Low Cost 3D Printing: My Experience

The Rocketry Forum

Help Support The Rocketry Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

rocketguy101

Well-Known Member
Joined
Jan 17, 2009
Messages
1,476
Reaction score
114
3D printing is a process that takes a CAD file and “prints” an actual part from some material like plastic. Industrial printers cost tens of thousands of dollars, putting them out of the reach of most hobbyists. In the last few years, low cost printers have shown up and modelers on rocket forums have been asking about them.

This report describes the experience I have had with a particular low cost printer. I am in no way an expert on 3D printing (or “rapid prototyping” as the industry calls it), nor are the statements I make about printer capabilities applicable to all printers.

A friend at work bought a “Mosaic” hobby 3D printer from MakerGear https://www.makergear.com/ to see what it could do. James is a certified SolidWorks professional, so he knows his way around SolidWorks (SWX) and has written many VB programs to create parts libraries for our employer. As soon as he found out I was into rockets, he asked if there was anything I wanted to try printing.

I have several rockets modeled in SolidWorks, but settled on testing the system with Chris Flannigan's beautiful D-Region Tomahawk he posted on Scaleroc. I used the SWX "scale" feature to scale the full-size model to fit a BT-60 tube (1.637 / 9.00 = 0.181888889).

I wanted to print the fin shroud, lug, fins and fin attachments as a single unit that would slide over the BT-60 tube, and the nose cone.

D-Region_isometric2.jpg

D-Region_fins.jpg

3D_Print_001a.jpg

3D_Print_002a.jpg
 
It turns out the Makergear printer has a 5" x 5" x 5" work area, so we had to make some adjustments. For one thing, we had to split the nose cone into two parts, and the whole fin assembly had to be broken down into separate parts.

James was doubtful that the Mosaic would be able to print the fine bolt head detail, so he removed them from the CAD model.

3D_Print_003a.jpg

3D_Print_004a.jpg

3D_Print_005a.jpg

3D_Print_006a.jpg

3D_Print_007a.jpg
 
After James cleaned-up my scaled SWX models, he followed these steps:

1. Model your concept in SolidWorks (or any good quality 3D software) and save the model to a .STL format.

2. The 3d Printer (Mendel/RepRap process) will use a print software similar to Pronterface.

a. Within Pronterface is another software that runs called Skeinforge. (some people use Slic3r). This software is used to "slice" the .stl file into printable layers of gcode.

b. Once the gcode has been generated, Pronterface is used to send the gcode to the printer.
 
The process to create the gcode ran about an hour per part. The actual printing process took 2 to 3 hours per part, depending on the complexity of the part. The results were both amazing, but a little disappointing. Amazing that we took a part that existed electronically and created something you can hold in your hand, but the surface finish was less than I had hoped for.

Those squarish pieces on the shoulder are supports created by the printer software to help support the wall during the print process. The printer feeds a plastic filament through a heated nozzle, applying it to either the table for the first layer, or the previous layer. Since it is molten as it leaves the nozzle (think hot glue gun) it can slump. There are very tiny ligaments holding the supports to the part which can be easily snapped-off. Sanding removes any trace of these. In fact one broke off while I was taking the pictures, leaving three still attached.

3D_Print_008a.jpg

3D_Print_009a.jpg

3D_Print_010a.jpg

3D_Print_011a.jpg
 
The nose tip turned out fairly well, but James discovered that as the print head got near the tip, it was moving so fast that it drug the molten material with it, leaving it skewed, looking like a soft-serve ice cream cone. I assured him I could fix that with some sand paper on my lathe. He came up with a fix on another rocket project, blowing through a straw to cool the layers as the print head got near the tip. That worked, and results looked great!

3D_Print_012a.jpg

3D_Print_013a.jpg

3D_Print_014a.jpg
 
We discovered the fins will not print well. Since there was detail on both sides of the fin, it could not be printed flat on the table. James tried two different directions of printing the fin standing up, neither turned out very well.

