Numerical analysis of additive manufacturing of maraging steel

The usage of Additive Manufacturing (AM) is very recently expanded with most of the major companies located in the vicinity of Europe and nearby. It is now possible to manufacture complex metallic geometries and the process is not only restricted to prototypes. Additive manufacturing is now making a revolution in most of the manufacturing companies for example automotive, aerospace industry, energy sector, medical instruments and tooling. It has advantage as any product can be built in minimum assemblies some time in 1 pass not requiring any assembly at all. In additive manufacturing a layer of fine powder is being melted and consolidated with the use of laser or electron beam. Then another powder layer is arranged and selectively melted on the previous layer. The process is repeated again and again in order to achieve the required height of built part. AM can be many types including EBM, LBM, DED and 3 D printing depending on the power source used and process of application of powder. The procedure of AM starts from a simple 3 D CAD file. The file is then converted to the another format compatible for additive manufacturing machine. This format is STL file. STL file converts the CAD model into small triangles of different sizes and stores their location as well as interior and exterior of parts. Based on the information stored in STL file, the model is divided into small slices of equal height. The part geometry is constant with one slice. The sliced data is then transfer to additive manufacturing system for part fabrication. In selective laser melting process power from laser selectively melts the metal and that forms a liquid melt pool. The melt pool suddenly solidifies as the laser passes that area. The successive melting of powder by consecutive tracks forms 1 layer. These layers are together consolidated one above other to form the solid geometry according to the data of CAD model (Figure 6.1). For strong bond of tracks and adjacent layer partial remelting of previous layer and overlapping of scan track is done. This also leads to higher densification and uniform properties with required degrees of isotropy. Some of the mostly used materials in this filed are AlSi10Mg, Ti6Al4V, stainless steel and maraging steel. Most common material among these is AlSi10Mg. The reason behind this popularity is its low weight combined with high strength and fracture toughness. Its main application is in automotive and aerospace engineering. A lot of research has been done related to AlSi10Mg, Ti6Al4V etc. While the area of maraging is still remain unexplored. This is the reason that maraging steel has been chosen for analysis in AM. The purpose of this thesis is to analyze the melt pool shape and size considering some common parameters of power, speed, spot size, penetration depth etc. In SLM cooling rate is fast so sudden solidification restricts formation of lath martensite in maraging steel. Large inclusions of titanium and aluminum combined oxides (TiO2:Al2O3) are present with size 10 to 20 µm. These are responsible for lowering the mechanical properties in aged conditions. As built maraging steel part is a ductile low carbon body centered cubic and has a martensitic structure. The relatively better strength and toughness of maraging steel is achieved by age hardening and aging. The aging process uniformly distribute the fine nickel rich intermetallic precipitates. These precipitates are the reason behind the strengthening of martensitic structure. And this precipitates also restricts the reversion of martensite to austenite and ferrite. During solidification elongated grains grows towards the center of the melt pool (Figure 6.6). Epitaxial growth occurs only at some places like center of melt pool. Far from the center small nucleated grains forms. Few grains can grow towards the center of melt pool. Equiaxed grains forms at the top of melt pool. Next scanning layer remelts again the previously solidified layer which results in the disappearance of equiaxed grains from the top of melt pool. This causes only grains remain are those on the top layer of last melt pool. There are various factors that affects the performance of additive manufacturing like the density increases with increasing scan speed and reaches a maximum value. The maximum point of density varies with different scan speeds. The density vs scan speed shows a parabolic pattern. For higher laser power, minimum lower surface roughness reaches at lower scan speed. If melt pool remains stable surface quality improves by increasing energy density. Generally, laser is applied with different scanning strategies. Surface roughness, porosity level, microstructure and heat buildup depends mainly on scanning strategy. Different scanning strategies are unidirectional scanning, bidirectional scanning, rotation of scanning direction between 2 layers, dividing the scanning