3D printing for HPR

[JohnCoker]

It's clear that additive manufacturing (3D printing) is the future of short-run fabrication. However, it's far from clear whether the mechanical properties of typical 3D printed parts are sufficient for high-power rocket construction.

I am working on a new article to document my experiments in trying to prove out 3D components:
https://www.jcrocket.com/printed-components.shtml

 

[Steve Shannon]

For twenty years I’ve been learning from your articles, John. I’m looking forward to seeing this one!

 

[Troy3003]

I appreciate all of your videos and writings as well. You may want to talk with Paul Trainer. I’m sure you have heard the name. He seems to have a very successful run at printed rocketry.

 

[SolarYellow]

I'd really like to see you test parts made with Siraya Tech Blu-Tough Nylon Black resin. It's ~1/4 the price of the Tough 2000 resin. I can't see myself spending $200/L for the Tough 2000, but $50/L is likely for fin cans and maybe low-end HPR ebays.

I'd also be interested in more complex parts. IMO, printing is the wrong technology for plates that can be easily lasered or CNC routered out of flat sheet. Doing things like NCs, ebays, and fin cans is where the value lies.

 

[OzHybrid]

It's clear that additive manufacturing (3D printing) is the future of short-run fabrication. However, it's far from clear whether the mechanical properties of typical 3D printed parts are sufficient for high-power rocket construction.

I am working on a new article to document my experiments in trying to prove out 3D components:
https://www.jcrocket.com/printed-components.shtml

The advantage of 3D printing is to be able to make things you could not easily do with conventional materials. So a complete motor mount with contact points for the motor rather than duplicating a MMT tube in plastic and integrating designed centering rings with ribs built in might be something you'd print. nosecone would definately be something you'd print. For straight centering rings only, plywood already has the advantage of multiple fiber direction orientation and good strength to weight. There are plenty of youtube fff material testing sites already out there. Testing printed centering rings as a direct equivalent of what you'd cut out of plywood is not a good test. It's not what you'd do if you were creating and printing in 3D.

 

[cwbullet]

Tuned in and can't wait to learn.

 

[Rschub]

There are many applications for 3D printed parts, I myself would not use them for any primary load bearing structure due to the inevitable scatter in actual strength that the manufacturing process produces by default.

Conventional materials are produced in controled conditions in huge quantities and tested for their physical properties. 3D printing is not that. Every machine, operator and material will be different from the others.

I am sorry, but in my opinion your test results will be only good for your process, only for as long as you don't change it. It will not necessarily be applicable to anyone else, except as an example of how to qualify 3D printed materials for oneself, which is a worthy goal.

 

[2-0turbo]

Interesting. I’ll look forward to your results.

 

[mrwalsh85]

The SLI team I mentor typically uses 3d printed centering rings for their subscale and full scale rockets. This year, they 3d printed the nose cone as well.

The 3D printed CR's have been in use for two plus years now, without failure. The caveat is that they use an aluminum thrust plate to transfer the stresses to the airframe, with an Aeropack retainer. I've been trying to egg them to make a fully 3d printed av bay sled to hold their electronics and to promote wire management... No luck yet. Maybe one of these years.

 

[bad_idea]

Like others have said, this isn't testing 3D printed components how I would use them with the current commodity technology (I suspect my thoughts on this might be very different a decade from now). Nonetheless, it will be interesting and valuable to see. On a directly personal level, it will be useful to be able to quantify under what conditions a simple print we make in our spare bedroom could replace a plywood piece we outsourced to a laser shop. On a different but more immediately important level, it will be good to have a better basis by which to evaluate the purely 3D-printed pieces we see turn up at the RSO, produced by those without a strong rocketry background, especially including many university teams. For many new rocketeers coming from a maker perspective, 3D printing is the answer to every question. Could be it's the answer to more than we might think.

 

[rharshberger]

Like others have said, this isn't testing 3D printed components how I would use them with the current commodity technology (I suspect my thoughts on this might be very different a decade from now). Nonetheless, it will be interesting and valuable to see. On a directly personal level, it will be useful to be able to quantify under what conditions a simple print we make in our spare bedroom could replace a plywood piece we outsourced to a laser shop. On a different but more immediately important level, it will be good to have a better basis by which to evaluate the purely 3D-printed pieces we see turn up at the RSO, produced by those without a strong rocketry background, especially including many university teams. For many new rocketeers coming from a maker perspective, 3D printing is the answer to every question. Could be it's the answer to more than we might think.

