1 Introduction

The utilization of additive manufacturing (AM) techniques to produce both industrial and individual products has witnessed a significant surge in the past 10 years [1]. Compared to conventional manufacturing methods, AM offers numerous benefits that make it highly attractive to various industries. AM provides a high degree of design freedom, allowing for complex and intricate geometries that are difficult or impossible to achieve using conventional methods [2]. This opens up new possibilities for product innovation and customization, enabling the production of unique and personalized items. One key benefit of AM is the reduction in assembling and postprocessing requirements. With traditional manufacturing, multiple parts may need to be produced separately and then assembled together. In contrast, AM enables the fabrication of complex structures as a single piece, eliminating the need for assembly and reducing overall production time and costs [3]. Furthermore, AM allows for shorter production cycles, eliminating the need for tooling and setup processes typically required in traditional manufacturing [4]. This makes it particularly advantageous for rapid prototyping and agile production, where quick iterations and fast turnaround times are crucial. Moreover, the ability of AM to produce small quantities at lower costs is a significant advantage [5]. Traditional manufacturing often involves high setup costs and economies of scale, favouring mass production. In contrast, AM enables the cost-effective production of small batches or even individual items without incurring substantial upfront expenses. This benefits industries such as aviation [6, 7], health care [8, 9], construction [10], consumer goods [11], and the automobile industry [12, 13], where customization and low-volume production are increasingly in demand. Given these intrinsic features and advantages, it is highly anticipated that the utilization of additive processes will continue to grow significantly in the coming years. As businesses and industries recognize the potential of AM to revolutionize manufacturing, investments in research, development, and infrastructure are expected to increase. According to market analysis, the global 3D printing market is expected to grow at a compound annual growth rate (CAGR) of 20.8% in the next 10 years [14].

AM technology is classified into four primary categories, depending on the form of raw materials utilized: powder, filament, solid layers, and liquid [15]. Within the filament category, one prominent technique is fused deposition modeling (FDM), which has been trademarked by Stratasys since 1989 [16], or fused filament fabrication (FFF). The operational principle of the FFF technique is straightforward. It involves heating a feedstock filament until it reaches a molten state and then extruding it through an extruder head. This molten filament is then deposited in successive layers onto a build plate, enabling the fabrication of components layer by layer. FFF technology offers several notable advantages [17], including a wide selection of available materials, polymers, and composites such as natural fiber-reinforced polymer composite (NFRPC), polylactic acid (PLA), polyethylene terephthalate glycol (PETG), polyamide (PA), and acrylonitrile–butadiene–styrene (ABS) [18, 19]. Furthermore, FFF has affordable maintenance costs, allows for multilateral and easy material replacement [20], enables rapid manufacturing of small components, does not involve the use of harmful or toxic substances, and achieves an overall tolerance of 0.1 mm. However, this manufacturing technology has certain limitations that must be acknowledged. These limitations include challenges with accuracy, the need for support structures when dealing with overhangs exceeding 45° or features such as bridges and protrusions [21], extended production times for parts with a large number of layers, the presence of visible seam lines between layers (known as the Z seam), the generation of residual stresses within parts, the possibility of warping or delamination due to temperature gradients during production, and the issue of high surface roughness. One of the initial user’s perceptions is significantly affected by the surface quality of the product. The surface’s roughness directly impacts a component’s