**3. Results and Discussion**

#### *3.1. Mass and Dimensional Changes*

The mass and dimensional changes of all specimens subjected to the different postprocessing techniques are summarized in Table 2. Differences in mass, height, width, and length are given as absolute values.


**Table 2.** Mass and dimensional changes as a function of the post-processing treatment.

Overall, solvent support removal, abrasive shot blasting, and shot peening have not induced substantial changes in the mass or the main dimensions of the treated parts. This means that there is no detectable solvent absorption or Ultem dissolution during the support removal process. In addition, abrasive mechanical post-processes, performed under the previously described treatment conditions, do not abrase the material enough to induce detrimental changes in their shape.

Regarding vapor smoothing, it is the only treatment that has induced a significant weight increase in the samples. This change is more pronounced with increased treatment times, indicating a gradual absorption of chloroform vapors. Considering the initial mass of the samples was around 6 grams, the 0.35-gram increase reported after 270 min of treatment represents a change of almost 6%. A point to note concerning the dimensional analysis of this particular treatment is that, during the first 120 min, there is a slight reduction in all dimensions due to an initial smoothing of the outer surface as adjacent Ultem filaments fuse. As time progresses, despite retaining their overall shape, samples expand or swell in all directions. With even more prolonged exposure, samples would probably start to completely melt and lose their shape, which would be highly undesirable.

In terms of dimensional accuracy, specimens subjected to thermal annealing suffered the most remarkable transformation: they all expanded in the building direction (an average of 0.5 mm in height) and contracted in the remaining directions (1 and 5 mm in width and length, respectively). This correlates with the theory proposed by Zhang et al. [30] concerning the relaxation of thermal stresses created in the building direction, which seems to be valid at temperatures higher than the material's Tg as long as exposure is maintained within the herein presented ranges. It is also noteworthy to mention that higher temperatures induce more significant dimensional changes.

Finally, ball burnished samples experienced a decrease in height (coinciding with the ball burnishing direction) and a moderate expansion in width and length. As demonstrated in previous publications [43], part of the applied burnishing force is used to allocate the material in the existing inner voids, resulting in surface densification. Therefore, the increase in width and length should not be attributed to an overall deformation of the part but to creating an overhanging edge due to the expansion of the pressed material on the upper layer.
