– Report 1 –
State of the art in Additive Manufacturing of Cemented Carbide
Ph.D. Coordinator, Ph.D. Student,
Prof. Dr. Eng. Nicolae Bâlc M.Eng. G?d?lean Rare? Vasile
TOC o “1-3” h z u 1.Introduction PAGEREF _Toc495992367 h 42.Study on Manufacturing Cemented Carbide PAGEREF _Toc495992368 h 52.1Introduction PAGEREF _Toc495992369 h 52.1.1Definition and Terminology PAGEREF _Toc495992370 h 72.1.2Composition PAGEREF _Toc495992371 h 72.1.3Properties PAGEREF _Toc495992372 h 92.1.4Application PAGEREF _Toc495992373 h 112.2Powder production PAGEREF _Toc495992374 h 132.2.1Tungsten powder production PAGEREF _Toc495992375 h 132.2.2Cobalt powder production PAGEREF _Toc495992376 h 152.2.3Tungsten carbide synthesis PAGEREF _Toc495992377 h 162.3Consolidation PAGEREF _Toc495992378 h 172.3.1Milling of carbides PAGEREF _Toc495992379 h 182.3.2Granulation PAGEREF _Toc495992380 h 192.3.3Green consolidation PAGEREF _Toc495992381 h 202.3.4Sintering of WC-Co hardmetals PAGEREF _Toc495992382 h 223.Study on Additive Manufacturing of Cemented Carbide PAGEREF _Toc495992383 h 253.1Research in polymer binders for SLS PAGEREF _Toc495992384 h 253.1.1Amorphous polymers PAGEREF _Toc495992385 h 263.1.2Semi-crystalline polymers PAGEREF _Toc495992386 h 273.1.3Investigations in polymer powders for SLS PAGEREF _Toc495992387 h 273.1.4Conclusions PAGEREF _Toc495992388 h 313.2Indirect SLS of metallic and ceramic powders PAGEREF _Toc495992389 h 313.2.1Principle of indirect SLS PAGEREF _Toc495992390 h 313.2.2Investigation in indirect SLS of metallic/ceramic powders PAGEREF _Toc495992391 h 323.2.3Conclusions PAGEREF _Toc495992392 h 363.3Direct SLS of Hard Metal powders PAGEREF _Toc495992393 h 373.3.1Principle of direct SLS PAGEREF _Toc495992394 h 373.3.2Investigation in direct SLS of WC-Co powders PAGEREF _Toc495992395 h 383.3.3Conclusions PAGEREF _Toc495992396 h 403.4SLM of Hard Metal powders PAGEREF _Toc495992397 h 413.4.1Principle of direct SLM PAGEREF _Toc495992398 h 413.4.2Investigation in SLM of WC-Co powders PAGEREF _Toc495992399 h 413.4.3Conclusions PAGEREF _Toc495992400 h 443.53D Printed Hardmetal: Binder Jetting PAGEREF _Toc495992401 h 453.5.1Principle of binder jetting PAGEREF _Toc495992402 h 453.5.23D printed hardmetal tools PAGEREF _Toc495992403 h 464.Conclusions PAGEREF _Toc495992404 h 485.References PAGEREF _Toc495992405 h 49
BIBLIOGRAPHY l 1031
IntroductionAdditive manufacturing (AM) technologies have the capability to create 3D parts, most of them from a powder bed, by adding contoured layers on top of each other, such as selective laser sintering (SLS), selective laser melting (SLM) and binder jetting, to mention a few. The biggest advantage that these technologies have over the conventional ones is the lack of limitations regarding the part form complexity. This advantage is named solid freeform fabrication and it refers to the fabrication of complex solid objects directly from a computer model without part-specific tooling or human intervention. In the case of hard materials like cemented carbide and ceramics, the concept of free-form fabrication has a great appeal.
The composite material cemented carbide is the focus of this Ph.D. research paper. Due to its unique mechanical properties, mainly high hardness and the limitations imposed by the conventional manufacturing methods such as die pressing and machining to form and shape the parts, makes cemented carbide suited for additive manufacturing technologies. The objective of this research paper is to manufacture 3D printed cemented carbide parts. To achieve this objective two separate technologies must be combined, additive manufacturing and powder metallurgy. Additive manufacturing has the purpose to generate the actual 3D parts and powder metallurgy has the purpose to generate the right powder composition used in the AM process and the final sintering step.
The first purpose of this research report is to offer an in-depth look at the state of the art in manufacturing cemented carbide parts through the conventional method of powder metallurgy. The studies presented in the first part of this report show the entire process of manufacturing cemented carbide starting from the powder production for each component, tungsten and cobalt powders, following with the tungsten carbide synthesis, milling of those powders, green consolidation and finally the actual sintering process.
The second purpose of this report is to study the methods and approaches used in additive manufacturing to fabricate cemented carbide parts. It was taken into consideration the researches made in the fields of indirect and direct selective laser sintering, selective laser melting and binder jetting. To have adequate powder properties and adequate powder composition for the AM process a close look was taken at the researches made in polymer powders used in AM.
All the studies and research papers presented in this report have the purpose to help and contribute to our approach in manufacturing cemented carbide green parts through indirect selective laser sintering.
Study on Manufacturing Cemented CarbideIntroductionSince its first appearance at the Leipzig Trade Fair in 1927 and until the present day, cemented carbide or hardmetal, is one of the most successful composite engineering material ever produced CITATION 3 l 1033 1. Its unique combination of strength, hardness and toughness makes it suitable for applications in machinery wear parts, mining/rock drilling and industrial tools, such as cutting tools and turning inserts CITATION 3 l 1031 1.
The Frenchman Henri Moissan first fabricated the carbide used as the base for hardmetal, tungsten carbide (WC), in the late 1890s but its technological and commercial importance will not be recognized until after two decades CITATION 5 l 1031 2. During the early 1900s, tungsten became an important metal for lamp filaments, with wire drawing been the process to produce them. The draw dies, the tools used in wire drawing process, were made out of two materials CITATION 5 l 1031 2CITATION 6 l 1031 3: high speed steel (HSS) and diamond. HSS at the time was revolutionizing the metalworking industry, was accessible and relatively low cost but unsatisfactory for drawing tungsten wire due to excessive wear. Diamond was very effective due to its high hardness, but very costly and in addition was difficult to obtain in the years during and after the First World War due to the world economic crisis CITATION 6 l 1031 3.
The need for a harder material, similar to diamond and cost effective as steel appeared with tungsten carbide been known to have the adequate hardness for such an application. The first attempt to fabricate draw dies from tungsten carbide was made in Germany, in 1914 by H. Voigtlander and H. Lohmann. They developed a fabrication process for hard carbide draw dies by sintering parts pressed from powders of tungsten carbide and/or molybdenum carbide and patented this process CITATION 5 l 1031 2.
