materials Article

Numerical Simulation and Experimental Study on Residual Stress in the Curved Surface Forming of 12CrNi2 Alloy Steel by Laser Melting Deposition Zhaoxing Cui 1,2 , Xiaodong Hu 1 , Shiyun Dong 2, *, Shixing Yan 2 and Xuan Zhao 3 1 2 3

*

School of Mechanical and Electronic Engineering, Shandong University of Science and Technology, Qingdao 266590, China; [email protected] (Z.C.); [email protected] (X.H.) National Key Laboratory for Remanufacturing, Army Academy of Armored Forces, Beijing 100072, China; [email protected] State Key Laboratory of Advanced Welding and Joining, Harbin Institute of Technology, Harbin 150001, China; [email protected] Correspondence: [email protected]

Received: 1 September 2020; Accepted: 25 September 2020; Published: 28 September 2020







Abstract: The performance and service life of the nuclear emergency diesel engine shaft made of 12CrNi2 alloy steel is very important for the safety of nuclear power. Laser melting deposition (LMD) is a challenging camshaft-forming technology due to its high precision, rapid prototyping, and excellent parts performance. However, LMD is an unsteady process under the local action of laser, especially for curved surface forming, which is more likely to generate large residual stress on components, resulting in cracks and other defects. At present, the stress research on LMD curved surface forming is relatively insufficient. In the present paper, material parameter testing, high-temperature mechanical properties analysis, single-track sample preparation, and heat source checks are conducted. At the same time, the ABAQUS software and the DFLUX heat source subroutine are used to compile the curved double-ellipsoidal moving heat source, and the effects of the temperature-dependent thermophysical parameters and phase change latent heat on the temperature field are considered. A three-dimensional finite element model is established to analyze the thermal stress evolution and residual stress distribution of multi-track multi-layer on a curved surface by LMD, and the effect of the scanning method and interlayer cooling time on the residual stress of the formed components is studied. The results show that with the increase in temperature, the strength of the material reduces, and the fracture morphology of the material gradually transitions from ductile fracture to creep fracture. The material parameters provide a guarantee for the simulation, and the errors of the width and depth of the melt pool are 4% and 9.6%, respectively. The simulation and experiment fit well. After cooling, the maximum equivalent stress is 686 MPa, which appears at the junction of the substrate and the deposited layer. The larger residual stress is mainly concentrated in the lower part of the deposited layer, where the maximum circumferential stress and axial stress are the tensile stress. Compared with the axial parallel lap scanning method, the arc copying lap scanning method has a relatively smaller maximum thermal stress and residual stress after cooling. The residual stress in the deposited layer is increased to some extent with the increase in the interlayer cooling time. Keywords: laser melting deposition; 12CrNi2 alloy steel; material parameters; thermal stress; residual stress

1. Introduction 12CrNi2 is a widely used low-carbon alloy carburized steel which is widely used in aerospace, electric power, petrochemical, marine, machinery, electronics, environmental protection, and other Materials 2020, 13, 4316; doi:10.3390/ma13194316

www.mdpi.com/journal/materials

Materials 2020, 13, 4316

2 of 25

industries. The addition of nickel and chromium alloy elements can improve the hardenability of the steel and the strength and toughness of the carburized layer [1–4]. It is the preparation material for the camshaft of nuclear power emergency diesel engines [5,6]. As an important energy source, nuclear power plays an important role in human development and sustainable development [7]. A camshaft as an important component of a nuclear power emergency diesel generator; its performance and service life play a vital role in nuclear power safety [8,9]. At present, the camshaft is basically formed by traditional forging, casting, or a combination of parts, but due to the complex shape of the camshaft and the difficulty of machining, the traditional technology cannot meet the needs of the development of the camshaft manufacturing industry [10]. Laser melting deposition (LMD) technology is a new type of laser additive manufacturing (LAM) technology which has the advantages of high precision, a fast forming speed, excellent parts performance, and small machining margin, so it has become a challenging technology for camshaft manufacturing [11–15]. During the LMD process, based on the basic principle of rapid prototype manufacturing, the metal powder sent synchronously is melted to liquid metal by the heat source of a high energy laser according to the set processing path. Then, the liquid metal is quickly solidified, and the metal parts are directly formed through layer by layer deposition. The application prospect is very broad. However, LMD is an unsteady, extremely cold, and hot transient process. In the process of manufacturing accumulation, a local heat input will inevitably lead to a non-uniform temperature field. The local thermal effect is directly manifested as the solidification of the melt pool, and the residual stress is easily formed during the subsequent cooling process [16–18]. As a kind of internal stress, residual stress can directly affect the static load strength, fatigue strength, stress corrosion resistance, and dimensional stability of the formed parts. Therefore, the effective prediction of temperature change in the deposition process and the residual stress after cooling has important guiding significance for the performance control of components. Numerical simulation is a convenient, fast, efficient, and economical method to predict the temperature and stress field of LMD, and it has been widely used [19,20]. Nazemi N et al. [21] established a finite element model of temperature history, microhardness value, and induced residual stress during the laser cladding of low/medium carbon steel plate P420 stainless steel powder, and studied the development of the residual stress of 10 single-track deposited specimens with different process parameters. Chew et al. [22] established a three-dimensional finite element model of the AISI 4340 steel laser cladding process, and simulated the residual stress distribution of single and multiple layers by laser cladding. Gong Cheng et al. [23] established a single-layer cladding model using the ANSYS software, simulated the temperature and stress field of 316L stainless steel during the cladding process, and studied the axial and horizontal residual stress distribution of the laser cladding layer. Alimardani et al. [24] proposed a three-dimensional transient finite element analysis method to simulate the thin-walled forming process of 304L stainless steel, and the transient temperature distribution of the melt pool and the real-time evolution of the stress field were obtained. Dai Deping et al. [25] used the Abaqus software to develop a nonlinear finite element method to simulate the temperature and stress field in the process of laser cladding. Using the developed calculation method, the temperature and stress field of the single-track single-layer cladding, single-track double-layer cladding, and singletrack ten-layer cladding of Inconel718 Ni-base alloy were numerically simulated. Ding et al. [9] established a multi-layer multi-track finite element model using the ABAQUS software, and studied the effect of different substrate preheating temperatures on the residual stress of a 12CrNi2 multi-layer multi-track specimen. Kang et al. [26] established a single-layer, double-track finite element model of 24CrNiMo to simulate the temperature field and stress field distribution in the forming process of LMD. Kiran et al. [27] established a single-track, multi-track finite element model of 316L to simulate the temperature field and stress field distribution. Gharbi et al. [28] studied the DMD process using a Yb-YAG disk laser and a widely used titanium alloy (Ti–6Al–4V) to understand the influence of the main process parameters on the surface finish quality. Qu et al. [29] fabricated a Ti-47Al-2.5V-1Cr intermetallic alloy by the laser melting

Materials 2020, 13, 4316

3 of 25

deposition (LMD) manufacturing process and studied the microstructure by OM, SEM, TEM, and XRD. Cottam et al. [30] investigated the effect of the deposition rate on the residual stresses formed during the laser cladding of Ti-6Al-4V powder onto a Ti-6Al-4V substrate. The above studies are mainly aimed at titanium alloy [31], nickel-based alloy [32,33], stainless steel, and other materials [34], but there are relatively few studies on 12CrNi2 high-performance alloy steel, which plays an important role in the national economy and national defense. At the same time, the research on the temperature and stress field of LMD is mainly based on single-track, thin wall, and plane models. However, the camshaft forming process includes cam, boss forming, and many other curved surface forming processes, and the research on cam curved surface forming is still insufficient. The stress distribution of LMD curved surface forming is mainly divided into axial stress, circumferential stress, and radial stress, which are different from the transverse stress and longitudinal stress of plane in both the distribution and magnitude. How to realize the simulation of the temperature field and stress field of LMD curved surface forming to predict the stress distribution; how to understand the change in the mechanical properties of materials under high temperatures and the change in the thermal stress of the LMD curved surface forming process; and how to choose the appropriate forming strategy are the difficulties. In this paper, numerical simulation was combined with the experimental method. Based on the thermoelastic plastic method, the ABAQUS finite element software was used to compile the curved surface moving double-ellipsoid heat source program. Considering the influence of the thermophysical parameters changing with temperature and the latent heat of phase transition on the temperature field, the finite element analysis model of the curved surface multi-layer multi-track LMD process was established. The influence of the thermal stress evolution, residual stress distribution, scanning mode, and interlayer cooling time on the residual stress of the curved forming component was analyzed, which provided a reference for stress–strain regulation and the actual LMD curved surface forming and camshaft manufacturing. 2. Finite Element Model 2.1. Geometric Model A schematic diagram of the curved surface forming by LMD is shown in Figure 1a. The round rod or round tube is chosen as the substrate constrained by a three-jaw chuck at one end and a thimble at the other. When the size of the substrate is long, support should be added in the middle of the substrate. In the forming process, the rotation of the substrate is driven by the three-jaw card; at the same time, the laser head moves in coordination under the control of a robot. Through the high-energy laser heat source, the metal powder sent synchronously to the robot is melted, quickly solidified, deposited layer by layer, and finally the metal component is formed. The forming strategy is shown in Figure 1b. At the beginning, the arc copying lap and S-scan mode were adopted. Figure 2 shows the multi-layer and multi-track finite element model of the curved surface formed by LMD. The size of the substrate is 80mm in outer diameter, 10 mm in thickness, and 120 mm in length. The width of the deposited layer is 8mm and the thickness is 1.2 mm, which is composed of two layers and six tracks. The arc length of the deposited layer is 21 mm. The DC3D8 element for heat transfer is used in the temperature field. The sweep technique is used to divide the mesh. The deposition layer is divided into hexahedron mesh, and the mesh size is 0.5 mm × 0.5 mm × 0.3 mm. The mesh length of the substrate near the deposited layer is 1 mm, and it increases gradually to a maximum of 20 mm far away from the deposited layer. The mesh size becomes larger gradually in the thickness direction far away from the deposited layer. The minimum mesh is 0.2 mm and the maximum mesh is 2 mm.

Materials 2020, 13, 4316

4 of 25

Materials 2020, 13, x FOR PEER REVIEW

4 of 26

(a)

(b)

Figure 1. Schematics of (a) the LMD curved surface forming and (b) scanning strategy.

Figure 2 shows the multi-layer and multi-track finite element model of the curved surface formed by LMD. The size of the substrate is 80mm in outer diameter, 10 mm in thickness, and 120 mm in length. The width of the deposited layer is 8mm and the thickness is 1.2 mm, which is composed of two layers and six tracks. The arc length of the deposited layer is 21 mm. The DC3D8 element for heat transfer is used in the temperature field. The sweep technique is used to divide the mesh. The deposition layer is divided into hexahedron mesh, and the mesh size is 0.5 mm × 0.5 mm × 0.3 mm. The mesh length of the substrate near the deposited layer is 1 mm, and it increases gradually to a maximum of (a)20 mm far away from the deposited layer. The mesh (b) size becomes larger gradually in the thickness direction far away from the deposited layer. The minimum mesh is 0.2 Figure Schematicsofof the LMD curved surface forming and scanning strategy. Figure 1. 1.Schematics (a)(a) the LMD curved surface forming and (b)(b) scanning strategy. mm and the maximum mesh is 2 mm. Figure 2 shows the multi-layer and multi-track finite element model of the curved surface formed by LMD. The size of the substrate is 80mm in outer diameter, 10 mm in thickness, and 120 mm in length. The width of the deposited layer is 8mm and the thickness is 1.2 mm, which is composed of two layers and six tracks. The arc length of the deposited layer is 21 mm. The DC3D8 element for heat transfer is used in the temperature field. The sweep technique is used to divide the mesh. The deposition layer is divided into hexahedron mesh, and the mesh size is 0.5 mm × 0.5 mm × 0.3 mm. The mesh length of the substrate near the deposited layer is 1 mm, and it increases gradually to a maximum of 20 mm far away from the deposited layer. The mesh size becomes larger gradually in the thickness direction far away from the deposited layer. The minimum mesh is 0.2 mm and the maximum mesh is 2 mm. (a) (b) Figure Figure 2. 2. Mesh Mesh model: model: (a) (a)overall overall model, model, (b) (b) partial partial model. model.

