An idea for increasing the Z-axis strength of printed parts

Hmm... not quite sure where to put this... so if a moderator wants to move it to a different forum, please feel free.

Anyway, I had an idea; and it might not be original, but as I've never seen it done before, I wanted to put it out there for people to kick around.

My understanding so far is that extruded plastic is stronger along the extrusion then it is between the printed layers; especially with fiber-fortified plastic. So, in order to boost the z-axis strength of a printed object, I had the idea of creating vertical circular holes (i.e, hollow columns that are the same diameter as the extruder nozzle, or a little bit smaller) in the object that span some number of layers, and then coming back and extruding plastic down the holes until it fills up the hole. This vertical section of plastic would then act a bit like a bolt or rivet, re-enforcing the object along the Z axis.

I imagine these vertical columns might be best placed at the vertexes of a honeycomb infill pattern, and along the edges of vertical perimeters; and staggered vertically. To allow air to escape it may be necessary to leave the sides off of the bottom layer of a column.

I figure that a larger nozzle (and perhaps an E3D Volcano style setup) might work better for this due to the increased thermal mass of the plastic that's being injected into the column. The number of layers that a single column could span would probably have to be limited as the plastic will cool off as it approaches the bottom of the column, but the exact number that is optimal will probably depend on a number of factors, including the plastic being used, what if any fillers are in it, the extruder setup, the layer height, bed and/or build volume temperature, etc. It may be beneficial to kick the head temperature up a few degrees before performing the column filling as well.

I'll try to work up a 3D illustration showing what I mean, but it might take me a while - I've got a few other things I've gotta monkey about with right now.


So... what do ya' all think? Genius? Idiocy? Done before? Somewhere in the middle?

Edited 1 time(s). Last edit at 02/24/2016 09:55PM by bdurbrow.

With my old makerbot I had a lot of issues with delamination... and I remember reading that solution somewhere but I ended creating the pieces in a way that the strength requirements are higher in the x y axis.

Now I noticed that the pieces printed by my prusa are much stronger, made me think about increasing the temperature of the makerbot, and that helped a lot!

---

My printers:
-Makerbot TOM (#5215, circa 2011), MK6 extruder, ABS 3 mm
-HICTOP Prusa i3 (modded for auto-level, thread screws), ABS/PLA 1.75 mm

About me:
[www.thingiverse.com]

The process you describe is very much the same as injection molding. I fear you will encounter that 3D printing extruders will not be able to push enough material out fast enough to fill up the hole with a usefull plug of plastic. The material will cool down way to fast and instead of a solid block you will end up with a coil of loosely connected strands that is even less stable than the normal printed parts.
Regarding strength, my experience with PC is that it has such a strong layer bonding that it usually doesn't make a difference, the parts i tested didn't break along the layers. To a lesser degree i encountered this with PETG as well.

---

[www.bonkers.de]
[merlin-hotend.de]
[www.hackerspace-ffm.de]

I make a lot of ABS parts and I always worried about the strength of layer adhesion until I got an all metal hotend and I could crank up the temperature. The difference in layer adhesion from 240c to 245c is amazing. I don't have a full scientific way to measure but I did print some test parts that I could hook a fish scale to and at 240c I could break the layer between bonds at around 25lbs of force yet when I printed the same item at 245c I was not able to break the part using the fish scale and it goes up to 70lbs pull.

Now while your method may sound proper one main issue will prevent it from working and that is that the deeper layers will be cooled down so injecting the hot plastic is likely to not stick to the cool layers hardly at all. You need both plastics to be hot to give a good bond. You could test this theory easily by taking an old print and allowing the hotend to drizzle plastic onto it without the hotend touching in. My guess is that once cooled you can fairly easily remove the new plastic.

I believe, the main obstacle in your plan is to print such small holes. My printers create a blobby mess with holes smaller than 2mm dia.

