Thread Tools
Sep 28, 2017, 06:45 PM
wob
wob
Happy If It Flies or Not
wob's Avatar
Discussion

Injecting Polyurethane Foam into single-walled 3d printed parts


Hi folks,

TLDR; I haven't found any threads (here or on the interweb) about strengthening 3d printed parts with polyurethane foam for RC aircraft so I wanted to post my experiences with it. If someone else is already doing this with 3d printed parts please point me in their direction!

A little history: My interest lies specifically in 3d printing ducts and frames for quadcopters and potentially 3d printed planes, so lighter/stronger is better (isn't it always?). I have experience in thermoforming and other simple fabrication methods but as I sit around looking at my EPO airplanes I wish I could mold parts like those but I don't have a half a zillion dollars laying around for the equipment. My dislikes with EPO and their ilk is they dent, ding, melt in the sun and have other issues, and molded EPO is definitely not strong enough for multicopter shroud use whatsoever beyond toy Air Hogs. This got me thinking about DIY polyurethane molding for various parts but again that requires molds and all that stuff. Also, one can't easily integrate complex features with molded parts (at least not in a DIY/hobby setting) which is why most EPO airplanes and toy shrouded drones still use connecting injection-molded parts and good old fashioned epoxy. Also, it seems what EPO really needs is a strong skin to protect the foam and create a torsionally stiff sandwich like a surfboard. This all got me thinking that if I could somehow thermoform some thin gauge parts (like throwaway container thin) and inject that with foam I would get really strong and lightweight parts. Again, this requires molds which are annoying unless I plan to make 1000 of the same thing. The ah-ha moment came when I realized I should try printing thin-walled fully hollow parts on the 3d printer and inject them with foam. Basically the 3d printed part becomes the mold. Google searching led me to people doing it for movie props or strengthening 3d printed knick-knacks but I want stuff that flies and survives crashes. BTW, my main motivation for ducts is that I'm super lazy and hate balancing and/or replacing props, so what I want is super strong full prop protection AND ducts to negate or benefit from the added weight of the prop protection.

I won't go into duct theory and whatnot here but I can post links to the relevant papers if anyone is interested. Anyway, to start the prototyping/testing process, I drew a quick small duct for tiny little 2.3" propellers (FT Gremlin, BabyHawk, etc) and printed it in PETG with the common .4mm nozzle, single-walled, no infill. A quick side note: Future ducts will be up to 4 or 5 inch props because I have some actual payload to lift with high quality hardware but I prefer to scale up after prototypes are complete rather than print for days and days.

Moving on, the hollow yet fragile duct -- with single walls and no infill -- weighs 6 grams, and due to the beneficial duct shape on my test rig, it gains a modest ~18 grams of thrust at approximate hover throttle with the same energy input as the unshrouded tests -- I have the spreadsheet somewhere but from memory the rough numbers are 30% increased efficiency or 40% improved thrust which isn't bad for a partial duct. Full throttle numbers were off the charts but I am really concerned with hover efficiency since that's all non-racer multirotors really do. This is all great, but a hollow duct (even printed in PETG) is quite easy to crush and break and snaps at some layers or even across layers as stress is applied. For comparison, I printed the same duct with the standard double walls at a modest 20% infill and it predictably became a heavy 15 gram tank that would cancel out any efficiency gains. Of course, it's quite strong with the infill but not-flyable. Also, infill isn't perfect with single-walled parts -- it still has to be used so sparingly that it leaves certain areas unprotected and breakable, hence going for foam.

Onto the foam experiment: I picked up some 2-part 2lb poly foam from TAP plastics and used a hobby syringe to inject it into the duct. I'm not looking for super crazy density so I injected about 10% of the total volume of the duct and it filled nicely without squirting everywhere. Now the duct weighs about 9 grams which leaves a total of a 9 gram actual thrust gain at the same throttle/current, meeting my goal of strong prop protection and the hopeful goal of higher efficiency than bare props. What surprised me is this thing is INCREDIBLY strong... stronger than I thought it would be: Squeezing it like a tennis ball I could not crush it in my hand and it then required two hands -- squeezing hard -- to finally destroy it. Not bad for a 9 gram object of that size. Spurred on by the positive test results I cobbled together some simple CAD shapes that resembled a quadcopter (hey there's only so many ways to connect 4 giant circles), printed them out, and came out with what I'm temporarily calling the "Blue Bomber". I didn't take into consideration the fact that I'm using the world's cheapest uncalibrated 3d printer so I forgot to design in some relief for the motors to be adjusted laterally for even prop clearance so I ended up trimming the slightly off-center props down which killed some efficiency from the bench tests but this was good enough for a prototype. Even at a porky 75 grams frame weight -- due to lazy design -- it flies precisely the same amount of time as the open prop version (basically a FT Gremin/BabyHawk). I'll consider this a success!

