3D printing for HPR

The Rocketry Forum

Help Support The Rocketry Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
It's easier to 3D print parts them cut them from plywood. (I expect many more people have 3D printers than CNC routers.)
I concur. Ease of printing vs cutting and routing is key, but it is also cleaner. No sawdust to clean up. Then again, you do waste a lot of filament getting it perfect,
 
I concur. Ease of printing vs cutting and routing is key, but it is also cleaner. No sawdust to clean up. Then again, you do waste a lot of filament getting it perfect,
Use the best method from everything YOU have available. While it's possible to completely print a rocket, and I've done it, you can achieve better performance and lighter weight by not restricting yourself to a single construction methodology.
Printed components can suffer from delamination after heavy landings. There is no easy way to identify the start of any cracks.
With conventional materials the failure modes are known. You can cut centering rings with a jigsaw, or 2 holesaws or a pair of scissors if you're using 1/16 ply. There is no necessity for a cnc router and I've used all those methods.

Lastly there is no point in printing a centering ring in some exotic material when the cost of that is excessive. So those results while interesting, are not practical for those who cannot afford those exotics by the cost of either the printer required or the ongoing exotics cost.
The majority of people will be printing in PLA, +, PETG or ABS.
 
I concur. Ease of printing vs cutting and routing is key, but it is also cleaner. No sawdust to clean up. Then again, you do waste a lot of filament getting it perfect,
I am more than happy to support people like Dragon Rocketry that provide laser cutting services, I understand the thinking behind trying to do everything in house, but I am wary of rabbit holes from long experience.
White Rabbit GIF
 
In the long run, something that would be very helpful would be a matrix of an object’s strength vs. printed wall thickness and infill percentage. Like it would be nice to know that a 1-layer wall with 30% infill is 90% of the strength of a 2-layer wall with 50% infill.

Thank you for doing this work!
 
I do a lot of 3dprinting for ebay components and also for jigs like for fins and things. It's been a game changer for me, if something doesn't turn out right then make an adjustment and re-print. Let the printer do the hard work.
I agree with you, recently made a few nosecone / av bays using up some 38mm leftover fiberglass tubes. Different vintages and different suppliers / wall thickness on the fiberglass tubes. Each had a slightly different I.D. I ended up making slight changes to the OD of the shoulders, 'customizing them' so they each fit a particular tube perfectly. Easy to do on a 3d printer. Just minor 'tweeks' of the two index features on the nosecone cup (light green part) to make it custom.
Mike K

1706645344657.png
 
You can cut centering rings with a jigsaw, or 2 holesaws or a pair of scissors if you're using 1/16 ply. There is no necessity for a cnc router and I've used all those methods.
Jigsaw and hole saws are much more work and produce poorly-fitting parts. (I've used these and other methods in the past.) Without CNC routing, I think 3D printing is the best hope for home-made components in the long term.
Lastly there is no point in printing a centering ring in some exotic material when the cost of that is excessive. So those results while interesting, are not practical for those who cannot afford those exotics by the cost of either the printer required or the ongoing exotics cost.
The majority of people will be printing in PLA, +, PETG or ABS.
I was curious about the range of printers I had access to (including through services). I will focus next on adjusting settings to print with the X1C in ABS, PETG and PLA. (There is no PLA in this first set because it's low Tg rules it out for use in CRs, but I am including PLA in the bulkhead samples.)
 
In the long run, something that would be very helpful would be a matrix of an object’s strength vs. printed wall thickness and infill percentage. Like it would be nice to know that a 1-layer wall with 30% infill is 90% of the strength of a 2-layer wall with 50% infill.
Agreed. It's quite expensive to get the testing done, but I will try to quantify some of these dimensions.
 
Jigsaw and hole saws are much more work and produce poorly-fitting parts. (I've used these and other methods in the past.) Without CNC routing, I think 3D printing is the best hope for home-made components in the long term.

Agree that jigsaw and hole saws are limited. And you're super-lucky if you happen to find two hole saws that make the precise ID and OD you need for centering rings. Even one is pretty unusual.

CNC routing at the low end can be cheaper than printing, but it doesn't have the versatility. On the other hand, I think it's way superior for things that can be built up out of flat stock. And it can be used for balsa, bass, plywood, composites, foam, even light aluminum work. It's kinda nice to be able to choose your material and make it into the shape you want, rather than trying to find a filament or resin with appropriate characteristics and then working to find all the settings to make it functional.

A proper shop should have both.
 
I've been printing centering "rings" like this for my interchangeable motor mount system. The main benefit is everything is concentric and perpendicular, i.e. not tilted. The round indentations are to clear rail button rivet nuts and the slots are for clearing the fin tabs. These are printed with very low infill so they are very light.