In order lawlaw of temperature and and stress, the data path ispath drawn In order to tostudy studythe thedistribution distribution of temperature stress, the extraction data extraction is up as shown in Figure 3. Figure 3 shows the plane diagram after the curved surface is expanded, drawn up as shown in Figure 3. Figure 3 shows the plane diagram after the curved surface is in which Node1 is the center of thepoint second track on the second deposited layer, pathlayer, 1 is onpath the expanded, in which Node1 is point the center of the second track on the second deposited of the deposited and path 2 isand on the upper of the substrate. paths are 1upper is on surface the upper surface of thelayer, deposited layer, path 2 is surface on the upper surface of Both the substrate. Materials 2020, 13, x FOR PEER REVIEW 5 of 26 in thepaths sameare plane as Node1 and perpendicular the scanningto direction. Both in the same plane as Node1 and to perpendicular the scanning direction.

(a)

(b)

Figure 2. Mesh model: (a) overall model, (b) partial model.

In order to study the distribution law of temperature and stress, the data extraction path is drawn up as shown in Figure 3. Figure 3 shows the plane diagram after the curved surface is expanded, in which Node1 is the center point of the second track on the second deposited layer, path 1 is on the upper surface of the deposited layer, and path 2 is on the upper surface of the substrate. Both paths are in the same plane as Node1 and perpendicular to the scanning direction.

Figure 3. Schematic diagram of the data extraction location. Figure 3. Schematic diagram of the data extraction location.

2.2. Finite Element Calculation Settings The initial calculation model is based on the process parameters of laser power 2000 W, scanning speed 5 mm/s, and powder feeding rate 11.6 g/min. Aiming at the finite element calculation

Materials 2020, 13, 4316

5 of 25

2.2. Finite Element Calculation Settings The initial calculation model is based on the process parameters of laser power 2000 W, scanning speed 5 mm/s, and powder feeding rate 11.6 g/min. Aiming at the finite element calculation setting of the temperature field, firstly, according to the material parameters obtained by the test, the relevant material properties are set for the model, and the time step is mainly divided into three parts—namely, the element life and death step, the heat source loading step, and the cooling step. The main type is heat transfer. At the same time, heat transfer by convection and radiation is achieved by setting the convection and radiation boundary conditions, and the undeposited area is temporarily suppressed using the model change function. Secondly, we use the self-defined heat source subroutine to load and set the initial ambient temperature at 25 ◦ C. Finally, we create a job to solve [35–38]. The finite element model of the stress field is the same as the temperature field model, but it is necessary to change the element type to 3D stress, and to set the prestress field as the odb file of the temperature field calculated before. Here, the right-end face of the model is restricted by five degrees of freedom, except the axial rotation, and the left-end face is imposed with axial and radial constraints. 2.3. Load and Boundary Conditions 2.3.1. Heat Source Model and Latent Heat As a kind of bulk heat source, a double-ellipsoidal heat source has been widely used in the calculation of temperature field, such as for welding and laser cladding, as it takes into account the inconsistent energy distribution caused by the movement of the heat source and the influence of the heat source on the depth direction. The laser energy is distributed in a certain volume and is applied to the nodes of the material model in the form of heat flux density. The double-ellipsoid heat source Materials 2020, 13, x FOR PEER REVIEW 6 of 26 model is shown in Figure 4.

Figure 4. Double-ellipsoid heat source model. Figure 4. Double-ellipsoid heat source model.

The heat flux inside the ellipsoid along the front half axis of the model is as follows: The heat flux inside the ellipsoid along the front half axis of the model is as follows: √   2 6 3 ff Q    x2 2 y y 2 z2   2 6 3 f Q  2x− 3 2 − 3 2  q(x, y, z) = √ f exp .z  −3 q x , y , z = π πabc f exp  −c3f 2 −a 3 2 b− 3 2 

(1) (1)

The heat flux inside the ellipsoid along the rear axis of the model is as follows: The heat flux inside the ellipsoid along √ the rear(axis of the model is) as follows: 6 3 fr Q y2 x2 z2 23 23 q(x, y, z) = 6√ 3 f Qexp −3 − −  2 x a2 y b2 .z 2  r b q x , y , z = π πabc exp −c3b −3 −3 .

(2) (2)

(

(

)

)

π π abc f

π π abcb



 

cf

cb 2

a

a2

b  .

 b 2 

where Q is effective thermal power; a, b, cf, and cb are heat source shape parameters; ff and fr are the proportional coefficients of the energy distribution of the anterior and posterior ellipsoid; ff + fr = 2. For the laser heat source, cf = cb= a, so ff = fr.

Q = P⋅η, where P is the laser power; η is the absorptivity of the material to the laser energy.

(3)

Materials 2020, 13, 4316

6 of 25

where Q is effective thermal power; a, b, cf , and cb are heat source shape parameters; ff and fr are the proportional coefficients of the energy distribution of the anterior and posterior ellipsoid; ff + fr = 2. For the laser heat source, cf = cb = a, so ff = fr . Q = P · η,

(3)

where P is the laser power; η is the absorptivity of the material to the laser energy. Therefore, the double-ellipsoidal heat source model can be expressed as: q(x, y, z) =

√ ( ) y2 6 3Pη x2 z2 exp −3 − 3 − 3 . √ c2 a2 b2 π πabc

(4)

The above formula is based on the Cartesian coordinate system, which is very suitable for the plane state of the substrate surface, but for curved surface forming the size of the heat source is different and the temperature field is very unstable due to the change in spatial coordinates (x, y, z). Here, the double-ellipsoidal heat source is improved, taking the cylindrical coordinate system as the benchmark, considering the change with the spatial position, and introducing the θ angle. The formula is as follows:   √ 2    y2 (z0 − R sin θ)2  6 3Pη  (x0 − R cos θ)  exp − 3 − 3 q(x, y, z) = √ −3 , (5)      c2 a2 b2 π πabc tv L − θ0 = − θ0 . (6) R R where L is the arc length of the heat source moving, and the size is the product of the scanning speed and the scanning time; t is the scanning time; v is the scanning speed; R is the curvature radius of the surface (when the surface is a cylindrical surface, R is the cylindrical radius); and θ0 is the angle between the line connecting the initial position to the center point of the surface and the positive direction of the X axis. In the LMD process, there is a solid phase to liquid phase to solid phase transition, during which the material will continue to absorb or release a large amount of heat called phase transition latent heat. Therefore, the latent heat of the phase transition must be considered when establishing the finite element model [39]. The method of dealing with latent heat in ABAQUS (2017, SIMULIA, Johnston, RI, USA) is the enthalpy method—that is, the method of enthalpy varying with temperature is used to define latent heat, and its expression is as follows: θ=

Z ∆H =

ρc(T )dT,

(7)

where ∆H is enthalpy, ρ is density, and c is the specific heat capacity. 2.3.2. Boundary Conditions In the process of solving, it is necessary to give the initial temperature distribution and the boundary conditions of heat transfer in order to obtain the appropriate calculation results—that is, the uniqueness of the solution. There are three main types of boundary conditions [40–42]: (1) The first kind of boundary condition: Referring to the initial temperature or temperature function on the boundary of an object is the initial temperature distribution of the model before calculation. For the LMD process, it is mainly the temperature of the powder and substrate: T|s = TS orT|s = TS (x, y, z, t),

(8)

Materials 2020, 13, 4316

7 of 25

where S is the boundary range of deposition forming, TS is the environment temperature, and TS (x, y, z, t) is the temperature function of the surface of the deposited specimen. (2) The second kind of boundary condition: Referring to the input of heat on the surface of the specimen and the known heat flux on the surface of the object is: ∂T ∂T ∂T kx nx + k y n y + kz nz = qs (x, y, z, t), (9) ∂x ∂y ∂z where nx , n y , and nz are the cosine of the normal direction outside the boundary, and qs (x, y, z, t) is the heat flux density function. (3) The third kind of boundary condition: This refers to the heat exchange between the material model and the surrounding environment, which mainly includes convective heat transfer and thermal radiation. Among them, the convective heat transfer can be expressed as: kx

∂T ∂T ∂T nx + k y n y + kz nz = h(Te − Ts ), ∂x ∂y ∂z

(10)

where h is the convective heat transfer coefficient, Te is the fluid medium temperature, and TS is the environment temperature. Thermal radiation can be expressed as: kx

∂T ∂T ∂T nx + k y n y + kz nz = σ0 ε(Te 4 − Ts 4 ), ∂x ∂y ∂z

(11)

where σ0 is the Stephen Boltzmann constant, and the value is 5.67 × 10−8 W·m−2 ·K−4 ; ε is the radiation heat transfer coefficient. In general, the boundary condition of the mixture of thermal convection and thermal radiation heat transfer is most often considered in the third kind of boundary condition, and its expression is as follows: ∂T ∂T ∂T kx nx + k y n y + kz nz = h(Te − Ts ) + σ0 ε(Te 4 − Ts 4 ). (12) ∂x ∂y ∂z 2.4. Stress Analysis (1) Stress–strain relationship In the process of LMD forming, the relationship between thermal elastoplastic stress and strain can be expressed as follows: (13) { dσ} = [D]{dε} − {C}dT, where {dσ} is the stress increment, { C} is the vector matrix related to the temperature, { dε} is the strain increment, [D] is the elastic-plastic or elastic matrix, and dT is the temperature increment. If the material is in the elastic zone: [D] = [D]e , (14)     ∂[D]−1 e (15) { C} = { C} e = [D]e { α} + { σ} , ∂T where T is temperature and α is the linear expansion coefficient. When the material reaches the plastic range:   f (σ) = f0 εp , T ,

(16)

where f0 is the yield stress function related to the temperature and plastic strain and f is the yield function.

Materials 2020, 13, 4316

8 of 25

At the same time, the mathematical expression of the plastic strain increment { dε} P of the material is as follows: ( ) ∂f , (17) dε} = λ { 0 p ∂σ where λ0 is related to the hardening criterion and the material used for deposition. (2) Solving process: The relationship between the LMD numerical solution {dε}e and { dδ} e can be expressed as follows: {dε}e = [B]{ dδ} e .

(18)

Finally, the stress increment { dσ} of the corresponding structural unit can be solved by the 9 of 26 Formula (13).

