3.1.

Deformation

The fundamental cause of LAM-induced distortion is the nonuniform heating and cooling of material during the LAM process, which produces elastic and plastic strains due to the mismatch of thermal expansions and constrictions between the melt pool zone and the surrounding zone in the parts. The influence of different laser scanning patterns on the deformation of the substrate is studied at this angle to reduce the deformation of the LAM fabricated parts and reduce the workload. A laser/powder deposition method was applied to the Shape Deposition Manufacturing System at Stanford University. The results show that both selective deposition and the use of a low CTE material such as INVART can significantly reduce the deformation of simple beams due to internal stresses.³⁹ Gao et al.⁴⁰ examined the impact of the deposition pattern on the substrate’s deformation during direct laser fabrication. Eight laser scanning patterns were designed to create a cylinder on an asymmetrical IN718 arc substrate, respectively. The findings indicate that the substrate deformation along the $z$ direction for all scanning patterns is greater than the two other directions. Residual stress-induced warping is one of the most pressing concerns in the direct deposition of fully dense metals, leading to undesirable losses in dimensional resistance.^41–43 Biegler et al.^44,45 introduced a novel method for measuring distortions experimentally and using the effects to validate numerical simulations. In conjunction with optical filters, in-situ distortions and periodic expansion and shrinkage are calculated by digital image correlation. When the experimental and numerical results are compared, the length direction of the sample has a strong agreement, while the height direction has quantitative deviations. Nie et al.⁴⁶ performed a numerical simulation of the laser hot-wire AM to obtain the temperature, stress, strain fields, and distortion. The findings demonstrate that temperature fluctuates consistently with layer-by-layer deposition and that the overall distortion increases steadily during deposition. An accurate thermal-mechanical model was developed to propose controlling distortion in a wire-fed electron-beam thin-walled Ti-6Al-4V freeform. The overall distortion along the $x$ axis was reduced to 0.12 mm using a current dynamic scheme. The distortions along the $x$ and $z$ axes were reduced to almost zero using a constant temperature limit at the bottom of the substrate.⁴⁷ Chen et al.⁴⁸ proposed a multiscale process simulation framework for effectively and reliably simulating residual distortion and stress at the part scale. The effectiveness of this proposed framework is shown by simulating a double cantilever beam and a canonical component of different wall thicknesses and comparing the results with experimental measurements, which shows excellent consensus. Ramos et al.²⁶ presented a novel scanning strategy implemented in the SLM manufacturing technology to reduce the residual stress produced during the buildup process. The scanning strategy implemented appeared to be the most effective method for reducing the component’s deformation. Foteinopoulos et al.⁴⁹ established a 3D thermal simulation method for the PBF process. The proposed method provides satisfactory and computationally low-cost results in terms of the strength of predicted thermal stresses and deformations. A new distortion control approach was developed and validated experimentally, based on optical 3D scan measurements.⁵⁰ The findings of the study revealed that distortion reduction in SLM is now applicable on industrial macro-scale parts. The produced distortion in SLM was compensated for by changing the original geometry with FE expected distortion.

