The Many Pitfalls of 3D Printing

If this is starting to seem like a computer blog, please bear with me. Three-D printing is really applied computer science, so you need delve into such arcane matters. Those of you that follow my blog know that I hate when computers take over my model railroad. That is one of the reasons I model the 1860s. I do not enjoy  wiring and dread the thought of maintaining a large fleet of locomotives with DCC technology. But, I can't deny that technology has made modeling a railroad set in this period possible as there is almost no commercial support. I have embraced computer aided drafting of track plans, laser cutting, DCC, battery power, microprocessors, LED lights, and now 3D printing to accomplish my goals.

So with that preface, let's look at another pitfall awaiting the modeler as he explores the dark, sticky morass of 3D printing. To date most of my 3D printing objects have looked incredibly detailed with nearly invisible layer artifacts. That is why I was puzzled by the facets visible in the first prints of the boiler I made. They required a significant amount of sanding to smooth.

Visible fastes on boiler sides 

I decided to see how hard it was to prepare the boiler for painting. I primed it with Rustoleum Filler Primer. I sanded it as described in my previous post.  I used 220-400-600 grits to get a nice smooth surface. Then I sprayed with a coat of Vallejo Black Primer which I had on hand. I don't like this primer as it doesn't seem to self level well, and you can not wet sand it. I don't plan to use it again. 

Paint test on an early version of the model
Next, I sprayed a coat of Future acrylic floor polish to get a nice glossy surface. When that dried, I sprayed  the final coat of Vallejo Natural Steel.  This is not the paint I plan to use for the final model but I had it on hand. The results were good in the areas that I was able to effectively sand. But against the boiler straps (which remain unpainted in the photo), and along the firebox, I could see artifacts of the print process. 

This got me wondering if there was something I was doing wrong in the 3D print work flow that created the facets. After some google searches and checking with the Fusion360 help forum I had the answer.

Non-computer jocks can skip the next two paragraphs. In my Fusion360 workflow, I had been using the "Export" command to save the artwork as a .stl file for 3D printing. For some unknown reason, Fusion360 uses a server to do that conversion. So, you have to send the file over the internet and it sends back the .stl file after about 3 or 4 minutes. This option does not allow any customization of the conversion settings. It is also slow as the file has to be sent over the internet. In this case the resulting .stl file was 75MB.

But, there is a better way. If you right click on the object in the Fusion360 component browser a pop up menu with several options opens up. Buried in that list is the option to "Save as stl." If you select this, another pop-up window opens with options to change the settings of the stl conversion, including the option for a "highest" quality file. The default was "medium" quality. Now that was obvious! (sarcasm alert)

For the small objects I printed so far, medium quality was good enough. But on the large boiler the medium setting was not good enough. So I saved the file as a high quality stl. The resulting file was 550MB, nearly 10 times larger. But it saved it nearly instantly. Win-Win.

So I set about printing the boiler again. After 9.5 hours on the printer I ended up with a much nicer product.  There are no visible printer artifacts as you can see in the photo below.

New boiler print on the highest quality settings
 The frame in these photos is a temporary one made with laser cut MDF. I have ordered a 3D printed part from Shapeways using SLA "Accura Xtreme." which according to their website is "Accura Xtreme is a gray and rigid acrylate-based plastic. It is 3D printed using a large format stereolithography (SLA) printer capable of producing small to large parts with high resolution and detail as well as smooth surfaces with limited layer lines. This material is well suited for challenging functional assemblies with a smooth, injection molded-like surface finish." Seems like a good option for the frame.  If that is not good strong enough or is unsightly, I will pursue other options. Once I have the frame, I will set up the motor and gear box and  try to get it to at least turn the front drive wheel.

In the meantime, I have other tasks to do like make the cow catcher, the cab, many sundry other parts,  and prepare the artwork for laser cutting the rods and fine details. Those will be sent to a commercial laser cutter than can handle cutting metal. 

Source: USMRR Aquia Line and other Model Railroad Adventures

Updated: December 4, 2020 — 12:08 am