Materials 2020, 13, x FOR PEER REVIEW

3. Experimental Materials and Methods 3. Experimental Materials and Methods 3.1. Materials and Deposition Processes 3.1. Materials and Deposition Processes The powder used in this experiment is 12CrNi2 alloy steel powder, and its chemical composition The in powder is shown Table 1 used [43]. in this experiment is 12CrNi2 alloy steel powder, and its chemical composition is shown in Table 1 [43]. Table 1. Chemical composition of the 12CrNi2 powder (quality components, wt%). Table 1. Chemical composition of the 12CrNi2 powder (quality components, wt%). Element C Fe Si Cr Ni Mn O

Element Content

C Content 0.12

Fe 0.12 Bal

Bal Si 0.34

1Cr

0.34

1

1.59

Ni0.57 1.59

Mn 0.008 0.57

O 0.008

After a lot of exploration and research from composition design to self-preparation, the powder After a lot of exploration and research from composition design to self-preparation, the powder was finally prepared by ultrasonic gas atomization equipment developed by the Institute of Metals, was finally prepared by ultrasonic gas atomization equipment developed by the Institute of Metals, Chinese Academy of Sciences. Chinese Academy of Sciences. A scanning electron microscope was used for observation, and the observation results are shown A scanning electron microscope was used for observation, and the observation results are in Figure 5. It can be seen from the figure that the powder has a good sphericity, and most of the shown in Figure 5. It can be seen from the figure that the powder has a good sphericity, and most of particles are regular spherical, which can form a better fluidity and is conducive to the powder’s the particles are regular spherical, which can form a better fluidity and is conducive to the powder's absorption of laser energy. absorption of laser energy.

(a)

(b)

Figure 5. Micromorphology Micromorphologyofof 12CrNi2 powder: (a) multiple (b)powder. single Figure 5. thethe 12CrNi2 alloyalloy steel steel powder: (a) multiple powder,powder, (b) single powder.

The laser melting deposition experiment adopts the laser additive manufacturing system, which is The laser deposition experiment adopts robot the laser additive manufacturing system, composed of a melting laser system, powder feeding system, system, and control system. Among which is composed of ais laser system, powder feeding robot system, fiber and laser. controlThe system. them, the laser system mainly the IPG YLS-4000 (IPG,system, Burbach, Germany) laser Among them, the laser system is mainly the IPG YLS-4000 (IPG, Burbach, Germany) fiber laser. The laser is output by a fiber with a 1mm core diameter, and the matching laser head is the PRECITEC YC52 (Precitec, Gaggenau, Germany) laser head. During the experiment, firstly, 80 sieve screens were used to screen the powder to filter impurities and large particles, and then the powder was put into an electric drying oven with a constant temperature. The heating temperature was set at 120 ℃

Materials 2020, 13, 4316

9 of 25

is output by a fiber with a 1mm core diameter, and the matching laser head is the PRECITEC YC52 (Precitec, Gaggenau, Germany) laser head. During the experiment, firstly, 80 sieve screens were used to screen the powder to filter impurities and large particles, and then the powder was put into an electric drying oven with a constant temperature. The heating temperature was set at 120 ◦ C for 3 h Materials 2020, 13,moisture x FOR PEER 10 of 26 to remove the inREVIEW the powder. Secondly, a muffle furnace is needed to preheat the substrate. Finally, the test can only be carried out on the premise of ensuring the normal operation of the protective gas and powder feeding system. 3.2. Testing of Material Parameters 3.2. Testing of Material Parameters 3.2.1. Testing of the Thermophysical Parameters of Materials 3.2.1. Testing of the Thermophysical Parameters of Materials In order to ensure the accuracy of the thermophysical parameters of the material, the parameter data In varying with temperature are obtained by experimental measurement Jmatpro order to ensure the accuracy of the thermophysical parameters of theand material, thesimulation. parameter According to with the experimental theexperimental sample size and test standards shownsimulation. in Table 2. data varying temperaturerequirements, are obtained by measurement andare Jmatpro Then, the scanning strategy of 90° interlayer rotation and size the process of the laserinpower According to the experimental requirements, the sample and testparameters standards are shown Tableof 2. ◦ 2000 W, scanning speedofof905 mm/s, and rotation the powder feeding rateparameters of 11.6 g/min arelaser adopted to Then, thethe scanning strategy interlayer and the process of the power deposit thethe sample. The final sample size isand obtained by wirefeeding cutting.rate of 11.6 g/min are adopted to of 2000 W, scanning speed of 5 mm/s, the powder deposit the sample. The final sample size is obtained by wire cutting. Table 2. The sample size and test standards of the different tested parameters. Table 2. The sample size and test standards of the different tested parameters.

The Sample Test Standards or Equipment SizeSample Size The Test Standards or Equipment The thermal conductivity and φ12.5 mm × The thermal conductivity and GB/T22588-2008 [44] ϕ12.5 mm × 2.5 mm GB/T22588-2008 [44] heat capacity specificspecific heat capacity 2.5 mm UAE/PH-TDWN010 automatic φ15ϕ15 mmmm × 35 Density × 35 mm UAE/PH-TDWN010 automatic true Density true density analyzer mm density analyzer Thermal expansion analyzer The thermal expansion coefficient φ3 ϕ3 × 50 mm Thermal expansion analyzer (Baehr, mmmm × 50 (Baehr, Pirmasens Germany) The thermal expansion coefficient mm Pirmasens Germany) Parameter Name

Parameter Name

3.2.2. Testing of Mechanical Properties at High Temperature 3.2.2. Testing of Mechanical Properties at High Temperature In order to test the high-temperature mechanical properties of 12CrNi2 alloy steel by laser melting In order to test the high-temperature mechanical properties of 12CrNi2 alloy steel by laser deposition, the high-temperature tensile specimens were prepared in conjunction with the testing melting deposition, the high-temperature tensile specimens were prepared in conjunction with the Center of University of Science and Technology Beijing. The high-temperature tensile tests were carried testing Center of University of Science and Technology Beijing. The high-temperature tensile tests out according to the GB/T 228.2-2015 [45]. The size of the sample is shown in Figure 6.

Figure Dimensional drawing Figure 6. 6. Dimensional drawing of of the the high-temperature high-temperature tensile tensile specimen specimen (unit: (unit: mm). mm).

3.3. Observation on the Morphology of the Molten Pool of the Sample 3.3. Observation on the Morphology of the Molten Pool of the Sample The process parameters of laser power 2000 W, scanning speed 5 mm/s, and powder feeding rate The process parameters of laser power 2000 W, scanning speed 5 mm/s, and powder feeding 11.6 g/min were selected to carry out a single-track deposition experiment. The sample was cut along rate 11.6 g/min were selected to carry out a single-track deposition experiment. The sample was cut the section by wire cutting, and the cut samples were inlaid by an automatic mosaic machine. Secondly, along the section by wire cutting, and the cut samples were inlaid by an automatic mosaic machine. the embedded samples were soaked and washed in acetone, ground to 2000 mesh with sandpaper, Secondly, the embedded samples were soaked and washed in acetone, ground to 2000 mesh with polished, and then etched with 4% nitric acid alcohol solution for 10 s to make a metallographic sandpaper, polished, and then etched with 4% nitric acid alcohol solution for 10 s to make a metallographic sample. Finally, the prepared metallographic samples were placed under a three-dimensional profiler to observe the cross-section morphology of the melt pool.

Materials 2020, 13, 4316

10 of 25

sample. Finally, the prepared metallographic samples were placed under a three-dimensional profiler 11 of 26 to observe the cross-section morphology of the melt pool.

Materials 2020, 13, x FOR PEER REVIEW

4. Results and Discussion 4. Results and Discussion 4.1. Test Results of Material Parameters 4.1. Test Results of Material Parameters 4.1.1. Thermophysical Parameters 4.1.1. Thermophysical Parameters The thermophysical parameters of the material are obtained as shown in Figure 7. It can be seen material are obtained shown in Figure 7. It can be seen fromThe the thermophysical figure that with parameters the increaseof inthe temperature, the specificas heat capacity increases at first and from the figureand thatthe with theconduction increase incoefficient temperature, the specific heatwhile capacity increases at first and then decreases, heat decreases gradually, the change in the thermal then decreases, and the heat conduction coefficient decreases gradually, while the change in expansion coefficient fluctuates slightly, showing a trend of decreasing at first and then increasingthe as thermal a whole. expansion coefficient fluctuates slightly, showing a trend of decreasing at first and then increasing as a whole.

(a)

(b)

(c) Figure 7. Physical parameters parameters of of12CrNi2: 12CrNi2:(a) (a)specific specificheat, heat,(b) (b)thermal thermalconductivity, conductivity, coefficient (c)(c) coefficient of of thermal expansion. thermal expansion.

4.1.2. Mechanical Mechanical Properties Properties at at High High Temperatures Temperatures 4.1.2. According to to the the GB/T GB/T 228.2-2015 According 228.2-2015 standard standard [45], [45], the the parallel parallel length length is is 40 40 mm, mm, the the extensometer extensometer standard distance length is 25 mm, the original standard distance length is 25 mm, standard distance length is 25 mm, the original standard distance length is 25 mm, and and the the parallel parallel section diameter 5 mm. In cooperation with the testing Center of University of Science and Technology section diameteris is 5 mm. In cooperation with the testing Center of University of Science and ◦ C are tested. According to the of Beijing, the tensile properties of the samples at 200, 400, 600, and 800 Technology of Beijing, the tensile properties of the samples at 200, 400, 600, and 800 ℃ are tested. tensile datatoobtained, thedata stress–strain are drawn curves as shown Figure According the tensile obtained,curves the stress–strain arein drawn as8.shown in Figure 8.

Materials 2020, 13, x FOR PEER REVIEW Materials 2020, 13, 4316 Materials 2020, 13, x FOR PEER REVIEW

12 of 26 11 of 25 12 of 26

Figure 8. Stress–strain curve of 12CrNi2. Figure 8. Stress–strain curve of 12CrNi2. Figure 8. Stress–strain curve of 12CrNi2.

Figure 88 shows shows that that with with the the increase increase in in temperature, temperature, the Figure the yield yield strength strength decreases decreases gradually, gradually, Figure 8 shows that with the increase in temperature, the yield strength decreases gradually, while the tensile strength increases at first and then decreases, and the elongation after fracture while the tensile strength increases at first and then decreases, and the elongation after fracture increases. while the tensile strength increases at first and then decreases, and the elongation after fracture increases. The in temperature and the plastic of The increase in increase temperature reduces thereduces strengththe andstrength improves theimproves plastic properties of properties the material. increases. The increase in temperature reduces the strength and improves the plastic properties of the material. The classical plasticity theory of metal materials is applied for the default plastic material the material. The classical plasticity and theory of yield metalsurface materials is applied defaultyield. plastic material characteristics of ABAQUS, mises is used to definefor thethe isotropic The plastic The classical plasticity and theory of yield metalsurface materials is applied forthethe defaultyield. plastic material characteristics of ABAQUS, mises is used to define isotropic The plastic deformation behavior of metal materials can be briefly described, as shown in Figure 9. In the case of characteristics of ABAQUS, and mises yield surface is described, used to define the isotropic yield. The plastic deformation behavior of metal materials can be briefly as shown in Figure 9. In the case of small strain, the material properties are basically linear elastic and the elastic modulus E is constant; deformation behavior of metal materials can be briefly asthe shown inmodulus Figure 9. E Inisthe case of small strain, the material properties are basically lineardescribed, elastic andsignificantly, elastic constant; when the stress exceeds the yield stress, the stiffness decreases and the strain of the small strain, the material properties are basically linear elastic and the elastic modulus E is constant; when theincludes stress exceeds the yield stress, strain. the stiffness decreases the significantly, and the strainbut of the material plastic strain and elastic After unloading, elastic strain disappears, the when theincludes stress exceeds the yield stress, the stiffness decreases significantly, and the strain of but the material plastic strain and elastic strain. After unloading, the elastic strain disappears, plastic strain is irrecoverable; if loaded again, the yield stress of the material will increase, which is the material includes strain and elastic again, strain. the After unloading, strain but the plastic strain isplastic irrecoverable; if loaded yield stress of the the elastic material will disappears, increase, which so-called work hardening. the plastic strain is irrecoverable; if loaded again, the yield stress of the material will increase, which is the so-called work hardening. is the so-called work hardening.

