Notes on Print Quality and Post-Processing

Once printed with finely tuned printing settings, the resulting parts generally have rough surface across the layers. The magnitude of the grooves between the layers is directly correlated to the deposition bead dimensions, which in turn depends on nozzle diameter. For continuous fiber printing, this thus relates to the tow-count of the deposited fiber. This surface roughness is undesirable for applications where flow of a medium along the printed part is desired. It is also a natural 3D printing part defect, and thus crack initiator which generally causes the weakening of the prints along the build direction, as was extensively discussed in section 2.3, and shown in figure 4.37. For this multitude of reasons, it is important to assess post-processing techniques for 3D printed parts with continuous fibers. A variety of post-processing or surface finishing methods are available, some are listed here in order of prevalence to the 3D printing world:

• Sanding or machining: Using mechanical abrasion tools to remove the rough layer from the part. This is generally done with sandpaper or rotary tools, and industrial environments often use CNC milling as a post-processing step. This

Figure 4.37: Post-processing can leave a mirror-like surface finish

can result in parts with tight tolerances, and can leave a mirror-like surface finish if desired. The drawback of this step is that material is removed, and for thin-walled applications, stiffness or strength requirements may render this method impossible.

• Solvent vapor treatment: Careful submersion into a solvent vapor chamber can smoothen the outer layer, eliminating ridges significantly. This can reduce a part’s surface roughness with minimal effort spent, and can produce a mirror- like finish. This method can very fast (solvent treatment of ABS parts with acetone vapor can be done in as little as 2 minutes) and can be tricky if aggres- sive solvents are used. Excessive exposure can cause dripping or deformation (and even disintegration) of the parts.

• Coating or solution application: Manual (or automated) application of a (polymer) finishing solution to fill in the ridges. Also used for adding material for sanding, machining or coloring/painting. This method is heavily used in the prototyping industry for demonstration parts or (scale) models.

In this research, each of these methods has been tested briefly to assess the applica- bility. The methods used were selected such that the treatment method resulted with

the composite in the most ”pure” state possible: without introducing other chemicals or colorants. This was a requirement originating from the aerospace ducting market.

4.4.1 Abrasive Post-Processing

An example of abrasive post-processing is shown in figure 4.38. The printed specimen was sanded using various grit sanding paper resulting in a smooth surface. This method would be a time-intensive and manual step, but could be automated to some extent.

(a) (b) (c) (d)

Figure 4.38: Effects of sanding: (4.38a) Original, (4.38b) after sanding with 120 grit paper (4.38c) after sanding with 220 grit paper and (4.38d) after sanding and polishing with 600 grit paper

4.4.2 Vapor Chamber Post-Processing

Polymers dissolve in certain solvents, and for PEI, we know it dissolves well in chloro- form from the research on the impregnation of the dry fibers and chemical resistance tables. Therefore, a chloroform vapor treatment has the potential to smoothen out sharp ridges in a printed part. In this post-process, chloroform is boiled in an enclosed chamber containing a specimen, and after a certain exposure time, the chloroform gas dissolves the surface of the part, which affects its surface roughness. If done correctly, the part surface finish is noticeably smoother. If the specimen is underexposed to the chloroform vapor, little or no change will be observed. If the specimen is overexposed,

thinner specimen will start to degrade while thicker specimen will show drip marks and eventually degradation will occur. Timing is therefore a critical aspect of this process. This process was used to smoothen some specimen, with a test setup that is

Figure 4.39: Chloroform vapor trial setup

schematically shown in figure 4.40. The best results were obtained when setting the

heater temperature to 55◦C and an exposure time (between opening and closing of

the chamber lid) of about 120 seconds. It should be noted that several parameters in this experiment were not measured, including concentration of chloroform particles (amount of liquid added in the chamber and chamber volume), so further testing would be required to accurately determine the optimal smoothing parameters. Some strength increase was noticed after post-processing due to the redistribution of PEI on the printed part. However, condensation on the part is an issue and accurate exposure time control are necessary, therefore, the setup described in figure 4.41 is proposed for future trials. In this setup, a fan ensures circulation of the vapor around the part to ensure overall equal exposure, and a separate heating and boiling reservoir with a fill valve allows for controlled exposure time. A release valve on the specimen chamber ensures safe removal of the lid as pressure may have built up during the test. The specimen is suspended from the lid on a porous mesh to ensure adequate airflow and prevent condensation or pooling on the cradle. Both the specimen cham-

(a) 0 seconds (b)_u120 seconds (c)_u180 seconds Figure 4.40: Chloroform vapor treatment trials: (4.40a) untreated specimen (4.40b) smoothened and (4.40c) overexposed specimen

ber and reservoir are heated to ensure the atmosphere contains adequate solvent and to prevent condensation.

Figure 4.41: Vapor chamber post-processing setup

4.4.3 Brush-on Post-Processing

The third post-processing step used was brush-on application of solutions and sol- vents. In this case, again chloroform is used. Three different mixtures were applied using a medium-stiffness paint brush (figure 4.42): pure chloroform, a solution of 3%

PEI (identical to the one used for impregnation) and a solution of ≈15% PEI and

chloroform. A rounded (hollow) cube was used such that each of the 4 outer sides had a similar surface finish before processing, eliminating printing process variables,

such as material or feedstock variation, as much as possible (figure 4.43). The appli- cation of the coating was done in a cross-hatch pattern, alternating with the grain and across the layers every 5 layers, to generate as equal of a coating as possible. The application was iterated until the part alteration was visual. The resulting surface

finishes are shown in figure 4.44. These surfaces were cut and cast for microscopy,

Figure 4.42: Post-processing through brush- on solution application

Figure 4.43: Untreated printed sur- face

(a) (b) (c)

Figure 4.44: Effect of various brush-on post-processing solutions: (4.44a) after appli- cation of 75 strokes of chloroform, (4.44b) after application of 20 strokes of 3% PEI

and chloroform solution and (4.44c) after application of 10 strokes of≈15% PEI and

chloroform solution

the resulting cross-sections are shown in figure 4.45. The results show how brushing with pure chloroform has little effect, other than slightly affecting part shine, even after extensive brushing operations. Once PEI is added to the brush-on solution, the

surface shows direct effect, and higher concentrations of PEI have more pronounced effects. The brush-on amount should not be exaggerated, as the microscopy shows that this coating contains a significant amount of air pockets.

(a) (b) (c)

Figure 4.45: Microscopic effects of various brush-on post-processing solutions: (4.45a) after application of 75 strokes of chloroform, (4.45b) after application of 20 strokes of

3% PEI and chloroform solution and (4.45c) after application of 10 strokes of ≈15%

PEI and chloroform solution