The effect of sample dimensions and process parameters (beam current, scan speed, offset focus and scan length) of electron beam melting (EBM) system on microstructure of the EBM built Ti-6Al-4V alloy was investigated. The microstructure of EBM built Ti-6Al-4V alloy consists of columnar grains of prior β phase. Inside the columnar grain, typical (α+β) structures, namely Widmanstätten α platelets and rod-like β phase formed on the interfaces of the fine α grains, are observed. Grain boundary α layer forms along grain boundary of prior β columnar grain. With the increase of thickness of the test slab, beam energy density and scanning length, the prior β columnar grain grows along the build direction and diameter of which increases. The columnar grain diameter also decreases with the increase in height of the test slab. With increasing the thickness of the test slab and beam energy density, α platelets get coarser. 

本文研究了试件尺寸和工艺参数（电子束强度，扫描速度，焦点偏移量和扫描长度）对EBM加工的Ti-6Al-4V微观结构的影响。通常，可以观察到EBM加工的Ti-6Al-4V的微观结构是由原始β相的柱状晶粒组成。在柱状晶粒内部观察到典型的（α+β）结构，即魏氏体α片和在细小的α晶粒的界面上形成的杆状β相。还发现沿原始β柱状晶粒的晶界形成α层晶界。随着试件厚度，电子束能量密度和扫描长度的增加，先前的β柱状晶粒的直径增大，并且生长的方向与加工方向一致。同时，柱状晶粒直径也随着高度的增加而减小。随着厚度和电子束能量密度的增加，α片会变得更粗大。

Purpose – Ti-6Al-4V is one of the most attractive materials being used in aerospace, automotive and medical implant industries. Electron beam melting (EBM) is one of the direct digital manufacturing methods to produce complex geometries of fully dense and near net shape parts. The EBM system provides an opportunity to built metallic objects with different processing parameter settings like beam current, scan speed, probe size on powder, etc. The purpose of this paper is to determine and understand the effect of part's thickness and variation in process parameter settings of the EBM system on surface roughness/topography of EBM fabricated Ti-6Al-4V metallic parts. Design/methodology/approach – A mathematical model based upon response surface methodology (RSM) is developed to study the variation of surface roughness with changing process parameter settings. Surface roughness of the test slabs produced with different parameter settings and thickness has been studied under confocal microscope. Response surface methodology was used to develop a multiple regression model to correlate the effect of variation in EBM process parameters settings and thickness of parts on surface roughness of EBM produced Ti-6Al-4V. Findings – It has been observed that every part produced by EBM system has detectable surface roughness. The surface roughness parameter Ra varies between 1-20 µm for different samples depending upon the process parameter setting and thickness. The Ra value increases with increasing sample thickness and beam current, and decreases with increase in offset focus and scan speed. Originality/value – Surface roughness is related to wear and friction property of the material and hence is related to the life time and performance of the part. Surface roughness is an important property of any material to be considered as biomaterial. The surface roughness of the material depends upon the manufacturing method and environment and hence it is controllable either during fabrication or by post processing. From the 1st order regression model developed in this study, it is also evident that sample thickness, scan speed and beam current have relatively more effect on roughness value then the offset focus. With the model obtained equation, a designer can subsequently select the best combination of sample thickness and process parameter values to achieve desired surface roughness.

In the current work an investigation of the microstructures of EBM built Ti-6Al-4V test bars has been performed using OM, SEM, TEM and XRD. It has been found that the prior β phase, that formed during the initial solidification, possesses a column shaped morphology with growing direction parallel to built direction. Typical (α+β) structures namely Widmanstätten α platelets with rod-like β phase formed on the interfaces of the fine α grains, have been observed in the columnar prior β grains. Grain boundary α phase was found to be formed around the boundaries of the columnar prior β grains. Different phases present in the parts, especially the BCC β phases have been characterized. The TEM/EDX results indicate very high V composition in the β phase. Results of TEM/SAED and XRD also revealed that a superlattice structure could be present in the β phase. Phase transformation sequence is discussed according to the processing history and the microstructures observed.

A systematic investigation of the microstructures of a series of magnesium based die cast magnesium‐aluminium‐rare earth (RE) alloys has been performed using X-ray diffraction analysis, scanning electron microscopy and transmission electron microscopy. The alloys had an aluminium content of 3‐4 wt-% and RE content of between 0·6 and 3·5 wt-%. The alloys were studied in the as cast condition and after aging at 200 and 250°C. Three kinds of binary aluminium‐RE phases were found in the alloys existing intergranularly. Al11RE3 was the predominant intergranular phase. In addition, Al3RE and Al2RE phases were found in the alloys with high RE content. The Al3RE phase became unstable when the alloy was aged at 200°C, while the Al11RE3 phase remained stable at 250°C. In alloys with a RE/aluminium weight ratio of <0·6, primary Al12Mg17 and aluminium enriched zones existed at the grain boundaries beside the Al11RE3 phase, both of which are considered to contribute to the reduction in creep resistance. The aluminium enriched zone transformed into Al12Mg17 precipitates when the alloys were aged. Therefore, to achieve improved creep resistance in Mg‐Al‐RE alloys, the RE/Al ratio should exceed 0·6.