esthetic appeal leading to potential reductions in sales [22]. Surface roughness pertains to the presence of small and irregular characteristics on a material’s surface, which can arise from different reasons, including material properties, external influences, and the manufacturing process [23, 24]. Surface roughness is commonly quantified using units of microinches or micrometres, and it can be assessed using different parameters that include root-mean-square roughness (Rq), average roughness (Ra), total roughness (Rt), peak height (Rp), maximum roughness (Rmax), average maximum height (Rz), and valley depth (Rv) [25]. Surface roughness plays a crucial role in numerous manufacturing and engineering applications due to its functional attributes of products and its influence on performance. Factors such as lubrication, wear resistance, and fluid flow are directly affected by surface roughness [26, 27]. As the utilization of 3D-printed components in specific product application processes continues to grow, the significance of achieving high-quality surface finishes becomes increasingly pronounced. Furthermore, surface roughness quality directly influences the mechanical characteristics of the manufactured components. In the FFF-AM, the manufactures parts’ surface roughness could be enhanced through one or both preprocessing and postprocessing approaches [28]. Preprocessing approaches involve selecting printing parameter levels and optimizing these parameters, which encompass manufacturing and structural factors. On the other hand, postprocessing techniques encompass chemical and mechanical finishing, as depicted in Fig. 1. Several researchers have focused on leveraging preprocessing techniques to minimize the surface roughness of components made using FFF. Akande [29] optimized infill density, printing speed, and layer thickness factors to improve dimensional accuracy and surface roughness of PLA components made using E3DP. He found that the layer thickness was the most significant factor affecting the dimensional accuracy and surface roughness. Vasudevarao et al. [30] investigated the effect of layer thickness, build orientation, road width, model temperature, and air gap on the parts’ surface quality fabricated using FDM. The research concluded that the layer thickness and the build orientation significantly impacted the surface quality, while the rest parameters had no significant impact. Alsoufi and Elsayed [31] conducted research to study the surface and roughness dimensional accuracy quality of various filament-based materials, including ABS, PLA, ABS+ , and PLA+ , and found that PLA+ displayed superior surface characteristics and was more precise. In contrast, ABS exhibited significant surface roughness, waviness, and primary behaviour. Pérez et al. [32] studied the influence of FDM’s wall thickness, printing path, layer thickness, printing speed, and printing temperature on the surface roughness value of PLA components and concluded that only layer height was the most influencing factor. Sammaiah et al. [33] studied the impact of layer height and infill density on the surface quality of ABS parts made using FFF technology. The findings from the experiment indicated that the surface roughness was higher when using a 20% infill density combined with a 0.26-mm infill layer height. In contrast, the ABS material demonstrates a favourable surface finish (lower surface roughness) when employing 100% infill density with a 0.06-mm infill layer height. In terms of the postprocessing methods, many researchers have investigated the effect of employing these processes on the printed components’ surface quality. Using acetone chemical treatment and aluminium coating, Nguyen [34] improved the surface quality of ABS samples fabricated by FDM technology. The conducted research found that the acetone chemical treatment reduced the surface roughness, while the aluminium coating increased the surface roughness. Moradi et al. [35] employed a carbon dioxide (CO2) laser to improve the surface quality and dimensional accuracy of components made of PLA-FDM by optimizing three parameters, including focal plane position, laser power, and laser cutting speed. Galantucci et al. [36] improved the surface quality of FDM samples by more than 90% using physical vapour deposition (PVD) and abrasive milling.