Well, when I get a chance I'll monkey around with it. An idea I had while thinking about the temperature issue was to have a separate extruder with a long tubular nozzle on it (well, perhaps 6 to 10 mm in length), to go down into the hole and apply hot material directly to the bottom, perhaps with the machine making a Z-axis move while applying the material. The nozzle could have a resistive coating on the outside of it to heat it up; with appropriate selection of resistive material it could double as it's own temperature sensor. The metal inner core could serve as the current path down to the far end of the resistive coating (with an electrically insulating layer in between the majority of the metal core and the resistive layer), and the current returning at the top of the resistive area to a brazed-on wire (or vice-versa - the resistor does not are about polarity, after all!). DIY options for that might include coating it with a vitreous glaze (basically ground-up glass), which by itself is an insulator, but could perhaps be mixed with graphite powder to make it somewhat conductive.... again, I'll have to monkey about with it. The glaze could possibly be fired by a propane torch. Materials for this kind of process are available from many art supply stores and online resellers. Other materials may be possible as well.

Pretty good idea, but such a long nozzle would lead to oozing.

Why not make the whole part hollow and fill it with 2k-resin afterwards? Would also print much faster.

Quote
o_lampe
Why not make the whole part hollow and fill it with 2k-resin afterwards? Would also print much faster.

Total cost of material used and weight are the primary reasons; process automation coming in a distant third. For some things, however, I have considered doing something like that; using what's known as "epoxy concrete" or "epoxy granite" - basically, epoxy filled with rock flour, sand, and rock aggregate up to about 6mm in size (or, alternatively, 3D printing a mould, smoothing it, and casting the item in epoxy granite).

My thoughts about this are primarily centered around fiber-reinforced material; where it's already a structurally anisotropic material owing to the fibers tending to align along the extrusion axis.

Oozing might be dealt with by a: turning off the heat on the extended end when the nozzle is not in use (I imagine that that small of an object won't have that much thermal mass, so the plastic will likely freeze in in the tube while the rest of the next set of layers is printing; and likewise will come back up to temperature rather quickly) and b: snapping a cover over it when it's not in use (a small RC servo might be a good way to accomplish that).

or you could just add some 3mm holes, and pass some little bolts through, obviously not as good as direct bonding and placement might interfere with fit. Or another head with some 'super glue' by super glue I mean a material that would aid bonding between layers...it might have to be developed or may already exist? Or a focused laser to fuse layers, or UV mortar.

Edited 2 time(s). Last edit at 02/27/2016 06:28AM by MechaBits.

Printing ABS @265C with a E3D volcano 1.2mm nozzle will have no issues with interlayer bonding if you print with 100% infill the part will be very strong. The blog post [e3d-online.com] states

Quote

the tabs on the part printed with a standard 0.40mm nozzle could be broken by hand if you really try. However the same features printed with Volcano were impossible to break with bare hands, and we really tried! Even with pliers we struggled to delaminate the layers.

I thought that that was just marketing blabla until I printed my first part with the volcano and tried to break it myself.

Edited 1 time(s). Last edit at 02/27/2016 06:27AM by Frans@France.

So what are the layer heights on the 1.2mm nozzle, I bet print speeds quite nippy.

Quote
MechaBits
So what are the layer heights on the 1.2mm nozzle, I bet print speeds quite nippy.

I did a 0.8 layer, speed was hard to tune, the filament is fed fast so there isn't much time to heat up to a good print temperature. As I said here [forums.reprap.org] "I might be able to push the volcano to 30mm/s, I failed at 35mm/s but it might be that replacing the Wade extruder for a full metal extruder would allow for some more pressure and thus have less issues with the lower plastic extrusion temperature."

aah..forgot about the details, when your talking 30mm is that just for extruder or is it linked to axis motion.

You would be better off just running gorilla glue down a hole.

Be careful though as it expands and could rupture the print.

Edited 1 time(s). Last edit at 02/27/2016 11:16AM by DRobs86.

I like the idea of dipping honeycombs in resin(a sweet solution), it has a certain appeal.

The trouble with bolts is that for a large print, there could be thousands of them; so placement would have to be automated - possible, but it leads to a complicated head and feeder design. To get good adhesion to the bolt though, just heat it before driving it in.

Quote
MechaBits
I like the idea of dipping honeycombs in resin(a sweet solution), it has a certain appeal.

Unfortunately, it also tends to attract bees...

So maybe this "Hybrid 3D printer: FDM + Polyjet" will be an idea? [forums.reprap.org] 3D print with holes and then use the motor driven syringe to deliver a "filler" to the holes. Using 2 part poxy and a mixing nozzle would do the trick.