On my quest for something so light that it will float into space, I am ordering some slightly less dense poly foam and am going to try printing with a .2mm and .3mm nozzle. My goal for this little prototype is 50 grams frame weight. (I'm sure the paint added 10 grams so I will be ordering some colored PETG).

I hope posting this experience helps others with printers and possibly future 3d printed plane design with semi-monocoque wing/fuse sections being possible... basically very similar to an EPO model but crazy strong and not as delicate as the pure 3d printed models (no disrespect to 3dLabPrint -- his stuff is amazing!!!). I've attached the pics of the past few weeks in no particular order... and I'm trying to figure out the best way to show strength in a picture. Maybe I'll stand on one of them :-)
Last edited by wob; Sep 28, 2017 at 06:53 PM.
Sign up now
to remove ads between posts
Sep 28, 2017, 07:15 PM
Retired CAD guy
birdofplay's Avatar
Very informative !

I always rekon'd that the [pastic would deform when the foam expanded.
So I never gave it a chance.

I remember TAP plastics in POrtland when I lived out on the west coast.

More info on light weight foams would be good to know, please :>}
Sep 28, 2017, 07:34 PM
wob
wob
Happy If It Flies or Not
wob's Avatar
Quote:
Originally Posted by birdofplay
Very informative !

I always rekon'd that the [pastic would deform when the foam expanded.
So I never gave it a chance.

I remember TAP plastics in POrtland when I lived out on the west coast.

More info on light weight foams would be good to know, please :>}
Thanks! I always thought it would deform, too, until I tried it. The key is knowing the volume of the part in advance and only filling as much as needed. The TAP foam "expands up to 30 times in open air" so I figured instead of 1/30th the volume I'd try 1/10th since this was closed space and I got very lucky on the first try. I used my finger over the fill holes until the foam got into all the nooks and crannies and then let the excess bleed out. The largest challenge is mixing then injecting before it sets off. I may try some double barreled syringes with the mixer tips on them for more control.

The strength of this 2lb foam is quite amazing and after 24 hours it is difficult to deform or break. For these purposes I think 1lb foam will be the sweet spot but I'll order a couple other sample weights to try as well, and will share them when I have time.

If I can get success with the .2mm nozzle then it may be feasible to print 3d printed wing/fuse sections at a similar weight as store-bought EPO models with far greater durability. As I mentioned in my original post, the 3d printed parts are really there to hold the shape until the foam sets up.
Sep 29, 2017, 08:41 AM
Retired CAD guy
birdofplay's Avatar
Large plane surfaces are going to more easily deform IMHO.

Those little circular parts are more rigid, like eggs.

BTW, did it fly ?
Sep 29, 2017, 09:41 AM
wob
wob
Happy If It Flies or Not
wob's Avatar
Quote:
Originally Posted by birdofplay
Large plane surfaces are going to more easily deform IMHO.

Those little circular parts are more rigid, like eggs.

BTW, did it fly ?
Agreed, I thought it would deform, too, however the foam is not flowing under any pressure yet quite evenly disperses. Also, the fuselage of the little copter has 5" straight sides (on the inside) and they did not deform. When I started this project I though I'd have to create holding fixture molds so parts didn't deform.

Yep, it flies great and can be bounced off of anything without breakage. Flight time is exactly the same as the open-prop version so ducts are working as expected. I'm very pleased how it turned out. I'll try to post a brief video if I can get my wife to film it.
Sep 29, 2017, 12:28 PM
Retired CAD guy
birdofplay's Avatar
SO ???

Vids and ...
Specs,( motors and controller) Files ( stl's), equipment list (Rx Pak size) ???
Sep 29, 2017, 01:41 PM
Registered User
stanordave's Avatar
keep us updated!
Sep 29, 2017, 07:57 PM
wob
wob
Happy If It Flies or Not
wob's Avatar
Quote:
Originally Posted by birdofplay
Vids and ...
Specs,( motors and controller) Files ( stl's), equipment list (Rx Pak size) ???

