Internal structure is affecting my perimeter quality! Halp

[nirakara]

I've designed a sword for a cosplay i'm working on which has an internal channel to thread an LED strip. Unfortunately my print has scarred and brittle walls... but only where there is internal structure. The same print has a column of just hollow material which prints beautifully, on every single layer in parallel with the messed up stuff. My best guess is too many retractions right now and i'm running some tests, but it's a slow process and maybe someone else has a better idea. I've posted pictures and the stl file here: http://www.thingiverse.com/thing:1707999 and I'm printing with T-Glase at 215 C / 70 C table. Again the hollow column is strong and beautiful so I don't want to mess with the any other settings, just slicing and maybe CAD. Thanks so much!

P.S. In the picture the left leg fell off, that's how bad the print is. It just has some goopy epoxy on it right now. And yes there was supports under the triangle but the bad quality continues above it for the rest of the 6" of height it has.

[nirakara]

Update: I forced S3D to get rid of most of the retractions for this part and it looks a bit better. At least it feels more structural (the old wall flaked a lot, this one less so), but the part with the internal structure still isn't perfect. There are striations that visually align with the supports and there's a lot more travel over these sections than really seems necessary. I've noticed my supports are more solid than they should be so it seems like i'm overextruding a bit; with all the extra movement it might be scraping over the layers. Still if there was a way to force less travel it would be helpful.

[nirakara]

Additional note: S3D FFF file is on that thingiverse link too. I'm going to try the "Random starting point" option to avoid the movement over the perimeter next. We'll see.

[nirakara]

Update: I have not tried the last two things I mentioned as I was playing with S3D to figure out what else I could do. I'm leaning towards a Z-lift move since nothing else I do in the software will avoid all the travel above the perimeter. I've also realized the internal structure (an extruded cut in solidworks) was printing on top of the perimeter wall the way that S3D was slicing it. It was hard to tell because it would just draw the path overlayed with the perimeter path without throwing any kind of error. I had CADed it to leave a 0.5mm wall (my nozzle is 0.35), so I upped it to 0.7mm and now it shows up as a second internal line inside of the perimeter. We'll see what happens.

[nirakara]

Well I think I've fixed it about as well as it can be fixed. Fixing the cad so the internal walls didn't overlap with the perimeter helped quite a bit (I'm looking at you S3D for this one) and I've increased the travel speed up to 150mm/sec and increased retraction distance too to keep it clean. Hopefully putting my notes on here will help someone too in the future.

[Derrick]

Well I think I've fixed it about as well as it can be fixed. Fixing the cad so the internal walls didn't overlap with the perimeter helped quite a bit (I'm looking at you S3D for this one) and I've increased the travel speed up to 150mm/sec and increased retraction distance too to keep it clean. Hopefully putting my notes on here will help someone too in the future.

You may want to try printing hotter. I think 235-240C may be better. If you have bad layer-layer adhesion it usually means you didn't extrude hot enough. OR you can print slower, which will also help the adhesion. I've never been able to print with that stuff below 225C and get a good stable part.

[nirakara]

Thanks, I'm seeing a little vertical separation and I'll try out hotter for that. I just didn't want to increase the temp when my main problem was leaving melted plastic walls. But now I've solved one problem I can start on another!