2015 Volume 32 Pages 253-263
Titanium and its alloys have been widely used for various industrial and medical applications because of their excellent characteristics of low density, high corrosion resistance and high biocompatibility. However, it is not easy to produce the more complicated shape and precise components with low cost because of their poor workability. Therefore, Direct Laser Forming (DLF) which is one of Additive Manufacturing (AM) techniques is desired to be a suitable and advanced processing technique for fabricating Ti alloy components. In this paper, DLF technique has been introduced to fabricate Ti-6Al-7Nb alloy parts for medical applications and Ti-6Al-4V alloy parts for aerospace and automobile by evaluating various properties.
Additive manufacturing (AM) technologies have finally hit the mainstream. Since 25 years of development as “rapid prototyping” techniques. AM techniques are a collection of manufacturing processes which join materials to make physical 3D objects directly form virtual 3D computer data. These processes typically build up parts layer by layer, as opposed to subtractive manufacturing methodologies which create 3D geometry by removing material in a sequential manner. In the technical community, an international consensus has coalesced around the use of “additive manufacturing” whereas in the popular press the technologies are known as “3D printing.” Every existing commercial AM machine works in a similar way. First a 3D CAD file is sliced into a stack of 2-dimensional planar layers. These layers are built by the AM machine and stacked one after the other to build up the part.
Today, there are seven different approaches to AM, and dozens of variants of these approaches. As most of these approaches were first patented in the late 80’s and early 90’s, in many cases the fundamental process patents have expired or are expiring soon—thus opening up the marketplace for significant competition in a way that was impossible over the past 20 years, due to intellectual property exclusivity1). ASTM International released the standard terminology in 2012 that classified AM technologies into seven broad categories. Below are quick summaries of the different types of 3D printing according to website of U.S. Department of Energy2).
In these seven approaches, Powder Bed Fusion by using laser or electron beam and Directed Energy Deposition are most useful for fabricating the metal components. Especially, Direct Laser Forming (DLF), namely Selective Laser Melting (SLM) or Direct Metal Laser Forming (DMLF), is one of the new AM techniques that emerged in the late 1980s and 1990s3). During the DLF technique, a product is formed by selectively melting successive layers of powder by the interaction of a laser beam. Upon irradiation, the powder material is heated and, if sufficient power is applied, melts and forms a liquid pool. Afterwards, the molten pool solidifies and cools down quickly, and the consolidated material starts to form the product. This process is repeated until the product is completed. The non-irradiated material remains in the building cylinder and is used to support the subsequent layers. After the process, the unused powder is sieved and can be reused. For some high reactive material such as Ti alloys, the process needs to be conducted under an inert argon or vacuum atmosphere.
A schematic illustration of typical DLF technique is depicted in Fig. 1. Compared to conventional manufacturing techniques, DLF offers a wide range of advantages, namely a lower time-to-market, a near-net-shape production without the need of expensive moulds, a high material utilization rate, direct production based on a CAD model, and a high level of flexibility. Moreover, due to the additive and layer-wise production, the DLF technique is capable of producing more complex geometrical features that cannot be obtained using conventional production routes. Some previous researches have concentrated entirely on the influence of the process parameters on the product properties such as the surface roughness and relative density4–8), or have investigated the obtained mechanical properties9–11) and the feasibility of the DLF technique for applications in, for example, the biomedical and aero nautical industries12–14).
A schematic illustration of DLF process.
In this paper, DLF technique is employed to form the complex shaped compacts using two types of titanium alloys: Ti-6Al-4V and Ti-6Al-7Nb. Ti-6Al-4V is the most widely used titanium alloy. It features excellent mechanical properties, such as high strength, low density and outstanding corrosion resistance. Therefore, Ti-6Al-4V has led to a wide and diversified range of successful applications which demand high levels of reliable performance in surgery and medicine, aerospace, automotive, chemical plant and other major industries. Especially, Ti-6Al-4V alloy offers the best all-round performance for a variety of weight reduction applications in aerospace, automotive and marine equipment15,16). The other titanium alloy is Ti-6Al-7Nb, which was developed as a biomedical replacement for Ti-6Al-4V alloy because Ti-6Al-4V contains vanadium (an element that has demonstrated cytotoxic outcomes when it is isolated). Ti-6Al-7Nb is a dedicated high strength titanium alloy with excellent biocompatibility for surgical implants. Used for replacement of hip joints, it has been in clinical use since early 198617–19). However, titanium and titanium alloys suffer from serious disadvantages of poor machining properties. It takes much time and cost to form Ti alloys as desired shapes of products.