That is a possibility... my sticking point with two part epoxies though is that it sets up in the nozzle after the print is done... requiring a disposable nozzle; or some sort of solvent flush system. Not an insurmountable obstacle... but definitely something to be aware of before heading in that direction. They tend to also be somewhat expensive...

It's too bad that photoset resins aren't more readily available... that would be perfect for this sort of thing.

For the heated nozzle, another possibility might be using a high temperature paint over the metal nozzle body; for example Rust-oleum Automotive High Heat claims to handle 2000F; and based on the MSDS the white color looks like it should be an electrical insulator; while the black one contains carbon black and might be a suitable resistor material (and if it's not out of the can, it could probably be fortified with graphite powder to make it one).

What about a specially designed first layer that would allow slow drain of acetone that is poured in the vertical holes after the print is done and it takes enough time in contact with the wall to dissolve it so that you get a better fusing between layers? You could use a syringe or some sort of pump and a tool head that goes to every hole at the end of the print and injects the acetone in the hole. If holes end on the model before the model is finished printing on the vertical, then a second nozzle of sorts on the same printhead could be employed, while the print is paused, to inject acetone and then let it drain, while the FFF head continues printing.

This way even slanted tubes can be designed in the model, or a script can be developed to post-process a model and insert slightly bent vertical tubular holes or at an angle, to better assist with the strengthening of the part.

This could be messy but it could be worth it. Enclosed printer with filters would be a good option.

I don't know of PLA and safe(r) solvents. This could be quite unhealthy overall, but with the proper precautions it could be done.

On another hand, for small pieces that you'd like to touch up after they are finished (manual process), the friction welding could be used for filament-sized vertical holes. I don't know about a way to automate this, even if you had a dremel head, it still has to have a filament advancing mechanism and a cutter so it could rotate the pieces.

Edited 1 time(s). Last edit at 02/29/2016 04:55AM by realthor.

---

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

Aceton vapor would be dangerous. But I like the idea about using welding. Why not ultrasonic weld the layers? No need for holes but just an extra "extruder" that welds at certain point the layers.

Quote
bdurbrow
That is a possibility... my sticking point with two part epoxies though is that it sets up in the nozzle after the print is done... requiring a disposable nozzle; or some sort of solvent flush system. Not an insurmountable obstacle... but definitely something to be aware of before heading in that direction. They tend to also be somewhat expensive...

there are some single component epoxies but those are normally much more expensive then the 2 parts. Mixing nozzles will indeed cost around usd1

Quote
Frans@France
Aceton vapor would be dangerous. But I like the idea about using welding. Why not ultrasonic weld the layers? No need for holes but just an extra "extruder" that welds at certain point the layers.

That's why I mentioned an enclosed, filtered setup. It should be standard for ABS anyway, I don't know which is more dangerous, Acetone or ABS particulate...

Friction Welding can only be done with shallow parts, maybe very shallow only
I thought about that but I assume the ultrasonic welding head itself will be a fortune. If that is not an issue for this particular application, then a roller toolhead following the nozzle while ultrasonic vibrating could be a solution. But I would assume it to be quite expensive.

From my POV the best solution would be a non-toxic solvent in the above acetone-to-tubes setup.

If not possible, then injecting a resin of sorts at the end of the print would be the only solution. But combining materials in this way is not very wise, not in an era where cradle-to-cradle design is a concept that might solve many problems with recycling and landfills and pollution overall, while preserving the resources. It makes such "composites" impossible to recycle. Even ift he resin and the plastic would be both biodegradable, that is an ultimate solution, when you decide that the recycling has degraded the material in such a way that wasting the embedded energy is an "acceptable" loss.

Edit: we have more and more laser engravers around so what about having the laser follow the nozzle and focused at the seam between the previous layers. This can be experimented with but it's definitely doable. A combination of incident angle, power and focus should beautifully fuse the layers.

Edited 1 time(s). Last edit at 02/29/2016 07:03AM by realthor.

---

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

As solvents go, acetone is one of the more innocuous in terms of toxicity... your body makes it, and has metabolic pathways for getting rid of it. Nonetheless, I still wear proper protective equipment when handling it.