Figure Real stress–strain stress–strain curve. curve. Figure 9. 9. Real Figure 9. Real stress–strain curve.

The are usually The data data obtained obtained in in the the uniaxial uniaxial tensile tensile experiment experiment are usually expressed expressed in in terms terms of of nominal nominal strain and nominal stress. Plastic strain and real stress are required when defining plastic material The data obtained in the uniaxial tensile experiment are usually expressed in terms of strain and nominal stress. Plastic strain and real stress are required when defining plastic nominal material parameters in they are to theories, and elastic strain and nominal stress.Therefore, Plastic strain real stressaccording are required when defining material parameters in ABAQUS. ABAQUS. Therefore, they and are converted converted according to relevant relevant theories,plastic and the the elastic modulus and yield strength of the material are calculated as shown in Figure 10, which shows that the parameters in ABAQUS. Therefore, they are converted according to relevant theories, and the elastic modulus and yield strength of the material are calculated as shown in Figure 10, which shows that elastic modulus and yield limit of the material decrease gradually with the increase in temperature. modulus yield strength of thelimit material are material calculateddecrease as showngradually in Figure with 10, which shows that the elasticand modulus and yield of the the increase in

the elastic modulus and yield limit of the material decrease gradually with the increase in temperature. temperature.

Materials 2020, 13, 4316 Materials Materials 2020, 2020, 13, 13, xx FOR FOR PEER PEER REVIEW REVIEW

12 of 25 13 13 of of 26 26

(b)

(a)

Figure Figure 10. 10. Physical Physical parameters parameters of of 12CrNi2: 12CrNi2: (a) (a) modulus of elasticity, (b) yield yield limit. limit. (a)modulus modulusof ofelasticity, elasticity, (b)

thethe surface morphology of the tensile An obviousAn necking phenomenon Figure 11 11shows shows surface morphology of thespecimens. tensile specimens. obvious necking ◦ C, and the fracture surface of the sample at 800 ◦ C is is observed in the specimens at 200, 400, and 600 phenomenon is observed in the specimens at 200, 400, and 600 ℃, and the fracture surface of the zigzag. at With inWith temperature, the surface oxidationthe of the sample is gradually sample 800the ℃ increase is zigzag. the increase in temperature, surface oxidation of theintensified, sample is and the length of the sample fracture gradually increased. gradually intensified, and theafter length of theissample after fracture is gradually increased.

Figure Figure 11. 11. Surface Surface morphology morphology of of the the samples. samples.

The fracture fracture morphology morphology was was further further observed observed by by scanning scanning electron electron microscope, microscope, as as shown shown in in The ◦ C are ductile fractures. Figure 12a,c shows that there are fiber Figures 12 and 13. Both 200 and 400 Figures 12 and 13. Both 200 and 400 ℃ are ductile fractures. Figure 12a,c shows that there are fiber zones, shear shear lip lip zones, zones, and and aa small small amount amount of of pore pore defects defects on on the the macroscopic macroscopic fracture fracture surface surface of of the the zones, ◦ ◦ samples at C. There inin the center of of thethe sample at 200 C, ℃, which is the samples at 200 200 and and400 400 ℃. Thereisisaalarge-sized large-sizedhole hole the center sample at 200 which is result of the exfoliation of unmelted powder or large-sized inclusions in the separation process of the the result of the exfoliation of unmelted powder or large-sized inclusions in the separation process of fracture sample underunder tensiletensile load. In Figure 12b,d, there arethere dimples. In addition, large number of the fracture sample load. In Figure 12b,d, are dimples. In aaddition, a large micropores and small dimples can be seencan in Figure which12b, is anwhich obvious feature of afeature micropore number of micropores and small dimples be seen12b, in Figure is an obvious of a ◦ C are smaller and shallower than those at accumulation fracture. The dimples of the samples at 200 micropore accumulation fracture. The dimples of the samples at 200 ℃ are smaller and shallower 400 ◦those C. than at 400 ℃.

Materials 2020, 13, 4316

13 of 25

Materials Materials 2020, 2020, 13, 13, xx FOR FOR PEER PEER REVIEW REVIEW

14 14 of of 26 26

(a)

(b)

(c)

(d)

◦ C, (c,d) ◦ C. Figure 12. 12. Fracture Fracture morphology morphology at at 200 200 and and 400 400 ◦℃: Figure C: (a,b) (a,b) 200 200 ℃, (c,d) 400 400 ℃.

(a)

(b) Figure 13. Cont.

Materials 2020, 13, 4316

14 of 25

Materials 2020, 13, x FOR PEER REVIEW

15 of 26

Materials 2020, 13, x FOR PEER REVIEW

(c)

15 of 26

(d)

◦ C. (c) (d) Figure 13. 13. Fracture Fracture morphology morphology at at 600 600 and and 800 800 ◦℃: Figure C: (a,b) 600 ◦℃, C, (c,d) 800 ℃.

Figure 13. Fracture morphology at 600 and 800 ℃: (a,b) 600 ℃, (c,d) 800 ℃.

C. The creep fracture Figure 13 shows the fracture morphology of the samples at 600 and 800 ◦℃. ◦ C, which is due to the creep phenomenon during tension at high temperatures. occursFigure atat600 and 800 60013 and 800 which is due toof the creep phenomenon at high shows the℃, fracture morphology the samples at 600 and 800during ℃. The tension creep fracture ◦ C are not Figure shows macroscopic fracture surfaces of the surfaces samplesduring at the 600 samples and 800 at temperatures. Figure 13a,c shows thatis the macroscopic fracture of 600 occurs13a,c at 600 andthat 800the ℃, which due to the creep phenomenon tension highand ◦ C has a necking phenomenon and the creep fracture has obvious smooth, in which the specimen at 600 temperatures. Figure 13a,c shows that the macroscopic fracture surfaces phenomenon of the samplesand at 600 800 ℃ are not smooth, in which the specimen at 600 ℃ has a necking theand creep plastic atplastic the time, which is at the result of which transgranular of creep 800 ℃deformation are not smooth, in same which the specimen 600 ℃ has athe necking phenomenon andtransgranular the creep fracture has obvious deformation at the same time, is the resultpropagation of the fracture hasofobvious plastic deformation at theissame time, is theatresult of but the transgranular crack. Meanwhile, there is no necking phenomenon in thewhich specimen 800 C, there isatobvious propagation creep crack. Meanwhile, there no necking phenomenon in◦the specimen 800 ℃, propagation of creep crack. Meanwhile, there is no necking phenomenon in the specimen at 800 ℃,the plastic deformation at the time of creep fracture, which is caused by the propagation of creep cracks but there is obvious plastic deformation at the time of creep fracture, which is caused by but there is obvious plastic deformation at the time of creep fracture, which is caused by the spreading along the grain boundary. propagation of creep cracks spreading along the grain boundary. propagation of creep cracks spreading along the grain boundary. 4.2. Experimental Experimental Results Results and and Model Model Verification Verification 4.2. 4.2. Experimental Results and Model Verification 4.2.1. Experimental Results 4.2.1. Experimental Results 4.2.1. Experimental Results Figure 14 shows the schematic diagram of single-channel deposition. A “binding zone” is formed Figure 14 shows the schematic diagram of single-channel deposition. A "binding zone" is Figure 14between shows the schematic layer diagram A "binding zone" is a at the interface the deposited andof thesingle-channel substrate, anddeposition. the formation of this part makes formed at the interface between the deposited layer and the substrate, and the formation of this part formed at the interface between the deposited layer and the substrate, and the formation of this part good metallurgical bond between the substrate and the deposited layer. The surface of the molten makes a good metallurgical bond between the substrate and the deposited layer. The surface of the makes a good metallurgical bond between the substrate and theadeposited layer.ofThe surface of At thethe pool is arc-shaped, and the substrate is partially melted to form certain depth penetration. molten pool isis arc-shaped, and the substrate isis partially partiallymelted meltedtotoform forma acertain certain depth of molten pool arc-shaped, and the substrate depth same time, due to the concentration of laser energy, the heat-affected zone after the action of theofheat penetration. At the same time, due to the concentration of laser energy, the heat-affected zone after penetration. the same time, due edge to theline concentration of laser energy, the heat-affected zone after source is very At small, and the white in the picture is the boundary of the heat-affected zone. the and the thewhite whiteedge edgeline lineininthe the picture is the boundary theaction actionofofthe theheat heatsource source is is very very small, small, and picture is the boundary of of The width of the heat-affected zone basically tends to a 0.5 mm range. the heat-affected heat-affectedzone zonebasically basicallytends tends a 0.5 mm range. the heat-affectedzone. zone.The Thewidth width of of the the heat-affected toto a 0.5 mm range.

Figure of: (a) (a)single-channel single-channeldeposition, deposition, cross-section of the Figure14. 14. The The schematic schematic diagram diagram of: (b)(b) thethe cross-section of the Figure 14. The schematic diagram of: (a) single-channel deposition, (b) the cross-section of the single-channeldeposition, deposition,(c) (c) the the cross-section cross-section morphology deposition. single-channel morphologyofofthe thesingle-channel single-channel deposition. single-channel deposition, (c) the cross-section morphology of the single-channel deposition.

