1. Introduction
While selective laser sintering (SLS) is used in additive manufacturing to produce parts with high surface quality, these parts have greater surface roughness and grain size than those produced by other polymeric additive manufacturing techniques
[1]. Although a smooth surface is not always preferred by consumers, surface roughness and grain size are properties of 3D-printed parts that consumers notice during tactile and visual assessments
[2][3]. The surface roughness of parts produced by SLS 3D printers results from the SLS process, which uses a high-power laser to sinter granular polymer powder into a solid structure. The surface texture quality of SLS-printed parts is affected by many parameters, including the preparation method, the equipment utilized (powder properties, machine setup and processing parameters), and the position and orientation of the part inside the build chamber
[1][4][5].
Polyamide (PA) is the most commonly used material in SLS and is mainly used in the form PA12, PA11, and PA6
[6]. Polyamide is also known as nylon, which indicates a synthetic polyamide, and the two terms are often used interchangeably. Researchers herein use PA11 powder or PA1101, which is a rebranded type of Rilsan Invent Natural (Arkema) from EOS
[7]. While PA is mainly manufactured from petrochemical sources, ongoing research is creating drop-in materials for PA and completely new biobased materials
[8]. Polyamide 11 (PA11) is a well-established biobased material
[9] formed by the polycondensation of
𝜔-aminocarboxylic acids to create a linear polymer chain with a characteristic recurring functional acid–amide group (
Figure 1) with alkyl group chains (R groups) from the reactants
[10]. Lately, the polycondensation of
𝜔-aminocarboxylic acids has become more popular due to an increasing interest in biobased aliphatic polyamides. PA11 is synthesized by using 11-aminoundecanoic acid from castor oil as a monomer and multifunctional agents
[8][11]. A PA11 powder is generated by either milling or grinding procedures, spray drying, or precipitation from solvents
[12]. The result is a semicrystalline thermoplastic, making it suitable for SLS printing purposes
[7][8].
Figure 1. Chemical structure of PA.
To be processed, the semicrystalline powder requires higher temperatures than alternative materials, but it provides higher tensile properties
[13]. The crystallization of PA11 starts with the cooling of the polymer chains. When synthesized, the chains appear to be disordered and amorphous, but as they cool down, they start arranging themselves into a repeating, ordered structure. The degree of crystallinity in PA11 can vary depending on several factors, including the cooling rate
[14][15], molecular weight
[16], and the presence of any additives
[17]. The PA11 powder is also known for its good dimensional stability under fluctuating humidity and good mechanical properties, including its high strength, stiffness, and abrasion resistance
[6][7].
The powder is important for the selective laser sintering process. The key factors of flowability and packing density during the SLS process, in particular the re-coating of the powder bed, are affected by the powder’s particle size distribution and shape (where more narrow and rounded particles are preferable)
[12][18][19], surface roughness and interparticle forces, and moisture and temperature
[19][20][21][22][23]. In SLS printing, after completing a part, a common procedure is to recycle the remaining unused powder. The reuse of aged polyamides is currently a significant area of research. In fact, to facilitate printing and ensure better dimensional stability, parts are printed slightly below the powder’s melting temperature, and this causes unsintered powder in the build chamber to undergo alterations in its thermal and mechanical properties, making its reuse challenging
[24][25]. Thus, before reuse, the old powder needs to be sieved and mixed with new powder in appropriate ratios
[24].
Printing in 3D with SLS is performed by using successive layers of powdered material delivered either by a blade from a hopper (powder reservoir) or by a roller from a powder feeder to the powder bed. The powder can be preheated to around 100 °C in the hopper/feeder to achieve the best flowability and packing density by reducing the Hausner factor
[26] below 1.25 and the granular Bond number below 100 in the unconsolidated material
[23][27]. The granular Bond number is the ratio of the interparticle forces to the contribution of gravity in two particles. Moreover, preheating the feeder/hopper reduces the temperature difference between the feeding system and the powdered layer so that the surface of the powder bed can sinter, which decreases the undercooling effect in the molten layer and prevents the sintered layer from curling or warping
[20][28][29][30][31].
Along with the preheating of the powder, the bed and the chamber are also preheated, but to a temperature right below the melting point (for PA11, the melting point is 201 °C, but the temperature is kept at 180 °C). This is performed to increase the energy efficiency of the printing process. The laser can then transform the material into its molten state by increasing the temperature by just a few degrees. This relaxes internal stresses in the part and prevents warping
[31][32]. In an optimal sintering window, the hysteresis between melting and crystallization inhibits the crystallization and the solidification of the layers until all of them are sintered and the powder is kept in a melt state with low viscosity
[22][30][31][33]. Schmidt et al.
[18] showed that if the sintering temperature is too close to the crystallization peak, premature crystallization occurs and the printed part curls and becomes distorted. If the sintering temperature is instead too close to the melting temperature, a loss in the accuracy of the part features occurs. On average, if there is a large difference between the onset melting temperature and the onset crystallization temperature, the crystallization of the polymer melt is reduced. This reduced crystallization helps to decrease the shrinkage of the printed parts, hence reducing their internal stresses
[34].
Keeping the part being printed at a constant temperature for the entire process is challenging. Strano et al.