Emax BabyHawk/FT Gremlin flight gear w/ 500mah 2S
Basically tiny RS1104 motors and a flight controller so small you might accidentally swallow it

Poorly edited poor light quality video with a little LOS and FPV below. Tried to get my wife to record a quick LOS flight for me but she said something about me being a nerdbomb and handed me our toddler (who incidentally loves the quadcopter). I wack a couple of branches (on purpose ha ha yeah right) to show the resilience of the frame. Those mishaps would have destroyed open props and I would have had to trudge out into the yard to retrieve it.

The files for this are a mess because it's a prototype and I could not publish them in good conscience without fixing them. They probably wouldn't even print without Cura's "fix horrible" setting. When I get into version 2 I'll make clean watertight cad files and throw 'em up on Thingiverse.

blue bomber (3 min 0 sec)
Last edited by wob; Sep 29, 2017 at 08:05 PM. Reason: added more pics
Oct 04, 2017, 01:02 PM
Caution:Makes sharp left turns
Troy's Avatar
That's a really cool project! I like the fact you tried it instead of listening to conventional wisdom. Knowing these materials, and having printed several parts with 1 perimeter (3D Lab Print designs) I would have expected the shells to bulge out due to internal pressure and exotherm heat. I've toyed with the idea of casting and molding foam frames similar to this but don't have the time to design and get it done. I did do a video at work on 3D Printing molds to cast foam and that was a good success. I made flexible landing pads and motor protectors for a simple wood-framed quad.

3D Printed Molds for DIY Drone Part 1 (8 min 44 sec)
Oct 04, 2017, 05:32 PM
wob
wob
Happy If It Flies or Not
wob's Avatar
Quote:
Originally Posted by Troy
That's a really cool project! I like the fact you tried it instead of listening to conventional wisdom. Knowing these materials, and having printed several parts with 1 perimeter (3D Lab Print designs) I would have expected the shells to bulge out due to internal pressure and exotherm heat. I've toyed with the idea of casting and molding foam frames similar to this but don't have the time to design and get it done. I did do a video at work on 3D Printing molds to cast foam and that was a good success. I made flexible landing pads and motor protectors for a simple wood-framed quad.
Thanks, Troy! I was happy to see that I'm already familiar with BJB -- I've watched many of your BJB videos over the years as I've looked into molding, casting, and so forth -- all your vids are an awesome wealth of information, so thank you for the time you and your crew put into them :-)

I'm a big believer that conventional wisdom stifles innovation and discovery and trying something for yourself is invaluable (even if you might think you know what the outcome will be). Prior to experimenting, I was quite sure that allowing the foam to vent freely out of the fill holes at top would keep pressure at zero and keep walls from warping, and it turned out I was right. Without pressure, the foam was even filling in those tiny little motor mount spokes and even into the infill within the motor mount plate, and we are talking tiny little parts, here. Regarding heat buildup -- I was concerned, too, but since there are no large volumes of solid foam in these designs there is little or no heat build up (no where near softening the 3d printed plastic).

With the little copter complete (printing version 2 in orange PETG now), I plan to do some simple airplane wing cross sections (like 3d lab print) but instead of internal printed bracing I have a hunch that the majority of the wing can remain hollow to save weight. Granted, the end result would only come close to 3dLabPrint's aircraft in weight if I can pull this off with .2mm single walled prints. I'm not holding my breath with .2mm since that is tiny and scaling down always seems to cause a ruckus. See attached mockup of a simple 3d printed hollow wing section for what I have in mind...
Last edited by wob; Oct 04, 2017 at 06:52 PM.
Oct 30, 2017, 05:20 PM
Caution:Makes sharp left turns
Troy's Avatar
Interesting. I would expect the foam to increase density due to the increased surface area and drag created as the foam expands. I found this out on a small wing for my buddy Bret's Micro Jet:
Electrolyte: molded wings using expanded foam part 1 (8 min 1 sec)
Nov 01, 2017, 12:45 AM
Registered User
This looks so great. You have a lot of space in the center, can you consider making up a runcam split version, where the view is prop free, but camera is still physically behind the ducts?