The breakthrough that lead to the modern technology of cemented carbides came from the R&D group at OSRAM Studiengesellschaft, Germany lead by Dr. Franz Skaupy in the early 1920s CITATION 6 l 1031 3. K. Schröter and A. Fehse, comprising this research group, conducted the first experiments by adding a relative small quantity of nickel to the tungsten carbide at Dr. Franz Skaupy suggestion. These experiments yielded quite varying results until it turned out, maybe by lucky coincidence, that only tungsten monocarbide with an exact stoichiometric carbon content must be mixed with the metal powder (Co or Ni) and after being pressed and sintered yielding a hard but not too brittle material CITATION 6 l 1031 3.
WIDIA hardmetal tools at the Leipzig Trade Fair 1927CITATION 6 l 1031 3
Following this research, a patent was granted under the name of “Schröter DRP 420.689″ in 1923 that lays the basis for modern cemented carbide manufacturing process. The hard material was first marketed in Germany as ”Widia”, acronym for WIe DIAmant = like diamond, in 1926. The main purpose for the new material was to replace the diamond dies for tungsten filament production but it was soon realized that the field of applications is much larger and cemented carbide would revolutionize the entire metal cutting and rock drilling industries.
At the end of World War II, the Allied Forces forced Krupp to publish almost all details of Krupp-WIDIA products consisting of: hardmetal compositions, techniques of production, quality control methods and all R&D product reports CITATION 6 l 1031 3. British Intelligence Objectives Subcommittee published all this and it soon became a very cheap and popular book on how to produce and test hardmetals, which was used worldwide and at least for the next 10 years.
Today, the rise of the worldwide consumption of hardmetals is partly due to the rising tendency of automation in metalworking: turning, milling, drilling, and forming had dramatically risen in the second half of the twentieth century CITATION 6 l 1031 3.
Worldwide production of hardmetals from 1930 to 2008CITATION 6 l 1031 3Definition and TerminologyIn the technical literature, ranges of composite materials consisting of a hard carbide particles bonded together by a metallic binder phase are named cemented carbides or hardmetals. The two terms are equivalent, but their use is dependent on region and language. In Germany and in the german literature only the word “hartmetall” is used and in European English literature and USA, “cemented carbides” is the term preferred CITATION 6 l 1031 3 CITATION 1 l 1031 4. Since their first appearance in 1927, in the machining industry, the term “cemented carbides” is associated with a hardmetal that contains a dominant WC-phase. “Hardmetal” is a more general term, used for the binder-free and cemented materialsCITATION 6 l 1031 3.
Other important composite materials from the same family of powder metallurgy, also developed in the early day of hardmetals 1920s-1930s, are the cermets. The term “cermet” contains the syllables “Cer” from ceramics and “met” form metals and describes a wide range of materials consisting of hard ceramic particles bonded by a metallic binder with properties superior to that attained by any one single component CITATION 6 l 1031 3 CITATION 1 l 1031 4.
CompositionCemented carbide is the result of powder metallurgy and liquid-phase sintering, resulting in a composite material consisting of at least one type of hard, wear resistant particles – the hard phase and ductile, softer metallic particles from the iron group of metals, acting as a matrix to hold the hard particles to form the bulk material – the binder phase.
In the early days, 1920s-1930s, the base for cemented carbide is formed by tungsten carbide (WC) as the hard phase, giving the composition its high hardness, high wear resistance and cobalt (Co) as the binder phase, acting as a medium for carbide grains to grow, forming the skeletal structure and also giving the composition its toughness CITATION 3 l 1033 1 CITATION 6 l 1031 3. Today, after 100 years, the base for cemented carbide remains the same, but with the purpose of improving the steel machining performance, cubic transition metal carbides like Titanium Carbide (TiC), Tantalum Carbide (TaC) and Niobium Carbide (NbC) were added to the basic WC-Co composition even as early as in the 1930s. WC does not dissolve any of these transition metal carbides, but forms a solid solution with these carbides, leading to a microstructure with an enhanced creep resistance, higher oxidation resistance and higher adhesive wear resistance against long-chipping materials. For applications that require high toughness, a higher binder content is chosenCITATION 6 l 1031 3.
The proportion of the hard phase is normally 80 % to 96 % from the total weight of the composite and its grain size averages between 0.2 µm (nano grades) up to 20 µm (extra coarse grades). Tungsten carbide (WC) is the major component of the hard phase with the rest of the hard carbides mentioned above (TiC, TaC, NbC) been the minor part and adding specific properties to the composition CITATION 6 l 1031 3.
The proportion of the binder phase is from around 4 % up to 20 %. Cobalt (Co) is the most common metallic binder and works best with WC, whereas nickel is not so commonly used, working best with titanium carbide (TiC) and chromium carbide (Cr3C2) CITATION 5 l 1031 2.
Microstructural features of a WC–Co (left) compared to a WC–TiC–TaC–Co HM (right)CITATION 6 l 1031 3Further productivity improvements in the metal working industry have been achieved around 1968/69, almost simultaneously at Sandvik in Sweden and at Krupp-WIDIA in Germany, by coating the hardmetal. Indexable inserts and the full hardmetal tools were coated with a thin wear-resistant film of titanium nitride (TiN), titanium carbide (TiC), titanium carbo-nitride (TiCN), titanium aluminium nitride (TiAlN), and aluminium oxide (Al2O3) by chemical vapor deposition (CVD) or physical vapor deposition (PVD). Nitride and carbonitride layers possess high hardness and high oxidation resistance, whereas the chemically inert Al2O3 layer serves as an oxidation-resistant heat barrier to avoid excessive heating and subsequent softening of the substrate CITATION 1 l 1031 4CITATION 6 l 1031 3.
Classification of tungsten carbide (WC) grain sizeCITATION 6 l 1031 3 CITATION 4 l 1031 5Nano Ultrafine Submicron Fine Medium Coarse Extra coarse
?0.2 0.2-0.5 0.5-0.8 0.8-1.3 1.3-2.5 2.5-6.0 ?6.0
A very important feature of cemented carbide is the potential to vary its composition so that the resulting physical and chemical properties can ensure maximum resistance to wear, hardness, transverse rupture strength (TRS), fracture toughness, compressive strength, corrosion, and oxidation. In addition, the wide variety of shapes and sizes that can be produced using modern powder metallurgical processing offers tremendous scope to design cost-effective solutions to many of the problems encountered in the industry.
Wear resistance – is defined as the attainment of acceptable tool life before tools need to be replaced. The nature of it is very complex and the wear rate depends on many variables. In the case of cemented carbide, the abrasive wear resistance is influenced by the cobalt content and the tungsten carbide grain size as shown in the diagrams below CITATION 6 l 1031 3.
Wear resistance to hardness Wear resistance to cobalt contentCITATION 6 l 1031 3Fracture toughness – defined as the ability of a material to absorb energy before fracture. Tests show that the property increases with increasing binder content, and with increased WC grain size. When taking a closer look at a microscope we find the following types of fracture CITATION 6 l 1031 3:
right698500Cleavage fractures – takes place in the carbide grains and increases with increased grain size.