Fig. 1
figure 1

Fishbone diagram of the surface roughness improvement processes in FFF

Full size image

Khan and Mishra [37] investigated the impact of three acetone vapour parameters, including temperature, time, and smoothening cycles, on the surface roughness quality of ABS parts. The conducted research concluded that the smoothing parameters influenced the surface roughness. Boschetto and Bottini [38] improved the surface quality of FFF components by employing barrel finishing (BF). The results showed a 70% of reduction in the profile parameter after applying the BF process. Hlinka and Weltsch [39] investigated the influence of the laser surface finishing on SAC305 lead-free solder paste and DC01 steel. Valerga Puerta A et al. [40] studied the influence of two sizes of corundum blasting (180 µm and 630 µm) with different incidence pressure and different exposure times on FDM parts made of PLA. It has been observed that as the pressure decreases and the exposure time increases, there is a notable enhancement in surface quality with a 77% in Ra and 70% in Rt, regardless of the corundum size.

Vinitha et al. [41] examined the impact of two-barrel burnishing parameter, spindle speed and depth of penetration on the surface roughness values of ABS-FFF samples. Initially, the parts exhibited a surface roughness of 1.28 µm before undergoing the burnishing process. However, by optimizing the burnishing parameters, they were able to achieve a significant reduction in surface roughness, reaching 0.11 µm. Specifically, this improved surface roughness was attained at a spindle speed of 1400 rpm and a depth of penetration of 0.35 mm. Singh et al. [42] investigated the impact of a vapour smoothing station (VSS) with different cycle times on the surface roughness of ABS-P400 parts printed using FDM technology, and they found that the chemical exposure cycle significantly influenced the surface roughness of the 3D manufacturing parts, with an 87% of contribution. Taşcıoğlu et al. [43] employed the vibratory surface finishing (VSF) to enhance the surface quality of PLA-FFF samples. The surface roughness of the fabricated samples was decreased by 66% using this postprocess. In a study conducted by Kumar et al. [44], the impact of two types of electroplating (copper and nickel) on the surface roughness of FDM-PLA components was investigated. Prior to the electroplating process, the surface roughness was measured to be 0.97 µm. However, after applying copper electroplating, the surface roughness was significantly reduced to 0.03 µm. Similarly, nickel electroplating yielded a surface roughness of 0.1 µm, demonstrating an improvement in surface quality for both electroplating techniques. Neosanding postprocessing, also known as ironing postprocessing in some software, is a cost-effective and straightforward technique that can be conveniently performed using the same printer employed for FDM. Despite its potential benefits, limited research has focused on investigating the effects of utilizing the neosanding process on the properties of E3DP components. Sardinha et al. [45] conducted a study to examine the impact of implementing neosanding on various layers of ABS parts manufactured using FDM technology, specifically focusing on warpage deformation. The specimens were divided into different series based on the neosanding process applied to them. The findings revealed that the lowest warpage deformation was attained when neosanding was applied solely to the second layer, resulting in an impressive reduction of 89%. In another study conducted by Griffiths and Leigh [46], the influence of three postprocessing techniques, namely burnishing, subtractive machining, and ironing, on the conductivity of PLA-FDM parts printed with silver nanoparticle ink using inkjet technology was investigated. The research aimed to evaluate the impact of these techniques on the average batch conductivity of the samples. The results revealed notable improvements in conductivity for the machined and ironed surfaces. Specifically, the average batch conductivity increased by 12.5% for the machined samples and by 71.8% for the ironed samples. Among the techniques studied, burnishing showed the most promising outcomes, with an impressive average batch conductivity improvement of 95.9%. In the study conducted by Sardinha et al. [47], the authors implemented the neosanding process as a postprocessing step for ABS samples produced using FDM technology. The results of their investigation revealed a notable reduction in distortion levels, reaching up to 33%. Additionally, applying the neosanding process led to a substantial decrease in the surface roughness value, with measurements decreasing from 23 to 9 µm, corresponding to a reduction of approximately 60%. Paz et al. [48] studied the influence of three postprocessing methods, including vapour polishing, plasma, and neosanding, on the electrical conductivity of graphene nanoplatelets (GNP)-reinforced nanocomposite samples made of ABS using FFF technology. The results showed that applying acetone vapour polishing resulted in a significant reduction in surface electrical conductivity, with a decrease in nearly one order of magnitude. Conversely, the plasma treatment of the parts led to an improvement in electrical conductivity. Neosanding postprocessing of the samples yielded surface electrical conductivities ranging from 10–7 to 10–5 S/sq.

In this study, we have focused on utilizing neosanding postprocessing as a means to enhance the surface roughness of printed PLA specimens employing FFF technology. Despite the limited existing research on this process and the lack of well-optimized factors, we aim to address this gap by conducting an independent investigation into the influence of three critical neosanding factors: neosanding spacing, neosanding speed, and flow rate. We aim to shed light on the efficacy of neosanding and provide insights into optimizing its parameters. By addressing the existing gaps in the literature, this research aims to contribute to a better understanding of the neosanding process and its potential for enhancing surface quality in additive manufacturing.

The manuscript adheres to a well-organized structure, commencing with Sect. 2, which comprehensively explains the neosanding postprocessing technique. Section 3 of the research paper outlines the experimental procedures utilized for fabricating the components and provides a detailed explanation of the measurement techniques employed. The findings and ensuing discussion are presented in Sect. 4, emphasizing notable insights and observations. Lastly, the concluding section succinctly summarizes the primary outcomes of this study, highlighting the key findings and their implications. All the specimens were printed using a desktop open-source 3D printer.

2 Neosanding process

The neosanding process, named after Neotko, a moderator on the Ultimaker forum [49], is based on the principle of clothes ironing. The neosanding process is a surface treatment technique that employs a specialized instrument with a heated and flat surface, like a heat gun or a soldering iron [48]. This tool is utilized to apply a combination of pressure and heat to the part’s surface. The primary objective of this technique is to improve surface quality by addressing imperfections, irregularities, and roughness that may be present [50]. By subjecting the surface to pressure and heat, the neosanding process has the ability to smoothen the surface of the component effectively. It helps to alleviate any visible flaws or rough patches that may have occurred during the FFF manufacturing process. Moreover, the application of pressure and heat aids in fusing together layers and minimizing gaps that could be present [51]. To implement the neosanding process, it is typically integrated into the slicer software used in the FFF manufacturing workflow. The software controls the movement of the heated extruder head in a back-and-forth motion over the targeted printed layer of the object. This controlled movement ensures that the pressure and heat are evenly distributed across the surface, facilitating the smoothing and fusion of the material.