4.2.2. The Heat Source Check 4.2.2. The Heat Source Check

Materials 2020, 13, 4316

15 of 25

Materials 2020, 13, x Source FOR PEER REVIEW 4.2.2. The Heat Check

16 of 26

In order order to to obtain obtain more more reasonable reasonable heat heat source source parameters parameters and and ensure ensure the the accuracy accuracy of of the the finite finite In element numerical simulation of the temperature field, the finite element model of the temperature element numerical simulation of the temperature field, the finite element model of the temperature field of of the the curved curved single single layer layer and and single single track track is is established. established. As Asshown shownin inFigure Figure 15, 15, the the substrate substrate field selects a round rod with the same size as the central shaft of the camshaft, with diameter of 80 mm mm selects a round rod with the same size as the central shaft of the camshaft, with diameter of 80 and length of 120 mm. The size of the deposited layer is consistent with the size of the single-track and length of 120 mm. The size of the deposited layer is consistent with the size of the single-track section prepared prepared under thethe laser power of 2000 W, the speedspeed of 5 mm/s, section underthe theparameters parameterswith with laser power of 2000 W, scanning the scanning of 5 and the powder feeding rate of 11.6 g/min. A quarter of the circumference of the central shaft is chosen mm/s, and the powder feeding rate of 11.6 g/min. A quarter of the circumference of the central shaft aschosen the arc as length deposited A proper simplification has been made that is the of arcthe length of thelayer. deposited layer. A proper simplification hashere. beenConsider made here. the single-track deposition on the curved surface be regarded thick-plate deposition, and the Consider that the single-track deposition on thecan curved surfaceascan be regarded as thick-plate size and morphology of the molten pool obtained are almost the same as those of the plate deposition. deposition, and the size and morphology of the molten pool obtained are almost the same as those of Therefore, using the experiment plate the deposition instead the curved deposition here only the plate deposition. Therefore,ofusing experiment of ofplate deposition instead of can the not curved save the experimental cost, but also obtain accurate verification results. The DC3D8 thermal analysis deposition here can not only save the experimental cost, but also obtain accurate verification results. unitDC3D8 is selected as theanalysis elementunit type,isand the process of aand laser power of 2000 W, scanning The thermal selected as the parameters element type, the process parameters of a speed of 5 mm/s, and powder feeding rate of 11.6 g/min are used for the simulation. The simulated laser power of 2000 W, scanning speed of 5 mm/s, and powder feeding rate of 11.6 g/min are used temperature-field pool cloud image is compared molten pool morphology of the the for the simulation. molten The simulated temperature-field moltenwith poolthe cloud image is compared with previous actual process experiment to verify the accuracy of the simulation. The simulated cloud molten pool morphology of the previous actual process experiment to verify the accuracy of the image of the The melt simulated pool is compared theof cross-section morphology of the melt in the simulation. cloud with image the melt pool is compared withpool theobtained cross-section experiment above to verify the accuracy of the simulation. morphology of the melt pool obtained in the experiment above to verify the accuracy of the

simulation.

(a)

(b)

Figure Figure 15. 15. Single-track Single-trackgeometric geometric model: model: (a) (a)overall overallmodel, model,(b) (b)partial partialmodel. model.

Figure 16 16 shows shows the the comparison comparison of of the the melt melt pool poolobtained obtainedthrough throughsimulation simulationand andexperiment. experiment. Figure ◦ C. The dotted dotted line line in in the the figure figure represents represents the the melting melting point point line line of of alloy alloy steel, steel, whose whose value value isis 1470 1470 ℃. The When the the temperature temperature is is higher higher than than this this temperature, temperature, the the substrate substrate and and powder powder melt melt to to form form aa melt melt When pool. The Thewidth widthof ofthe themelt meltpool poolobtained obtainedby bythe theexperiment experimentisis3.46 3.46mm mmand andthe thedepth depthisis1.04 1.04mm. mm. pool. The width of the 3.32 mm andand thethe depth of the pool pool is 0.94ismm, The the simulated simulatedmelt meltpool poolis is 3.32 mm depth of melt the melt 0.94 with mm,errors with of 4% and 9.6%, The experiment and simulation fit well, and the data in shown Table 3. errors of 4% andrespectively. 9.6%, respectively. The experiment and simulation fit well, and are theshown data are The causes thecauses errorsof may as follows: (1)as thefollows: high-temperature thermophysicalthermophysical parameters are in Table 3. of The thebeerrors may be (1) the high-temperature difficult to measure directly due to directly the limitation conditions. the same time, thesame data time, obtained parameters are difficult to measure due toof the limitation At of conditions. At the the by the Jmatpro simulation and experimental test have errors. Error exists the data obtained bysoftware the Jmatpro software simulation and experimental test (2) have errors. (2) between Error exists between the heat transfer conditions in the numerical simulation and the actual experimental environment; (3) the effects of the material gasification and melt pool flow on the heat transfer are ignored in the simulation.

Materials 2020, 13, 4316

16 of 25

Materials 2020, 13, x FOR PEER REVIEW

17 of 26

heat transfer conditions in the numerical simulation and the actual experimental environment; (3) the Materialsof 2020, x FOR PEER REVIEW and melt pool flow on the heat transfer are ignored in the simulation. 17 of 26 effects the13, material gasification

Figure 16. Comparison of the melt pool obtained by the simulation and experiment. Table 3. Comparison of the size of melt the melt obtained by simulation the simulation experiment. Figure 16. Comparison of the the melt pool obtained by and experiment. Figure 16. Comparison of poolpool obtained by the the simulation and and experiment. Table 3. Comparison melt pool obtained the simulationValue and experiment. Name of the size of the Experimental Valueby Simulation Error Table 3. Comparison of the size of the melt pool obtained by the simulation and experiment. The width of melt pool/mm 3.46 3.32 4% Name Experimental Value Simulation Value Error

The depth of melt pool/mm Name The width of melt pool/mm The width of melt pool/mm The depth of melt pool/mm 4.3. ThermalThe Stress Evolution depth of melt pool/mm

1.04 Value Experimental 3.46 3.46 1.04 1.04

0.94 Value 9.6% Simulation Error 3.32 4% 3.32 4% 0.94 9.6% 0.94 9.6%

Figure 17 shows a schematic diagram of the thermal stress induced by the temperature 4.3. Thermal Stress Evolution 4.3. Thermal Stress Evolution gradient. In the process of LMD, a laser beam rapidly heats the powder on the upper surface of the Figure 17 shows a schematic diagram of the thermal stress induced by the the temperature substrate at room temperature. The high-energy will quickly melt togradient. form a Figure 17 shows a schematic diagram ofheat the source thermal stress induced bypowder the temperature In the process of LMD, a laser beam rapidly heats the powder on the upper surface of the substrate melt pool, and with the heat source moving, the melt pool will rapidly cool and solidify. Because gradient. In the process of LMD, a laser beam rapidly heats the powder on the upper surface of the the at room temperature. The surface high-energy heat source quickly melt theispowder to temperature form a melt laser beam upper rapidly and will the heat conduction slow, the substrate atheats roomthe temperature. The layer high-energy heat source will quickly melt the powder to form a pool, andrises with the heat source increase moving,inthe melt pool will rapidly cool solidify. Because the gradient temperature, thewill strength ofcool theand material decreases at the melt pool, andsharply. with theWith heatthe source moving, the melt pool rapidly and solidify. Because the laser beam heats the upper surface layer rapidly and the heat conduction is slow, the temperature same time, and the expansion of the top heating layer is limited by the elastic compression of the laser beam heats the upper surface layer rapidly and the heat conduction is slow, the temperature gradient rises sharply. With the increase in strengththe of the decreases at the lower material. When the yield of temperature, the material the is topmaterial layer will be plastically gradient rises sharply. With the strength increase in temperature, thereached, strength of the material decreases at the same time, and the expansion of the top heating layer is limited by the elastic compression of the compressed, inducing strain of andtheproducing stress. In elastic the absence of external same time, and the expansion top heatingcompressive layer is limited by the compression of the lower material. When the yield strengthcan of the material is reached, the top will be plastically mechanical constraints, reverse occur in the direction away fromlayer the laser beam. In the lower material. When the yieldbending strength of the material is reached, the top layer will be plastically compressed, inducing strain and producing compressive stress. In the absence of external mechanical subsequent process, the and upper layer of the plastic compression begins contractofand bend compressed,cooling inducing strain producing compressive stress. In the to absence external constraints, reverse bending can occur in the of direction away from the laser beam. In the subsequent at a certain angle to one side in the direction the laser beam, but due to the limitation of the mechanical constraints, reverse bending can occur in the direction away from the laser beam. Inbase the cooling process, the upper of layer the plastic compression begins to contract bend at a certain material at the bottom, thelayer upper thatoftends to bend is constrained byand thecontract substrate the subsequent cooling process, the upper layer the plastic compression begins to andatbend angle toto one side in tension, the direction ofthe the lower laser beam, butsubjected due to thetolimitation ofTherefore, the base material at the bottom layer pressure. tensile at a certainproduce angle to one sidewhile in the direction of theislaser beam, but due to the limitation of thestress base bottom, the upper layer that tends to bend is constrained by the substrate at the bottom to produce can be formed in the deposited layer, and compressive stress is mainly formed in the lower part of material at the bottom, the upper layer that tends to bend is constrained by the substrate at the tension, while the lower layer is subjected to pressure. Therefore, tensile stress can be formed in the the deposited layer. tension, while the lower layer is subjected to pressure. Therefore, tensile stress bottom to produce deposited layer, and compressive stress is mainly formed in the lower part of the deposited layer. can be formed in the deposited layer, and compressive stress is mainly formed in the lower part of the deposited layer.

Figure 17. Thermal stress induced by the temperature gradient. Figure 17. Thermal stress induced by the temperature gradient.

Figure 18 shows the temperature field distribution when scanning to the center of the second track ◦ C, and of theFigure first layer. The peak thisdistribution time is 2347 the front is denser the 18 shows the temperature field when scanning toisotherm the center of thethan second Figure 17. Thermalat stress induced by the temperature gradient. back end in first the scanning which is due to the inconsistent offront the melt pool caused by track of the layer. Thedirection, peak temperature at this time is 2347 ℃,energy and the isotherm is denser than the back18end in the direction, which is display due when to the inconsistent energy of the melt pool the movement of the laser. By setting thefield temperature limit to the melting temperature of alloy Figure shows thescanning temperature distribution scanning to the center of the second ◦ C,movement caused by of the laser. By melt setting thecan temperature display limit to the steel the width and depth of the pool shown in the field cloud track at of1470 thethe first layer. The peak temperature at this time is be 2347 ℃, and the temperature front isotherm ismelting denser temperature ofend alloy steel at 1470 ℃, thehas width depth of melt poolenergy canthe beoffitting shown the picture. the early the simulation beenand fitted with thethe experiment, and isingood. than theIn back instage, the scanning direction, which is due to the inconsistent the melt pool temperature field cloud picture. In the early stage, the simulation has been fitted with the caused by the movement of the laser. By setting the temperature display limit to the melting experiment, the fitting temperatureand of alloy steel is atgood. 1470 ℃, the width and depth of the melt pool can be shown in the temperature field cloud picture. In the early stage, the simulation has been fitted with the experiment, and the fitting is good.

Materials 2020, 13, x FOR PEER REVIEW Materials 2020, 13, 4316 Materials 2020, 13, x FOR PEER REVIEW

18 of 26 17 of 25 18 of 26

Figure 18. Temperature field distribution when scanning to the center of the second layer of the first Figure 18. Temperature field distribution when scanning to the center of the second layer of the layer. Figure 18. Temperature field distribution when scanning to the center of the second layer of the first first layer. layer.

