Printing Paul's Model of Mars

[Flack]

In this thread, I will document my effort to 3D print a physical version of Paul's 3D model of Mars.

Background: In September of 2017, Pinback asked Tdarcos if he could create a 3D model of the planet Mars to place next to a virtual roller coaster. Tdarcos made and built the model, but neglected to mention in advance that he expected to be compensated for his work. After delivering the model, he asked Pinback for $10 in exchange for his work. In response to this, Pinback sent Tdarcos $10. Just kidding. Instead, Pinback referred to Paul's model as "an orange ball on a stick," Tdarcos congratulated Pinback on "stiffing a cripple" declaring him "dead to me," and Pinback called Tdarcos "an ungrateful, clueless asshole," because everybody on this BBS, including me, is insane.

Because $10 means more to Tdarcos than the rest of us, I emailed Paul in October and asked if I sent him the $10 he felt Pinback stiffed him, would he allow me to print out a physical copy of his model. It took Paul five weeks to respond to my initial email and another three to send me his PayPal address, but after almost eight weeks I was able to pay Paul and officially begin this project.

[Flack]

Here are the steps I think I need to complete: convert the model to a format I can work with, import the model into TinkerCad, export the model from TinkerCad into Cura, generate the necessary gcode, print out the model, and possibly paint the model.

The first part of the project is converting the model Tdarcos made into something that can be printed.

The file Paul uploaded was in .DAE format. My 3D printer (a Creality CR-10), and essentially all 3D printers, print using gcode. The program I use to generate gcode, Cura, does not recognize .DAE files, but I know it uses .STL files. A quick Google search returned this free online file converter which was able to convert the .DAE file to an .STL file.

With the model now in .STL format, I was able to easily import it into TinkerCad. The model looks a little different without the Mars texture applied to it, but I won't be printing that part. From here, I should be able to export the model into Cura.

[Flack]

And, almost immediately, I hit a snag.



In this screenshot from Cura, you can see the program has highlighted the underside of "Mars" in red. The red indicates a portion of the model that cannot be printed in real life. Because 3D printers print objects from the bottom up, there is nothing to support the underside of the sphere. Turning the model on its side does not fix the problem.

Immediate solutions appear to be either cutting the model in half down the middle and gluing the two pieces together, or printing the whole thing in three parts (orange ball, stick, base) and then assembling them.

3D printing can best be described as a series of hurdles and headaches. More to come after I figure out what to do next!

[Flack]

I imported the .STL file back into TinkerCad and worked on splitting it. Surprisingly, TinkerCad has no "split" feature. Instead, I had to import the model twice, deleting the left half of the model the first time and the right half the second time.



During this process I discovered that the pole supporting the sphere isn't perfectly centered. It's only off 1mm; not detectable to the naked eye, but showed up when splitting the model exactly down the center according to the virtual grid. I don't know if this was the case in the original .DAE file, or a slight aberration that was introduced by converting the model to an .STL file. Because I plan to glue the two parts back together, I don't think this will matter in the end.



This is a good stopping point for the night.

[Flack]

After splitting the model, I was able to import it into Cura. If a model is printable, Cura will calculate an estimated print time and how much filament will be required. With the base scaled to 180mm square, Cura is estimating a print time of 4:07, using 38.24m or 114g of filament. For reference, the current spool of filament I'm using costs $16/kilogram, meaning this model at this size will use approximately $1.82 worth of PLA filament.



A view from underneath the model shows a possible issue -- some lines where the support pole connects to the base and the planet show slight yellow lines, which in theory denotes that they are not connected. This may just be an issue with the way Cura is displaying the model. At this point, there's really only one way to find out.

[Flack]

15 minutes into the print process, things are looking... questionable.



While I can see the shape beginning to take place, I can also visually see the gap between the shaft and the base that appeared in Cura. I don't know why Cura isn't seeing it as one solid model.

All I can do now is wait 4 hours and see what happens.

[Ice Cream Jonsey]

I'll be official ly back in 2 days. Will Mars be ready? Will Earth exist??

[Flack]

I'll be official ly back in 2 days. Will Mars be ready? Will Earth exist??

Well, we'll have something!



The model continued to print, but for some reason, things just didn't look right. Again, 3D printing is a fickle process. It's hard to tell if there's something wrong with the model itself or something is going wrong with the print itself. Sometimes a different of 5-10 degrees can be the difference between a smooth print and a rough mess.



I generated a simple 20% fill for the interior of Mars.



I ended up with a lot of stringing on these. That's definitely not an issue with the model, but rather something to do with the retraction speed and/or temperature.

Next step is letting everything cool down and seeing if I can get it off the print bed without destroying it.

[Jizaboz]

Hm it does look close. Maybe you can "sand" the rough edges somehow? Not sure how well that works with that material..

[pinback]

I'll give you ten bucks for it.

[Flack]

I'll give you ten bucks for it.

I predetermined you would be paying me $20 for it -- sorry.

[Billy Mays]

You could add tassels and turn them into bookends?