NSF Phase II: Inconsistent Extrusion of PLA

We have been doing plenty of GBX testing at re:3D since my last forum post, but I've fallen behind on posting. Let's get back on track.

Background:

One of the tasks of the NSF Phase II grant is to characterize printability of different size distributions of plastic regrind. To do this, PLA was sifted and extruded to determine whether different particle sizes require an increased extrusion rate. Sifters were 3D printed with varying hole sizes, and PLA granulate was sifted into 5 categories: 0-2mm, 2-3mm, 3-4mm, 4-5mm, and over 5mm. Calibration cubes were printed to determine the optimal extrusion rate for each size distribution. This was done by comparing actual layer widths of the print to the expected layer width from Simplify3D.

Results:

Unfortunately, layer width values varied wildly across calibration cubes printed with the same settings, indicating inconsistent extrusion. To diagnose the problem, tall cylinders with 1 perimeter shell were printed out of the 3-4mm sifted PLA granulate (Figure 1).

Figure 14: PLA cylinders printed in vase mode with 1 perimeter shell. When the translucent PLA is placed against a light, the variance in layer width becomes visible.

The cylinder printed at 195⁰C/190⁰C/170⁰C (bottom/middle/top heat zones) shows inconsistent layer widths for the bottom centimeter of the print before it recovers and prints more consistently. However, even after recovery, more subtle extrusion inconsistencies proliferate throughout the print, resulting in visible stripes when light passes through the varying thickness of the translucent PLA.

To test if decreasing the temperature improved extrusion consistency, temperatures were reduced at 5C intervals. The extrusion inconsistency at the print base disappeared at 185⁰C/180⁰C/160⁰C. However, the “stripes” still appeared. Reducing the temperatures further results in reduced surface quality starting at 175⁰C/170⁰C/150⁰C. Therefore, the minimum print temperatures are 180⁰C/175⁰C/155⁰C, at which point the calibration cylinder still has the stripes.

What could have caused this?

All these tests were performed with a version of the extruder barrel with empty space between the bottom of the compression screw and the nozzle below. Melted plastic accumulated in this space before extruding out the nozzle. This may have caused the extrusion inconsistency. A shorter extrusion barrel was designed and manufactured, reducing the empty space in the barrel.

Another possibility is inconsistent feeding of plastic regrind into the top of the extruder. The above tests were performed with consistently sized regrind sifted between 3 and 4mm, but other factors can influence feeding, such as the exposed fluting on the pellet screw and the feed tube geometry.

Next post:

Until extrusion is more consistent within the same print, accurate extrusion data cannot be obtained to compare different plastic particle size distributions. Therefore, the above tests were repeated with the new, shorter extruder barrel, and the results will be in the next post!