[35] found that the preheating energy is proportionally related to the number of sintered layers and the build orientation of the part. The build chamber plays an important role in the sintering of the part. Melt inhomogeneities are verifiable along the
z-axis (up) due to the nature of the process
[22][36] and in the
𝑥𝑦 plane (bed)
[37][38]. The latter are probably due to the heating devices being unable to uniformly cover the entire planar surface with the same temperature distribution. The lack of melt homogeneity affects the part being printed; for example, inconsistencies in cooling and crystallization rates can cause warping effects in the lower area of the part
[39]. In addition, an increase in the average molecular weight of the PA11 can positively shift the crystallization toward lower temperatures and the melting point towards higher temperatures, which hinders the particles in coalescing, hence disrupting the surface quality of the part
[40].
Once the sintering process has created a layer of the part, the powder bed is lowered by a height that corresponds to the layer thickness and a new powdered layer is deposited. The unsintered powder (powder cake) remains in place and provides structural support, which means printing supports are not required and unconventional printing directions are permissible, whereas they are generally prohibited by other 3D printing systems. For each layer deposited, the printing parameters, such as the bed temperature and removal chamber temperature, can vary. The removal chamber is physically separated from the powder by a steel plate. This is important for the cooling process. In fact, the removal chamber can normally be kept at a fixed temperature so the layer and its powder cake can cool down homogeneously without curling while the subsequent layer is processed
[41]. However, there are different ideas on how to adjust this temperature to increase the efficiency and improve the cooling conditions for the part
[42].
Once the part has been printed by stacking layers upon layers and it has a temperature theoretically equal to the removal chamber temperature, it is left to cool down homogeneously surrounded by its powder cake until an extraction temperature has been reached. When the piece is removed, the excess powder is moved to a recycling container for reuse. The cooling phase is another critical step in the SLS printing process, where crystallization is the main phenomenon; refer to Amado
[31] for more insights on this matter.
2. Sintering Process
The main step in the SLS system is the sintering process
[10][36][37][43]. The packing density and the viscosity in the molten state at the processing temperature are the most important parameters that define the sintered density
[22][31]. Variations in the processing temperature change the viscosity and thus the rate of densification. The ratio between the viscosity of the melt and the surface tension has a big impact on the coalescence of the polymer powder particles. A higher ratio reduces the coalescence rate of particles during the sintering time, while a lower ratio stimulates the formation of droplets that thwart the homogeneity of the layer
[31]. The coalescence of polymer powder particles is the main constituent of the sintering process. When the polymer powder particles are above their glass transition or melting temperature, they form necks to decrease their total surface area (
Figure 2).
Figure 2. Schematic representation of the coalescence process. A droplet of radius 𝛼0 coalesces with another droplet and ends with the radius 𝛼𝑓. During the process, the particle radius 𝛼 and the angle of the intersection 𝜃 and the radius of the neck x change as a function of time t.
Frenkel
[44] was the first to describe the neck formation phenomenon, Pokluda et al.
[45] modified Frenkel’s findings by considering the variation in the radius of the particles over the whole SLS process, while Bellehumeur et al.
[46] and later Scribben et al.
[47] incorporated the viscoelastic nature of the molten polymers to obtain a more precise description of the process. The neck formation is activated by the energy density in the particles caused by the incident light, which is parameterized by the laser power and speed. The energy density leads to good or poor bonding between the layers. If the energy density is too low, delamination can occur between the layers, while if it is too high (low-speed, high-power laser) warping and curling are more frequent. In the case of non-adjusted combinations, balling is also possible (high-speed, high-power laser)
[33][48]. Moreover, the energy density also affects the microstructure of the sintered part. In the case of a higher energy density, the cooling rate and crystallinity degree might cause the part to shrink more, resulting in a lower porosity as the pores become smaller or are absorbed
[48]. This was confirmed by Kozior
[5], who showed that a reduction in the energy density leads to higher stress relaxation and overall improved quality of the surface texture. However, Wang et al.
[49] found that higher laser power leads to lower shrinkage of the part. This is due to the increased sintering width and depth caused by a higher power, which creates a higher sintered area, thus decreasing the heat exchange between the sintered area and the rest of the sintered layer.
The laser beam is steered to the point of the powder bed to be sintered. This is performed by a galvanometer scanning technology that is coupled with a beam expander and beam collimator to control the energy density, the diameter and the spot size of the laser beam, see
Figure 3. As a laser source, CO
2 and Nd:YAG lasers are mainly used, but CO
2 is usually preferred as it performs better in SLS systems. However, EOS has recently designed the CO laser beam that is used, which seems to have a smaller light spot, higher processing efficiency, and larger processing range. The laser used irradiates the powder with a radial Gaussian distribution. Xin et al.
[50] introduced a ray-tracing model that considers the attenuation of the laser energy in the powder bed, which can be described as a series of scattering effects. In this way, under the assumption of spherical particles, they found a strong dependency of the scattering on the distribution of the laser energy, resulting in variations of the temperature in different grains. Thus, as found experimentally also
[51], there is a correlation between the angle of incidence and the laser irradiance and thus the sintering performance.
Figure 3. Schematic representation of the SLS technology.
This entry is adapted from the peer-reviewed paper 10.3390/polym15132967