Therefore, in this paper, net-shaping technology, DLF technique is applied to Ti alloys for efficient processing; also it is progressed in vacuum atmosphere to prevent the oxidation of Ti alloys. This research was aimed at challenging the three dimensional titanium alloy compacts with more complex shapes, which are often used in bio-medical, aerospace and other field.
In this section, the honeycomb structure with array of hexagons is taken for an example in order to investigate the effect of parameters of DLF technique on the features in vertical plane, which perpendiculars to the scanning plane. The gas-atomized Ti-6Al-7Nb alloy powder (mean particle size: 26 μm) was employed to fabricate the honeycomb structure with dimension of 6.5 × 6.5 × 6.5 mm. The sizes of the hexagon edge in scanning plane were designed as 600 μm. Fig. 2 shows the schematic diagram of the honeycomb compact.
Schematic graph and sizes of (a) a honeycomb compact and (b) a hexagon cell in honeycomb compact.
The specimens were manufactured by an ytterbium fiber laser with 300 W maximum laser power, 50 μm beam diameter and the continuous wave. To prevent the oxidation of the specimens, the scanning cab was evacuated by the vacuum pump at first, and was filled by pure argon gas. As a result, the scanning was performed at 600 Pa of argon pressures.
In order to investigate the effect of characteristics of laser beam on the morphology and microstructure of Ti-6Al-7Nb alloy, two types of laser scanning process (A and B) were proposed in this section. The experimental parameters are listed in Table 1. For process-A, the high power of laser beam of 280 W was employed and the beam scanned the powder layer only one time. Instead, the lower power of laser beam of 20 W was employed in process-B, and a pre-sintering (one time) and the repeating scanning (20 times) was employed to achieve the smooth scanning-track.
| Process-A | Process-B | |
|---|---|---|
| Laser Power (W) | 280 | 20 |
| Scan Speed (mm/s) | 80 | 80 |
| Number of Scanning | 1 | 20 |
| Powder Layer Thickness (μm) | 80 | 100 |
When the proportional shaped specimens were fabricated by DLF using CAD data, the morphology of surface was discussed based on the images of Scanning Electron Micrographs (SEM). The measurement for density, compression test and tensile test were carried out as well to evaluate the mechanical properties of specimens.
The precision in vertical plane is one of the important factors to describe the morphology of a formed component. Fig. 3 shows the SEM images of formed parts by the process-A and process-B, which parameters are showed in Table 1. Both the cross-sectional views in XY plane and in XZ plane are investigated to present the characteristics in vertical plane. Obviously, it was found that a higher density are obtained by process-A. The voids appear in the view of XZ plane, even the number of scanning repetition is 20 times. The reason can be explained that the energy of 20 W of laser beam is not sufficient for the full melting of Ti-6Al-7Nb alloy powder. Therefore the higher power is helpful to increase the relative density of the formed specimens.
Images of cross-sectional views in XY plane and XZ plane for the formed specimens by process-A and process-B in Table 1.
Some tests were carried out in order to investigate the properties of the honeycomb structures. Fig. 4 shows the testing results of density and some mechanical properties such as compressive strength and Young’s modulus of the real human bone and specimens which are fabricated following the process-A and process-B as shown in Table 1. In Fig. 4(a), the density of specimens and human bone are exhibited and compared. Both the values of the specimens by DLF technique are higher than those of human bone.
Testing results for specimens by process-A, process-B and the referenced values of human bone: (a) density measurement, (b) compressive strength and (c) Young’s modulus.