Shear fractures – takes place in the binder and increases with the binder content.
Fracture toughness as a function of the Co content for different WC grain sizesCITATION 6 l 1031 3.
The strength of a hardmetal is controlled by the properties of the hard phase, the binder phase and the carbide–carbide and carbide–binder interfaces. The strength of the hard phase is dependent on the grain size as well as the chemical composition and stoichiometry. WC has a rather narrow stoichiometry window for carbon, which calls for high stability during the production process. The binder of choice for WC-based hardmetals is cobalt, although nickel and its alloys are prevalent for applications requiring corrosion resistance CITATION 6 l 1031 3.
left111823500right111061500Hardness – it increases with decreasing binder content and decreasing grain size and is measured for hardmetals using the Vickers indentation method according to EN 23 878 (ISO 3878). The hardness range extends roughly from that of tool steels, 700HV30 up to 2200 HV30. Hardness decreases with increasing temperature due to increasing plasticity CITATION 1 l 1031 4.
Hardness relative to cobalt contentHardness relative to temperature
Compressive strength – is one of the most important properties of cemented carbides. Compressive strength values of 4–8 kN/mm-2 make hardmetals one of the strongest materials available; witness their use in diamond manufacture as anvils and in hot rolls for metallic materials.
right5080The compressive strength increases with decreasing binder content and decreasing grain size. A carbide grade with a small WC grain size and low binder content has a typical compressive strength approaching 7 kN/mm-2.
Compressive strength relative to cobalt contentCITATION 6 l 1031 3 .
ApplicationThe combination between tungsten carbide (WC) and cobalt (Co), using liquid-phase sintering to form the base for cemented carbide, provides the essential and unique mechanical properties like strength, hardness and toughness critical to modern industries. Because cemented carbide can vary its composition so that the resulting physical and chemical properties can be adapted to a wide range of applications in different industries, it has become over the past several decades the most dominant tool material today. As an example, approximately 70% of metal cutting tools are been made from WC, with 20% of high-speed steels, and 10% of titanium carbide based cermets and other advanced ceramic materials CITATION 6 l 1031 3.
Although the use of cemented carbide is very diverse, the majority of applications can be classified into three major categories: machining, rock drilling, and wear parts CITATION 6 l 1031 3.
Machining – is a manufacturing process in which a cutting tool is used to cut away material to leave the desired part shape CITATION 5 l 1031 2. Machining is one of the most important manufacturing processes. The Industrial Revolution and the growth of the manufacturing-based economies of the world can be traced largely to the development of the various machining operations CITATION 5 l 1031 2. The majority of cutting tools are for machining of metals, they are also used for other materials such as plastics, wood, composites, and ceramic preforms. The main types of cutting tools are tipped tools, designed with replaceable inserts, for use in lathes, milling heads, rotary drilling tools and cutting tools made from full hardmetal used in end mills, drills, tapers, reamers, mainly in smaller dimensions. There are two unique phenomena regarding metal cutting:
Cutting tip temperature – during the machining process the temperature between the cutting tool tip and the work piece can be as high as 1000°C putting a stringent requirement on high-temperature strength, or hot hardness of the cemented carbide (WC-Co) CITATION 6 l 1031 3. To improve this property titanium, tantalum, chromium, and vanadium carbide can be added to WC-Co system CITATION 6 l 1031 3.
Cutting tool and metal chips interaction – can lead to chemical adhesion and wear of the tool, especially at the high temperatures at the cutting tip. A special wear phenomenon called “crater” wear is usually the result of the chemical interaction between the cutting tool and the workpiece. Titanium carbide and tantalum carbide are usually added to WC–Co to reduce crater wear CITATION 6 l 1031 3. The modern machining industry is aiming for higher and higher productivity so the durability of a cutting tool is not as important as the ability of a tool to cut at extremely high speed, thus prioritizing the properties optimization of WC–Co CITATION 6 l 1031 3.
Rock Drilling – is the second largest market segment that uses WC tools and includes those used for oil and natural gas exploration, mining, construction, and all other industrial operations that need to excavate or drill through natural rocks, concrete, and masonry. In this case, there is no chip forming, the temperatures are much lower compared with metal machining, but due to high compressive strength of the rock formations, special properties are required. The rock drilling tools take advantage of the high compressive strength, high hardness, high bending strength, and moderately high fracture toughness of WC–Co materials. There is no other material that has the same combination of mechanical properties, making WC–Co the only viable material for such industrial operations CITATION 6 l 1031 3. The carbide grades selected for these applications are with high cobalt content and coarse grains, in order to achieve a high toughness at the expense of wear resistance.
Wear Parts – are the third major segment of industrial applications for WC–Co. Common characteristics of wear parts, distinguished from metal cutting and rock drilling, are those involving relative motion of the carbide tool surface and a mating surface or particles. Carbides, in this case, are not directly involved in cutting or crushing of the mating material. Typical examples of this category include metal forming dies, powder forming dies, wire drawing dies, plastic extrusion dies, metal working roll mills, abrasive fluid handling nozzles, mechanical seals, paper or plastic cutting tools, food processing tools, and special applications, such as ammunitions for fire arms CITATION 6 l 1031 3.
The application range of straight grade cemented carbide CITATION 6 l 1031 3Powder productionThe base material for manufacturing cemented carbide are hard refractory carbides and a metal binder, both of them in powder form. The most common used refractory carbide phases are WC, Tic, TaC and less common carbide phases are niobium carbide (NbC), vanadium carbide (VC), chromium carbide. The primary metal binder is cobalt, although nickel and other iron-based elements are used CITATION 1 l 1031 4. In the following pages, the production of those raw materials is presented.
Tungsten powder productionTungsten carbide (WC) is a compound composed of one atom of tungsten and one of carbon CITATION 6 l 1031 3. To achieve it you must go through a set of important steps involving powder metallurgy. The first step starts with the extraction of tungsten (W) from mineral ores found in the Earth`s crust. The primary source of tungsten are the mineralsScheelite (CaWO4) and Wolframite (FeMn)WO4 CITATION 1 l 1031 4. A second source of tungsten are scraps from industries that use extensively cemented carbide, like cutting tools, rock drilling and wear parts. There are two types of scraps. Hard scraps consist of worn tools such as cemented carbide cutting and drilling tool bits and have a low amount of impurities. Soft scraps come from the rejects in the manufacturing process of tungsten products such as wires, coils, powders, turnings and do not have a defined shape.
Wolframite ore (left) and different types of tungsten carbide scraps (right)
The objective of the extraction process is to convert the tungsten contained in the ore concentrate to the intermediate compound, usually tungstic acid or ammonium paratungstate CITATION 1 l 1031 4.
Ammonium Paratungstate (APT) is the main tungsten compound used in the manufacturing of tungsten metal powder. It is made via a hydrometallurgical route and during the last 100 years, there have been significant scientific and technological developments in the processing of APT. Ammonium paratungstate (APT) is a white crystalline salt of ammonium and tungsten, with the generally accepted formula (NH4)10(H2W12O42)·4H2O CITATION 6 l 1031 3.