Figure 19 shows the cloud map of the thermal stress distribution during deposition. At the Figure 19 shows the cloud mapisofinthe during stress deposition. At the position of laser action, the material thethermal meltingstress state, distribution and the equivalent is veryAt small Figure 19 shows the cloud map of the thermal stress distribution during deposition. the position of laser action, the material is in the melting state, and the equivalent stress is very small and and approaches 0 MPa. When the laser moves forward far away from the region, the temperature position of laser action, the material is in the melting state, and the equivalent stress is very small approaches 0 MPa. When the laser moves forward far awaysolidifies from theand region, the temperature there there gradually and melt pool forward gradually Meanwhile, the and approaches 0decreases, MPa. When thethe laser moves far away from theshrinks. region, the temperature gradually decreases, and the melt pool gradually solidifies and shrinks. Meanwhile, the shrinkage will shrinkage will bedecreases, preventedand by the the melt solidified nearby, which will a certain constraint there gradually pool metal gradually solidifies and exert shrinks. Meanwhile, the be prevented by the solidified metal nearby, which will exert a certain constraint effectdepends and gradually effect and gradually form thermal stress. The magnitude of the thermal stress mainly on the shrinkage will be prevented by the solidified metal nearby, which will exert a certain constraint form thermalgradient. stress. The magnitude of the thermal stressismainly depends on the stress temperature gradient. temperature When the temperature gradient toothe large, the thermal will exceed effect and gradually form thermal stress. The magnitude of thermal stress mainly depends on the the When the temperature gradientand is tooproduce large, theplastic thermaldeformation. stress will exceed thethe yield limit of decrease the material yield limit of the material With gradual in temperature gradient. When the temperature gradient is too large, the thermal stress will exceed the and produce plastic deformation. With the gradual decrease in temperature, the thermal stress will temperature, the thermal stress will gradually become stable, and the residual stress after cooling is yield limit of the material and produce plastic deformation. With the gradual decrease in gradually become stable, and the residual stress after cooling is finally formed. finally formed. temperature, the thermal stress will gradually become stable, and the residual stress after cooling is finally formed.

(a)

(b)

(a) distribution during deposition: (a) (b) point of the first layer Figure 19. Thermal stress (a) scan scan to to the center of track19. track 2. 2. (a) scan to the center point of the first layer 2; (b) scan tostress the end of the first layer of track Figure Thermal distribution during deposition: of track 2; (b) scan to the end of the first layer of track 2.

curve of of thethe deposition process of Node1. When the Figure 20 20 shows showsthe thethermal thermalstress stressevolution evolution curve deposition process of Node1. When heat source moves to the selection point, the thermal stress approaches zero and then gradually the heat source movesthe to the selection point, the thermal zero andof Node1. When Figure 20 shows thermal stress evolution curve stress of the approaches deposition process to about 110 MPa with the departure of the heat source. Due to the remelting effectgradually between increases to about 110 MPa with the departure of the heat source. Due to the effect the heat source moves to the selection point, the thermal stress approaches zero andremelting then adjacent tracks, when this position is melted by the adjacent heat source again, the thermal stress between whenwith this position is melted by the adjacent the thermal increasesadjacent to abouttracks, 110 MPa the departure of the heat source.heat Duesource to theagain, remelting effect approaches 0 again, then increases gradually with the decrease in temperature and finally stabilizes at stress approaches 0 again, then increases gradually with the decrease in temperature and finally between adjacent tracks, when this position is melted by the adjacent heat source again, the thermal about MPa after stabilizes at about 350 MPa after stress 350 approaches 0 cooling. again, then cooling. increases gradually with the decrease in temperature and finally

stabilizes at about 350 MPa after cooling. 4.4. Residual Stress Distribution In the initial stage of LMD, the substrate is at a low temperature, and the upper powder is melted by the high-energy heat source rapidly, resulting in a very large temperature gradient in the bonding area. With the progress of laser deposition, the temperature of the substrate rises gradually and the heat accumulates, leading to a gradual decrease in the temperature gradient until the temperature gradient at the top is the smallest. The greater the temperature gradient is, the greater the residual stress is. Therefore, the residual stress is greatest in the region where the deposited layer and the substrate are combined.

Materials 2020, 13, 4316 Materials 2020, 13, x FOR PEER REVIEW

18 of 25 19 of 26

Figure 20. Thermal stress evolution during the deposition of Node1.

4.4. Residual Stress Distribution In the initial stage of LMD, the substrate is at a low temperature, and the upper powder is melted by the high-energy heat source rapidly, resulting in a very large temperature gradient in the bonding area. With the progress of laser deposition, the temperature of the substrate rises gradually and the heat accumulates, leading to a gradual decrease in the temperature gradient until the temperature gradient at the top is the smallest. The greater the temperature gradient is, the greater the residual stress is. Therefore, the residual stress isduring greatest the region where Figure 20. Thermal Thermal stress evolution evolution during thein deposition of Node1. Node1. the deposited layer Figure 20. stress the deposition of and the substrate are combined. Figure 21 mapmap of the stress distribution after cooling. maximum equivalent Figure 21isisa cloud a cloud ofresidual the residual stress distribution after The cooling. The maximum 4.4. Residual Stress Distribution stress is 686stress MPa, which at theappears junctionat of the substrate andofdeposition, a large temperature equivalent is 686 appears MPa, which junction substrate where and deposition, where a In the initialformed. stage of LMD, the substrate is at a lowexpansion temperature, and the upper expansions powder is gradient is easily Due to the difference in the thermal coefficient, different large temperature gradient is easily formed. Due to the difference in the thermal expansion melted by the high-energy heatcooling source rapidly, resultingresulting in a veryinlarge temperature gradient in the and shrinkages of metal after will be formed, greater and deformation. coefficient, different expansions and shrinkages of metal after cooling will stress be formed, resulting in bonding area. With the progress of laser deposition, the temperature of the substrate rises gradually The circumferential residual stress is symmetrically relative to the scanning direction greater stress and deformation. The circumferentialdistributed residual stress is symmetrically distributed and the heat and accumulates, leading to aisgradual decrease in the time, temperature gradient until the (ROT plane), the maximum stress 719 MPa. At the same the axial residual stress is relative to the scanning direction (ROT plane), and the maximum stress is 719 MPa. At the same temperature gradient at the top is the smallest. The greater the temperature gradient is, the greater symmetrically relativeistosymmetrically the ROZ plane,distributed and the maximum is 751 MPa. Furthermore, time, the axialdistributed residual stress relativestress to the ROZ plane, and the the residual stress stress is. Therefore, thelarger residual stress is greatest in theresidual region where the deposited the axial residual is slightly than the circumferential stress. The reason islayer that maximum stress is 751 MPa. Furthermore, axial residual stress is slightly larger than the and the substrate are combined. axial deformation is restricted axial constraints, whichdeformation makes it difficult for materials to constraints, expand and circumferential residual stress.byThe reason is that axial is restricted by axial Figure 21 isgreater a cloud map stress of the residual stress distribution after cooling. The maximum shrink, and thus which makes it difficultresidual for materialsistogenerated. expand and shrink, and thus greater residual stress is equivalent stress is 686 MPa, which appears at the junction of substrate and deposition, where a generated. large temperature gradient is easily formed. Due to the difference in the thermal expansion coefficient, different expansions and shrinkages of metal after cooling will be formed, resulting in greater stress and deformation. The circumferential residual stress is symmetrically distributed relative to the scanning direction (ROT plane), and the maximum stress is 719 MPa. At the same time, the axial residual stress is symmetrically distributed relative to the ROZ plane, and the maximum stress is 751 MPa. Furthermore, the axial residual stress is slightly larger than the circumferential residual stress. The reason is that axial deformation is restricted by axial constraints, which makes it difficult for materials to expand and shrink, and thus greater residual stress is (b) generated. (a)

(a)

(c)

(b)

Figure 21. Residual stress distribution after cooling: (a) mises stress, (b) circumferential stress, (c) axial stress.

Figure 22 shows the nephogram of residual stress in the direction of the ROZ section. The greater residual stress is mainly concentrated near the heat-affected zone, where the maximum circumferential stress and axial stress are tensile stress, while compressive stress is formed not far below. The overall residual stress of the deposited layer is relatively small. (c)

Figure 21. 21. Residual Residual stress stress distribution distribution after after cooling: cooling: (a) (a) mises mises stress, stress, (b) (b) circumferential circumferential stress, stress, (c) (c) Figure axial stress. axial stress.

Materials 2020, 13, 4316

19 of 25

(b) (b)

(a) (a)

(c) (c) stress distribution distribution in in the the section section direction: direction: (a) (a) mises mises stress, stress, (b) (b) circumferential circumferential Figure 22. 22. Residual Residual stress Figure Residual stress, (c) axial stress. stress, (c) axial stress.

Figure23 23shows showsthe theresidual residualstress stress distribution curve of the the extracted path and path 2. Figure Figure Figure distribution curve of of the extracted path 1 and pathpath 2. Figure 23a 23 shows the residual stress distribution curve extracted path 11 and 2. 23a shows shows that along along path 1, the thestress misesand stress and circumferential circumferential stress on the the layer deposited layer shows that along path 1,path the mises circumferential stress on stress the deposited increase at 23a that 1, mises stress and on deposited layer increase at first and then decrease. The maximum residual stress appears near the last path, and the first and then decrease. The maximum residual stress appears near the last path, and the residual stress increase at first and then decrease. The maximum residual stress appears near the last path, and the residual stress on both both sides sides isThe relatively small.isThe The main reason is isthe that this position is the the last last on both sides is on relatively small.is main reason thatmain this position last position of deposition, residual stress relatively small. reason is that this position is position of deposition, deposition, where aa and largetemperature heat accumulation accumulation and temperature gradient are formed. formed. where a large heat accumulation gradientand are formed. However, the are stress in the position of where large heat temperature gradient However, the stress in the former position of deposition will be partially released due to the former position of deposition be partially due to the and post-thermal of However, the stress in the will former positionreleased of deposition willremelting be partially released dueaction to the remelting and and post-thermal post-thermal action of post-deposition. post-deposition. Furthermore, the circumferential stress shows shows post-deposition. Furthermore, the circumferential stressFurthermore, shows tensilethe stress, and the maximum value remelting action of circumferential stress tensile stress, and while the maximum maximum value is is about about 305 305 MPa, while whilecompressive the axial axial stress stress gradually changes is aboutstress, 305 MPa, the axial value stress gradually changes from stress to tensile stress, tensile and the MPa, the gradually changes from compressive stress to tensile stress, and the maximum stress is about 90 MPa. and maximum stress to is about MPa.and the maximum stress is about 90 MPa. fromthe compressive tensile90 stress,

(a) (a)

(b) (b)

stress distribution distribution on on aa path path perpendicular perpendicular to to the the scanning scanning direction: direction: (a) Figure 23. 23. Residual Residual stress Figure (a) path 1, 1, (b) path 2. (b) (b) path path 2. 2.

Figure 23b shows the residual stress distribution curve of path 2 on the upper surface of the substrate. The change trend of the curve is similar to that of the residual stress curve in welding—that is, the circumferential residual stress in the deposited region shows a large tensile stress, with a maximum value of about 455 MPa. Extending from the deposited layer to both ends, the residual stress decreases gradually and finally tends to compressive stress, of which the maximum value is about 160 MPa.

