选区激光熔化（Selective Laser Melting, SLM）作为增材制造（Additive Manufacturing, AM）方法中的一种，由于其成形精度较高，能够直接制造形状复杂的构件，已经在航空航天、医疗和某些高端制造领域得到较为广泛的应用。然而，在现有的约5500多种合金材料中，只有为数不多的材料可用于SLM，而且一般都要经过热处理才能达到锻造的性能。因此，当前SLM材料制备技术的主要研究方向，一是设计适合SLM特点的新材料，另一个则是探索适合现有合金材料的新工艺，特别是能否不经过热处理而使材料或构件的性能达到甚至超过锻造的水平。然而，由于SLM的成形过程十分复杂，许多工艺参数都会对材料的力学性能产生影响。为了从纷繁的各种影响因素中提取主要因素，探索其背后的规律，本文利用无量纲参数及无量纲工艺图作为实验设计的工具，以316L不锈钢材料为研究对象，围绕着如何进一步提升SLM材料的力学性能，对SLM成形件的工艺参数、微观组织及力学性能进行了较为系统地研究。本文的主要研究内容及结论如下：

（1）利用无量纲参数和无量纲工艺图总结了分散在文献中的各种工艺数据，缩小了工艺参数的选取范围，设计了考虑交互作用的正交实验，对SLM工艺过程中的几个重要参数进行了显著性分级，分析了显著性顺序的结果，得到了目标变量分别为表面粗糙度（R_a）、致密度（ρ）及硬度（HV₁）时的最佳工艺参数组合。方差分析结果表明激光功率对所有目标变量的影响都是显著的，而其它参数及其交互作用对目标变量的影响程度则因目标变量的不同而有所变化，因此实验设计时应以更关心的目标变量作为工艺参数的选取依据。

（2）研究了工艺参数对SLM-316L不锈钢微观组织的影响，结果表明：以致密度 > 99%为指标，SLM材料具有一定普适性的工艺窗口范围是无量纲等效能量密度E₀^*介于1.1–2.5之间；无量纲功率q^*对控制SLM过程中的最高温度有重要作用，提高q^*值会减少大缺陷（> 50 μm）的数量；大缺陷（特征尺度> 50 μm）的数量随E₀^*的增加而减小，当E₀^*超过2.33时，大缺陷数量增加，这与致密度的变化相对应；当致密度超过99%时，缺陷的形貌接近于球形。冷却速率随着E₀^*的增加大致呈现单调下降的趋势，且估算的冷却速率量级介于10⁵ K/s到10⁷ K/s之间。当E₀^*从1.49 增加到5.25时，胞结构尺寸从~0.35 μm单调递增到~1.5 μm。

（3）熔池动力学数值模拟结果揭示了两种熔化模式对熔池形状和材料微结构的影响。匙孔模式下熔池内的流动较热导模式的流动更为剧烈。热导模式下熔池内形成了两个对称向外的环流，而匙孔模式下熔池内除了形成一个向外的环流, 还在熔池底部形成一个向下的环流。热传导模式下熔池内最大温度梯度方向逐渐向打印方向偏转，形成<001>织构；而匙孔模式下，熔池内侧向的最大温度梯度占据主导地位，形成<101>织构。

（4）使SLM 打印态的316L不锈钢拉伸性能超过锻造的两种工艺途径是：在热传导熔化模式下，利用较低的能量密度参数提升熔池内的冷却速率，控制胞结构尺寸~0.43 μm；而匙孔熔化模式下，控制熔池深宽比~2.5，形成<011>织构。本文热传导模式下的材料力学性能为σ_y = 584 MPa, σ_UTS = 773 MPa和ε_f = 46%，而匙孔模式下的材料力学性能为σ_y = 596 MPa, σ_UTS= 762 MPa和ε_f = 42%。两种模式下的试样位错密度对SLM-316L不锈钢屈服强度的贡献非常大，在对材料屈服强度进行预测时，建议使用修正后的Hall-Petch关系： 

σ_y=σ₀ + kd^-0.5 + MαGbρ^0.5。

（5）研究了SLM-316L不锈钢的跨尺度层级结构（熔池、晶粒及胞结构）在持续应力作用下直至发生破坏过程中的演变特点。结果表明熔池边界、晶界及胞结构的边界是材料的三个薄弱区域，会在应力的作用下持续发生变形直至失效。材料的这三个薄弱区域导致打印态粗大的柱状晶随着应变的增加而逐渐发生“撕裂”，使得晶粒尺寸随着应变的增加而减小。总的来看，SLM-316L不锈钢的这种变形机制导致初始晶粒的细化并且伴随着形变孪晶的形成，使材料的加工硬化范围提高，韧性提高。

（6）研究了热处理及拉伸速率对SLM-316L不锈钢力学性能的影响。结果表明：873 K热处理后胞结构仍然稳定存在，因而力学性能与打印态基本相同，随着热处理温度的提高，跨尺度的层级结构逐步消失，微观组织逐步粗化，各向异性特征减弱，材料力学性能降低。热处理有助于提高材料的加工硬化能力，加工硬化指数从打印态的0.13增加到固溶处理后的0.33。由于SLM-316L不锈钢微观组织的各向异性特点，用Hollomon关系拟合得到的加工硬化指数n低于材料发生颈缩对应的真应变，当材料热处理温度达到固溶温度时，n值最大且基本满足Hollomon关系；SLM-316L不锈钢的断后延伸率对应变速率非常敏感，虽然SLM-316L不锈钢的率敏感系数（m = 0.017）远大于传统316L粗晶试样（m = 0.006），表现出明显的率强化特性，但是材料内部缺陷的存在使得其在高应变率条件下加速扩展，可能使材料在未完全发挥力学性能潜能之前而过早失效。