Physically, however, it's a different story: that stuff likes to burn... it's not something I'd like to have in close proximity to a hot print head, or anything that might generate a spark (connectors and relays come to mind).

Quote

I don't know which is more dangerous, Acetone or ABS particulate...

Well, do you want to die in a fire now, or of cancer later?

---

The parts for an ultrasonic head can be had inexpensively off of eBay; I actually have some here on my workbench. My intent was to ultrasonically boost the extrusion head; hoping that the extra energy would help promote layer adhesion. The difficulty I'm having with the design is coming up with a horn that will propagate ultrasonic energy while preventing the transducer material from getting overheated - piezoelectric materials really don't like getting hot; the manufacturer I bought the transducer discs from suggested 80C as the maximum that it could tolerate in operation. I also suspect that the ceramic cartridge heaters might not be too happy being subjected to ultrasonic vibration; something more forgiving may be a requirement for such an extrusion head. Also, the hot end parts will need to be brazed together, as ultrasonics tends to loosen threaded connections unless they are glued down (and typical thread lockers that would work just fine on the cold end of an ultrasonic head will break down on the hot end - in fact, that's how you remove red Loctite; you get it over 287C before applying torque). Home Depot however, does carry an aluminum brazing rod that should do the trick...

Another possibility for adding extra energy to the print area is to adapt a hot air gun, such as is commonly used for BGA rework (if you've ever had a game console die on you, and had somebody "re-ball" the GPU, that's what they were doing to get the GPU off of the motherboard). Parts for these are readily available on eBay; and could be integrated into a 3D-printed housing designed to attach to the extrusion head. A custom nozzle would need to be fabricated to direct the hot air around the extruder's nozzle, but that shouldn't be too hard - a few minutes with some tin snips and some needle-nose pliers should do the trick. They are also quite lightweight, so adding that to the print head shouldn't require a major mechanical upgrade.

What neither an ultrasonic head nor a hot air setup will do, however, is tackle the issue of the anisotropy of fiber-filled materials; for that you pretty much need to extrude while moving up in addition to laying down material in the x-y plane.

---

As for recycling composites; I would suggest considering a pyrolysis process; like this:

[en.wikipedia.org]

Discover magazine had several articles about the process, the first being from 2003:

[discovermagazine.com]

Once converted back to oil, reforming it into new plastic should be rather straightforward (I am not, however, an organic chemist, so my knowledge of the subject is somewhat limited).

Quote
bdurbrow

Quote

I don't know which is more dangerous, Acetone or ABS particulate...

Well, do you want to die in a fire now, or of cancer later?

Exactly my point: both are so dangerous you don't really want to play with that.

---

Building an ultrasonic head adjacent to the hotend(s), just like the induction probe, might avoid some of the problems you describe. However I don't know about the vibrations being transmitted through the plastic carriage and the liniar rods/whatever. If you say it's inexpensive I believe you.

About another heater blowing on the layers...hmmm, isn't the layer fan actually supposed to coll those off?

I still think the laser could be used when not engraving/cutting paper/playing with the cat/etc.

---

Quote
bdurbrow
As for recycling composites; I would suggest considering a pyrolysis process; like this:

[en.wikipedia.org]

Discover magazine had several articles about the process, the first being from 2003:

[discovermagazine.com]

Once converted back to oil, reforming it into new plastic should be rather straightforward (I am not, however, an organic chemist, so my knowledge of the subject is somewhat limited).

Pyrolysis can hardly be called recycling. It just pumps more energy to the already embedded energy from creating the filament/resin+ making an object out of it and so on.
Other processes are not widely used and until they do, we're still thrashing the earth.

---

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

Quote
bdurbrow
Physically, however, it's a different story: that stuff likes to burn...

exactly and heavier then air so it will build up on grond level making a spectacular event when ignited by some fireplace.

At the risk of veering off-topic, I'll just say that the pyrolysis process can be quite eco-friendly if it's fed with a clean energy source... like solar, for example. And because what you are trying to generate is heat; you don't have to go thru an expensive and low-efficiency photovoltaic step... some mirrors and a black paint job on the reactor cell will do the trick just fine.

But whatever you do, DON"T USE COAL!!!