3D_Print_015a.jpg

3D_Print_016a.jpg

3D_Print_017a.jpg

3D_Print_018a.jpg
 
The fin attaching rails were, to me, the coolest part! Since the printed fin did not work out, I decided to use balsa with a paper cover to obtain the sharp leading edge.

3D_Print_019a.jpg

3D_Print_020a.jpg

3D_Print_021a.jpg

3D_Print_022a.jpg

3D_Print_023a.jpg
 
The shroud ended up being so thin (~1 filament thick), it really taxed the limits of the filament printing process…there are holes and gaps along the seam where the print head starts and stops each print layer.

In the photo you can see another support ring that the printer software creates that is removed with a combination of cutting and sanding.

I was not able to get the shroud to fit over the BT-60 completely—I was afraid it was going to split open, so I slid it on part way, then dry fitted the rail and fin to get a feel for what it would look like completed!

3D_Print_024a.jpg

3D_Print_025a.jpg

3D_Print_026a.jpg

3D_Print_027a.jpg
 
Instead of designing the shroud to go over the tube, I redesigned it with a shoulder to fit in the BT-60 (and a built-in centering ring), and glued the lower motor mount paper centering ring into it.

3D_Print_028a.jpg

3D_Print_029a.jpg

3D_Print_030a.jpg

3D_Print_031a.jpg
 
As mentioned previously, I turned the tip of the nose cone on my wood lathe to get rid of the “curly-q” (now looking at the primered part photo, it looks like it needed a little more work :) ). I then thinned down some Bondo Glazing and Spot Surfacer putty w/ lacquer thinner to fill in the little voids and the seam running down the part.

After sanding, it was interesting to see the pattern formed by the printer as it created the part. I then sprayed on a heavy primer called "Mr. Surfacer 1000" https://www.swannysmodels.com/Surfacer.html -- I let that dry thoroughly before sanding.

3D_Print_032a.jpg

3D_Print_033a.jpg

3D_Print_034a.jpg

3D_Print_035a.jpg
 
The fin rails were glued to the bottom shroud, the motor mount assembled, and everything was glued into the main body tube. I brushed on "Mr. Surfacer 500", sanded a bit, and applied some of the thinned Bondo to fill in larger voids.

3D_Print_036a.jpg

3D_Print_037a.jpg

3D_Print_038a.jpg

3D_Print_039a.jpg
 
The fins are balsa cores with some cardstock skins glued on to get that sharp edge on the leading edge of the fin. I filled the edges w/ some Elmer's Carpenters Filler and apply a coat of CA (superglue) to the paper to stiffen it. When the CA dries out, the paper feels like plastic!

3D_Print_040a.jpg

3D_Print_041a.jpg
 
A little more sanding and finishing on the paper tubes (I used thinned Elmer's Carpenters Filler for that) then an overall coat of primer.

3D_Print_042a.jpg

3D_Print_043a.jpg

3D_Print_044a.jpg

3D_Print_045a.jpg
 
Here is the finished model! According to Rocksim it will fly 588 ft on a D12-5 or 1015 ft on an E9-6. Hopefully it will fly at our next club launch, if the winds cooperate!

3D_Print_046a.jpg

3D_Print_047a.jpg

3D_Print_048a.jpg

3D_Print_049a.jpg

3D_Print_050a.jpg
 
As I stated earlier, any of the statements I made about printer capabilities are applicable only to the MakerGear “Mosaic” low end printer. While I was a little disappointed in the surface finish and resolution, it is still very cool to be able to take a CAD model and make something you hold in your hands (the industry calls this “art to part”). There are two other members of my rocket club at work, and when James brings in our parts, we gather in his cubicle and giggle like school girls over the parts. With these hobbyist printers, it is not quite “click and print” … yet. There is some work involved in the conversion, and creating the gcode. A true gcode programmer might be able to tweak the program and get better results than we did.

The filament material used with this printer costs about $15 per pound. The nose weighs 1.53 oz. as-printed. The second shroud weighs 1.47 oz. The fin rails weigh about 0.04 oz. each. I am not sure how a professional printing bureau would charge for these parts as there is the conversion, coding, and actual printing time involved too.
 
