.com/what-are-togetherfarm-blocks/



* Could Simpliframe be of any help? https://www.youtube.com/watch?v=W63rXTSYOhQ. Mybe not but then again you might see it differently.

* Roll-to-Shape angled modules that when rolled in one direction could snap together. Inspiration: http://gizmodo.com/5195275/roll-up-chair-takes-ikea-to-the-woods



Edited 8 time(s). Last edit at 11/04/2015 08:21AM by realthor.

---

RepRap Lander concept on Concept Forge
RepRap Lander concept on RepRap Forums
My Things, mostly experimental stuff

Quote
J-Max
It should be at last better than any chinese acrylic frame.
But will it be realy innovative facing the existing Holliger and Tantillus Repraps ?
Just awnser some questions :
Why those Repraps did not succeed ?
And why the Prusa I3 and the Kossel did ?
++JM

People don't buy printers based on any engineering principles. If they did they'd be looking for stable mechanical structure from the start. They buy printers based on price, and fantasies about maximum print size, as promised (but usually not delivered) by the bed size.

Those didn't succeed because 1) no one was offering kits and people are too lazy to source the parts themselves, 2) they had small bed sizes and people want bigger beds because they want to make bigger prints.

The i3 succeeded in spite of its poor design because it had a $300 price tag and a bigger bed (never mind if you can't print anywhere but the center of it).
The kossel succeeded because some people like deltas and could by kits of ready-made (printed-ugh!) parts cheaply.

I think it is pointless to print the frame structure when structural materials are widely available and far exceed the performance of anything that will ever be printable with plastic. Insisting on printing the frame of the machine dooms it to mediocre print quality. If the goal is to make something that squirts out plastic in the rough shape of Yoda or a tug boat, keep fantasizing about printing the machine's frame, and keep producing mediocre printers.

What's next, printing the bed plate?

Edited 1 time(s). Last edit at 11/04/2015 08:12AM by the_digital_dentist.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

Quote
the_digital_dentist
I think it is pointless to print the frame structure when structural materials are widely available

3D printers will be soon readily available using mass production and economies of scale and proprietary solutions and patented parts. People will buy those and there will be nothing different then the situation from which reprap ingnited. With that in mind reprap will be pointless at that point but for the people that will stick to the philosophy based on other aspects like ethics, morale and de-centralization, empowering the maker and the communities. At least this is what I am sticking to. And this is what this thread is about.

I kind of aggree on the readily availability of structural materials and for convenience and speed one could go buy some panels, cut them to size and maybe incorporate them into 3D printed frames that snap together? But then again who makes those panels? Are they made locally? Are they open source or are they patented?

I am talking about more than costs and time here. I don't want to convince anybody to aggree with me but for the ones that do, let's try to push this to a possible solution. There are plenty of threads on reprap forums about whatever one can fancy/aggree with.

So, to sum up what we have so far in terms of ideas already presented, which ones have gained some aggreement? I wish I could make a poll of some sorts to record votes for the different ideas.

Edited 2 time(s). Last edit at 11/04/2015 08:47AM by realthor.

---

RepRap Lander concept on Concept Forge
RepRap Lander concept on RepRap Forums
My Things, mostly experimental stuff

Hi guys,

The Kawai Tsugite 3 way joint is known to be weak :
[www.youtube.com]

The simpliframe is not far away to my "simple shape connectable" thought.
Plydron is patented, but it's just a base for the start.
Its hinges do not fit our requirements and it's definitely too thin.
IMHO we must stick to a very basic principle with minimum number of parts for assembly.
The quick is interseting only if you need to unfold or dismantle often.
Plastic parts and a superglue bottle can be enough.

As I said at the start, I'm still dubitative with the need printing some parts
that can be easily replacable by a block of wood or any sheet material.
So I agree with some of The Digital Dentist's words.
Even with 2020 extrusions, I consider the printed corner parts to be the weak points of the frame.

I think plastics or mix material's applications are wisely more interesting.
In other words, what unusual applications can we develop for our needs and our machines ?
I don't think working on frames can be realy usefull.
But bearings, springs, foldability stuff, connectors, mecanicals, can be more usable.
And also materials we can print on for further applications.

For example, Instead of printing honneycomb pannels,
can't we print on foam to make usable and connectable pannels at low cost ?

++JM

Edited 1 time(s). Last edit at 11/04/2015 09:26AM by J-Max.