It will be interesting to see Johns take/results based on his choices of equipment, materials, processes, post processing (if any), etc.

To get the most out of 3D printed parts structurally USUALLY some form of post print process is needed to make sure the layers are as homogenous as possible.

I have seen some fantastic 3D printed rockets both good and bad techically and structurally andcthe only way to tell them apart in most case was to fly them.

 

[bad_idea]

To get the most out of 3D printed parts structurally USUALLY some form of post print process is needed to make sure the layers are as homogenous as possible.

My post-processing so far has been "slather all surfaces with fiberglass."

 

[charrington]

I'm very excited to see the results of the HP MJF printer. My friend works at HP and has access to one of them and I have felt the parts produced by it. Do not think for one second they mimic the typical filament 3d printed parts you are used to. These are incredibly strong and robust parts.

 

[cwbullet]

My post-processing so far has been "slather all surfaces with fiberglass."

Probably not needed if you do the right infill direction and %.

 

[thzero]

Probably not needed if you do the right infill direction and %.

Weakness part on FDM prints is the layer bonding. But it also depends on the thickness of the material. Say with 3d printed tubes, they are pretty thin walled, so only so much you can do and they are more prone to crack along the layer lines. Same with couplers. Of course you can always add thickness, but my guess is that only helps so much and adds weight. Or slather it with fiberglass, but that adds weight.

Material choices with "hobby grade" resin printers aren't as high, although with the post curing and finer lines, maybe you get a better more resilient part and less prone to crack on the layer lines. Certainly more of a PITA to prototype with. So would be curious to see results. I noticed there was some ABS like resin, that at some point I'd try out. Use the FDM for prototyping, resin for the "production" part? Also most resin printers aren't very big either, so limits things a bit or you're needing to slice up models into sections then join them later.

I'll say that thus far the 54mm-66mm fully 3d printed rockets have been flying quite well on up-to-Gs - G-64 is biggest I've used in the 66mm thus far. Am going to try a H128 on it. My biggest concern is really the couplers and that they could crack along the layer lines.

Inverted Pursuit labs has at least one 3d printed rocket that had run with a J in it (https://inverted-pursuits-lab.square.site/product/screwball/4?cp=true&sa=true&sbp=false&q=false) although he's also done the "slather with fiberglass" techniquie.

I know I have originally embarked on the slather fiberglass approach - thats what I did with the airfoiled shaped fins on The Joker (aka my former L2 project). And the first 2.6" LOC body tubed rocket with 3d printed fin/nose cone I did the same. Added a crap ton of weight. Frankly, I think its overkill on the mid sized rockets *if* you have an optimized printer - I don't think my original printer I ever really good optimized well and the prints weren't as good so more inclined to failure IMO thus the fiberglass.

 

[thzero]

Lit will be useful to be able to quantify under what conditions a simple print we make in our spare bedroom could replace a plywood piece we outsourced to a laser shop.

print out a centering ring and compare -100% infill
print out a flat plate fin and compare -100% infill

Which is basically what Jon is doing. With the fins you could probably rig up some torquing tests as well as some flutter tests.

 

[Rschub]

The question is not "How strong is it" The real question is "How strong does it need to be" If you don't know the answer to the second question, the first question is irrelevant.

The 3D printed CR's have been in use for two plus years now, without failure. The caveat is that they use an aluminum thrust plate to transfer the stresses to the airframe, with an Aeropack retainer.

Loads follow the path of greatest stiffness, Aluminum is an order of magnitude stiffer than plastic. The plastic will just deflect and the aluminum ends up taking all the load. Those centering rings might as well be made from styrofoam for all the load they see from the motor.

Most rockets that I have seen are overbuilt by quite a lot. But the loads increase exponentially as the airspeed increases. At some point you will find the "speed of plastic". Until we have a load calculator as accessible as rocksim, flight testing will be the measure of success.



I don't want to throw cold water on this project, I have used and will use 3D printed parts on my rockets, just understand the issues.