2" can fly ok with 1080p 60fps

GOPRO with WINGS - 106 grams! (2" Quad + Runcam Split) (4 min 33 sec)



Quote:
Originally Posted by wob
Hi folks,

TLDR; I haven't found any threads (here or on the interweb) about strengthening 3d printed parts with polyurethane foam for RC aircraft so I wanted to post my experiences with it. If someone else is already doing this with 3d printed parts please point me in their direction!

A little history: My interest lies specifically in 3d printing ducts and frames for quadcopters and potentially 3d printed planes, so lighter/stronger is better (isn't it always?). I have experience in thermoforming and other simple fabrication methods but as I sit around looking at my EPO airplanes I wish I could mold parts like those but I don't have a half a zillion dollars laying around for the equipment. My dislikes with EPO and their ilk is they dent, ding, melt in the sun and have other issues, and molded EPO is definitely not strong enough for multicopter shroud use whatsoever beyond toy Air Hogs. This got me thinking about DIY polyurethane molding for various parts but again that requires molds and all that stuff. Also, one can't easily integrate complex features with molded parts (at least not in a DIY/hobby setting) which is why most EPO airplanes and toy shrouded drones still use connecting injection-molded parts and good old fashioned epoxy. Also, it seems what EPO really needs is a strong skin to protect the foam and create a torsionally stiff sandwich like a surfboard. This all got me thinking that if I could somehow thermoform some thin gauge parts (like throwaway container thin) and inject that with foam I would get really strong and lightweight parts. Again, this requires molds which are annoying unless I plan to make 1000 of the same thing. The ah-ha moment came when I realized I should try printing thin-walled fully hollow parts on the 3d printer and inject them with foam. Basically the 3d printed part becomes the mold. Google searching led me to people doing it for movie props or strengthening 3d printed knick-knacks but I want stuff that flies and survives crashes. BTW, my main motivation for ducts is that I'm super lazy and hate balancing and/or replacing props, so what I want is super strong full prop protection AND ducts to negate or benefit from the added weight of the prop protection.

I won't go into duct theory and whatnot here but I can post links to the relevant papers if anyone is interested. Anyway, to start the prototyping/testing process, I drew a quick small duct for tiny little 2.3" propellers (FT Gremlin, BabyHawk, etc) and printed it in PETG with the common .4mm nozzle, single-walled, no infill. A quick side note: Future ducts will be up to 4 or 5 inch props because I have some actual payload to lift with high quality hardware but I prefer to scale up after prototypes are complete rather than print for days and days.

Moving on, the hollow yet fragile duct -- with single walls and no infill -- weighs 6 grams, and due to the beneficial duct shape on my test rig, it gains a modest ~18 grams of thrust at approximate hover throttle with the same energy input as the unshrouded tests -- I have the spreadsheet somewhere but from memory the rough numbers are 30% increased efficiency or 40% improved thrust which isn't bad for a partial duct. Full throttle numbers were off the charts but I am really concerned with hover efficiency since that's all non-racer multirotors really do. This is all great, but a hollow duct (even printed in PETG) is quite easy to crush and break and snaps at some layers or even across layers as stress is applied. For comparison, I printed the same duct with the standard double walls at a modest 20% infill and it predictably became a heavy 15 gram tank that would cancel out any efficiency gains. Of course, it's quite strong with the infill but not-flyable. Also, infill isn't perfect with single-walled parts -- it still has to be used so sparingly that it leaves certain areas unprotected and breakable, hence going for foam.

Onto the foam experiment: I picked up some 2-part 2lb poly foam from TAP plastics and used a hobby syringe to inject it into the duct. I'm not looking for super crazy density so I injected about 10% of the total volume of the duct and it filled nicely without squirting everywhere. Now the duct weighs about 9 grams which leaves a total of a 9 gram actual thrust gain at the same throttle/current, meeting my goal of strong prop protection and the hopeful goal of higher efficiency than bare props. What surprised me is this thing is INCREDIBLY strong... stronger than I thought it would be: Squeezing it like a tennis ball I could not crush it in my hand and it then required two hands -- squeezing hard -- to finally destroy it. Not bad for a 9 gram object of that size. Spurred on by the positive test results I cobbled together some simple CAD shapes that resembled a quadcopter (hey there's only so many ways to connect 4 giant circles), printed them out, and came out with what I'm temporarily calling the "Blue Bomber". I didn't take into consideration the fact that I'm using the world's cheapest uncalibrated 3d printer so I forgot to design in some relief for the motors to be adjusted laterally for even prop clearance so I ended up trimming the slightly off-center props down which killed some efficiency from the bench tests but this was good enough for a prototype. Even at a porky 75 grams frame weight -- due to lazy design -- it flies precisely the same amount of time as the open prop version (basically a FT Gremin/BabyHawk). I'll consider this a success!