Moreover, many scholars have studied the influence of laser scanning patterns on residual thermal stresses and distortion using FE analysis. It was shown that choosing a proper scanning pattern can significantly reduce the material deformation and residual stress and can enhance the model accuracy.^{11,20,51–58} Figure 4 shows the LSFed parts’ distortions using the raster, offset-in, offset-out, and fractal deposition patterns. Concerning as-built component warping and porosity, a newly established scan strategy is compared experimentally with its commercial equivalent. A significant decrease in warping and porosity resulted from the new approach.⁵⁹ Li et al.⁶⁰ created an FE model for quick prediction of SLM fabricated components distortion using four laser scanning strategies. Results showed that the horizontal scanning pattern produced the slightest longitudinal deflection. In contrast, the crosswise direction produced the biggest deflection. The DED-induced distortions and residual stresses were analyzed using a 3D thermo-mechanical FE model.⁶¹ It was found that different scanning techniques used during the preheating process significantly impacted distortion. Some modifications of the actual scanning sequence have been proposed to reduce the computational expense of the simulation framework without compromising quality. Klingbeil et al.⁶² experimented on plate-shaped specimens provided by two methods of direct metal deposition. The experimental findings indicated that both the material deposition process and the deposition direction may significantly influence subsequent warping. Nickel⁶³ explored how the pattern used to deposit a layer affects the substrate’s warpage and analyzed the inter-layer surface defect known as the Christmas Tree Step. Their findings revealed that bolting the substrate down during deposition has a major impact on a component deflection for the elastic-perfectly plastic model. It is necessary to understand how distortion accumulates during AM in a machined surface to design techniques to reduce stress and distortion. Denlinger et al.⁶⁴ performed a distortion measurement on titanium and nickel alloy with different interlayer dwell times. This study’s findings showed that the addition of dwelling time during the deposition process allows for additional cooling, which results in lower distortion and residual stress. Kruth et al.⁶⁵ illustrated the benefit of the processing parameters by SLM to obtain maximum density metal parts. Proper scanning patterns have greatly minimized thermal deformations and the beneficial effect of vaporization to limit the balling effect. Qian et al.⁶⁶ developed a helix scan strategy and progressive scan strategy for SLM technology to reduce the deformation of manufactured layers. From his research, it was found that the helix scan strategy minimizes the deformation of the manufactured layers.

Fig. 4

Exhibitions of LSFed part deformation induced by different deposition patterns (a) raster parallel to the $z$ direction, (b) raster parallel to the $x$ direction, (c) offset-in, (d) offset-out, and (e) fractal.⁵¹

3.2.

Temperature Field and Thermal Analysis

The LAM process involves complex physical and chemical reactions, such as microstructure, changes in segregation, thermophysical properties, rapid melting and cooling, and many other processes. The uneven distribution of the substrate’s temperature field during the laser irradiation causes the cladding part’s deformation. The temperature field should be distributed evenly to solve this issue during the process. Under the current engineering application technology conditions, it is challenging to use experimental methods to measure heat transfer and temperature field distribution in the molten pool. In comparison, the numerical simulation of the temperature field, establishing a model, and determining boundary conditions using a computer can obtain more accurate thermophysical parameters, mass and heat flow fields, as well as other important information. The numerical analysis of the temperature field is based on the actual situation of the process. Figure 5 exhibits the temperature field contours for different laser scanning patterns.

Fig. 5

Temperature field distribution plot for three scanning strategies: (a) line scan, (b) 15-deg rotate scan, and (c) 90-deg rotate scan.⁶⁷

The non-uniform distribution of temperature and the rapid thermal cycle can lead to thermal distortion and dimensional errors in the LC process.^68,69 So, understanding and monitoring thermal behavior during the LAM process is essential.⁷⁰ Gong et al.⁷¹ analyzed the planar LC temperature field with a length direction, a width direction, and a simulated helical scanning path. Results showed that the laser scanning path could significantly affect temperature distribution. Foroozmehr and Kovacevic⁷² and Asikainen et al.⁷³ established a model for evaluating the influence of the scanning pattern on the final distribution of stress, and their findings revealed that the scanning pattern could alter the process temperature history, and as a result, the final properties of the deposited material may change. Ghosh and Choi⁷⁴ analyzed the influence of deposition patterns on induced thermal stresses. It was found that choosing a proper deposition pattern can reduce thermal mismatch, resulting in less warping. From the consistency of the layer viewpoint, the analysis of temperature distribution within the metallic layer during metal laser sintering is essential.^75–81 Denlinger and Michaleris⁸² proposed an approach for modeling stress relief effects at high temperatures during laser DED processes with various inter-layer dwell periods. In their findings, the thermal behavior of workpieces is found to be highly reliant on dwell time. In the powder layer, the high temperature produced results in significant distortion of the component and caused thermal and residual stresses. Song et al.⁶⁷ proposed a combination of FE simulation and experimental verification to analyze the influence of scanning strategies on thermal behavior and residual stress distribution. The temperature distribution simulation results show that scanning strategies significantly influence the temperature field. Patil et al.⁸³ conducted a numerical experiment to investigate the scan pattern’s impact and the time needed to cool down the powder around the melt pool. His research explores the significance of hatch patterns for the thermal history of successive subdomains. The FE formulation for heat transfer in a material with isotropic thermal properties has the following governing equation:

A coupled thermo-mechanical simulation was developed for two different scan strategies to predict residual stress. The findings showed that the scan vector length directly impacts temperature oscillations. Due to the shorter time between neighboring scans, the unidirectional approach has a more consistent minimum temperature.²⁵ An experimentally determined surface convection was integrated into a DED AM thermo-mechanical model.¹⁸ The residual stress measurements and in-situ deflections suggest that the measurement-based convection model generates more precise stress measurements in all situations. Dai and Shaw⁸⁴ performed a 3D simulation of finite elements to investigate the temperature field using a moving laser beam to process several material components. The distribution of temperature, transient stress, residual stress, and distortion of a multi-material part have significant effects on laser processing conditions and material properties, especially thermal conductivity and thermal expansion. The numerical simulation was carried out using the ANSYS code.^85,86 Bian et al.⁸⁷ investigate the effect of laser power and two scanning techniques (stripe scanning and chessboard scanning) by SLM on the residual stress distribution in 316L steel. The implementation of stripe scanning (rather than chessboard scanning) and an increase in laser power from 160 to 200 W typically increased tensile residual stress in the region of interest. Furthermore, compared with switching between two scanning methods, the increase in laser power from 160 to 200 W tended to have a more significant effect. Many attempts have been made to develop the LC process from the 1D model to the 2D model, including thermal and structural analysis.^88,89 The multilayer problem modeling is equally important because the thermal correlations of subsequent layers affect the temperature gradients that regulate heat transfer and thermal stress growth. An advanced simulation technique known as element birth and death is used in modeling the 3D temperature field in several layers in a powder bed.⁹⁰ The results demonstrated that heated regions are subjected to rapid thermal cycles correlated with corresponding thermal stress cycles. The variation in heat dissipation at different substrate locations was investigated to prevent defects such as over-burning and collapsing at the LC layer’s boundary.⁹¹ Over-burning and collapsing regions are predicted and experimentally validated for different direction cladding and the same side cladding. The findings show that the scanning path has a significant impact on boundary over-burning and collapsing. The cladding consistency of the boundary can be improved with the same direction cladding and different side cladding without affecting machining performance, material utilization, or process parameter rationality.

A 3D transient heat transfer model was used to reliably measure the transient temperature field for residual stress and distortion simulation.⁹² The results demonstrated that decreasing the layer thickness can significantly reduce the residual stress. Alimardani et al.⁹³ proposed a model to analyze the influence of the workpiece being preheated and clamped to the positioning table. Results showed that preheating enhanced efficiency by eliminating both the thermal stresses and the settling time in the first layer to create a stable melt pool. Ma and Bin⁹⁴ Ren et al.^95,96 investigated the impact of laser scanning patterns on temperature, residual thermal stresses, and distortion in LAM. An effective 3D thermal history evaluation FE framework was designed to predict temperature field evolution for arbitrary scanning patterns. The scanning patterns evaluation framework used to predict the thermal history is shown in Fig. 6.

Fig. 6

Scanning pattern evaluation framework scheme.⁹⁵

In laser heating, the heat transfer process is a very complex issue. During the laser heating process, the energy source comes primarily from laser beam radiation, latent phase change heat, and deformation heat. Latent heat is released when the metal content is melted or solidified, which will affect the heat transfer process. Furthermore, the material’s plastic deformation can produce deformation heat, which is typically very low compared with the other two heat sources.⁹⁷

3.3.