There are two hydrometallurgical methods for APT production. The first one and the oldest one is made by taking scheelite ore and processing it by acid digestion. During this digestion reaction, scheelite (CaWO4) is converted into tungstic acid. Tungstic acid, thus obtained, was dissolved in ammonium hydroxide and the solution after purification was crystallized to APT via evaporation crystallization CITATION 6 l 1031 3. In the case of wolframite ore, a similar method is used. At first, the mineral ore is digested in an alkali solution and the result of this chemical reaction is a sodium tungstate solution. From this solution, scheelite (CaWO4) is precipitated. After scheelite is obtained, the process continues with the same procedure explained above.
Decomposition by HCl
Decomposition by HCl
Tungstic acid H2WO4
Tungstic acid H2WO4
Dissolution in NH4OH
Dissolution in NH4OH
Tungsten Oxide WO3
Tungsten Oxide WO3
The General flow diagram for the processing of scheelite ore by the acid process CITATION 6 l 1031 3.
In the modern methods of hydrometallurgical processing of APT, all the tungsten feeds are digested in sodium hydroxide (or sodium carbonate) at high temperature. After purification steps, the sodium is separated from the sodium tungstate solution via liquid ion exchange (LIX) and finally converted into APT. In some plants, mainly in China, the separation of sodium and other impurities from sodium tungstate solution is made using solid ion exchange (SIX).
26871672199500548490315826700413562376035100278449475352700143336673305500413386917998000278307218399800144016217923100Ore/scrap oxide (NaOH)
Ore/scrap oxide (NaOH)
Filtration and Purification
Filtration and Purification
General flow diagram for the LIX process CITATION 6 l 1031 3
Tungsten Oxides: Yellow and Blue Oxide are the next step in the process of producing tungsten metal powder. In this step the ammonium paratungstate (APT), the base of this process, is converted into different tungsten oxides: tungsten yellow oxide (WO3) and tungsten blue oxide (TBO) CITATION 6 l 1031 3. Tungsten yellow oxide is produced by the calcination of APT in the air, typically in rotary furnaces operating at 500-700 °C CITATION 6 l 1031 3. Blue oxide is made using the same process of calcination but in this situation under the exclusion of air, with temperature above 450 °C CITATION 1 l 1031 4. After obtaining the utmost important tungsten oxides, the next step is to convert them into tungsten metal by a process of reduction.
3807726342233800319357634223380024455173413125001760220342426100Hydrogen Reduction is the predominant industrial method in use today for the conversion of tungsten oxide to tungsten metal. It is utilizing multi-zone, multi-tube horizontal pusher-type furnaces where the oxide powder is loaded into metal boats and then stoked into the furnace via corrosion-resistant tubes CITATION 6 l 1031 3. Reduction temperatures typical for commercial production range from 700 to 1000 °C. The boats progress through the furnace and they are subjected to a flow of excess hydrogen that may be co- or countercurrent to the direction of the travel of the boats. The hydrogen serves as the primary reduction species and acts to remove the water vapor generated during the reduction process CITATION 6 l 1031 3. Production of tungsten powder of any desired size, between 0.5 and 15 µm is carried out by varying the process parameters. The important reduction parameters are temperature, time, hydrogen flow rate, height of oxide powder mass and apparent density of the oxide. However, not only grain size, but also other powder characteristics like, grain size distribution, agglomeration, apparent density, compacting behaviour, are also changed as a consequence CITATION 1 l 1031 4. The reduction sequence for commercially produced tungsten metal is generally accepted to proceed as WO3 WO2.9 WO2.72 WO2 W metal CITATION 6 l 1031 3.
Cobalt powder productionFor cemented carbides, based on tungsten carbide (WC), cobalt (Co) is the most important metallic binder and it can be found in the ores of other metal in earth`s crust. Mineral ores like sulphides, arsenides, oxides, and hydroxides are used in the production of cobalt powder. Currently there are three industrial methods of producing cobalt powder CITATION 1 l 1031 4.
The first method used to produce cobalt powder was by reducing cobalt oxide in hydrogen environment. This processing method has some drawbacks for the cemented carbide industry; it cannot produce the essential ultra-fine cobalt powders and the level of impurities is relative high CITATION 1 l 1031 4.
The second method uses pyrolysis of a cobalt salt, such as cobalt oxalate, which is itself made by reacting oxalic acid and cobalt chloride to produce these ultra-fine powders of around 2 µm in size. During the milling process, together with the tungsten carbide, the cobalt particles break down even more finely, up to 0,001µm, which allows intimate mixing and coating on the tungsten carbide particles CITATION 1 l 1031 4.
The third method and the newest one is called Polyol Cobalt and has been developed on an industrial pilot scale for the production of spherical monodisperse and non-agglomerated cobalt powder CITATION 1 l 1031 4. The process is based on the reduction of cobaltous hydroxide by a mixture of ethylene-glycol and diethylene-glycol. Micron size and sub-micron size cobalt powder can be obtained only by controlling certain process parameters like, reaction temperature or by adding to the reactive medium foreign metal nuclei (palladium, silver) in order to induce heterogeneous nucleation CITATION 1 l 1031 4.
Tungsten carbide synthesisThe next step in achieving cemented carbide, after the process of producing tungsten powder, is the synthesis of tungsten carbide (WC). Tungsten forms two distinct carbide compounds: a monocarbide, WC, and a subcarbide, W2C. Of the two phases, WC has the greatest industrial interest due to its use in hardmetal applications. The method of producing tungsten carbide depends on the desired size of the particles and the end application. The most common carburization methods are carburization in static bed roast or rotary furnaces, the Menstruum process and reduction/carburization of WO3 CITATION 6 l 1031 3. Other methods have been developed, like fluidized bed carburization and chemical vapor reaction.
The standard method of producing tungsten carbide is referred to as “traditional carburization and it is used for the most WC powder types. It consists of pushing a graphite boat of blended tungsten metal and carbon powders through a furnace at elevated temperatures, 1200–2000 °C, in a reducing atmosphere. Carburization takes place in a diffusion-controlled process via a shrinking core mechanism CITATION 6 l 1031 3.
The carburization process begins with the formation of a thin surface layer of WC. After the initial WC formation, carbon diffusion into the particle results in the formation of W2C, which grows inward to consume the tungsten metal core. As the reaction progresses, the outer layer of WC will grow inward following and eventually consuming the W2C. This will take place until as the tungsten metal core is consumed CITATION 6 l 1031 3.