On my quest for something so light that it will float into space, I am ordering some slightly less dense poly foam and am going to try printing with a .2mm and .3mm nozzle. My goal for this little prototype is 50 grams frame weight. (I'm sure the paint added 10 grams so I will be ordering some colored PETG).

I hope posting this experience helps others with printers and possibly future 3d printed plane design with semi-monocoque wing/fuse sections being possible... basically very similar to an EPO model but crazy strong and not as delicate as the pure 3d printed models (no disrespect to 3dLabPrint -- his stuff is amazing!!!). I've attached the pics of the past few weeks in no particular order... and I'm trying to figure out the best way to show strength in a picture. Maybe I'll stand on one of them :-)
Last edited by sonic_boom; Nov 01, 2017 at 12:54 AM.
Nov 02, 2017, 02:43 PM
wob
wob
Happy If It Flies or Not
wob's Avatar
Quote:
Originally Posted by Troy
Interesting. I would expect the foam to increase density due to the increased surface area and drag created as the foam expands. I found this out on a small wing for my buddy Bret's Micro Jet:
Agreed. Thanks for sharing your video. I imagine that the larger the part (more surface area) the more dramatic the problem would present itself. This one of several reasons why I am trying to source a reusable dual barrel syringe with a disposable static mixer nozzle to simultaneously mix/inject the foam which should gain valuable time to rotate the part and coat most of the inside, as is done in roto-molding. I am assuming that coating the interior and keeping the mixture moving as much as possible while the foam is still in a liquid-ish state would result in a more uniform foam density and even better penetration into the small features as long as a tiny bit of back pressure can be maintained. Troy, do you know of any 2 part poly foam that has a longer pot life -- and is less vicious than the typical 2lb foam -- prior to setting off? A nice long pot life would do wonders for getting more consistency with this injection method.

Side note: I bought a few empty dual barrel syringes on amazon with static mixing nozzles but was not happy to find that they are entirely disposable which is not what I wanted, since I am doing a lot of tests. I am going to try and permanently attach the sealing pistons to the plunger so they can be re-used.

Update: Lack of free time and new printing problems has stalled my foam injection tests for now. I'm printing with a .3mm nozzle and getting great looking light prints that are 25% lighter than my original tests but now the prints are no longer water tight as they were with the first quadcopter frame. Seems to be a mixture of layer height, retraction, and other settings are creating hundreds of invisible pinholes in the print. Or... it could be that .4mm is the smallest I can feasibly go. Not sure yet.
Nov 02, 2017, 02:50 PM
wob
wob
Happy If It Flies or Not
wob's Avatar
Quote:
Originally Posted by sonic_boom
This looks so great. You have a lot of space in the center, can you consider making up a runcam split version, where the view is prop free, but camera is still physically behind the ducts?

2" can fly ok with 1080p 60fps
Thanks for your comment, sonic. My interest lies more with exploring this fabrication methodology rather than designing frames for all the different motors/props/cameras/running gear combos out there. I've too little time and there are too many to keep up with.
Feb 12, 2018, 02:43 PM
Overgrown Child
Any success with the fixed wing design? This is something that I’m VERY interested in trying myself. Originally I was thinking I would just use the expanding “Great Stuff” foam, though that’s a single use can I believe. I’ll have to check out the epoxy-style foams.


Quick Reply
Message:

Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Discussion Making a foam cored wing using 2-part rigid polyurethane foam in the mould. AvB Composites Fabrication 356 Jul 05, 2018 07:28 AM
Discussion 3D printed CG locator, fixed, single digital scale kcaldwel Electric Competition Soaring-F5J/ALES/e-Soaring 41 Jul 30, 2017 03:18 PM
Discussion Getting back into RC - Going to build a 3D printed drone markag Beginner Multirotor Drones 17 Apr 28, 2017 08:06 AM
Discussion About polyurethane foam SniperScope Scale Kit/Scratch Built 3 Nov 11, 2016 03:26 PM
Build Log Foam, printed tissue and polyurethane; 15-inch Wedell-Williams K5083 Scratchbuilt Indoor and Micro Models 6 Nov 20, 2015 09:23 AM