The neosanding process, like any other processing technique, contains various controllable factors that influence its outcomes. These factors include neosanding spacing, neosanding speed, and flow rate [52]. Neosanding spacing represents the gap between adjacent neosanding paths. This particular factor is influenced by the extruder head diameter. Figure 2 visually presents a schematic illustration of the neosanding spacing. Increasing the neosanding spacing results in reduced overlap between neosanding passes. Neosanding speed refers to the velocity at which the heated extruder head traverses the target layer, which can be set independently from the printing speed [53]. Flow rate indicates the quantity of filament added during neosanding relative to the predefined layer thickness, expressed as a percentage [54]. Without a flow rate, the neosanding process cannot be carried out.

Fig. 2
figure 2

Schematic of neosanding spacing

Full size image

3 Experimental procedure

3.1 Specimen material, FFF printer, and neosanding factors

The process of creating a 3D model of the sample commenced using SolidWorks. Subsequently, the 3D part was converted into a STereoLithography (STL) file format and transferred to PrusaSlicer, a versatile open-source software designed for slicing 3D models. PrusaSlicer is known for its compatibility with various 3D printers. Using this software, the model was sliced into layers, and the resulting G-code file was generated. The G-code file was then sent to the 3D printer for printing. The samples were produced using a desktop 3D printer that conforms to open-source standards. In this study, the dimensions of the test specimens used in the experiment were 30 mm in length, 20 mm in width, and 2 mm in thickness, and Fig. 3 illustrates the sample manufacturing process. The specimens were printed using black ecoPLA filament, with a diameter of 1.75 ± 0.05 mm tolerance manufactured by 3Djake [55], a biodegradable and environmentally friendly polymer [35]. The Creality Ender 3 v2 printer was employed for building the specimens. The core of this study was to investigate the neosanding process factors. To ensure consistency, the printing parameters were kept fixed, as outlined in Table 1. The neosanding process factors were varied across four different values, as presented in Table 2. The selection of values was based according to the default settings of the Slicing software, a deliberate choice aimed at enabling the systematic investigation of each individual factor in a separate manner [56, 57].

Fig. 3
figure 3

Sample manufacturing process

Full size image
Table 1 Test specimen printing conditions
Full size table
Table 2 Neosanding postprocessing factors and their values
Full size table

To thoroughly investigate the impact of each neosanding process factor on the surface roughness value, the study analysed them individually. For each factor under examination, the remaining factors were kept at their default values, which were set to 0.1 mm for neosanding spacing, 20 mm/s for neosanding speed, and 15% for flow rate. This methodology allowed for a focused examination of the specific influence of each factor on the outcomes of the study. A preliminary experimental run was conducted before applying the neosanding postprocessing technique, followed by ten experimental runs focusing on different neosanding factors. Each run consisted of printing three samples, resulting in a total of 33 specimens. It is important to note that runs 5 and 11 were identical to run 1. Therefore, there was no need to repeat these specific experimental runs.

3.2 Methods of measuring the surface roughness

Historically, surface measurements were conducted through tactile or visual means, which inherently introduced subjectivity and limited accuracy [58]. To overcome these limitations and achieve objective and more precise results, two distinct methods of instrumentation were employed. The first type employed tactile sensing to analyse the vertical characteristics of the surface, whereas the second type aimed to replicate the visual capabilities of the human eye by examining the surface’s lateral structure or spatial information. These advancements in instrumentation have significantly enhanced the measurement accuracy and objectivity in surface analysis.

Later, instruments have been emerged as valuable tools for measuring surface roughness. These instruments are categorized into two main types: stylus profilometer and confocal profilometry measurement methods. For the stylus profilometer measurement method, a diamond-tipped probe called a stylus is employed to detect variations in height caused by surface topography. This stylus is linked to a bar equipped with a transducer on the other side, capturing and amplifying any signal changes. The transducer employed in this method is commonly based on the inductance principle and offers a wide range of sensitivity, ensuring durability in its operation [59]. In the confocal profilometry measurement method, various mechanisms are frequently employed to collect topographical data. These instruments encompass optical interferometry, laser triangulation, confocal microscopy, and focus detection. Moreover, scanning probe microscopy (SPM), which comprises over twenty different devices categorized according to the interaction of the probe with the surface, is also utilized [60]. The aforementioned methods have been extensively utilized in the previous research [61,62,63,64,65]. In this study, we performed measurements of the average surface roughness (Ra) on the topmost surface of PLA-FFF specimens. The measurements were carried out perpendicular to the neosanding path, illustrated in Fig. 2, using both techniques. The surface roughness of the specimens was assessed at ambient temperature and in accordance with the ASME B46.1 (2019) standard [66].

