Thread Tools
Oct 07, 2017, 08:53 AM
Registered User
Thread OP
Discussion

Could foam core be replaced by a 3D printed "square" comb profile ?


This morning woke-up thinking of 3D printed molds. Maybe would have been better to stay asleep but...

This looked to me interesting to investigate as I have the printer and so machine time is not an issue, material is cheap, accuracy can be not bad...

Searching the net I fould at least 2 references about composite molding and 3D printing:
http://web.ornl.gov/sci/manufacturin...D-Printing.pdf
https://e3d-online.com/blog/2017/03/...fold-moulding/

The 3D printed mold would be used in conjunction with mylars, to keep the smoot finish.

But molds can be useless if I don't have the foam core that goes with it. Could it work to replace the foam core by a 3D printed honeycomb printed to the airfloil shape (that could come with a 3D printed spar or reservation for a carbon tube)?

Just to tinker, I looked at the weight it could achieve. I arrived to 12g when the foam equivalent is 7.5g. Not a show stopper.

I tweaked the slicer to print my panel without walls, to have only the infill, at 5%.

Here is the resulting proof of concept. I believe that the fact I printed only the result is responsible of part of the oozing. I also had to print supports which went out better than I expected, but not perfect.

Did somebody investigated this direction? Any hope ?

JMF
Sign up now
to remove ads between posts
Oct 07, 2017, 01:02 PM
Registered User
Ward Hagaman's Avatar
Very interesting! How well does epoxy bond to the plastic?

My gut tells me that you would need a mold though...I can picture the mylar pressing in between the honeycomb cells.
Oct 08, 2017, 12:56 AM
Registered User
Thread OP
Yes, the bits on the table would be ideally:
- a 3D printed mold fir the top part of the wing,
- using mylars to cope with the 0.1 to 0.2mm steps in the 3D printing
- be able to rely on the tooling I have, see below

My thinking aims toward being able to build an ALES composite wing, on par with vacuum baged foam core wing as per weight and strengh, with the tools I have:
- (cheap) 3D printer,
- access to laser cutter,
- no hot wire stuff (as I don't have)
- no CNC

I would like to be a bit more acurate than plain vanilla vacuum bagging, at the expense of a bit increased cost.

Possibility to reduce the hand on working time (the real life time you need to build the thing, not the time the 3D printer works for you).

JMF
Oct 09, 2017, 03:28 AM
Registered User
We have some interests in common.
I started working on 3d printed wings last year (my "big project" is SynerJ 4m in 3 technologies: composite, balsa and 3D printed).
After I compute the 3D model of the wing I can create 3 types of output: stl files for molds / plugs / cores, dxf files for lasercut and st files for 3d printing.
The last one is a little bit more complicated. In my program I created segments of 10 cm of wing and these files were used for import in OpenSCAD.
The idea was to create a 3D structure similar to the one used in Kraga Kodo (but it was developed before I knew about Kodo).
You can find some ideas and progress reports on my Romanian RC forum:
http://www.rhcforum.ro/topic/68986-p...4m-tiparit-3d/
I've done also some tests fo printing the structure with PLA with 0.4 , 0.3 and 0.2 mm nozzles. My goal was to get a printed wing under 1 kg. Not so easy - from the set of 3 criteria (weight / strength / time to print) you just have to pick 2, all 3 are not easily obtained in the same time.
I was preparing to build a set of 10-12 3d printers (FDM technology ...molted plastic) in order to be able to print the whole wing in a reasonable amount of time (with a cost around 150 EUR per printer) but then I found that with a similar cost I can build a 3D resin printer.
I started working first with UV curable resins, you can see an early prototype in the Romanian forum topic (the picture with orange PLA parts). It's a "duplex" printer (arduino is controlling 2 separate XY axis mechanics with a common Z axis, the goal is to print in the same time a segment and the mirrored image). The 405 nm UV lasers are pointing to the bottom of resin cuve, printing area for this prototype is 26 x 3 cm.
Then I started working on a printer with daylight resin and LCD screen. It's similar to Photocentric 3D LiquidCrystal HR:https://photocentric3d.com/liquidcry...v=9b6a28c805e0
Same iPad retina screen (20x15 cm with XY resolution of 0.096 mm or 96 microns), same build volume (20x15x25 cm). Non-proprietary software (www.nanodlp.com) and hardware (raspberry pi + arduino/ramps+nextion touch screen). Compared to the quite expensive commercial printer (2000+ EUR) mine is a lot cheaper (200 EUR for materials). Plus some extra features (wired/wireless net, USB port, control via browser or touchscreen).
Yesterday I finished assembling the first LCD screen printer (I have parts for 6 UV laser and 6 LCD screen printers).
Last picture is from the prototype doing a "dry print" run for the 3D lattice structure.
The final goal is to print segments of wing with external resin "foil", with this internal structure and with channels for carbon tube spars.
Beside this 9.7" iPad screen printer I have in work bigger screen printer but with lower resolution (2 pieces of 13.3" full D area 30x16.5 cm and a bigger 24" using a 4K UHD monitor ...this one will be similar LiquidCrystal Pro but of course at 1/10 of the price). If the result of printing with these higher size / lower resolution screens will be satisfactory I will sell the iPad screen printers.
PS sorry for the finger prints and hot glue traces on the red cover: it was a quick and dirty assembly last night in order to see if the pieces fit together).
Last edited by ender67; Oct 09, 2017 at 04:25 AM.
Oct 09, 2017, 03:21 PM
Registered User
Thread OP
Hi Ender,

