DETERMINATION OF PROCESS PARAMETERS FOR ELECTRON BEAM SINTERING (EBS)

Presented at the COMSOL Conference 2008 Hannover Institut für Werkzeugmaschinen und Betriebswissenschaften Prof. Dr.-Ing. M. Zäh Prof. Dr.-Ing. G. Reinhart DETERMINATION OF PROCESS PARAMETERS FOR ELECTRON BEAM SINTERING (EBS) Hannover, November 6th 2008 European Comsol Conference Dipl.-Ing. Markus Kahnert

Outline 1. Introduction Electron Beam Sintering (EBS) 2. Objective of the present work 3. Mathematical model - Overview - Heat Source Model - Material Properties 4. Numerical Model 5. Simulation Results and Discussion 6. Conclusion 12.11.2008 Seite 2

Electron beam sintering Electron Beam Sintering electron beam generator electron beam sintered layer building volume with metal-powder part building plate building volume building plate new layer of metal powder preheated area divergence of the electron beam focused electron beam sintered layer re-coating preheating scanning with different strategies 12.11.2008 Seite 4

Electron Beam Sintering (EBS) Potentials of the electron beam High beam power and energy density Fast, massless beam deflection Quasi-simultaneous beam exposure Uniform insertion of energy Open configurable beam figures Mechanical pumps Beam generator Vacuum chamber 12.11.2008 Seite 5

Objective of the present work Process stability Metal parts production Two major deficiencies can be observed - Balling - Delamination Process cannot be examined due to accessibility and resolution of thermography unit Balling Understanding of the process needs to be enhanced Analysis of the beam-material interaction by means of FEA Delamination 12.11.2008 Seite 7

Mathematical Model Overview 80 60 heat conductivity [W/(mK)] temperature Temperatur [K] 40 20 Validation 0 273 773 1273 1773 Temperatur [K] Zeit time [s] Mathematical Model Powder properties FEM- Simulation Result analysis Knowledge to use in real process 12.11.2008 Seite 9

Mathematical Model Heat Source Model Horizontal intensity according to a Gaussian distribution Vertical intensity follows a 2nd degree polynom Power efficiency: 78% Superposition Penetration depth at given accelerating voltage and powder density: 62 µm Moving Heat Source at constant velocity time=0.002 isosurface: heat source [W/m³] 12.11.2008 Seite 10

Mathematical Model Temperature Dependence of Material Properties Thermomechanical properties of metal powders are not common knowledge due to trade secrets and multitude of powders 80 60 40 heat conductivity [W/(mK)] Experiments and analytical models available Heat capacity is similar to that of massive material Heat conductivity and emissivity modelled and validated 20 0 273 773 Temperatur 1273 [K] 1773 T [K] ε 333 0,78 emissity of metal powder 481 0,8 569 0,81 680 0,83 755 0,86 1208 0,94 1396 0,98 12.11.2008 Seite 11

Numerical Model Numerical Model melt line solid material hatch distance h beam spot diameter d b trace width powder layer thickness model length electron beam (I b, U a, v s ) Square to be meleted powder (t l, υ p ) y z x Scanning pattern powder material Part Model solidified material Thermal analysis Meshed model 12.11.2008 Seite 13

Simulation Results and Discussion Objective of the Simulation Calculation of the size and shape of the melt pool Simulation of the heat distribution within one melt line Starting Conditions and Process Parameters Model parameter Variable Value Process parameter Variable Value EB penetration depth S 6.2 10-05 m accelerating voltage U a 100 kv trace width 2.0 10-04 m beam current I b 1mA model length 2.0 10-03 m beam spot diameter d b 200 µm powder layer thickness 1.0 10-04 m layer thickness t l 100 µm Material parameters Material Density Value 316L (1.4404) Powder: 3.4 g/cm³ Massive: 7,9 g/cm³ preheating temperature scan speed hatch distance υ p v s h 1353 K 0.1 m/s 0.1 mm Grain Size 20-63 µm 12.11.2008 Seite 15

Simulation Results and Discussion Heat distribution within one melt line Simulation of the transient temperature distribution in the center of a melt line at various depths Complete melting of a layer at point z=-1*10-4 m (= layer thickness) After the powder has reached melting temperature, the layer thickness decreases proportional to its density increase temperature [K] @ various depths T S = 1699 K x=0 m; y=1*10-3 m; z=0 m x=0 m; y=1*10-3 m; z=-1*10-4 m x=0 m; y=1*10-3 m; z=-2*10-4 m 12.11.2008 Seite 16

Simulation Results and Discussion Calculation of the size and shape of the melt pool Calculation of the Size and shape of the melt pool according to real process parameters L/W-ratio is 3.8, no balling observed in comparison: 2.1 in laserbeambased processes (SLM, DMLS, etc.) Possible reasons: - preheating temperature - vacuum in the process area - different beam-material interaction Research efforts to be done time=0.002 s isosurface: temperature [K] L/W=3.8 v Beam =const. electron beam liquid mm system boundary solid phase 1 preheated powder phase 2 heat input phase 3 liquid phase 4 phase change phase 5 solid 12.11.2008 Seite 17

Conclusion Manufacturing of EBS parts only possible with considerable process knowledge Simulation carried out in order to evaluate necessary beam power and other process parameters FE-Modelling of the energy input based on a mathematical abstracted heat source and a discretized powder bed is an effective tool Differences between Laser- and Electron Beam-based processes discovered Model can be transferred to alternative additive layer manufacturing technologies 12.11.2008 Seite 19

Institut für Werkzeugmaschinen und Betriebswissenschaften Prof. Dr.-Ing. M. Zäh Prof. Dr.-Ing. G. Reinhart Kontakt: Dipl.-Ing. Markus Kahnert Telefon +49 (0) 821 / 5 68 83-33 Telefax +49 (0) 821 / 5 68 83-50 E-Mail: markus.kahnert@iwb.tum.de Adresse iwb Anwenderzentrum Augsburg Beim Glaspalast 5 86153 Augsburg www.iwb-augsburg.de

Requirements in production technology Mega-trends: Demand for individualization of products increases ( mass customisation ) Products become more complex Life cycles shorten customized, complex products mold insert for injection dies engine component innovative technologies for an economic manufacturing of complex parts (powder based manufacturing methods SLM, DMLS, EBS) wide range of processable materials (high quality steel, titanium alloys, aluminum alloys) source: ConceptLaser source: DaimlerChrysler efficient numerical methods for virtual design of manufacturing processes Finite- Element-Analysis (FEA) powder based manufacturing processes 12.11.2008 Seite 21