3.2.1 Stylus profilometer method measurement

The stylus profilometer method is a widely used technique for measuring surface roughness in various industries and research fields. It provides valuable insights into the topographical characteristics of a surface by utilizing a mechanical stylus that scans the surface profile [67]. This method is based on the principle of stylus contact with the surface, where the stylus tip follows the contour of the surface, detecting minute variations in height. The stylus profilometer has some potential benefits in surface roughness measurement. It provides a level of precision, allowing for the characterization of surface features and deviations. Moreover, it offers a wide range of measurable parameters, such as Ra, Rz, and Rsk. In the stylus profilometer technique [68], a sharp-tipped diamond stylus is employed to move over the surface under a controlled measurement force. The vertical movements of the stylus are detected and recorded, allowing for the calculation of surface roughness parameters. The resulting data provide valuable information on the surface quality, finish, and texture of the material being analysed. The stylus profilometer method has found widespread applications in various sectors, including engineering, manufacturing, quality control, and materials science. It plays a key role in assessing the surface integrity of parts, evaluating the effectiveness of manufacturing processes, and ensuring adherence to specified standards and requirements. In this study, the Mitutoyo SURFTEST 301 was employed for the stylus profilometer measurement method to measure the surface roughness of our FFF-printed PLA specimens. Prior to conducting the measurements, the Mitutoyo SURFTEST 301 instrument was calibrated following the user’s manual guidelines. This calibration process involved using a reference specimen, provided by the manufacturer, has been done to ensure the accuracy and reliability of the measurements, and Fig. 4 shows the calibration results. The following stylus profilometer method parameters were employed to measure the surface roughness:

  • Measurement method: differential inductance method.

  • Probe type: diamond stylus probe.

  • Stylus radius: 2 µm.

  • Measurement accuracy: ± 2 µm.

  • Cut-off length/sampling length: 0.8 mm/4 mm for surface roughness values ranging from 0.1 to 2 µm and 2.5 mm/12.5 mm for surface roughness values ranging from 2 to 10 µm [66].

  • Measurement speed: 0.5 mm/s.

  • Traverse speed: 1 mm/s.

  • Measurement force: 4 mN.

Fig. 4
figure 4

Surface roughness profile of the reference specimen after calibration

Full size image

3.2.2 Confocal profilometry method measurement

Confocal profilometry is an advanced non-contact measurement technique that has gained substantial attention in surface metrology. It offers precise and accurate measurements of surface topography and roughness without any need for physical contact with the tested part [69]. This method employs a confocal microscope, which utilizes a focused laser beam to illuminate the tested part and collect the reflected light. By analysing the intensity and position of the reflected light, confocal profilometry provides detailed three-dimensional surface information with high resolution. The confocal profilometry method can provide several advantages, such as eliminating the risk of damaging delicate or sensitive components due to there being no direct physical contact involved. This makes it particularly appropriate for measuring fragile or soft materials. Moreover, the non-contact nature of confocal profilometry allows for quick and efficient measurements, providing high-throughput analysis of large surface areas. Furthermore, the high precision and accuracy of this technique make it ideal for applications that demand precise surface characterization.

In this investigation, the Keyence VR-5000 was used for the confocal profilometry method. Keyence VR-5000 is a cutting-edge surface roughness measurement instrument developed by Keyence Corporation, a global leader in the field of industrial automation and measurement technologies [70]. The used Keyence VR-5000 were:

  • Measurement method: optical interferometry.

  • High magnification camera up to 160x.

  • Measurement range: 206 mm × 104 mm.

  • Measurement accuracy: ± 0.5 µm.

  • Cut-off length/sampling length: 0.8 mm/4 mm for surface roughness values ranging from 0.1 to 2 µm and 2.5 mm/12.5 mm for surface roughness values ranging from 2 to 10 µm [66].

  • Display resolution: 0.1 µm.

  • Working distance (WD): 75 mm.

  • Image pickup device: 1′′ 4 million pixels monochrome complementary metal oxide semiconductor (CMOS) camera.