I'm impressed by your dedication and the advanced ideas you investigate. On my side, I would try to keep the un matched high performances of composite materials, but trying to benefit from modern tools to reduce labor time and increase accuracy.

Looking at different things, I discovered that Stratasys had published several papers about using 3D printing to build molds/Tooling. I'm impressed, especially by their shell mold:
http://blog.stratasys.com/2017/05/25...osite-tooling/

The published a 65 pages white paper on their techniques. I downloaded it, and have to go through:
http://www.stratasys.com/landing/composite-tooling

For me, it looks promising as it could solve:
- my personal CNC limited time access,
- the problem of the CNC tool material: MDF wraps, Corian is expensive, aluminium is not for the amateur....

If such molds where cheap, could hollow molded wing be the way to go ?

JMF
Oct 09, 2017, 03:24 PM
Registered User
Thread OP
By the Way, ender, I came back to the STL files you had provided to me to make my experiments with the 3D printer :-)

XFLR2STL is back,

JMF
Oct 10, 2017, 02:26 AM
Registered User
Yes, hollow molded is the way to go. There was a poll/debate on another section of the forum for solidcore vs hollow molded with interesting results: for the weight difference between solidcore and rohacell/airex sheets you can afford (both weight & price) an internal layer of carbon/kevlar/glass composite.
The only advantage of solidcore is that you can finish the wing in one session...but on the other hand you need an expensive milled core. And (maybe) you don't need a vacuum installation.
My work was initially done in 2 directions: composite and 3d printed. For composite the problem was to create an accurate mold with usual hobby tools. As you already saw it's quite hard to accurate mill a solidcore (you need a vaccuum bed or you need to do a lot of tricks to substitute it) and CNC time is expensive. I know how to build an inexpensive CNC using 3d printed parts, steppers, bearings, belts and lead screws) but it's a lot of work to mill everything and you need a dedicated shop (and to be honest sometimes this is a problem, especially in my home country where a lot of people are living in block of flats in 1or 2 bedroom apartments). But you can find space on your desktop for a 3d printer.So in the last months I focused on 3d printers.
I spent a lot of time printing sections of the wing with different nozzles, infills, settings...and I reached the conclusion that FDM printing is nt suitable for wing but it may be suitable for molds. However if you want to do some thermo-curing for composite PLA is not suitable (if you leave a PLA part in your car in summertime it's ruined) and for ABS you need a special printer (with a good heated bed and better with a closed printing space).
I still wanted to have a 3d printed wing so I focused on resin printing but the costs of 2 resin 3d printer was off my budget (and printing space was limited) so I started designing and building my own printer. I found out that there are actual even cheaper to build compared to FDM 3d printer (in the 150-200 EUR/USD range).
My idea was to replace the solidcore with a light lattice structure 3D printed with resin and to cover the surface either with a cheap 3d carbon like cover (like the one used in cars) or even with carbon/resin (but for this you need a mold 3d printed in FDM because resin is too expensive to be used in molds).
Oct 10, 2017, 06:04 AM
Registered User
Thread OP
I always try not to cross-post, but I'll make some exception here to keep related data in the same place:

At the moment I have some thinking around 3D printed molds that may apply to CNC too. Stratasys has a design guide where they state that 3D printing + Adhesive FET (Tooltec CS5) achives good surface finish without any sanding (they give the figure but I don't have it here).