The three stages of carburization process:
Stage 1 (initial layer formation):W + C = WC
Stage 2 (W core consumption):WC + W = W2C
Stage 3 (W2C consumption):W2C + C = WC
Accurate carbon control is very important in hardmetal production and is exercised over the carbon content of the WC and other carbides used in the manufacturing process. A slight deficiency of carbon results in the formation of the extremely embrittling ‘eta’ phase, a double carbide of W and Co, while an excess of carbon gives fine flakes of free graphite which weakens the hardmetal and lowers the resistance to abrasive wear. Complete elimination of either of these is possible only if the carbon content of the sintered alloy is controlled within narrow limits. For example, the carbon limit for a WC-Co containing 6% Co is between 5.68-5.9%, a range of only 0.11%. Even within this range, it has been shown that carbon variation can have important effects on strength, and in practice with modem manufacturing methods, closer limits with a range of only 0.06% is maintained CITATION 1 l 1031 4.
Schematic representation of process flow for making WC powder CITATION 6 l 1031 3.
The final part in achieving cemented carbide is the consolidation process. It involves numerous operations, from the composition selection, which is the most important aspect in determining the final properties of the product, to green part making to obtain the desired shape and finally the actual sintering process. All of this operation will be described in the pager below. To achieve a final sintered product that will meet cost targets and all subsequent process requirements, powder selection is critical. Particle size and quality of hardmetal powders and the metallic binder are very important in achieving these objectives. Using recycled powder may reduce costs and save energy but a close look at the scrap sources must be taken.
These sources of powders may already have cubic carbides, binder metal, or other components, along with a relatively higher oxygen content due to prior processing CITATION 6 l 1031 3.
Milling of carbidesThe process of transforming tungsten carbide powder (WC) and cobalt powder (Co) in cemented carbide products like cutting tools starts with milling of those two components. The main objective of powder milling is particle size reduction according to CITATION 1 l 1031 4. Other important roles are to provide deagglomeration, uniform mixing of various component powders and ensure that every carbide particle is coated with cobalt in such that subsequent processes of green body consolidation and sintering can be accomplished with success.
The steps of the process start with the powders based on mix calculation, milling media and solvent that are charged in the mill with the organic binders. The mill is set to run for a predetermined length of time and after a sample is taken to determine if appropriate mixing and particle size reduction has taken place. If the milled powder meets the criteria, the powder is ready for the next stage of drying. Otherwise, the milling process continues or the powder is scrapped if it cannot be reworked CITATION 6 l 1031 3.
The two main phenomena that occur during the milling of carbide are the temperature increase and oxidation. Due to this reasons the milling takes place in an organic liquid such as acetone, hexane and alcohol, just to mention the most common CITATION 1 l 1031 4. This mixture of tungsten carbide powder, cobalt powder and the organic liquid form a slurry. The most common technologies for powder mixing are ball milling and attritor milling.
Ball milling, as the name suggests, carry a charge of balls made from carbide, stainless steel, porcelain of 2.5-3.0 times the weight of the powder charge in a horizontal hollow cylindrical vessel, half filled with an organic solvent and additives CITATION 1 l 1031 4 CITATION 6 l 1031 3. Ball mill diameter, media size and mill revolutions per minute (rpm) control the process of powder mixing and particle size reduction.
Attritor milling is a relatively high energy milling process requiring less processing time (15h) compared with ball milling (100h) making it more effective. When it comes to particle size the efficiency of ball milling stopes at around 2 µm particles but attritor milling can produce particle size less than 2 µm CITATION 6 l 1031 3 CITATION 1 l 1031 4. Using a vertical hollow cylinder filled with spherical media, powder and solvent, arms that extend from a central rotating shaft powder mixing is provided. Figure 13 presents the schematic representation of these two processes.
Schematic representation of ball milling (left) and attritor milling (right) CITATION 6 l 1031 3.
GranulationThe powder obtained after drying of milled carbide slurry from the operation above is very fine, non-free flowing and has low apparent density. For the pressing process to have good results, a technique of “granulation” is necessary making the powder agglomerates looser, coarser and nearly spherical with good flow and fill properties. Other basic quality requirements for granulated powder include, bulk density, granule shape and size distribution, granule density, all within specified limits. These properties determine the process control achievable in subsequent powder consolidation processes. The old conventional method of granulating the powder is the vacuum drying process. The modern method of granulation, the spray drying is more convenient and economical compared to vacuum drying for the production of granulated carbide powders for pressing compacts.
Spray drying process involves pumping the powder slurry through a nozzle to aerate and create a spray of the slurry in a counter or cocurrent flow of hot gases, such that droplets of the slurry are dried in flight before falling to the bottom of the vessel. It is critical to maintain slurry uniformity from the time slurry is discharged from the mill to the time it is actually sprayed. This is done by maintaining the slurry in a state of constant agitation in a feed tank prior to spraying.
Often wax or other additives may be added to the slurry prior to spraying, thus slurry in the feed tank must be kept in constant agitation to mix the ingredients and avoid any sedimentation or separation of different ingredients of the powder. The slurry is pumped into the nozzle using a pump or pressurization of the feed tank.
Schematic process flow for the spray drying process CITATION 6 l 1031 3.
Green consolidationThe next powder processing step is green-forming the granulated powder into desirable shapes so that they can be sintered to full density forming cemented carbide products. Green compacts are prepared by pressing loose powder mass using an external pressure. This gives shape to the compacts and dimensional control. Generally, compaction pressure is in the range of 21 – 42 kg/mm2, sufficient to impart green strength and no less than 60% green density for all hardmetal compositions CITATION 1 l 1031 4. Although hardmetal attains almost complete densification during liquid phase sintering regardless of green density for compacts with less than 60% theoretical density, percent shrinkage is high, dimensional control is difficult, and, hence, one prefers to have a high green density of the compacts prior to sintering CITATION 1 l 1031 4.
For green compaction of hardmetals, either a single or a double acting press can be used with the double action press having an advantage over the single action regarding uniform density distribution and hence more uniform shrinkage CITATION 1 l 1031 4. As the parts size increases the action force of the presses moves from mechanic to hydraulic. To high compaction force can produce uneven density within the pressed compact, which may lead to cracking when the force is released.
Isostatic pressing – for some cemented carbide products is necessary to apply the pressure uniformly, from all directions, which will result in a uniform distribution of density all throughout the compact and the sintering properties will be more uniform compared to the compacts pressed in rigid dies. Another advantage of isostatic compaction over conventional
compaction in rigid dies is that no die-wall friction is present and pressure loss is minimized, whereas in compaction using rigid dies, there is a perceptible pressure loss due to die-wall friction. The disadvantage will be that all the powders made by isostatic pressing need a presintering cycle, followed by shaping operations and finally sintering. This method is only used when there are a small number of parts, relatively large and with complex shapes not suited for rigid dies CITATION 1 l 1033 4.
A schematic diagram of a cold isostatic press CITATION 6 l 1033 3.
Extrusion – for the cutting toll industry that specializes in tools like drills, reamers, boring tools, deep-hole drills and so one, the need to obtain cemented carbide rods that have a high length to diameter ratio appears. Using this method working diameters ranging from 0.5 mm to 10 mm with lengths varying from 10 mm to 100 mm and in some cases even longer can be obtained. The hardmetal powder does not exhibit plastic flow even at high pressure therefore a plasticizer is used which has a low yield stress. The common plasticizers used are polyvinyl alcohol, poly ethylene glycol, starch solution, paraffin wax, synthetic resins to mention a few CITATION 1 l 1033 4.
