For a network project at my church, I printed up a bunch of these nifty Ethernet cable organizing clips designed by Murray Clark on Thingiverse. They come in a variety of sizes, with or without holes to fasten them down with cable ties or screws. You just squeeze one side to open the "jaw" and slip one or two cables in on the other side.
I didn't notice that they were designed for CAT5 cable, and our project used mostly CAT6 which has extra shielding. So when I tried to use them, they were too tight, and the cables didn't really fit in. That's OK, even though the files are compiled .STL and I can't really manipulate them. MatterControl has a Scale feature, so I figured out how to use that. I took some measurements, did some experiments, and eventually settled on scaling up by 8% in the two dimensions that affect the hole diameters. (No need to scale up in the third dimension, because thicker is not necessarily better.) I had printed the CAT5 clips in white, so I decided to print all the CAT6 clips in blue so I could tell them apart.
But when I removed the first clips from some test cables, they left deep slits in the insulation. What? Not good! Why was that happening? Well, I had noticed that many of my prints had a little "flange" on the first layer or two. I thought that was just the way things worked. It was annoying on my bottle cap project, because it was rather sharp when I twisted the cap on and off. Here's a pic in profile with a contrasting background. Look at the flange on the left side, and you can see how it could make a nice little razor knife inside each cable hole:
A little research showed me that this is called "elephant foot" and there are definitely ways to deal with it. It's caused by the first layer getting squashed by the upper layers. That can happen because of one or more of the following:
- Bed is too hot, so the lower layer doesn't firm up and the weight of the next layers squashes it.
- First layer is too thick.
- The nozzle to bed distance is not accurate, so the calculations are off. That could foil efforts to adjust #2.
I tried varying the bed temp and it seemed to have little effect, but I left it lower. I tried varying both #2 and #3 and saw progressive improvement. I settled on making the first layer the same height as the rest (the default was 50% thicker... not sure why) and adjusting the Z offset. I also turned off a setting that was turning off the fan for the first two layers (double negative there).
As I reduced the "elephant foot" I found I could reduce the scaling as well: that flange was significantly intruding on the hole.
At this point I'll mention that I keep track of the settings for all my print jobs in an Excel spreadsheet. That's because there are so many variables in 3D printing that it would be really easy to lose track of what works and what doesn't. I'm a software specialist, and have done lots of systematic testing in my career, and I've found it's important to either vary only one thing at a time, or if that's not possible, keep good records so you can go back over them. Here's a sample showing the parameters I was varying:
I had some trouble with the Z offset, as the manual and software help didn't make it clear whether to go positive or negative. My first tries were the wrong direction, causing the nozzle to be right on the bed. That did a good job of fooling me into thinking I had a nozzle jam. Eventually I reversed that setting and it went much better.
The final settings resulted in nearly zero "elephant foot". I don't have a good picture of the final result... they're all down at the project site now. The pic below kind of shows it. But here's also a fun picture of how this can all go wrong. I printed a batch of 12 clips at once (of various sizes and shapes) and because I didn't refresh the hairspray on the bed (or didn't have enough in that corner), one of the clips came loose. Being in the corner, the printing direction caused it to kind of stay in the area, so the nozzle kept coming back to it and adding to the mess, instead of making a total mess of the whole job.