4 Results analysis and discussion

The average roughness values, along with their standard deviations, were measured and are presented in Table 3 as a summary of the findings. All three factors of the neosanding process were individually investigated in this study. The default settings of these parameters were utilized. As depicted in Fig. 5, the results demonstrate a significant reduction in surface roughness through the implementation of the neosanding postprocess. Utilizing the stylus profilometer method, the initial surface roughness of 9.63 ± 0.97 µm was significantly reduced to a final value of 1.62 ± 0.1 µm, representing an impressive reduction percentage of approximately 83%. Similarly, the utilization of confocal profilometry resulted in a significant decrease in surface roughness, reducing it from 9.79 ± 0.96 µm to 2.02 ± 0.18 µm, which corresponds to a reduction percentage of approximately 79.5%.

Table 3 Experimental findings obtained from the conducted measurements
Full size table
Fig. 5
figure 5

Surface roughness results before and after the neosanding process

Full size image

In this study, the focus was on systematically investigating the influence of individual neosanding factors on surface roughness. The authors intentionally designed the experiments to analyse the effects of neosanding spacing, speed, and flow rate separately. The aim was to understand how each factor independently contributes to the surface quality improvement through the neosanding process. While the discussions may have alluded to broader trends, it is important to note that the findings pertain to the specific conditions tested. The authors acknowledge that there might be interactions among these factors, but the approach enabled the isolation of their individual impacts. As such, the insights gained from the study provide valuable information into optimizing the neosanding process for enhanced surface roughness without delving into potential interactions between factors.

4.1 Impact of neosanding speed on the surface roughness

The initial four experimental runs in this investigation explored the impact of neosanding speed on the surface roughness value for the printed components, as depicted in Fig. 6. Findings from both the stylus profilometer and confocal profilometry measurements indicated that higher neosanding speeds led to higher surface roughness values. Specifically, the stylus profilometer method exhibited a 34% rise in surface roughness, with values increasing from 1.62 ± 0.1 µm to 2.17 ± 0.18 µm. Correspondingly, the confocal profilometry method concluded a slight increase in the surface roughness value, with a marginal rise of 24% from 2.02 ± 0.18 µm to 2.5 ± 0.27 µm.

Fig. 6
figure 6

Impact of neosanding speed on the surface roughness value

Full size image

As the neosanding speed increases, the movement of the heated extruder head accelerates, causing uneven heating distribution and decreased precision in control. Consequently, this can contribute to a rougher surface finish. Moreover, higher neosanding speeds tend to decrease the quantity of molten filament added to fill the gaps between the deposited filaments, resulting in a rougher surface. This phenomenon arises due to the reduced duration available for the extruder head to deposit material when operating at higher speeds, leading to the presence of undulations or irregularities on the surface. As the neosanding speed increases, the movement of the heated extruder head becomes faster over the surface of each layer, leading to uneven thermal distribution and reduced precision in control. Consequently, this can have an impact on the surface finish, resulting in a rougher texture.

4.2 Impact of neosanding spacing on the surface roughness

Runs five to eight were conducted to investigate the influence of increasing neosanding spacing on surface roughness measurements. The findings revealed a clear positive correlation between neosanding spacing and surface roughness. Figure 7 illustrates the significant impact of neosanding spacing on surface roughness values. At a 0.1 mm of neosanding spacing, both stylus profilometer and confocal profilometry measurements yielded roughness values of 1.62 ± 0.1 µm and 2.02 ± 0.18 µm, respectively. These values increased more than two times when the neosanding spacing was increased to 0.2 mm and further increased to 4.8 ± 0.32 µm and 5.23 ± 0.27 µm at neosanding spacing of 0.3 mm for both stylus profilometer and confocal profilometry measurements, respectively. Notably, stylus profilometer and confocal profilometry measurements exhibited an approximately fourfold increase when the spacing reached 0.4 mm with a value of 6.11 ± 0.73 and 6.44 ± 0.83, respectively. The increase in surface roughness with the increase in neosanding spacing can be attributed to the reduction in the overlap between successive neosanding lines. When the neosanding spacing transitions from 0.1 to 0.2 mm, the overlap between adjacent neosanding lines decreases from 0.3 to 0.2 mm. Once the neosanding spacing reaches 0.4 mm, the overlap between the neosanding lines will be zero. Neosanding spacing refers to the distance separating consecutive neosanding lines, while the overlap degree is an indication of how many times the heated extruder head will move over the printed path. Both overlap and neosanding spacing have an impact on the surface quality and mechanical properties of the printed specimens. A larger neosanding spacing can contribute to a higher surface roughness value and potentially compromised mechanical characteristics; however, it can also lead to reducing in printing time. On the other hand, a smaller neosanding spacing can lead to a smoother surface and enhanced printed parts characteristics.