Titanium and its alloys especially Ti-6Al-4V is an attractive biomaterial due to their excellent biocompatibility. Electron Beam Melting (EBM) is one of the Solid Free Form Fabrication (SFFF) methods to build 3D solid and near-net shape objects for medical implants and aerospace industry. EBM system utilizes a high energy electron beam to selectively melt a powder layer according to CAD file in a vacuum chamber. EBM system can manufacture complex 3D geometries requiring no or very little machining before use. The EBM systems are energy and material efficient. The microstructures and surface properties of objects produced by EBM can be influenced by the setting of different processing parameters in the EBM system. In this study solid slabs of Ti-6Al-4V, approximately 5 x 5 cm with various thicknesses were produced with different sets of processing parameters such as beam current, offset focus, scan speed and scan direction. The effects of these parameters on surface roughness, surface morphology and microstructure of slabs have been evaluated by using confocal microscopy, SEM /EDX and optical microscopy. The samples for optical microscopy and SEM were prepared by using standard metallographic methods. Microstructures of Ti-6Al-4V alloy produced by EBM usually consist of columnar grains as shown in Figure 1. These grains always grow parallel to build direction. Layers of different contrasts were observed in the samples where the layer interface is perpendicular to the build direction as shown in figure 2. The growth of columnar grains and appearance of layers with different contrasts were observed irrespective of the parameter values. These two observed phenomena can be attributed to the partial reheating / re-melting of the solidified layer by the electron beam during the melting of subsequent layer. The diameter of individual grain and density of grains are not uniform and usually decreases with increase in build height. Upon cooling from the β-transus temperature, more or less continuous α-layers were found to form along the prior β grain boundaries. In the EBM produced Ti-6Al-4V alloy the β -phase was found to be in rod-like geometry, with a size of 0.05-0.1μm in diameter, imbedded in the α-plates. It was observed that the high value of off set focus can cause porosity in the sample. For example the resultant porosity could be up to 11%. On the other hand where the value of offset focus is relatively small no such phenomenon was observed. Reconstruction of 3D surface topography and roughness coefficient (Ra) were computed by using images taken from confocal microscope and novel computer program “COMSTAT’’ [A.Heydorn et al (2000)]. Figure 3 shows a 3D reconstructed surface of the EBM produced sample. The Ra is computed by using the equation below: Where Lfi is the thickness of ith point, Lf is the mean thickness and N is the number of measurements. It has been observed that the value of Ra is processing parameters dependent. A sample with bigger thickness or higher current values tends to have relatively higher values of Ra. The scan speed and scan direction can also influence the surface morphology and microstructures of the EBM produced alloys.

Ti-6Al-4V alloy is an attractive biomaterial. The current work evaluates the microstructures of the solid and net-shape Ti-6Al-4V alloy produced by Electron Beam Melting (EBM) system using SEM/EDX and optical microscope. The microstructures are influenced by the cooling rate, processing parameters of the EBM system and re-heating of the existing layer during the melting of subsequent layers. Layer structure and columnar grains have been observed, with growing direction parallel to the built direction. The interior of these grains consists of alternating α / β phases. The β phase in the colonies resembles rod shape embedded in the α platelet. Along the grain boundaries more or less continuous α layers were observed. In comparison to solid samples uneven surfaces and pores were seen in the net shape structure. Microhardness evaluation of the EBM produced alloys was also carried out and compared with conventionally produced alloys.

A systematic investigation on microstructures of magnesium based die-cast Mg-Al-RE alloys (AE alloys) has been performed by XRD, SEM and TEM. The alloys are with content of Al around 4wt% and various content of rare earth (RE). Samples of these alloys are in as-cast condition as well as in aged condition at 200°C and 250°C. The intergranular microstructure of the alloys has been studied in details. Three type binary Al-RE phases were found intergranularly. Al11RE3 is predominant intergranular phase in the as cast alloys. Al3RE particles and small amount of Al2RE phase were found in alloys with high RE content. Thermal stability of the Al-RE phases in AE alloys was suggested to decrease in sequence: Al2RE→ Al11RE3 →Al3RE. The Al/RE ratio of the die cast alloys determined their phase constitutions. Promising AE alloys for creep resistance is suggested to have an Al/RE ratio not higher than 1.8.