Microstructure and Mechanical Properties

Many studies about LAM in manufacturing structures or components using different materials, such as stainless steel, titanium alloys, aluminum alloys, and nickel-base superalloys, have been published. The mentioned studies have concerned the investigations of the metal workpieces’ microstructures and mechanical properties or samples fabricated by LAM. For AM, microstructures and mechanical properties are key factors studied to monitor and enhance the consistency of the final product.^98–102 Rare data about the influence of laser beam scanning paths on the material’s microstructures and mechanical properties exists in literature until now. Many scholars have studied the impact of scanning patterns on microstructure and mechanical properties in LBAM; their findings clearly state that the scanning pattern has a significant effect on microstructure and mechanical properties of LBAM fabricated components.^15,103–107 So, to control its microstructures and mechanical properties, it is essential to study the differences in structures and properties of LSF structures or components deposited with different laser beam scanning paths. Figure 7 shows the microstructures of LSF Inconel 718 samples after heat treatment with varying directions of scanning. The distribution of recrystallized grain sizes reflects the allocation of residual thermal stress. A previous study¹⁰⁸ proved that columnar dendrites expand epitaxially along the deposition direction, turning to distribute equiaxed grains after high-temperature recrystallization unevenly. A more uniform allocated grain size in the samples indicates that its residual thermal stress distribution is more consistent.¹⁰⁹ Specific treatments are required to obtain optimal mechanical properties due to the particular microstructure resulting from the SLM method. Different microstructural characteristics in ferritic and austenitic steel DED FGM sections were examined.¹¹⁰ The epitaxial grain growth along the building direction creates an equiaxed grain structure in both parts. In the bidirectional scan, the ferritic steel component has a preferentially inclined grain structure. At the same time, it becomes less favored in orthogonal and island scans due to the combination of scanning strategies between interlayers. Ali et al.¹¹¹ investigated the impact of scanning and rescanning strategies on residual stress formation and mechanical properties of SLM fabricated parts. It was found that the 90-deg alternating scanning strategy produced the least amount of residual stress. Still, there was no apparent connection between mechanical properties. Since the temperature distribution and cooling rate determine the sintered material’s properties, numerical simulations allow SLM to estimate the optimal conditions for producing objects with the preferred microstructure and mechanical properties.¹¹² Sheng et al.¹¹³ established a 3D multitrack and multilayer thin-wall to enhance the quality of thin-wall metal parts provided by the direct laser deposition shaping process using the element birth and death method.

Fig. 7

Recrystallized microstructures of LSF Inconel 718 alloy. (a) Deposited with single direction raster scanning and (b) deposited with cross direction raster scanning.¹⁰³