Figure 23b shows the residual stress distribution curve of path 2 on the upper surface of the substrate. The change trend of the curve is similar to that of the residual stress curve in welding—that is, the circumferential residual stress in the deposited region shows a large tensile stress, with a maximum value of about 455 MPa. Extending from the deposited layer to both ends, the residual stress decreases gradually and finally tends to compressive stress, of which the Materials 2020, 13, 4316 20 of 25 maximum value is about 160 MPa. At the same time, the axial residual stress is mainly tensile stress, which is small in the deposition area and reaches a maximum value at about 630 MPa at the junction between the deposited layerresidual and thestress substrate. Withtensile the distance away from thein deposited layer, area the At the same time, the axial is mainly stress, which is small the deposition stress decreases gradually.value at about 630 MPa at the junction between the deposited layer and the and reaches a maximum substrate. With the distance away from the deposited layer, the stress decreases gradually. 4.5. The Influence of Scanning Mode on Stress 4.5. The Influence of Scanning Mode on Stress The camshaft formed by LMD is divided into two steps: the first step is to form the central shaft, Thesecond camshaft by LMD divided twoplatform steps: theon first is to shaft. form the shaft, and the stepformed is to deposit theiscam and into convex thestep central Forcentral the curved and thedeposition second step is tocam deposit the camplatform, and convex the central shaft. For the curved surface of the and convex twoplatform forming on methods can be considered: an arc surface deposition of the cam and convex platform, two forming methods can be considered: arc copying lap and an axial parallel lap. As shown in Figure 24, the main difference between thean two copying is lap and ansingle-track axial parallel lap. Asisshown Figure difference between two methods that the direction not theinsame, so24, thatthe themain length of the single trackthe is not methods is that single-track direction is not the same, so that the lengthofofthe themethod single track not the same. The arcthe length of this layer is selected as the single-track length of theisarc the same. The arc length of this layer is selected as the single-track length of the method of the copying lap, and the lap direction is parallel to the motion friction direction of the cam, which arc is copying lap, and the the lap formability direction is and parallel to the friction direction of the cam, which is beneficial to improve service lifemotion of the cam. In the axial parallel lap-forming beneficial improve the formability and thickness service lifeand of the In the axial parallel mode, mode, the to single-track length is the cam thecam. arc length is short. Thelap-forming cam thickness is the single-track length is the cam thickness and the arc length is short. The cam thickness is chosen as chosen as the single-track length of the axial parallel lap forming, and the arc length is shorter. the single-track length of the axial parallel lap forming, and the arc length is shorter.

Figure 24. Top view of the forming method of cam: (a) arc copying lap, (b) axial parallel lap. Figure 24. Top view of the forming method of cam: (a) arc copying lap, (b) axial parallel lap.

Figure 25 shows the residual stress distribution of path 1 on the surface of the deposited layer Figure 25 shows the residual stress distribution of path 1 on the surface of the deposited layer with different scanning modes. The mises stress, circumferential stress, and axial stress increase at with different scanning modes. The mises stress, circumferential stress, and axial stress increase at first and then decrease along the positive direction of the Z axis. A large residual stress is formed first and then decrease along the positive direction of the Z axis. A large residual stress is formed with the axial parallel lap scan on the deposited layer because, compared with the arc copying lap with the axial parallel lap scan on the deposited layer because, compared with the arc copying lap scan, the axial parallel scan produces more deposition trajectories and the turning and pause times scan, the axial parallel scan produces more deposition trajectories and the turning and pause times of the laser heat source, so that the heat source is unstable, which results in a larger cooling rate and of the laser heat source, so that the heat source is unstable, which results in a larger cooling rate and temperature gradient, and finally a greater stress. temperature gradient, and finally a greater stress. Figure 26 shows the thermal stress change curve of the first track central point of the first layer with different scanning modes. During the movement of the laser heat source, the thermal stress of this point is constantly changing and is reduced with each approach of the heat source, which is consistent with the mechanical properties at high temperatures. Meanwhile, as the heat source leaves, the thermal stress increases again. Compared with the axial parallel scan, the maximum thermal stress in the process and the residual stress of arc copying lap scan are relatively small.

Materials 2020, 13, 4316

21 of 25

Materials 2020, 13, x FOR PEER REVIEW

22 of 26

(a)

(b)

(a)

(b)

(c) Figure 25. Stress distribution of the different scanning methods on path 1: (a) mises stress, (b) circumferential stress, (c) axial stress.

Figure 26 shows the thermal stress change curve of the first track central point of the first layer with different scanning modes. During the movement of the laser heat source, the thermal stress of this point is constantly changing and is reduced with each approach of the heat source, which is (c) temperatures. Meanwhile, as the heat source consistent with the mechanical properties at high leaves, the 25. thermal stress increases again. Compared with the axial parallel the maximum Stress distribution ofofthe scanning methods onon path 1: (a) mises stress, (b) Figure Stress distribution thedifferent different scanning methods path 1: scan, (a) mises stress, thermal stress in the process and thestress. residual stress of arc copying lap scan are relatively small. circumferential stress, (c) (c) axial stress. (b) circumferential stress, axial Figure 26 shows the thermal stress change curve of the first track central point of the first layer with different scanning modes. During the movement of the laser heat source, the thermal stress of this point is constantly changing and is reduced with each approach of the heat source, which is consistent with the mechanical properties at high temperatures. Meanwhile, as the heat source leaves, the thermal stress increases again. Compared with the axial parallel scan, the maximum thermal stress in the process and the residual stress of arc copying lap scan are relatively small.

Figure 26. Thermal stress stress curves curves of of the the first first track track central central point point of the first layer with different 26. Thermal scanning modes.

4.6. Effect of Interlayer Cooling on Residual Stress 4.6. Effect of Interlayer Cooling on Residual Stress Figure 27 shows the residual stress distribution of path 1 on the surface of the deposited layer Figure 27 shows the residual stress distribution of path 1 on the surface of the deposited layer with the action of the interlaminar cooling time. The change trend of residual stress is about the with the action of the interlaminar cooling time. The change trend of residual stress is about the same. The mises stress and circumferential stress increase at first and then decrease along the Z axis, Figure while the stress decreases and track then increases. With the first increase the different interlaminar 26. axial Thermal stress curves at of first the first central point of the layerinwith cooling time, modes. the maximum stress on the deposited layer gradually increases, while the maximum scanning mises stress increases from 335 to 531 MPa, and the maximum circumferential stress increases from 4.6. to Effect Interlayer Cooling on Residual Stress 301 503ofMPa. The residual stress on the upper surface of the deposited layer increases to a certain extent with the increase in the interlayer cooling time, which is mainly because, with the increase Figure 27 shows the residual stress distribution of path 1 on the surface of the deposited layer with the action of the interlaminar cooling time. The change trend of residual stress is about the

same. The mises stress and circumferential stress increase at first and then decrease along the Z axis, while the axial stress decreases at first and then increases. With the increase in the interlaminar cooling time, the maximum stress on the deposited layer gradually increases, while the maximum mises stress increases from 335 to 531 MPa, and the maximum circumferential stress increases22from Materials 2020, 13, 4316 of 25 301 to 503 MPa. The residual stress on the upper surface of the deposited layer increases to a certain extent with the increase in the interlayer cooling time, which is mainly because, with the increase in in interlayer cooling time, temperature ofinitial the initial decreases and the re-deposition thethe interlayer cooling time, the the temperature of the layerlayer decreases and the re-deposition forms forms a larger temperature gradient so that the stress increases. However, in the actual forming a larger temperature gradient so that the stress increases. However, in the actual forming process, process, a certain interlaminar cooling is necessary. Continuousdeposition depositionproduces produces heat ensuringensuring a certain interlaminar cooling is necessary. Continuous heat accumulation, which increases the size of the melt pool and is not conducive to forming. accumulation, which increases the size of the melt pool and is not conducive to forming.

(b)

(a)

(c) Figure 27. Stress times of of path 1: (a) mises stress, (b) Figure Stressdistribution distributionofofthe thedifferent differentinterlayer interlayercooling cooling times path 1: (a) mises stress, circumferential stress, (c) (c) axial stress. (b) circumferential stress, axial stress.

5. 5. Conclusions Conclusions The The main main results results are are as as follows: follows: (1) The strength of the materialdecreases decreasesand andthe theplastic plasticproperty propertyincreases increases with increase (1) The strength of the material with thethe increase in in temperature. In the forming process of the LMD curved surface, a large instantaneous thermal temperature. In the forming process of the LMD curved surface, a large instantaneous thermal stress stress will be generated where the temperature gradient is large, and whenthe thethermal thermalstress stress is is higher will be generated where the temperature gradient is large, and when higher than than the the strength strength of of material material at at this this temperature temperature the the thermal thermal cracks cracks are are easily easily generated. generated. The The necking necking ◦ phenomenon 600 the fracture surface of of thethe sample at phenomenonappeared appearedininthe thesamples samplesatat200, 200,400, 400,and and 600C, ℃,and and the fracture surface sample ◦ 800 C was sawtooth. With the increase in temperature, the surface oxidation of the sample is gradually at 800 ℃ was sawtooth. With the increase in temperature, the surface oxidation of the sample is intense, and the post-fracture length of the sample gradually. fractures occur at 200 gradually intense, and the post-fracture length of increases the sample increasesDuctile gradually. Ductile fractures ◦ C, and creep fractures occur at 600 and 800 ◦ C. and 400 occur at 200 and 400 ℃, and creep fractures occur at 600 and 800 ℃. (2) and guarantee forfor thethe simulation. The sizesize of (2) The The test testmaterial materialparameters parametersprovide providethe thepremise premise and guarantee simulation. The the meltmelt poolpool obtained by the numerical simulation of the curved single layer and single channel of the obtained byLMD the LMD numerical simulation of the curved single layer and single is compared with the melt pool size of the experimental test. The errors of the width and depth of melt pool are 4% and 9.6%, respectively, and the simulation fits well with the experiment. (3) The thermal stress of the LMD curved surface forming is mainly divided into axial stress, radial stress, and circumferential stress, which are different from the stress in the plane state in distribution and

Materials 2020, 13, 4316

23 of 25

magnitude. During the LMD process, the thermal stress varies with time. The maximum equivalent stress after cooling is 686 MPa, which appears at the junction of the substrate and deposition layer. The circumferential residual stress is symmetrically distributed relative to the scanning direction (ROT plane) and the maximum stress is 719 MPa, while the axial residual stress is symmetrical relative to the ROZ plane and the maximum stress is 751 MPa. The larger residual stress is mainly concentrated in the lower part of the sedimentary layer, where the maximum circumferential stress and axial stress are tensile stress, while the compressive stress is formed not far below. The performances and crack defects of the components are easily effected by tensile stress, a harmful stress, which should be reduced or eliminated as much as possible. (4) Compared with the axial parallel lap scanning mode, the maximum thermal stress and the residual stress after cooling are relatively smaller in the arc copying lap scanning mode. Meanwhile, taking into account the advantages of being less time-consuming and having a lower path turning time, this is an ideal forming method. The increase in the interlayer cooling time can increase the residual stress on the upper surface of the deposited layer to some extent, which needs to be avoided. However, certain interlayer cooling is helpful to reduce heat accumulation and facilitate the forming of components. Therefore, a small interlayer cooling time is selected to obtain a good formability and produce a small residual stress, which is an optimal strategy. Author Contributions: Data curation, S.Y., X.Z., and X.H.; funding acquisition, article guidance and modification, S.D.; investigation, Z.C. All authors have read and agreed to the published version of the manuscript. Funding: This research was funded by the National Key R&D Program of China (No. 2016YFB1100205), The National Natural Science Foundation of China (No. 51705532), and the National Key R&D Program of China (No. 2017YFB1105002). Conflicts of Interest: The authors declare no conflict of interest.

References 1. 2.

3. 4.

5.

6. 7.

8.

9.