Hardmetal powder extrusion is made in an extrusion press, using either a piston-type or a screw-type mechanism to push a powder/binder mixture through a die. The most common die shape is round, but a variety of cross-sections can be extruded CITATION 6 l 1033 3.
Schematic representation of a piston type extrusion press CITATION 6 l 1033 3.
Sintering of WC-Co hardmetalsThe term sintering was initially used to describe mineral agglomeration by heating, from the “cinder” concept CITATION 6 l 1033 3. By the early 1900s, along with the first development of cemented carbide, sintering was used to describe the heat treatment of metal powders to form solid bodies. By the 1920s, cemented carbides were sintered with a bonding transition metal (Co, Fe, or Ni) and in the late 1930s, ceramic firing began to be termed sintering CITATION 6 l 1033 3.
Sintering occurs when tight packed particles are heated near the melting temperature and due to this heating, the phenomenon of atomic motion occurs CITATION 6 l 1033 3. Atoms vibrate around their lattice sites and as the temperatures are getting higher induce more vibrational amplitude and increase the vacant lattice sites, making atomic diffusion along volume, surface, vapor, and grain boundary pathways accelerate at higher temperatures CITATION 6 l 1033 3.
According to CITATION 6 l 1033 3 “Sintering involves the reduction in surface energy initially by the growth of interparticle bonds and later in the cycle by microstructure coarsening. The surface energy per unit volume depends on the inverse of the grain size, so smaller grains have more energy and sinter at lower temperatures”. Not all surface energy is available for sintering, thus
sintering is favored when the surface energy is larger than the grain boundary energy. Liquid phases lubricate and enhance sintering, especially since diffusion rates in liquids are 100–1000 times higher than in solids and also external pressure supplements the sintering event by supplementing the capillary stress with an external stress CITATION 6 l 1033 3.
The sintering methods that are utilizing external pressure to ensure full density are as follows CITATION 6 l 1033 3:
Hot pressing in a heated die with uniaxial pressure, good for high solid contents.
Spark sintering with electric discharge heating through the hot pressing punches.
Hot isostatic pressing (HIP) in a flexible (glass, metal) container for complex shapes.
Sinter-HIP: vacuum sintering is followed by HIP in one cycle.
Due to their high meting temperature, hard materials like carbides, borides and nitrides, are resisting sintering and specific problems appear. The compounds can decompose before they are sintered, the technology for reaching very high temperatures can be quite expensive and although the sintered structures are very hard, they are also brittle. To improve sintering and to add toughness and ductility the hard phase is mixt with a metallic phase that melts, spreads, and cements the hard particles together. This is done by liquid-phase sintering (LPS) for almost all hard materials, especially WC-Co system.
Liquid-Phase Sintering (LPS) provides a fast sintering process with complete densification without the need for external pressure. About 90% of the commercial sintered products are formed using a liquid during the heating cycle CITATION 6 l 1033 3 CITATION 1 l 1033 4. As CITATION 6 l 1033 3 CITATION 1 l 1033 4 point out fast sintering is due to the enhanced atomic diffusion in the presence of liquid phase, which ultimately facilitates material transport. In cemented carbide systems, the hard phase makes up for the majority of the composition and in spite of this due to the metallic binder phase and liquid phase sintering the resulted composite material has a ductile behavior.
Reduction of surface energy is the main driving force for densification CITATION 1 l 1033 4. The minimum criteria for a successful liquid phase sintering are, a low temperature, solubility of the solid in liquid and liquid wetting of the solid grains CITATION 1 l 1033 4. In the case of the WC-Co system, the liquid phase is cobalt and satisfies all the three basic conditions required for liquid phase sintering. Cobalt is used as a binder metal for cemented carbides because of its good wettability and the solubility of tungsten carbide.
According to CITATION 6 l 1033 3 the liquid phase sintering process has four main stages:
Solid state heating – where prior to the liquid formation the solid state sintering happens and is responsible for more than 50% of the total densification CITATION 1 l 1033 4.
Rearrangement – when the first liquid is formed, grain surfaces are dissolved, while capillary force from the wetting liquid pulls the solid grains together and gives a bust in densification that is paced by heat transfer rates and the enthalpy needed to form the liquid CITATION 6 l 1033 3.
Solution-reprecipitation – the solid grains undergo dissolution into the fresh liquid and the transport from the smaller grains to the larger grains is made. This results in the growth of the large grains at the expense of the small grains. Grain shape accommodation ensures a tight fitting together of the grains to eliminate pores CITATION 6 l 1033 3.
Solid skeletal sintering – with the liquid reaching saturation, the contacting solid grains grow sinter bonds, giving a rigid three-dimensional (3D) solid skeletal structure with liquid dispersed in the spaces between solid grains. Further densification of the solid skeleton structure occurs slowly in the final stage of LPS.
Schematic representation of liquid phase sintering stages CITATION 6 l 1033 3.
Form the mineral ores used to produce tungsten and cobalt powder, to carburization of tungsten, obtaining tungsten carbide (WC), then forming the WC and Co powder mixture and finally using liquid phase sintering, the process of obtaining cemented carbide is a very complex one and it involves a large number of technologies that depend on one and other.
Study on Additive Manufacturing of Cemented Carbide
Additive manufacturing (AM) is the term used to describe all nonconventional fabrication methods and according to CITATION 2 l 1031 6 the term “3D Printing” will replace all the names, including additive manufacturing, to become the generally accepted generic term for layer technology.
Powder metallurgy (PM) is the term to describe the technology that is used to fabricate cemented carbides or hardmetals. According to CITATION Bro15 l 1031 7 additive manufacturing is nothing new to the powder metallurgy industry because powder metallurgy is the ‘classic’ additive manufacturing system and until now the only commercial method for making hardmetals. The goal of this research is to successfully combine additive manufacturing and powder metallurgy to create 3D printed cemented carbide bodies. To achieve this goal selective laser sintering (SLS), an application of AM, was chosen to produce green bodies from tungsten carbide and cobalt powder bonded together by a polymer binder that will be sintered in cemented carbide using the classic sintering process.
This chapter describes the research and studies done in the field of laser sintering (LS), direct and indirect, in polymer binders for LS and the latest successful method for 3D printing cemented carbide with the purpose of learning and adding value to our approach.
Research in polymer binders for SLSThe main objective of this papers is to generate a green body comprised from hardmetal powder (tungsten carbide WC + cobalt Co) bonded together by a polymer binder (polyvinyl alcohol). The SLS machine (Sinterstation 2000) has a low power CO2 laser, maximum 39 W, which is unable to generate enough heat to affect the tungsten or the cobalt powder in any way. At this stage, the material from the composition responsible for the strength and the integrity of the green body is the polymer. Due to its crucial role in the success of the research an analysis of polymers used in laser sintering applications must be made.
