*3.3. Surface Analysis*

Even though obtaining a combination of lower *Ra* and *Rz* values is typically enough to affirm an improvement in the surface finish of a part, microscope imaging provides a deeper insight into the reasons behind surface changes and their nature. In this sense, Figures 3–5 present a series of optical microscope images and SEM micrographs aimed at the direct comparison between the macroscopic and microscopic state of the treated samples. Images of an untreated or pristine part have also been added in Figure 2.

**Figure 2.** Digital microscope (**a**,**b**) and SEM micrographs (**c**) of untreated Ultem parts.

Figure 3a reveals that, compared to the pristine case, Ultem has lost some of its characteristic shine after being submerged in a mixture of 1,4-dioxane and toluene for 4 h. This tonality change could be explained by the presence of an Ultem residue that has been partially dissolved and non-uniformly redeposited along the interfilament space

(some zones are more coalesced, but some are not), as the more magnified central and right images (Figure 3b,c) show.

**Figure 3.** Digital microscope (**a**,**b**,**d**,**e**) and SEM micrographs (**c**,**f**) of the chemically treated Ultem parts.

As the roughness analysis has revealed, a drastically different surface is obtained when Ultem is vapor-smoothed in the presence of chloroform (Figure 3d–f). Microscopy results display a smooth surface where the additively manufactured nature of the part is almost inappreciable as Ultem filaments from the outer surface have completely melted and resolidified, forming a uniform layer. This result confers perspective and validates the effectiveness of the other analyzed chemical treatment: In comparison with vapor smoothing, Ultem's surface is much less affected despite being in direct contact with a chemical mixture capable of dissolving a support material with a similar chemical structure to Ultem's.

Regarding the visual inspection of the thermally annealed samples in Figure 4, the overall shape of the specimen is lightly modified as the corners appear more rounded, and the upper edge of the part has lost its perpendicularity with the other edges. Nonetheless, magnified images (Figure 4b,c,e,f) demonstrate that the air gap between adjacent filaments has diminished in the upper face and has almost completely disappeared in the lower face. Taking this into consideration, further experiments should be conducted at a slightly lower temperature (lower than 210 ◦C, but higher than the material's Tg) to try to reduce dimensional changes while still achieving better filament adhesion. Ideally, the presence of a solid or liquid media that enters in direct contact with the totality of the part during the thermal treatment would help achieve an even smoother surface, as detected in the inspection of the lower side of the thermally annealed part.

**Figure 4.** Digital microscope (**a**,**b**,**d**,**e**) and SEM micrographs (**c**,**f**) of the thermally treated Ultem parts.

Figure 5a–c demonstrates how the characteristic rounded shape of FFF manufactured parts is modified due to the ball burnishing process, as predicted from the roughness analysis. Ultem filaments appear flattered, and the overall state of the filament's surface, not damaged. On the contrary, the surface of the mechanically treated parts using abrasive treatments shown in Figure 5d–l appears to be radically modified. While the average surface roughness or overall dimensions of the treated parts do not reveal such intense changes, the surface of a single filament has been deeply affected.

Samples that have been shot blasted with white corundum (Figure 5d–f) have suffered the most remarkable changes. The impact of small, non-rounded beads made of a material as hard as corundum has eroded the surface and completely changed its morphology. The Ultem part's external filaments show cracks and numerous irregularities; Ultem has lost its shine, and the part appears more mattified. Nevertheless, the presence of filaments looks more blurred, meaning that the overall aspect of the part is more uniform. This can be regarded as a positive outcome, as a specific surface morphology can be desirable to promote adhesion of further coatings that want to be applied to the part.

In the case of abrasive shot blasting with glass beads (Figure 5 g–i), the surface is not as eroded as with white corundum due to the softer nature of the used abrasive and its rounded shape, but the surface still shows noticeable alterations in the form of small protrusions generated by the beads' impact. It should be remembered that initially spherical glass beads tend to break down into smaller and more protruded parts during the shot blasting process. Something to mention is that the right corner of the treated part shown in the macroscopic image Figure 5g has been more damaged than the rest of the surface, revealing one of the main drawbacks of these abrasive techniques: the need to automate or very precisely control the exact time, incidence angle, and distance of the abrasive gun. While in the case of chemical or thermal treatments a slight time deviation in the duration of the treatment is perfectly acceptable, in the case of abrasive shot blasting a few seconds' deviation can be detrimental to the final result.

**Figure 5.** Digital microscope (**a**,**b**,**d**,**e**,**g**,**h**,**j**,**k**) and SEM micrographs (**c**,**f**,**i**,**l**) of the mechanically treated Ultem parts.

Finally, microscopy images from the shot peened samples (Figure 5j–l) show a shinier surface due to the presence of metallic residue from the stainless-steel beads. Interestingly, despite the longer duration of this treatment (10 min versus 10 to 20 s in the case of abrasive

shot blasting), it has resulted in a more uniform surface due to its more controlled nature (shot peening was performed in an automated chamber and with a higher distance between the gun and the samples). In addition, the impact of spherical metallic beads has induced pressing of the part's external filaments, which denotes the application of compressive stresses that could be beneficial for the fatigue life of the parts. Considering the simplicity of this process and the insignificant affectation of the parts' dimensions, shot peening's effect on the mechanical properties of FFF Ultem parts is a key point that should be addressed in future works.