What about this process:
- 3D printed mold,
- stratify the surface with a thin layer of vacuum bagged fiberglass, with mylar on top of the fiberglass ?

It would be an "automated" sanding/painting process, getting best of the mold overall shape and mylar to bridge the gaps.

Could it fly ?

JMF
Oct 10, 2017, 06:05 AM
Registered User
Thread OP
Quote:
Originally Posted by ender67
Yes, hollow molded is the way to go. There was a poll/debate on another section of the forum for solidcore vs hollow molded with interesting results: for the weight difference between solidcore and rohacell/airex sheets you can afford (both weight & price) an internal layer of carbon/kevlar/glass composite.
Ender, could you provide the link? It looks interesting...

JMF
Oct 10, 2017, 07:59 AM
Registered User
https://www.rcgroups.com/forums/show...-or-solid-core
Just some thoughts about this "3d printed molds" thing.
3D printers have (relative) good XY accuracy (usually 1/40 to 1/160 mm) but on the other hand you need to be aware of certain limitations.
First of all: usually the mechanics is with belts on XY and lead screw on Z. Belts have a certain elasticity (sometimes we even mount tension springs on them) and in most designs there are a significant mass to move around (printing bed or printing head with extruder). Moving masses have inertia and produce vibrations (I can see a ripple effect at LE when I printed the tests segments). You need to slow down the printing as much as possible in order to increase accuracy. Another idea is to use a cube/corexy design with bowden in order to minimize the moving masses. I would go so far as to recommend using lead screw with anti-backlash nuts (double nuts coupled with spring) also for XY.
Second: slicers usually assume that the width of the extruded plastic is equal to the nozzle diameter but this is not true (some slicers have a parameter for this correction). In time nozzles increase their diameter (that's why some noozzles are hardened steel) plus there is a natural tendency of molen plastic to expand once it leaves the nozzle (it is pushed with force and it just expands.... this expansion increases when you print with a higher temperature to increase adhesion between layers).
Third: there is a certain variation of printing quality inside the same spool. You print "better" with a new spool due to 2 factors: the spool is "fresh" (limited exposure to humidity) and there is less tension in the filament.
And last: the "Z factor". First layers will be "crushed" on the bed in order to force the adhesion - this will lead to first layer (sometimes even the second) with larger dimensions. You need to rpint with relative small layers (0.1-0.15mm ...maximum 0.2 mm) in order to minimize the ripple effect of layers. But your worst enemy are bent lead screws and eoors in alignment and wobblying ...they all lead to large XY errors (even if the section has good XY dimensions it is a very bad thing if it is displaced by several tenths of mm).
For composites it is adviseable to cure the resin in mold at an increased temperature and if you want to do this you can forget about PLA and do the printing with ABS or PETG. ABS is more common but large pieces tend to warp (even with a heated bed) if you don't have a cosed case for 3d printer.
About finishing the surface: you can do a light sanding and resin coating (repeat if necessary). You can even do acetone polish of surface with ABS. also paining is possible.
One more thing: you don't need a high infill ratio for printing the mold. You need to position the pieces together and suggest to put them on rod or lead screws. These are cheap but sometimes with inconsistent diameter and are bent.You need to adjust the positioning hole diameter to match as possible the diameter of rod/screw and when you position the pieces you have to put them on a flat surface.
Just FIY: for the time I abandoned the idea of making a mold out of 6 mm plexiglass laser cut because of 2 reasons. First it is extremely expensive (even if I got some cheap plexiglass as a leftover from a big project). More important: the lasercut in a 6 mm plexiglass is not straight (like an I) but more like a V trench. You have a kerf of at least 0.5 mm (maybe more) so you still end doing a lot of sanding. But at least you can estimate these errors and end up with an accurate mold.


Quick Reply
Message:

Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Discussion Injecting Polyurethane Foam into single-walled 3d printed parts wob 3D Printing 33 Dec 15, 2019 08:44 AM
Build Log MPPCNCFC - 3D printed 4-axis-foam-cutter-hardware Kyrill CAD/CAM 43 Jun 16, 2019 04:12 AM
Discussion 3D printed fuselage flies with foam wings. basicguy 3D Printed Planes 19 Oct 26, 2017 09:12 AM
Yippee! The Mooch's replacement could be Stephen Miller Broken Wings Life, The Universe, and Politics 15 Aug 06, 2017 05:50 PM
Discussion Wing profile for a 3D printed solar glider clementt Hand Launch 26 Feb 04, 2017 12:35 AM