Based on the study of mechanical components’ characteristics, appropriate scanning route planning impacts the cracking failure behavior of thin-wall metal components. Kruth et al.¹¹⁴ assessed various aspects of SLM manufactured parts with different scanning strategies. Their findings revealed that the scanning strategy could hinder the surface quality, mechanical properties, and grain direction in the SLM parts. Many scholars developed improvements in the grain structure and texture formation by SLM with various scanning techniques. It has been identified that texture control and an apparent grain growth mechanism can be achieved by changing the scanning strategy with a suitable selection of laser power and scanning speed.^{105,115–118} Yang et al.²³ proposed a fractal scanning path for selective laser sintering (SLS). His research showed that SLS samples generated along a fractal scanning path have enhanced physical efficiency. Compared with linear scanning, the grain sizes were more uniform and completely sintered using fractal scanning. Morgan et al.¹¹⁹ adopted a scanning strategy that creates laser re-melted components of stainless steel (316L) with $<1\%$ porosities while retaining the principle of rapid prototyping. Yasa et al.¹²⁰ investigated the reasons for elevated edges in SLM. Different scan strategies and different hatching and contour parameters were evaluated to minimize the edge effect. In addition, for specific additive layer manufacturing components, fatigue loading is crucial and needs to be extensively investigated. Leuders et al.¹²¹ carried out a detailed microstructure and defected analysis to create deep mechanical property-microstructure relationships. According to the analysis, micron-sized pores have the greatest impact on fatigue strength, while residual stresses significantly impact fatigue crack growth. Mangour et al.¹²² studied the SLM process with different laser-scanning techniques to generate cylindrically shaped components from 316L stainless steel. It was found that the desired morphological and crystallographic textures of SLM-processed components characterized by either anisotropic or isotropic properties can be achieved by choosing the correct scanning strategy. Dimitrov et al.¹²³ and Thijs et al.¹²⁴ establish the effect of various scanning techniques and process parameters on the material efficiency of the SLM-manufactured Ti6Al4V parts. Their preliminary study reveals that the significant difference in the temperature distributions, temperature gradients, and cooling rates attributed to the building strategy used will result in substantial differences in the residual stresses and the microstructure of the components generated. Zhang and Shang¹²⁵ analyzed two filling path generation algorithms, namely, the parallel reciprocating scan subarea and the parallel scan contour. This algorithm improved the efficiency of implementation, increased the effectiveness of scanning, and further improved the inner forming consistency of metal pieces. Microstructure determines mechanical properties, and some scholars have studied the influence of process parameters on the microstructure of SLM 316L.^126,127 Investigation of the microstructure and shape of the tracks for process-parameter adjustments may provide valuable information for the creation of a comprehensive strategy for the production of SLM components with standardized microstructure and properties.²⁹ Current research indicates that microstructural flaws inherent in AM processes must be managed and minimized to allow for full use of AM fabricated parts as structural components.

4.1.

Mechanical Properties and Microstructure

LAM fabricated part’s microstructure and mechanical properties vary greatly depending on the build direction and other directions.^109,130 Compared with numerical simulation and analysis, several essential data types, such as the microstructure and internal defects in LAM, are challenging to obtain by online monitoring. In the meantime, traditional equipment is easily disrupted by the surrounding environment. Therefore, more sophisticated equipment and technology, such as high-speed x-ray imaging and spatial resolution acoustic spectrum, must be used to obtain the required data.

4.2.

Deformation

One of the biggest challenges in the LAM process is the deformation of the LAM fabricated part. Scanning patterns can have a greater effect on the structure and deformation of large components. A summary of different scanning patterns used in LBAM to analyze the impact of scanning patterns on deformation, stress, temperature, and mechanical properties is shown in Table 1. It has been seen in several studies that proper scanning patterns can significantly reduce deformation, but the parameters still need to be optimized. Parameters such as laser power, scanning speed, and powder feed rate significantly impact deformation and part quality.^33,87 In future studies, along with different scanning patterns, optimized process parameters should be implemented to minimize deformation and part quality issues. Some scholars used a laser displacement sensor (LDS) to predict the distortion in AM.^64,133 However, there was some irregularity in the data caused by a temporary power buildup on the LDS. Optimizing these issues in the sensor is also a great challenge for obtaining accurate data.

Table 1

Summary of scanning patterns.

4.3.

Surface Quality

Another issue encountered in LAM is the staircase effect or layering error in the fabricated pieces. Surface imperfections and defects such as staircase effects caused by layer-by-layer deposition methods, balling effects, and insufficient fusion result in a notably irregular microstructure. The surface roughness can significantly degrade the performance of laser additive-produced components, limiting their potential applications.¹³⁴ Fatigue performance, wear and scratch resistance, and dimensional precision can be adversely affected by surface defects. Further research is needed to see how LAM’s scanning strategy affects the produced parts’ consistency, surface roughness, and material strength with new material powders.