 

[GuyNoir]

This might be of interest:

 

[mtnmanak]

I may be the last person on TRF that does not have a 3D printer (finally fixing that - ordering an X1 Carbon for my first printer), so, mostly, I have no idea what I am talking about and look forward to John's tests/results.

One of the applications I am very intertested in is being able to make molds for composite layups. In the past, it has been very difficult to find/make good transitions, for example. I am hoping that I will be able to progress rapidly in my 3D printing prowess and then have an easy way to do layups for parts I always wanted to have/make, but couldn't before. We have a thread going on about getting Nike Smoke NC's, for example. Seems like it would be easy to print out the necessary mold and layup my own NCs now.

Will defer to all the 3D printing SMEs, though, since I am not even qualified to call myself a newb yet.

 

[cwbullet]

Weakness part on FDM prints is the layer bonding. But it also depends on the thickness of the material. Say with 3d printed tubes, they are pretty thin walled, so only so much you can do and they are more prone to crack along the layer lines. Same with couplers. Of course you can always add thickness, but my guess is that only helps so much and adds weight. Or slather it with fiberglass, but that adds weight.

All depends on you settings in the slicer.

 

[thzero]

I may be the last person on TRF that does not have a 3D printer (finally fixing that - ordering an X1 Carbon for my first printer), so, mostly, I have no idea what I am talking about and look forward to John's tests/results.

One of the applications I am very intertested in is being able to make molds for composite layups. In the past, it has been very difficult to find/make good transitions, for example. I am hoping that I will be able to progress rapidly in my 3D printing prowess and then have an easy way to do layups for parts I always wanted to have/make, but couldn't before. We have a thread going on about getting Nike Smoke NC's, for example. Seems like it would be easy to print out the necessary mold and layup my own NCs now.

Will defer to all the 3D printing SMEs, though, since I am not even qualified to call myself a newb yet.

Sure, sounds like a great idea... and no worries about loads, etc. since it's just being used as a tool

 

[JohnCoker]

TL;DR: I'm planning to test three hypotheses:

1. 3D printed parts are weaker than those made of traditional materials.
2. Traditional materials produce parts that are stronger than they need to be.
3. By optimizing the material, design and printing techniques, we can make printed parts strong enough.

For those asking about alternate materials or suggesting printer tuning, I'm starting with default materials and settings for all samples because I want the baseline analysis to be widely applicable. Tuning will most likely be necessary, but I want to get to that later.

Note that I'm not a 3D printing hobbyist; to me it's a tool for producing parts. I'm sure that enthusiasts have souped up their machines to print parts that have properties unmatched by anyone else, but I think it would be irresponsible to recommend something that isn't easily achievable by people who simply buy printers to make parts.

If that turns out not be the case, then I will consider hypothesis #3 to be disproven. However, I am confident that there is middle ground.

 

[SolarYellow]

Weakness part on FDM prints is the layer bonding. But it also depends on the thickness of the material. [snip]

I'll say that thus far the 54mm-66mm fully 3d printed rockets have been flying quite well on up-to-Gs - G-64 is biggest I've used in the 66mm thus far. Am going to try a H128 on it. My biggest concern is really the couplers and that they could crack along the layer lines.

Seems logical to orient the layer lines parallel to the airframe rather than transverse. Might have some more supports and infill to clear out, but at least the layer lines won't be asking the thing to cleave cleanly apart.

Loads follow the path of greatest stiffness,

I've known that for a long time, but have never seen it put so succinctly. Thanks.

 

[thzero]

Seems logical to orient the layer lines parallel to the airframe rather than transverse. Might have some more supports and infill to clear out, but at least the layer lines won't be asking the thing to cleave cleanly apart.

Sure... if someone has reliably printed with a qualify hobby/consumer grade FDM printer tubes in a horizontal instead of vertical and come out perfectly round, that would be good to know. Or even at a 45 degree angle. Resin, probably far easier, but less volume of a print. I've tried, and not had satisfactory results. Maybe if I finally get the AMS setup and use water soluble supports, might work better.

Fins, yeah, typically I try and print either flat, or with the layers either horizontal or preferred along so the grain is along the leading edge.