In Fig. 4(b) the compressive strength of specimens by DLF technique are compared with the strength of human bone (21–116 MPa)20). The compressive strength is one of the important parameters which present the capability of artificial material for implanting. The specimen by process-A shows highest value, and instead the specimen by process-B shows the similar value with real bone. Furthermore, the elastic modulus along X, Y axis is shown in Fig. 4(c) and compared with the reported value of real human chancellors bone (1.2–4.6 GPa)20). For the purpose of medical implanting, the Young’s modulus should match the value of human bone very well, at least is in this same grade. From the results in Fig. 4(c), porous structure by process-B shows a corresponding value with the real bone.
Artificial bones and implants should have not only the same mechanical properties as real bone but also the biocompatibility and high growth potential for bone cells. To improve the osseointegration, honeycomb structure is a possible simple and effective structure. Porous micro structure helps bone cells to grow into, so that it can be used as scaffold. It is reported that the bone cells are cultured well if the size of the unit cell structure is about 300 μm21), which is similar scale as a micro structure of cancellous bone. The hole diameter of the lower right sample in Fig. 5 is close to this dimension.
Honeycomb structures of Ti-6Al-7Nb fabricated by the present laser-forming process. Length below each photo shows the designed edge length of hexagonal walls, and the scan speed and number of the irradiation times are shown in the left of the images.
Mouse osteoblast cell line, MC3T3-E1 cell, was used for culturing experiment on the present titanium alloy compacts. Honeycomb structures with hole diameter about 700, 500 and 300 μm were prepared to culture the osteoblasts. The structures were settled in plastic dishes containing α-MEM (α-minimum essential medium) supplemented with 10 % FBS (fetal bovine serum). The cells were applied on them at 200 cells/mm2 and cultured at 37 °C in a CO2 incubator. Fig. 6 shows the SEM images of cell-cultured samples after 14 days and 28 days. The cells appear to the upper surface of honeycombs, and the extracellular matrix (ECM) is also observed. In the finest structure with 300 μm holes, ECM formed chord-like structure inside the holes, which could promote proliferation of osteoblasts. Fig. 7 shows the magnified view around the hole after 21 days culture. More ECM is observed in the smaller pore structures than in the larger one.
SEM micrographs of honeycomb structures, in which osteoblasts were cultured for 14 days or 28 days.
SEM image of osteoblasts on the honeycomb structures after 21-day culture. (a)700 μm hole diameter, (b) 300 μm).
So many researches of this process (DMLF) with Ti alloys were reported22–25). Main issue is to improve the relative density and accuracy of final products by DMLF process. That is, improvement of adhesion property between layers and the roughness of surface, for that, optimization of laser parameters, laser scanning strategy and re-melting of surface by laser have been studied and reported26). In this study, we focused on the metal powder feeding, especially feeding layer height (FLH). The reason is; in the same laser scanning condition, the property of final product is strongly affected by the amount of melted metal pool mainly decided by feeding layer height. In this point, the effect of feeding layer height on the final product should be preliminary investigated before laser parameter optimization. Therefore, this study presents the effect of FLH on the mechanical properties and surface roughness of laser formed Ti alloy parts
The material used for this study is Ti-6Al-4V alloy powder (Osaka Titanium technologies, TILOP64). This powder, produced by a gas atomization process, is spherical with a particle size under 45 μm. Mean particle size is 34.4 μm. All specimens were manufactured on the in-house developed DMLF machine by our research group. This DMLF installation is equipped with a ytterbium fiber laser (IPG YLR-300 SM), which produces a laser beam with a wavelength of 1070 nm and can reach a maximum power of 300 W in continuous mode. For the laser used in this research, the spot diameter is about 50 μm. The forming chamber is first evacuated and then filled with an inert argon atmosphere to reduce the oxidation during laser forming. At first, to investigate the effect of laser parameters on the surface roughness, single layer test was performed. Rectangular single layer (5 × 5 mm) was irradiated in various laser parameters and optimized laser condition was decided. Secondly, multi-layer specimens were prepared under the optimized laser condition. The forming conditions for single- and multi-layer specimens are shown in Table 2. The surface morphology and roughness was estimated by laser microscope (OLYMPUS OLS 4000). Multi-layer specimens (bar-type tensile specimen) were prepared in various FLH from 80 μm to 250 μm. Relative density of the tensile specimens was measured by image analyzer. Five pictures of cross-section for each specimen were used for image analyzing and the average was evaluated. The etched cross-section was observed in optical microscope. Oxygen content was estimated by oxygen analyzer (LECO TC-500SP). Tensile strength were also assessed for each specimen.