A polymer is composed of many large molecules, named macromolecules, that result in long chains CITATION Oss10 l 1031 8. Due to their molecular structure, polymers have unique and versatile properties, thus making them the most sought-after material today. Today, parts and components that were made from traditional material like wood, metal, ceramic or glass are manufactured from different types of polymers CITATION Oss10 l 1031 8. Polymers are divided in three categories: thermoplastics, thermosets and elastomers CITATION Oss10 l 1031 8.
They are all made of large molecules, with those molecules been in the first case crosslinked, meaning that physical links interconnect the polymer molecules (e.g. vulcanized rubber) and in the second case non-crosslinked, meaning that the molecules can move freely along their neighbors CITATION Oss10 l 1031 8. For this research, the focus will be on thermoplastic polymers. They have the property to solidify after cooling and take on two types of structure: amorphous and semi-crystalline CITATION Oss10 l 1031 8. The study will be made with a close look at their properties, finding the most suited type of polymer for laser sintering applications.
Amorphous polymers have their molecular structure arranged in a random pattern after solidifying. Generally, amorphous thermoplastics are transparent and have a glass like appearance. This is due to their random structure and the characteristic size of the largest ordered region similar to a carbon-carbon bond CITATION Oss10 l 1031 8. Amorphous polymers have two main states when considering their shear modulus as a function of temperature.
The first one is the glassy state, where the material exhibits hard and brittle properties. For example, this can state can go as low as -160 °C up to 80 °C in the case of polystyrene, one of the most common amorphous polymers. As the temperature increases, there is not a clear transition between solid and liquid, thus the temperature that divides the two states in an amorphous polymer is referred to as the glass transition temperature (Tg). The second state is reached when the polymer is above the Tg and it is called the entropy elastic region. At this point, the material can be more easily deformed, such as is done during the thermoforming process. In order for the material to flow, its temperature must be above a softening temperature (Ts), at which point it behaves like a viscoelastic fluid CITATION Oss10 l 1031 8. Examples of common amorphous polymers used in industry include polystyrene, polymethyl methacrylate, polycarbonate, polyvinyl chloride.
Polymer molecular structure: amorphous (left) and semi-crystalline (right)
Semi-crystalline polymersSemi-crystalline polymers have more order in their molecular structure compared with the amorphous polymers. During the cooling process, they harden when the molecules begin to arrange in a regular order below the melting temperature (Tm) CITATION Oss10 l 1031 8. Not all the molecules in the structure are transformed into ordered regions, thus small amorphous regions remain in the structure, thus the name semi-crystalline. The degree of crystallinity in a typical thermoplastic can vary from 40% up to 80% depending on the grade and the manufacturing process. The density and strength of semi-crystalline polymers increases with the degree of crystallinity according to CITATION Oss10 l 1031 8 when comparing low and high density polyethylene. Compared with the amorphous polymers they have a high melting point not a low one, they are translucent and not transparent, they have high shrinkage not low one, good chemical resistance and are harder. Some of the most common semi-crystalline polymers are polyethylene, polypropylene, polyamide and polyvinyl alcohol.
Investigations in polymer powders for SLSLooking at the global market share for polymer powders, according to CITATION Sch16 l 1031 9 the laser sintering (LS) powders are only a niche market compared with the worldwide consumption of polymer powders in the injection molding (IM) industry. The annual global consumption of polymers is around 290 million tons per year CITATION Sch16 l 1031 9 and only a very small fraction of this quantity makes up the laser sintering (LS) applications, around 1900 tons per year. The main reason is the very sophisticated combination of polymer properties necessary for successful LS applications. For a polymer to be applicable for laser sintering, we have to look at its properties and distinguishing between extrinsic and intrinsic.
Extrinsic properties refer to the polymer powder properties that can be controlled during the production process and include the particle shape and particle size. According to CITATION Dav14 l 1031 10 the powder must have adequate flow to produce a smooth layer when displaced by a roller and according to CITATION Sch16 l 1031 9 this characteristic is influenced by the particle shape. Thus, in the ideal case the particle must be spherical. To determine the flowabillity of a powder we have to look at the Hausner ratio (HR). HR is calculated by dividing the tap density to the bulk density and according to CITATION Sch16 l 1031 9 a HR < 1.25 equals free flowing powder and a HR > 1.4 means fluidization problems. Particle shape affects powder flowabillity, powder bed density, part surface roughness, and final part porosity CITATION Dav14 l 1031 10.
A particle size distribution (PSD) analysis is necessary for a successful laser sintering process. PSD is measured by laser diffraction systems CITATION Sch16 l 1031 9 and it shows the volume distribution of particles forming the powder according to their size. An example of a PSD diagram can be seen in the figure 19.
Particle size distribution diagram for Duraform EX powder CITATION Dav14 l 1031 10Uniform particle size distribution is crucial for the laser sintering process. Large particles tend to require more laser energy to melt than small particles, so large variations in particle size can result in complete melting of small particles and incomplete melting of large ones leading to a phenomenon called “coring” and may result in part porosity according to CITATION Dav14 l 1031 10. Studying both CITATION Sch16 l 1031 9 and CITATION Dav14 l 1031 10 reports a favorable particle sire distribution is between 20µm and 90µm.
Intrinsic Properties are determined by the molecular properties of the polymer, they cannot be influenced easily and include thermal properties, rheological properties (surface tension, melt viscosity) and optical properties (reflection, absorption, transmission). The most important property of a polymer during the LS process is its thermal behavior. For this, we have to look at the degree of crystallinity for each polymer. As presented above, regarding their crystallinity polymers can be amorphous or semi-crystalline resulting in different thermal properties that can benefit or hurt the LS process. The semi-crystalline polymers offer the best thermal properties for laser sintering CITATION CYa11 l 1031 11.
The part bed temperature (Tb), a process parameter during laser sintering, must be kept 2-4 °C below the melting temperature (Tm) of the respective polymer, with the laser adding the necessary extra degrees to sinter the particles and then the layers together CITATION Dav14 l 1031 10. The crystallization temperature (Tc) of the polymer must be also taken into consideration. If the part bed temperature gets as low as the Tc the resulted parts will be curled and distorted CITATION Sch16 l 1031 9. The temperature domain between the crystallization temperature (Tc) and the melting temperature (Tm) is called the “sintering window” CITATION Sch16 l 1031 9.
The sintering window of a specific polymer can be determined by a differential scanning calorimetry (DSC), been crucial to a successful sintering process. A theoretical representation of a DSC curve for a polymer can be seen in the figure 20.
Theoretical DSC curve with the sintering window specific for a LS process CITATION Sch16 l 1031 9According to CITATION Sch16 l 1031 9 other intrinsic properties like a low zero viscosity (?0) and a low surfaces tension (?) of the polymer during melting are necessary for successful LS processing. These requirements are not met in the case of amorphous polymers because their viscosity is still high above the glass transition temperature (Tg) resulting in brittle and instable parts CITATION Sch16 l 1031 9. Other important intrinsic properties for the polymer powders are the optical properties. During the LS process is important that the polymer powder absorbs enough laser energy in order to melt the particles and also important is to be able to transmit enough energy to the previous layer causing the layers to bond together CITATION Sch16 l 1031 9.