 

[JohnCoker]

As a side note, I'm also pursing a "just try it" approach with my best guesses. I built a 3" Nike-Smoke with everything 3D printed except the tubes and motor retainer. Of course there I took advantage of the ability to print things that can't be easily cut with a CNC router such as rings with integrated gussets and a 1-piece fin can.

It was during that process that I decided I wanted to know more about how 3D printed parts compared to traditional ones. The RSO/TAP/L3CC community is familiar with traditional materials, so if we could put 3D printed parts in context, it will help their acceptance.

These parts are printed in ABS on a Bambu X1C with 50% infill. Probably PETG would be a better choice as it is more easily printed, but ABS has been my go-to material so I wanted to start there.



I actually reprinted the fin can in 3 colors after this photo, and that's what I ended up using, but the picture below appears to be the only one I have of all the parts together.

 

[thzero]

What motors? Otherwise, looks pretty good to me.

 

[DragonRocketry]

I may be the last person on TRF that does not have a 3D printer (finally fixing that - ordering an X1 Carbon for my first printer), so, mostly, I have no idea what I am talking about and look forward to John's tests/results.

One of the applications I am very intertested in is being able to make molds for composite layups. In the past, it has been very difficult to find/make good transitions, for example. I am hoping that I will be able to progress rapidly in my 3D printing prowess and then have an easy way to do layups for parts I always wanted to have/make, but couldn't before. We have a thread going on about getting Nike Smoke NC's, for example. Seems like it would be easy to print out the necessary mold and layup my own NCs now.

Will defer to all the 3D printing SMEs, though, since I am not even qualified to call myself a newb yet.

Once you print a plug for making a mold you will understand why there is no one doing hand layups anymore. I have 2 printed nose cone that I am working on to make molds from and I got a lot of time in them just getting them smooth enough to start the process of laying up the mold.

 

[rharshberger]

As a side note, I'm also pursing a "just try it" approach with my best guesses. I built a 3" Nike-Smoke with everything 3D printed except the tubes and motor retainer. Of course there I took advantage of the ability to print things that can't be easily cut with a CNC router such as rings with integrated gussets and a 1-piece fin can.

It was during that process that I decided I wanted to know more about how 3D printed parts compared to traditional ones. The RSO/TAP/L3CC community is familiar with traditional materials, so if we could put 3D printed parts in context, it will help their acceptance.

These parts are printed in ABS on a Bambu X1C with 50% infill. Probably PETG would be a better choice as it is more easily printed, but ABS has been my go-to material so I wanted to start there.

View attachment 612714

I actually reprinted the fin can in 3 colors after this photo, and that's what I ended up using, but the picture below appears to be the only one I have of all the parts together.

Actually John, I think ABS is a better choice than PETG from the standpoint of bonding and post processing options/methods. PETG neither bonds well nor does it mechanically finish well (sand, file, etc) like ABS. You can also wipe a tin layer of solvent on ABS and it will chemically bond imperfections, PETG forget about it. ABS though has its downsides too, its more difficult to print with, temperature stable and vented enclosures or environment.

Nice looking print BTW.

 

[SolarYellow]

Actually John, I think ABS is a better choice than PETG from the standpoint of bonding and post processing options/methods. PETG neither bonds well nor does it mechanically finish well (sand, file, etc) like ABS. You can also wipe a tin layer of solvent on ABS and it will chemically bond imperfections, PETG forget about it. ABS though has its downsides too, its more difficult to print with, temperature stable and vented enclosures or environment.

Nice looking print BTW.

So, essentially, painting an ABS filament print with this stuff should make it significantly stronger and reduce the tendency to split between layers?

https://plastruct.myshopify.com/products/ppc-2
Or if there's another solvent that could be purchased in a quart can at any decent hardware store outside CA for a lot less $/oz, that would be even better...

 

[mtnmanak]

Once you print a plug for making a mold you will understand why there is no one doing hand layups anymore. I have 2 printed nose cone that I am working on to make molds from and I got a lot of time in them just getting them smooth enough to start the process of laying up the mold.

Interested in seeing your work on this! Once again, I have no experience, so have been speculating. My thought process was not a real mold, per se, but more of a thin shell to hold the composite layup. The plastic internal shell would just remain in the part. Probably worth starting another thread, if there isn't one already.