Selective laser melting (SLM), as one of the additive manufacturing (AM) methods, has been widely used in aerospace, medical and some high-end manufacturing industry due to its high forming accuracy and its ability to directly produce complex components. However, among the existing ~5500 alloys, only a few of them can be available for SLM. In addition, the post-printing treatment is always needed to achieve mechanical performance as the forging parts. Therefore, the main research directions of the current SLM material preparation technology are the following. One is to design new materials suitable for SLM, and the other is to explore new processing suitable for the existing alloys, such as whether the mechanical performance of SLMed materials can reach or even exceed the corresponding forging component without heat treatment.Nevertheless, the mechanical properties of the SLM-produced parts are affected by many process parameters involved in the SLM process. In order to extract the main factors from various influencing factors and explore the laws behind it, normalized quantities and normalized processing diagram are used as tools for experimental design. The process parameters, microstructure, and mechanical properties of SLM-processed parts are systematically investigated centering on how to further improve the mechanical properties of SLMed 316L stainless steel (316L SS). The main research items and conclusions of this dissertation are as follows:

(1) Dimensionless parameters and normalized processing map are used to summarize various process parameters scattering in literatures with an aim of further narrowing the range of the selection of process parameters. Orthogonal experiment is designed with the consideration of interaction of several important parameters in SLM, and Analysis of Variance (ANOVA) has been performed to rank the order of statistically significance of the parameters on response variables, such as surface roughness (R_a), hardness (HV₁) and density (ρ). After parameter optimization, the best combination of process parameter has been achieved in terms of R_a, HV₁, and ρ. The results of ANOVA show that laser power is a pronounced significant factor for all the response variables, while the significance level of other parameters and their interactions on response variables vary with different response variables, hence process parameters should be selected based on the response variables that are much more concerned.

(2) The effect of process parameters on the microstructure of SLMed 316L SS is studied. The results show that the processing window that may be applicable to general case for achieving high density part (> 99%) is E₀^* value ranging from 1.1 to 2.5. The q^* plays an important role in controlling maximum temperature in the SLM process. Increasing q^* value will reduce the number of large defects (> 50 μm). The number of large defects decreases with the increase of E₀^* but increases if E₀^* exceeds 2.33, corresponding to the variation in density. The morphology of the defect is nearly spherical when high-density (≥ 99%) parts are achieved. The cooling rate tends to monotonously decrease with increasing E₀^* and ranges in the order of 10⁵ K/s to 10⁷ K/s, while the primary dendrite spacing monotonously increases from ~0.35 μm to ~1.5 μm when E₀^*increases from 1.49 to 5.25.

(3) Numerical simulation reveals the influence of two melting modes on the shape and microstructure of melt pool. The fluid flow in the keyhole mode is more intense than that in the heat conduction mode. In heat conduction mode, two symmetrical outward circulations are formed in the melt pool. In the keyhole mode, in addition to an outward circulation, a downward circulation is also formed at the bottom of the molten pool. The direction of the maximum temperature gradient in the melt pool is gradually deflected to the build direction at conduction mode, resulting in the formation of<001>-texture, whereas in the keyhole mode, the lateral maximum temperature gradient in the melt pool is dominant, resulting in the formation of<101>-texture.

(4) Two ways of further improving the mechanical properties of as-SLMed 316L SS are verified by tensile tests. One is utilizing a relatively lower energy density under the heat conduction mode, the cooling rate in the melt pool is increased resulting in the cell size refined to ~0.43 μm. In the keyhole melting mode, the<101>-texture is formed by controlling the aspect ratio of the melt pool to ~2.5. The mechanical properties of the conduction mode samples are σ_y = 584 MPa, σ_UTS = 773 MPa, and ε_f = 46%, while those under keyhole mode are σ_y = 596 MPa, σ_UTS = 762 MPa, and ε_f = 42%. The contribution of dislocation density to yield strength of SLMed materials is significant, and the modified Hall-Petch relation is recommended for the prediction of the yield strength of a SLMed material:σ_y=σ₀ + kd^-0.5 + MαGbρ^0.5.

(5) The evolution of the hierarchical structure (i.e., melt pool, grain, and cell structure) of SLMed 316L SS under stress is studied. The results show that melt pool boundary, grain boundary, and cell structure are the three weak regions in SLMed material, and it will continue to deform until high tensile stress tears them apart. The three weak regions of the material lead to the refinement of the grain size with the increase of strain. In a word, the deformation mechanism of a SLMed material results in the subdivision of original grains and deformation-induced twins which improves the ductility by controlling the deformation process to go through a longer period of work hardening period.

(6) The effect of heat treatment and strain rate on the mechanical properties of SLMed materials are studied. The results show that the cell structure remains stable after 873 K heat treatment, hence the mechanical properties are not changed compared with the as-built samples. With the increase of heat treatment temperature, the hierarchical structure of the material gradually disappears, and the microstructure of the material show a gradually coarsened trend but the anisotropic microstructure weakened, resulting in the decrease of mechanical properties. Heat treatment is helpful to improve the work hardening ability of the SLMed material. The work hardening exponent (i.e., n) of the material increases from 0.13 to 0.33 with increasing heat treatment temperature. Due to the anisotropy of SLMed microstructure, the n value obtained by Hollomon relation is lower than the corresponding true strain where necking is occurred, however, the n value reaches its maximum and basically satisfies the Hollomon relation when solution heat treatment is applied. The elongation to failure of SLMed material is very sensitive to strain rate. Although the strain rate sensitivity of SLMed material (m = 0.017) is much greater than that of the traditional counterparts (m = 0.006), showing obvious rate strengthening characteristics, the existence of internal defects in the material make it accelerate its propagation under the condition of high strain rate, which may cause the material to fail prematurely before it can fully realize its potential of mechanical properties.