It turns out there is a whole world of hobbyist printers, mostly home-built units. A recent addition to the “under $1000” line is the Printerbot ($550 - $700 depending on model) from https://printrbot.com/

Here are some other 3D printer links from the “Hack a Day” site and others (sent to me by guys at work)…

https://hackaday.com/?s=3D+printer+reprap

https://hackaday.com/2010/08/29/from-repstrap-to-reprap-a-3d-printer-is-born/

https://hackaday.com/2012/02/08/yet-another-3d-printer/

This is a really high resolution printer that uses light curing resin
https://hackaday.com/2012/04/05/3d-printer-with-insane-accuracy-uses-a-dlp-projector/

https://store.makerbot.com/

https://www.geek.com/articles/gadgets/printrbot-an-all-in-one-3d-printer-kit-for-499-20111121/

https://www.kickstarter.com/projects/printrbot/printrbot-your-first-3d-printer?ref=video

Some well-known high end industrial printers are:

Z-Corp
https://www.zcorp.com/en/home.aspx

Stratasys
https://www.stratasys.com/

Objet
https://www.objet.com

A good reference to CAD and rapid prototyping is “Desktop Engineering” magazine – I get it free at work, and I think you can look at online content for free too.
https://www.deskeng.com/
 
While I was a little disappointed in the surface finish and resolution, it is still very cool to be able to take a CAD model and make something you hold in your hands

It was pretty cool to read about it, too! I had wondered about the ability to "print" components. Wow, a lot of work goes into the programming, and then cleaning up the pieces afterwards. But the technology has come a long way, and I can imagine a world where you design a rocket in Rocsim, send the file to a printer, and out pops a completed rocket. Our grandchildren will think that's how it has always been, and us old folks will all complain about the loss of construction skills in modeling. Thanks for sharing. Sather
 
Cool review!

I worked with a StratusSys machine (pro machine) 10 years ago as an undergrad and it was pretty impressive then. From your pics and descriptions looks like the Mosaic is on par or a little better with where they were then. We used it for all kinds of things and found that wing sections had to be printed sliced down the span, (so each layer was an airfoil shape) if they were printed sliced from leading edge to trailing edge they would delaminate in the wind tunnel if they were uncovered (glass or carbon) also the vertical printing lead to a lot cleaner airfoil shape. One thing we learned very quickly to improve finsih was to set the STL save parameters much tighter than the CAD software's default, reduce the "polygoning of the shape"

If I remember right we saw some of the same tip curling that you did but it was more the head nicking the previously laid layer than a heat thing. I forget how we adjusted it though as that was shortly after we got the machine.
 
Success!!! Two beautiful flights today. 1st flight on a D12-5 (nearly missed it in the pic!). The 2nd on an E9-6. Very straight, nice recovery.

DSC_4716_CropMR.jpg

DSC_4705MR.jpg

DSC_4706MR.jpg

DSC_4715_CropMR.jpg
 
Here are the results of my printing the Tomahawk files on a MakerBot Replicator. Layer height is set to 0.2mm, but all other settings are default. The first pic is the shroud and nose cone parts. That's the way I printed the bottom of the nose cone, with no support required. In the assembled nose cone, there is a considerable gap between the parts, due to no support of the shoulder step of the top piece. The gap isn't so bad that it couldn't be filled pretty easily. Other than that fill, the shroud and nose cone wouldn't require more than a light sanding and primer. Third pic is the fin, on it's raft and support structure. Fourth is the fin in the fin holder. The light is at the angle to show the layering at it's worst; actually, you can barely feel most of it. Again, some sanding and some light fill/primer and I believe it would be totally usable. Time to print all of it (with four fins) would run about 20-24 hours. Taking a WAG at the plastic cost, it might run about $3-$5. I'll have to go look and see if I have a BT-60 somewhere to try the fit. I still intend to try printing the thin shroud.