Qian, P.; Shiyun, D.; Xueliang, K.; Shixing, Y.; Ping, M. Effect of process parameters on formability of laser melting deposited 12CrNi2 alloy steel. Young Sci. Forum 2018, 10710, 1071045. Zhou, Y.; Chen, S.; Chen, X.; Cui, T.; Liang, J.; Liu, C. The evolution of bainite and mechanical properties of direct laser deposition 12CrNi2 alloy steel at different laser power. Mater. Sci. Eng. A 2019, 742, 150–161. [CrossRef] Hongwei, K.; Zhihong, D.; Wei, Z.; Yujiang, X.; Xiao, P. Laser melting deposition of a porosity-free alloy steel by application of high oxygen-containing powders mixed with Cr particles. Vacuum 2019, 159, 319–323. Dong, Z.; Kang, H.; Xie, Y.; Chi, C.; Peng, X. Effect of powder oxygen content on microstructure and mechanical properties of a laser additively-manufactured 12CrNi2 alloy steel. Mater. Lett. 2019, 236, 214–217. [CrossRef] Wang, K.B.; Liu, Y.X.; Zhao, X.; Lin, J.J.; Lv, Y.H. Microstructure and mechanical property of 12CrNi2 high strength steel fabricated by laser additive manufacturing. IOP Conf. Ser. Mater. Sci. Eng. 2018, 423, 012077. [CrossRef] Weng, Y.Q.; Yang, C.F.; Shang, C.J. State-of-the-art and development trends of HSLA steels in China. J. Iron Steel Res. Int. 2011, 46, 1–10. ´ Yumashev, A.; Slusarczyk, B.; Kondrashev, S.; Mikhaylov, A. Global indicators of sustainable development: Evaluation of the influence of the human development index on consumption and quality of energy. Energies 2020, 13, 2768. [CrossRef] Chen, S.Y.; Wang, R.X.; Ma, J.; Liang, J.; Cui, T.; Liu, C.S. Effect of scanning speed on microstructure and properties of 12CrNi2Re alloy steel prepared by laser additive manufacturing. In Proceedings of the 2017 International Conference on Materials Science and Biological Engineering (ICMSBE 2017), Yinchuan, China, 9–10 August 2017. Ding, C.; Cui, X.; Jiao, J.; Zhu, P. Effects of substrate preheating temperatures on the microstructure, properties, and residual stress of 12CrNi2 prepared by laser cladding deposition technique. Materials 2018, 11, 2401. [CrossRef]

Materials 2020, 13, 4316

10. 11.

12. 13. 14. 15. 16. 17. 18. 19. 20.

21. 22. 23. 24.

25.

26. 27.

28.

29. 30. 31. 32.

24 of 25

Qian, P.; Shiyun, D.; Xueliang, K.; Ping, M.; Shixing, Y. Effect of preheating on the microstructure and properties of laser melting deposited 12CrNi2 alloy steel. Chin. J. Eng. 2018, 40, 1342–1350. Liu, F.; Lin, X.; Leng, H.; Cao, J.; Liu, Q.; Huang, C.; Huang, W. Microstructural changes in a laser solid forming Inconel 718 superalloy thin wall in the deposition direction. Opt. Laser Technol. 2013, 45, 330–335. [CrossRef] Niu, F.; Wu, D.; Ma, G.; Wang, J.; Zhuang, J.; Jin, Z. Rapid fabrication of eutectic ceramic structures by laser engineered net shaping. Procedia CIRP 2016, 42, 91–95. [CrossRef] Tamanna, N.; Crouch, R.; Kabir, I.R.; Naher, S. An analytical model to predict and minimize the residual stress of laser cladding process. Appl. Phys. A 2018, 124, 1–5. [CrossRef] Gan, Z.; Yu, G.; He, X.; Li, S. Numerical simulation of thermal behavior and multicomponent mass transfer in direct laser deposition of Co-base alloy on steel. Int. J. Heat Mass Tranf. 2017, 104, 28–38. [CrossRef] Gao, W.; Zhao, S.; Wang, Y.; Zhang, Z.; Liu, F.; Lin, X. Numerical simulation of thermal field and Fe-based coating doped Ti. Int. J. Heat Mass Transf. 2016, 92, 83–90. [CrossRef] Qian, T.T.; Liu, D.; Tian, X.J.; Liu, C.M.; Wang, H.M. Microstructure of TA2/TA15 graded structural material by laser additive manufacturing process. T Nonferrous Metal. Soc. 2014, 24, 2729–2736. [CrossRef] Liang, H.; Wei, T.; Wenhe, L.; Chao, Z. Study of thermal-mechanical coupling behavior in laser cladding. Laser Optoelectron. Prog. 2014, 51, 091401. [CrossRef] Li, C.; Yu, Z.B.; Gao, J.X.; Han, X.; Dong, X. Numerical simulation and experiment of thermo-elastic-plastic-flow multi-field coupling in laser cladding process. China Surf. Eng. 2019, 32, 124–134. Dang, Y.X.; Qi, W.J.; Lu, L.L. Research status and development trend of numerical simulation of laser cladding technology. Hot Work. Technol. 2016, 45, 23–27. Farahmand, P.; Kovacevic, R. An experimental–numerical investigation of heat distribution and stress field in single-and multi-track laser cladding by a high-power direct diode laser. Opt. Laser Technol. 2014, 63, 154–168. [CrossRef] Nazemi, N.; Urbanic, J.; Alam, M. Hardness and residual stress modeling of powder injection laser cladding of P420 coating on AISI 1018 substrate. Int. J. Adv. Manuf. Technol. 2017, 93, 3485–3503. [CrossRef] Chew, Y.X.; Pang, J.H.L.; Bi, G.J.; Song, B. Thermo-mechanical model for simulating laser cladding induced residual stresses with single and multiple clad beads. J. Mater. Process. Technol. 2015, 224, 89–101. [CrossRef] Gong, C.; Wang, L.F.; Zhu, G.X.; Song, T.L. Numerical simulation of residual stress in 316L stainless steel cladding layer by laser additive manufacturing. Appl. Laser 2018, 38, 402–408. Alimardani, M.; Toyserkani, E.; Huissoon, J.P. A 3D dynamic numerical approach for temperature and thermal stress distributions in multilayer laser solid freeform fabrication process. Opt. Laser Eng. 2007, 45, 1115–1130. [CrossRef] Dai, D.P.; Jiang, X.H.; Cai, J.P.; Lu, F.G.; Chen, Y.; Li, Z.G.; Deng, D. Numerical simulation of temperature field and stress distribution in Inconel718 Ni base alloy induced by laser cladding. Chin. J. Lasers 2015, 42, 0903005. Xueliang, K.; Shiyun, D.; Hongbin, W.; Shixing, Y.; Xiaoting, L.; Binshi, X. Effect of laser power on gradient microstructure of low-alloy steel built by laser melting deposition. Mater. Lett. 2020, 262, 127073. Kiran, A.; Hodek, J.; Vavˇrík, J.; Urbánek, M.; Džugan, J. Numerical simulation development and computational optimization for directed energy deposition additive manufacturing process. Materials 2020, 13, 2666. [CrossRef] Gharbi, M.; Peyre, P.; Gorny, C.; Carin, M.; Morville, S.; Le Masson, P.; Carron, D.; Fabbro, R. Influence of various process conditions on surface finishes induced by the direct metal deposition laser technique on a Ti-6Al-4V alloy. J. Mater. Process. Technol. 2013, 213, 791–800. [CrossRef] Qu, H.P.; Wang, H.M. Microstructure and mechanical properties of laser melting deposited γ-TiAl intermetallic alloys. Mater. Sci. Eng. A 2007, 466, 187–194. [CrossRef] Cottam, R.; Thorogood, K.; Lui, Q.; Wong, Y.C.; Brandt, M. The effect of laser cladding deposition rate on residual stress formation in Ti-6Al-4V clad layers. Key Eng. Mater. 2012, 520, 309–313. [CrossRef] Miao, X.; Wu, M.; Han, J.; Li, H.; Ye, X. Effect of laser rescanning on the characteristics and residual stress of selective laser melted titanium Ti6Al4V alloy. Materials 2020, 13, 3940. [CrossRef] Li, J.; Wang, H.M. Microstructure and mechanical properties of rapid directionally solidified Ni-base superalloy Rene’41 by laser melting deposition manufacturing. Mater. Sci. Eng. A 2010, 527, 4823–4829. [CrossRef]

Materials 2020, 13, 4316

33. 34.

35. 36. 37. 38. 39. 40. 41.

42. 43.

44. 45.

25 of 25

Schnier, G.; Wood, J.; Galloway, A. Investigating the effects of process variables on the residual stresses of weld and laser cladding. Adv. Mater. Res. 2014, 996, 481–487. [CrossRef] Nazemi, N.; Urbanic, R.J. A numerical investigation for alternative toolpath deposition solutions for surface cladding of stainless steel P420 powder on AISI 1018 steel substrate. Int. J. Adv. Manuf. Technol. 2018, 96, 4123–4143. [CrossRef] Amine, T.; Newkirk, J.W.; Liou, F. Numerical simulation of the thermal history multiple laser deposited layers. Int. J. Adv. Manuf. Technol. 2014, 73, 1625–1631. [CrossRef] Yong, Y.; Fu, W.; Deng, Q.; Chen, D. A comparative study of vision detection and numerical simulation for laser cladding of nickel-based alloy. J. Manuf. Process. 2017, 28, 364–372. [CrossRef] Zhang, J.; Song, J.L.; Li, Y.T.; Chen, F. Numerical simulation of the temperature field in multi-layer powder-feeding laser cladding forming. Appl. Mech. Mater. 2012, 1867, 1418–1423. [CrossRef] Hua, L.; Tian, W.; Liao, W.; Zeng, C. Numerical simulation of temperature field and residual stress distribution for laser cladding remanufacturing. Adv. Mech. Eng. 2014, 6, 413–422. [CrossRef] Ding, L.; Li, M.X.; Huang, D.Y.; Jang, H.Y. Numerical simulation of temperature field and stress field to multiple laser cladding Co coatings. Appl. Mech. Mater. 2014, 2839, 382–387. [CrossRef] Kovacevic, R.; Zhang, Z.; Farahmand, P. Laser cladding of 420 stainless steel with molybdenum on mild steel A36 by a high power direct diode laser. Mater. Des. 2016, 109, 686–699. Shi, Q.M.; Gu, D.D.; Dai, D.H.; Du, L.; Ma, C.L.; Xia, M.J. Relation of thermal behavior and microstructure evolution during multi-track laser melting deposition of Ni-based material. Opt. Laser Technol. 2018, 108, 207–217. Zhanyong, Z.; Liang, L.; Le, T.; Peikang, B.; Jing, L.; Liyun, W. Simulation of Stress Field during the Selective Laser Melting Process of the Nickel-Based Superalloy, GH4169. Materials 2018, 11, 1525. Xuan, Z.; Yaohui, L.; Shiyun, D.; Shixing, Y.; Xiaoting, L.; Yuxin, L.; Peng, H.; Tiesong, L.; Binshi, X.; Hongsheng, H. The martensitic strengthening of 12CrNi2 low-alloy steel using a novel scanning strategy during direct laser deposition. Opt. Laser Technol. 2020, 132, 106487. GB/T22588-2008. Determination of Thermal Diffusivity or Thermal Conductivity by the Flash Method; Standardization Administration of China Standardization Administrat: Beijing, China, 2008. GB/T 228.2-2015. Metallic Materials-Tensile Testing-Part 2: Method of Test at Elevated Temperature; China Iron and Steel Association: Beijing, China, 2015. © 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).