| Laser power, W | Laser power, W | Scan pitch, μm | Feeding layer height, μm | |
|---|---|---|---|---|
| Single layer | 40–200 | 25–240 | 100 | 33.4 |
| Multi layer 1 | 120 | 50 | 100–250 | |
| Multi Layer 2 | 260 | 80 | 80 |
Fig. 8 shows the representative surface morphologies of single layer specimens with forming conditions. Scan direction is from left to right on the picture. When low laser and scan rate, melted powder forms ball shape and they are independently scattered on the substrate (Fig. 8a). If scan rate increases, balls start to be connected to each other and form connected larger ball shape (Fig. 8g). If laser power increases, these start to wet on the substrate (Fig. 8b), finally flat surface appears on the substrate with improved surface roughness (Fig. 8c) because of full-melting of powder.
Surface morphologies of single layer specimens by various laser parameters.
Fig. 9 shows the green-scale mapping of average surface roughness (Sa). Dark green means high value and bright green is low value. In the condition of low laser power and low scan rate that results independent balls on substrate, surface roughness is 50–58 μm. As the increase of laser power, surface roughness increases due to the connection between balls and partial wetting on substrate, however, if flat surface appears in high laser energy density (high laser power and low scan rate), surface roughness improves as 12 μm. Fig. 10 shows the surface roughness of multi-layer specimens in various FLH. Surface roughness of multi-layer specimens increased in compare with that of single layer specimens, however in the case of 100 μm FLH, the increasing is suppressed under 100 μm which is almost one third of other high FLH values.
The green-scale mapping of single layer surface roughness (Sa, μm) measured by laser microscope; dark green indicates high value and bright green indicates low value.
Surface roughness in different FLH.
Fig. 11 shows the relative density of multi-layer specimens fabricated in various FLH. The specimens of 100 μm FLH which has low surface roughness shows almost full density, 99.8 %. The others show low density and high deviation. From these points, to improve relative density and surface roughness, small FLH is very effective in DMLF process. In Fig. 12, tensile strength in term of various FLH was shown. In 100 μm FLH, high tensile strength 1083 MPa that is higher than ASTM value was obtained. In high FLH over 150 μm, in spite of low density, quite high tensile strength was obtained.
Relative density of multi-layer specimens fabricated by different FLH.
Tensile strength of multi-layer specimens fabricated by various FLH (dashed line shows the value of ASTM standard, 950 MPa).
Fig. 13 shows the cross sectional micrographs of tensile specimens. Laser scan direction is perpendicular to the pictures and scanning order is from right to left at 100 μm scan pitch. Low density specimens are explained by the creation of large pores. They are diagonally aligned in the front view (Fig. 13b, c and d). These pore pattern was also observed in Refs27). It was found that the slope angle of pore alignment is dependent on the hatch spacing: the higher the spacing, the higher the slope angle. When the hatch spacing equals the melt pool width28). Fig. 14 shows the cross section of tensile specimen manufactured in 200 μm FLH. The upper image is surface layer showing rough surface and the bottom image shows inside of specimen. Surface of specimen has wave pattern formed by inhomogeneous melting and solidification of powder. The angle of the wave coincides with the angle of inside diagonal pore. This means that the valley space of wave pattern remains as pore because the powder existing there is pulled into the melted pool and solidifies on the top area of wave. That is, high FLH causes rough surface having wave pattern and this makes the unique pores structure, therefore to control and improve the surface roughness and density, small FLH which results flat surface is strongly necessary.
Cross-section of multi-layer specimens fabricated in various FLH; (a) 100 μm, (b) 150 μm, (c) 200 μm, (d) 250 μm.
The relationship between surface roughness and diagonally aligned internal pores.