Gary

nc pieces and shroud.jpg

nose cone together.jpg

fin on raft support.jpg

fin and holder.jpg
 
Here are the results of my printing the Tomahawk files on a MakerBot Replicator. Layer height is set to 0.2mm, but all other settings are default. The first pic is the shroud and nose cone parts. That's the way I printed the bottom of the nose cone, with no support required. In the assembled nose cone, there is a considerable gap between the parts, due to no support of the shoulder step of the top piece. The gap isn't so bad that it couldn't be filled pretty easily. Other than that fill, the shroud and nose cone wouldn't require more than a light sanding and primer. Third pic is the fin, on it's raft and support structure. Fourth is the fin in the fin holder. The light is at the angle to show the layering at it's worst; actually, you can barely feel most of it. Again, some sanding and some light fill/primer and I believe it would be totally usable. Time to print all of it (with four fins) would run about 20-24 hours. Taking a WAG at the plastic cost, it might run about $3-$5. I'll have to go look and see if I have a BT-60 somewhere to try the fit. I still intend to try printing the thin shroud.

Gary

WOW! Those turned out nicely! I showed this to James and he was impressed too. He was puzzled why the supports on the nose tip did not print. Do you think the "grain" on the fin will cause any problems? James printed one fin this way, and another with the rear edge of the fin being on the table (i.e. the "grain" more or less parallel to the L.E. of the fin, like a balsa fin).

You can see it in your print: our original plan was to print the fin and the fin rail as a single unit. Later James removed the rail from the fin and asked me how wide did I want the opening--I said 1/8" so there is a slight imprint of the rail on the side of the fin -- makes it easy to align the fin on the rail now :)

Does your printer software allow a "sparse" setting? James printed that fin shroud so the exterior was "dense" and the interior was "sparse" to reduce the weight of the part. My finished model did not require any nose weight based on RockSim.
 
Hi rocketguy101 - just an FYI that you should be getting much better print quality from the Mosaic. This print is almost a year old but your prints should look a lot more like this. You can find more examples here

Fins - side.jpg

Rick
 
Hi rocketguy101 - just an FYI that you should be getting much better print quality from the Mosaic. This print is almost a year old but your prints should look a lot more like this. You can find more examples here

View attachment 82770

Rick

That's pretty cool Rick!

How long did it take to print and have you flown it yet?

Greg
 
Hi rocketguy101 - just an FYI that you should be getting much better print quality from the Mosaic. This print is almost a year old but your prints should look a lot more like this. You can find more examples here

Rick
Hi Rick, welcome to the forum! James said you emailed him after seeing these posts here. Perhaps you could provide some tips on how to get better performance? He really likes your printer and we have discussed possible tweaks. I hope this thread doesn't come off as "dissing" your product. I just wanted to point out to fellow rocket modelers that there is some work involved w/ these low cost printers--that it's not quite to the point of hitting "Print Screen" and magic happens.

I think its fantastic that this technology is becoming affordable!! And I hope you see there is interest here on the forum (there are a couple other 3D printing threads here) so we hope you stick around and perhaps provide some pointers!
 
Hi Rick, welcome to the forum! James said you emailed him after seeing these posts here. Perhaps you could provide some tips on how to get better performance? He really likes your printer and we have discussed possible tweaks. I hope this thread doesn't come off as "dissing" your product. I just wanted to point out to fellow rocket modelers that there is some work involved w/ these low cost printers--that it's not quite to the point of hitting "Print Screen" and magic happens.

I think its fantastic that this technology is becoming affordable!! And I hope you see there is interest here on the forum (there are a couple other 3D printing threads here) so we hope you stick around and perhaps provide some pointers!

No offense taken - I was just bummed by the print quality. We have an IRC channel where you can chat with other printer operators. You can also post questions/STLs to the MakerGear google group. There is a substantial and growing community of printer owners willing and able to provide assistance.

@GregGleason - that was just a print that I did for fun a while ago - never printed the rest of it :tongue:

Rick
 
Been having lots of fun with the printer. Pic is some NC's, nozzles for a scale static model, and a Black Brant forward launch lug. SD card for size compare.

RocketPrintFun.jpg
 

Latest posts

Back
Top