Fig. 7
figure 7

Impact of neosanding spacing on surface roughness values

Full size image

4.3 Influence of flow rate on the surface roughness

The influence of the flow rate on surface roughness was investigated in the latter part of this study. Runs nine to twelve specifically examined how the flow rate affected the surface roughness values. Both the stylus profilometer and confocal profilometry methods demonstrated a positive relationship between the surface roughness and flow rate, depicted in Fig. 8. In both employed measurements methods, the lowest surface roughness result was achieved at a flow rate of 5%, with values of 1.31 ± 0.12 µm and 1.84 ± 0.2 µm for the stylus profilometer and confocal profilometry methods, respectively. Since the flow rate increased from 5 to 20%, the roughness value obtained by the stylus profilometer technique gradually escalated from 1.31 ± 0.12 µm to 1.89 ± 0.14 µm. The increase in surface roughness with higher flow rates can be attributed to the deposition of additional material on the surface during the FFF process. When the flow rate is increased, more material is extruded and deposited, resulting in a thicker layer being added to the surface of the printed part. This thicker layer can introduce irregularities and imperfections, leading to an increase in roughness value. Therefore, when the flow rate increases beyond an optimal point, the excessive material deposition contributes to a gradual rise in surface roughness values measured by the stylus profilometer method. On the other hand, the confocal profilometry technique indicated a change in roughness value from 1.84 ± 0.2 µm to 2.19 ± 0.15 µm within the identical range of flow rate variations. Notably, when accounting for the standard deviations, the roughness responses of 1.84 + 0.2 µm and 2.19–0.15 µm intersect at approximately 2.04 µm. This intersection suggests a consistent roughness response across varying flow rate settings.

Fig. 8
figure 8

Influence of flow rate on surface roughness measurement

Full size image

4.4 Sensitivity analysis

Sensitivity analysis is a valuable technique for assessing the impact of uncertainties associated with input parameters on the output of a model. It provides a quantitative measure of how variations or changes in the model’s input parameters influence the resulting output [71]. This approach entails determining the output by evaluating a limited number of input values that encompass the potential range of the input variable [72]. While this method does not directly account for the variability in the output resulting from variations in the inputs, it does enable an assessment of how the range of input values affects the output [73]. In certain instances, mathematical techniques can aid in identifying the most influential inputs [74]. As can be observed in Fig. 9, the neosanding spacing was the most significant factor affecting the surface roughness value in both measurements methods. According to the given equations and R-square values in Fig. 9, the coefficient for spacing was 14.61 and 14.26 for the stylus profilometer and confocal profilometry, respectively. These results indicate that changes in the neosanding spacing have a significant impact on the surface roughness value for both measurement methods. For neosanding speed and flow rate, both measurement methods illustrate a negligible effect of these two factors on the surface roughness value.

Fig. 9
figure 9

Sensitivity analysis of neosanding factors on surface roughness measurements using the two measurement methods: A, D neosanding spacing, B, E neosanding speed, and C, F flow rate

Full size image

4.5 Stylus profilometer and confocal profilometry measurements

Based on the presented findings, both measurement techniques displayed a similar pattern regarding surface roughness measurements. Though, it is noteworthy that the confocal profilometry method consistently yielded higher findings compared to the stylus profilometer method in all experimental runs, and these findings agree with prior research findings [75, 76]. The disparity between the stylus profilometer and confocal profilometry measurement methods became more pronounced as the surface roughness declined. When examining surface roughness above 2 µm, the greatest difference of 15% was observed between the two methods, as approved in a previous study performed by Patel and Kiran [77]. This discrepancy can be explained by the fact that the stylus profilometer instrument operates as a cut-off filter, selectively eliminating roughness lower than the 2-µm tip radius. The findings of this study further support and demonstrate the impact of the stylus profilometer’s filtering mechanism on the recorded measurements.