M5-10 A TEM Study of Microstructural Change in Mg Alloys during Creep Wei L.-Y. Wei K. and Warren R. Division of Materials Science, School of Technology and Society Malmö University, 205 06 Malmö Sweden liu-ying.wei@ts.mah.se Due to their low densities and a good combination of castability and different properties, magnesium alloys have been attractive materials for automotive industry for weight saving. Die-cast AZ91 is one of the most widely used alloys. It exhibits excellent die castability, good strength properties and a fair ductility but its mechanical properties, especially the creep resistance, decrease on exposure at temperature above 120ºC [1]. Softening and coarsening of the interdendritic phase β-Al12Mg17 in AZ91 can readily occur at elevated temperatures due to its low melting temperature (Tm = 437°C) [2]. Rare earth in magnesium alloys usually brings about alloy strengthening through the formation of thermally stable intergranular phases and the complete suppression of the β-Al12Mg17 phase formation. In the present work tensile creep behavior of the die-cast magnesium alloys AZ91 and AE42 and their microstructural change during plastic deformation in creep have been studied. The objective of the work is to study the active creep mechanisms on the basis of the creep data in combination with the microstructural investigations. The SEM micrographs in Fig. 1 (a-b) illustrate the as-cast microstructures of the two die cast alloys showing a typical dendritic solidification structure with a fine-scale grain size. Segregation of Al adjacent to the dendrite boundaries can be seen clearly in AZ91D. Phases with a bright contrast and surrounded by high Al-segregated eutectic α-Mg are β-Al12Mg17 phase. Segregation of Al seems not obvious in AE42. The dominant interdendritic phase is Al11RE3. Mn-containing phases also formed during solidification in both AZ91 and AE42. The TEM micrographs in Fig. 2 (a-c) illustrate the structural changes during creep in AZ91 and AE42. Fig. 2 (a) shows the formation of denuded zones around the dendrite boundaries and the formation of discontinuous precipitates Al12Mg17. TEM/XEDS revealed that the Al content in the denuded zones decreased after creep test, from the original 5.5% to 2.5% (wt%). Fig. 2 (b-c) illustrates the formation of dislocation sources in the (0001) basal planes under a relatively high stress and the movement of dislocations from the sources towards the dendrite boundaries. The existence of the supersaturated Al-segregated zones adjacent to the dendrite boundaries is the main reason for the recovery of AZ91 at operating temperatures above 100ºC. Under loading plastic deformation occurred by the diffusion of Al atoms along the boundaries from compressive to tensile grain boundaries that caused the formation of the denuded grain boundaries and as a consequence the formation of the discontinuous precipitates Al12Mg17. Dislocation climb process also operated and tended to become a dominant process at stress above 60MPa.The low level of Al in AE42 suppresses the formation of Al supersaturated zones adjacent to the dendrite boundaries that, combining with the formation of the thermally stable interdendritic phase Al11RE3 (Tm~640ºC), provides AE42 a much better creep performance. Diffusion accompanying with the dislocation climb also occurred and became more dominant at a relatively high stress. Figure 1 SEM micrographs of the die cast alloys (a) AZ91D; (b) AE42. Fig. 2 TEM micrographs of the crept samples, (a-b) AZ91 after creep test at 100ºC/100MPa (n=3), (c) AE42 after creep at 175ºC/60MPa (n=2-7). References: 1. I.J. Polmer, 2nd Mg conference, DGM Informationsgesellschaft, Germany (1992), 201 2. M.S. Dargush, G.L. Dunlop, K. Pettersen, in: W. Higgins (Ed.), Trans. of 19th Intern. Die Casting Cong., North American Die Casting Association, Rosemont, IL, 1997, 131–137. Acknowledgements: Norsk Hydro ASA is gratefully acknowledged for providing the test bars

Tensile creep behaviour of the die-cast magnesium alloys AZ91 and AE42 has been studied at temperatures between 85oC and 200oC and at stresses in the range from 30 to 100 MPa. Microstructural investigations, mainly by TEM, have been performed on the selected crept samples to characterise the microstructural change during the plastic deformation, which reveals several phenomena related to the creep process including formation of dislocation sources, denuded zones around grain boundaries and microvoids, and changes in the nature of intermetallic phases. The active creep mechanisms have been discussed on the basis of the creep data in combination with the microstructural change during creep.

The as cast microstructures of several magnesium alloys in the AM series have been characterised, including the dendritic structure, Al segregation and the Mn containing phases. Two types of Mn containing particles were found and studied in the alloys mainly by using TEM. Type I is of equiaxed morphologies and a higher Mn/Al ratio in the composition. The flower shaped and dendritic type II phase has a lower Mn/Al ratio. The equiaxed type I could originate in the particle residues in the melts retained in the as cast alloy and the type II might form around a small type I particle. Electron diffraction has determined that both type I and type II have bcc crystal structure with lattice parameter a=0·91 nm. The higher Mn content in an alloy may lead to a higher amount of Mn containing particles.