area into small islands called island scanning. Oxygen reacts with molten metal at high temperature and forms metal oxide. The metal oxide is mainly found at the edges because of the continuous melting at top and bottom. Most of the metal oxides in additive manufacturing are found in aluminum as they are hard to break. In order to avoid the trapping of oxygen and oxide formation, it is recommended to use high incident heat which will melt the oxide film on the previous layer. Better scanning strategies and scanning parameter, use of noble gases like argon can significantly control the formation of oxide. By increasing the energy per unit length P/v, melt pool volume become bigger and viscosity reduces. While decreasing P/v, recoil pressure become significant and distort scan tracks. By further increasing the scan speed energy given is not enough to fully melt powder and partially melt the substrate. Which results in instability and formation of droplets. Relative density decreases with increasing scan speed. The reason of this decrease in density is the reduction of energy supplied per unit area by laser. Although this reduction is not significant at lower scan speed. Various analysis shows that by increasing the layer thickness from 30 µm to 40 µm results in 25 % reduction in scanning time while very little reduction in density. By much lowering scan speed, reduction in density is obtained due to balling effect. The main purpose of this thesis is the modeling of melt pool, estimation of temperature profile, maximum temperature and melt pool depth in maraging steel additive manufacturing components. Study and simulation is done using APDL part of ANSYS analysis software. Before starting the analysis material properties such as elastic modulus, young modulus, density, initial temperature, thermal conductivity, specific heat, coefficient of thermal expansion etc. are entered into the software either using user interface or ANSYS codes. On the basis of input data, the problem is being solved and further post processing is done in order to view the results, plot the necessary graphs. In the present study two element types PLANE55 and SOLID70 are chosen. PLANE55 (Figure 9.2) is designed for steady state or transient thermal analysis and having single degree of freedom temperature at each node. Similarly, the same conditions apply to SOLID70 (Figure 9.2) except that it is used for 3D analysis. In present analysis a small cuboidal shape model is chosen to do the analysis. Model is having the dimensions of x = -0.6e-3 mm, y = 1.25e-3 mm, z = -0.8e-3 mm. For performing an analysis in ANSYS it is necessary to divide the part into small meshes. The main purpose of meshing is to perform efficient, accurate and micro analysis. Whole load which is being applied on body is distributed in each mesh and analysis is done in every mesh. Convergence, accuracy and speed is dependent on meshing. In present analysis, multizone method is used. As it divides the geometry into small cubes. Finer mesh takes more time for analysis. To avoid time consumption, coarse mesh is used on part of body with little interest while finer mesh is done on the area of interest. The maximum number of nodes formed are 39401 and elements are 36818. In this case the power input is taken as 80, 100, 120 W combined with scan speeds of 400, 600, 800, 1000 mm/s. The thickness of 1 layer is 30 µm. When laser is incident upon the powder layer it penetrates till certain depth into the powder as well as previously formed layer. The effect of this penetration leads to the re melting of previously formed layer of powder which results in melt pool depth of approximately 3 layers thick generally. So in this study penetration depth is assumed as 100 µm. The diameter of incident laser or spot diameter is taken as 100 µm. The properties of maraging steel are fixed and don't change with temperature. While during analysis temperature dependent properties are taken into account. The melt pool shape, size and temperature distribution is shown in the following figure 9.10 to figure 9.15. Analysis is done using 3 layers of maraging steel as including more layers will increase the simulation time. . The temperature and melt pool behavior is simulated for 3 layers and combination of power and speed. Table 9.12 shows the achieved temperature and melt pool depth result of simulation based on these combinations of power and speed. The results of temperature as well as melt pool depth versus layers is being plotted. Figure 9.10 to figure 9.12 shows the behavior of temperature. While figure 9.13 to figure 9.15 shows the behavior of melt pool depth. It has been found that temperature and melt pool depth increases with addition of further layers. Also both of these found to be decreasing by increasing the speed. This may be attributed to decrease in value of energy per unit volume available by increasing the speed while keeping other parameters constant.