4.6 Optical profilometry analysis

Surface roughness assessment of the top surface specimens was conducted using optical profilometry before and after the implementation of the neosanding process. The findings, illustrated in Figs. 10 and 11, demonstrate a substantial improvement in surface roughness as a result of the neosanding process, Fig. 10, provides a clear visual representation of how the neosanding process minimizes the gaps between the deposited filaments and flattens the layer lines, and consequently positively influences the surface roughness of the 3D-printed specimens. The topographical magnification presented in Fig. 11 further highlights the influence of various neosanding factors on surface roughness. Notably, it can be observed that the neosanding spacing, see Fig. 11e–h, exhibited the most substantial impact on surface roughness value. With a spacing of 0.1 mm, three repetitions of the neosanding path were observed, whereas increasing the spacing resulted in a decrease in the number of repetitions. At a spacing of 0.4 mm, no repetitions were observed. The neosanding speed, Fig. 11a–d, emerged as the second most influential factor, while the flow rate exhibited the least significant impact on surface roughness.

Fig. 10
figure 10

Magnified view of the optical micrograph of the highest surface of printed specimens: a before neosanding and b after the neosanding process

Full size image
Fig. 11
figure 11

Impact of the values of neosanding factors on the surface texture: a–d: runs 1–4, e–h: runs 5–8, and i–l: runs 9–12

Full size image

The analysis depicted in Fig. 11 provides further insight into the influence of various neosanding factors on surface roughness. Notably, the neosanding spacing, Fig. 11e–h, was found to have the most pronounced effect on surface quality. At neosanding spacing of 0.1 mm, it is observed that the neosanding path was repeated three times, indicating that the tool traversed the same area multiple times. However, as the neosanding spacing increased, the number of repetitions decreased. At neosanding spacing of 0.4 mm, no repetitions were observed, suggesting that the tool covered a larger area without repeating the same region. This demonstrates that the neosanding spacing plays a crucial role in determining the surface texture. Furthermore, the neosanding speed, Fig. 11a–d, emerged as the second most influential factor. It can be observed that higher neosanding speeds resulted in rougher surface textures, whereas lower speeds led to smoother surfaces. This highlights the importance of controlling the speed parameter during the neosanding process to achieve the desired surface roughness. On the other hand, the flow rate exhibited the least significant impact on surface roughness (Fig. 11). Minor variations in flow rate did not lead to significant changes in surface quality compared to the neosanding spacing and speed. However, it is important to note that even though the flow rate had a lesser impact, it still contributes to the overall optimization of the neosanding process. Understanding these influences and optimizing the neosanding parameters is essential for achieving the desired surface quality in PLA-FFF parts.

5 Conclusion

This study has explored the application of postprocessing techniques, explicitly focusing on neosanding, to enhance the surface finish of extrusion-based 3D printing parts. By investigating the factors influencing surface roughness and employing different measurement methods, valuable insights have been gained regarding the effectiveness and limitations of neosanding in additive manufacturing. The findings presented in this study contribute to the broader objective of achieving superior surface quality in extrusion-based 3D printing components. In this final section, we summarize the key findings as follows:

  • The application of neosanding postprocessing has proven highly effective in reducing surface roughness by approximately 80%, highlighting its potential for enhancing the surface finish of extrusion-based 3D printing parts.

  • The investigation of neosanding process factors, including neosanding spacing, neosanding speed, and flow rate, has emphasized their significant influence on surface roughness. Proper optimization of these factors is crucial for achieving desired surface quality and minimizing roughness.

  • The stylus profilometer measurement method exhibited an impressive 86.5% reduction in surface roughness value, while confocal profilometry measurement achieved about 81% reduction. These results demonstrate the effectiveness of both measurement techniques in quantifying the surface roughness improvement achieved through neosanding.

  • Optimal surface roughness can be achieved by considering a comprehensive combination of preprocessing and postprocessing factors, paying meticulous attention to detail, and thoroughly understanding the factors that affect surface roughness.

  • Stylus profilometer measurement, while effective, has drawbacks, such as the capability of causing grooves or scratches on parts made of soft materials during the measurement process. Careful handling and consideration of the material properties are necessary when employing this method.

  • Confocal profilometry measurement, on the other hand, offers a non-destructive approach that provides rapid and precise results, making it a valuable alternative for surface roughness evaluation in additive manufacturing.

  • It is important to note that neosanding postprocessing, while effective, does have some drawbacks, such as longer printing times, and can be applied only on flat surfaces. These considerations should be taken into account when implementing neosanding in practical applications.

Finally, this study contributes to the advancement of surface finish optimization in additive manufacturing through the investigation of neosanding postprocessing. The insights gained from this research can inform the development of improved postprocessing techniques and facilitate the production of high-quality 3D-printed parts across various industries.