In microstructure, acicular martensitic phase due to rapid cooling and α + β phase were observed in all specimens (Fig. 15). In 100 μm FLH, martensitic phase showed epitaxial growth in direction to layering direction29). Fig. 16 shows oxygen content of tensile specimens in different FLH. Oxygen content slightly increased compared to raw material. However, this is not a concerning value for the decreasing of mechanical property30).
Acicular martensitic phase of multi-layer specimen by 100 μm FLH.
Oxygen contents of multi-layer specimens.
Fig. 17 shows the tensile test results of as fabricated and as heat-treated compacts. After heat treatment, the elongation increased, however, it is not reached to 14 % of wrought materials. Fatigue strength of direct laser formed compacts was evaluated using tensile test specimen. Tension and tension fatigue testing was adopted with stress ratio R of 0.1 and test rate as 30 Hz. Fig. 18 shows the S-N curve of heat treated laser formed compacts. The compacts show around 300 MPa of fatigue strength even near full density. Normally, wrought materials show higher than 500 MPa, thus, the fatigue strength of direct laser formed compacts are need to improve.
Tensile strength and elongation of DLF specimens; (a) as fabricated, and (b) as heat-treated.
S-N curve of DLF specimens.
At first, the following conclusions were obtained, as a result of examining various characteristics of the multi layered and honeycomb structured compacts in order to fabricate the medical devices with the titanium alloy powders, in the multi-layered compacts of Ti-6Al-7Nb alloy powders. Their tensile strength was 580 MPa. On the other hand, in the honeycomb structured compacts their strength was depended on the size of hole. In the suitable size of hole (300 μm) for bone in growth, their compressive strength was around 400 MPa which was still high as compared to that of human bone (170 MPa). The honeycomb structures showed good mechanical compatibility with real bone, and had superb biocompatibility. Osteo-blasts were cultured on the present honeycomb structures for 28 days. As a result, osteoblasts proliferated most on the structure with 300 μm holes.
Also, single layer and multi-layer specimens were fabricated by DMLF process to investigate the effect of feeding layer height (FLH) on the relative density and surface roughness of final products. From single layer experiment, when the scan rate increases, ball shapes which is formed by melted powders were connected with others and it become larger due to the partial melting of metal powder. In high laser power, metal powder melts fully and wets on the substrate (previous layer), finally flat surface appears. From multi layer experiment, small FLH was strongly effective for improving the relative density and mechanical properties. Moreover, surface roughness could be improved by small FLH because of full powder melting. In this study, 70 μm in Rz, 99.8 % in density and 1080 MPa in tensile strength were obtained by introducing small feeding layer height. Finally, by optimum fabricating parameter, 1130 MPa of tensile strength, 9 % of elongation and around 300 MPa fatigue strength were obtained.
The author sincerely thank to same lab. Staff, Dr. H.G. Kang, Dr. F. Tsumori and Dr. T. Osada for their valuable contribution to this study, and also to Dr. Kurata in Dep. Mechanical Eng. of Kyushu Univ. for his kind helping in culturing experiment.
Hideshi Miura
;%0A%09%09%09newWindow.document.open();%0A%09%09%09newWindow.document.write('<img src=%22./Graphics/32_2015017_19.jpg%22>');%0A%09%09%09newWindow.document.close();%0A%09%09)
Hideshi Miura is Professor of Mechanical Engineering at Kyushu University in Japan and head of Kyushu University Education & Research Center of Manufacturing. He received his BS, MS, and PhD degrees from Kyushu University. He is also a Former President of Japan Society of Powder & Powder Metallurgy (JSPM) and held the 2012 P/M World Congress at Yokohama in Japan as Vice–chairman. He has truly pioneered the development of metal injection molding process and direct leaser forming for various metals in Japan. He published about 300 articles, 1 book (Translated the Powder Metallurgy Science written by R. M. German), 12 patents, and 20 edited books. He received more than 20 Awards from JSPM, Japan Society of Mechanical Engineering (JSME), Japan Institute of Metals and Materials (JIM), Iron and Steel Institute of Japan (ISIJ) etc.