1706743638867.png
I forgot to mention I have a laser cutter, metal lathe, CNC milling machine but 3D printing is my preference for making parts.
 
Last edited:
OK, I've got the first batch of numbers back comparing printer settings (all printing ABS on the Bambu X1C). I want to do a couple more batches with higher number of wall loops to find the fall-off point, then I will publish a more complete update. I'm also curious whether the pattern comparison would change at 50% density.

Screenshot 2024-03-13 at 3.18.57 PM.png
 
Very good results. Although I have not tested it to this degree, this is what I am seeing on the field. Then again, I probably would not fly 3D printed rockets and parts with most Warp 9 motors.
 
Very good results. Although I have not tested it to this degree, this is what I am seeing on the field. Then again, I probably would not fly 3D printed rockets and parts with most Warp 9 motors.
I will. Send me one w/a 38mm motor mount. Your choice of an H999 or an I1299.
Watch the very end of this (after the still shots)

 
OK, I've worked my way up and down the range, and settled on a reasonable "98/54 plywood CR replacement."

TL;DR:
  • ABS, 50% infill, gyroid pattern, 8 wall loops, 8 top/bottom layers
  • 4mm base, 12 ribs 6mm high (10mm overall)
For all the gory details, including an OpenSCAD file you can use to create CRs of any size:
https://www.jcrocket.com/printed-components.shtml#hypothesis3
 
I agree with you, recently made a few nosecone / av bays using up some 38mm leftover fiberglass tubes. Different vintages and different suppliers / wall thickness on the fiberglass tubes. Each had a slightly different I.D. I ended up making slight changes to the OD of the shoulders, 'customizing them' so they each fit a particular tube perfectly. Easy to do on a 3d printer. Just minor 'tweeks' of the two index features on the nosecone cup (light green part) to make it custom.
Mike K

View attachment 627251
Love the design. That would be a good one to upload.
 
OK, I've worked my way up and down the range, and settled on a reasonable "98/54 plywood CR replacement."

TL;DR:
  • ABS, 50% infill, gyroid pattern, 8 wall loops, 8 top/bottom layers
  • 4mm base, 12 ribs 6mm high (10mm overall)
For all the gory details, including an OpenSCAD file you can use to create CRs of any size:
https://www.jcrocket.com/printed-components.shtml#hypothesis3
Great summary. I have started making all my PETG and ABS centering rings and Bulk Plates with this sort of design.

Have you considered testing bulkheads or plates?
 
@JohnCoker I really like the thoughtful and thorough approach to testing - there's a lot to learn in reviewing the results.

This is a bit of a thread steal question, but it was prompted by the failure mode (tube buckling) that was observed in some of your tests.
Since I recently got back into rocketry, I've been wondering if there's not a stronger way to mount CRs in body tubes. As opposed to fillets. That is, would gluing in a short (1-2cm) length of tube coupler in front of the CR add anything to a build. I've done it on my Super DX3 build but (feel free to laugh at me) I haven't launched an HPR (I'm Level 0!). The reason I added the ring was that by following the force coupling from the motor to the tube it seemed that the short coupler section would add to the shear strength. But an epoxy fillet might be stronger.

This is probably not an operational consideration in the G-H range but as I understand it as folks have put stronger motors in their rockets they sometimes fail.

It does seem to me that one could make a printed CR cone-shaped (cone point facing aft) and this would make the force transmission from the motor compressive rather than shear.

I like Mike's (MAC Performance) aluminum bulkhead idea (which it sounds like you've implemented - maybe you came up with it) in which the motor retainer is close-coupled to the bulkhead, avoiding any axial force transmission in the paper tube. The reverse cone might allow a lighter, stronger, stiffer couple from the motor to the body tubes and fins.

And I'm really jonesin' for a MarkForged 3D printer and a spool of Onyx CF....
 
@JohnCoker thanks.

"It increased the breaking force from 580 to 611 lbf"
Yay
"but less than I expected."
I guess a lot of rockets have been launched and recovered without this rings, and so a breakthrough in radically increasing strength is... expected.

"Yeah, I considered a double-cone shape, but then I thought it would be harder to print. (The ribbed CRs don't need supports.)"
I was thinking a single cone but I guess if you make the CR thick enough, there is in effect a compressive thrust through the ribs. Thanks for the explanation and the data.
 
"Yeah, I considered a double-cone shape, but then I thought it would be harder to print. (The ribbed CRs don't need supports.)"
I was thinking a single cone but I guess if you make the CR thick enough, there is in effect a compressive thrust through the ribs. Thanks for the explanation and the data.
Yeah, we could do testing of more shapes as well. I'm going to pause this series of tests here for the moment.

I also did tests with the ribs up and down, and it didn't make a significant difference.
 
Back
Top