The thermal differences mentioned above between amorphous, semi-crystalline polymers have a great impact on the selective laser sintering process parameters and the resulting properties of the parts made with this method. Part properties as relative density, tensile strength and dimensional accuracy that are very important for the success of this research are studied by C. Yan et al. in the paper the paper CITATION CYa11 l 1031 11. The experiments are made using a HRPS-III SLS system made by HUST China in a nitrogen atmosphere with a typical amorphous polymer, polystyrene (PS) and a typical semi-crystalline polymer, nylon-12 (PA12).
The sintering window mentioned above been crucial for the laser sintering process is described in this paper CITATION CYa11 l 1031 11 for both amorphous and semi-crystalline polymers. In the case of amorphous polymers is between Tc and Tg and for semi-crystalline polymers is between Tc and Tm. In this temperature window problems as part warpage, curling will be avoided, stress relaxation, good flowabillity will be achieved and the part is in a supercooled liquid state during the layer building process CITATION CYa11 l 1031 11.
The relative density of the parts obtained using semi-crystalline polymers tend to have 50% higher density compared with the amorphous polymers, achieving almost full density CITATION CYa11 l 1031 11. The motivation for this observation offered by CITATION CYa11 l 1031 11 is that amorphous polymers have a higher viscosity that semi-crystalline ones during the SLS process. As mentioned, the Tb is set around the Tg for amorphous polymers, thus the viscosity decreases slowly. For semicrystalline polymers, the Tb is set around Tm resulting in a much lower viscosity.
The tensile strength for the nylon-12 (semi-crystalline) polymer is up to 300% higher compared to the polystyrene (PS) at the same laser energy density. This is due to the existence of crystallites in semi-crystalline polymers resulting in superior mechanical properties compared to that of amorphous ones at the same molecular weight and structure. In addition, the higher density in the case of nylon-12 parts greatly contributes to better tensile strength, this value is close to the tensile strength value of the fully dense parts. CITATION CYa11 l 1031 11.
The observations made by CITATION CYa11 l 1031 11 regarding the dimensional accuracy of the SLS parts are as follows: The average dimensional error in X and Y direction of the parts made with nylon-12 is 2.6 higher than that of the parts made with polystyrene (PS). In the Z direction, the average dimensional error is 3.9 times higher in the case of nylon-12 parts. The conclusion that can be drawn from this result is that parts made using amorphous polymers provide a better dimensional accuracy compared with parts made using semi-crystalline polymers. The author of CITATION CYa11 l 1031 11 attribute this result to two reasons: the volume shrinkage caused by phase transition and the volume shrinkage caused by sintering densification.
The phase transition volume shrinkage of amorphous polymers in smaller than that of semi-crystalline polymers because of their high viscosity at the Tg, resulting in a bad powder flowabillity with the sintering of just the adjacent particles causing many voids. Semi-crystalline polymers powders are loosely packed, have a good flowabillity around Tm, allowing them to sinter in near-fully dense pats, but with a large volume shrinkage caused by sintering densification. Considering the two factors above, it can be concluded that the volume shrinkage of a semi-crystalline polymer is larger than that of an amorphous polymer during the SLS process.
ConclusionsFollowing the study of the research papers CITATION Sch16 l 1031 9 CITATION Dav14 l 1031 10 CITATION CYa11 l 1031 11 regarding the adequate polymer powder suited for laser sintering applications, the following conclusions are drawn:
– Spherical shaped particles provide the best flowabillity of the powder necessary during the laser sintering process.
– Uniform particle size distribution provides the best results and it can be measured using a laser diffraction system.
– The part bed temperatures (Tb) of amorphous polymers and semi-crystalline polymers should be kept close to glass transition temperature (Tg) and initial melting temperature (Tm) respectively. Measured by differential scanning calorimetry (DSC) and combined with trial and error experiments to determine Tb for the SLS process.
– Amorphous polymer parts have low relative densities with many voids inside. Semi-crystalline polymer parts have high relative densities close to fully dense parts.
– The strength of the amorphous polymer parts is very low because of their low density. For semi-crystalline polymer sintered parts been close to full density the resulted strength is also closed to that of the fully dense parts.
– The dimensional accuracy of the SLS parts of amorphous polymers is higher than that of semi-crystalline polymer SLS parts at the same processing parameters.
Indirect SLS of metallic and ceramic powdersPrinciple of indirect SLSSelective laser sintering (SLS) as all layer based additive manufacturing technologies starts from a 3D CAD model that is sliced into equal thickness using computing power and under the action of a laser the 3D model is constructed from a powder base layer by layer. In the case of indirect SLS, the raw material for this process consists of structural metallic or ceramic powders that are coated or mixt with a polymer used as binder material. During the sintering process, the laser is used in such a way that the binder material melts and forms a bond between the structural particles without actually melting them. The melting of the material binder is exhibited by its high infrared absorption and low melting point CITATION 7 l 1031 12.
The process produces a green part which is bounded together only by the polymer and requires to be processed in a high temperature furnace to remove the polymer and sinter the part, creating metal-metal or ceramic-ceramic bonds, depending on the case. The porosity of the green parts obtained is low, around 45%, therefore is the need of post processing techniques to further densify the parts to achieve a better mechanical properties CITATION 7 l 1031 12.
Using coated powders will result in green parts with higher strength and also the coated powders are more homogenous than the mixed powder, therefore avoiding problems with segregation that occur in mixed powders. One of the advantages of indirect SLS is that only the binder needs to be melted, although there are some limitations to, in the variety of the materials that can be used and the parts present poor mechanical proprieties, thermal proprieties and accuracy CITATION Mon02 l 1031 13.
Investigation in indirect SLS of metallic/ceramic powdersCeramic materials such as alumina (aluminum oxide – Al2O3), zirconia (zirconium oxide – ZrO2), tungsten carbide (WC) and metallic materials such as tool steel are hard to be fabricated and shaped by conventional machining methods. Indirect SLS is a viable additive manufacturing method that offers a high degree of free-form fabrication and as shown by research papers in this field such as CITATION Khu13 l 1031 14 CITATION Khu131 l 1031 15 CITATION KSh11 l 1031 16 has a degree of success. To quantify the success of this method we have to look at the properties of the green and sintered parts such as density, porosity, homogeneity of the structure and degree of thermal cracking to name some of them. Due to the working principle of a commercial laser sintering machine presented above and the fact that in indirect SLS only the polymer binder is affected by the laser beam the resulted green parts have densities only between 20 – 40 %. Thus, post-processing operations are required to improve the green parts density and respectively improve the final sintered part density.
Studying the research done in this field of indirect SLS metallic or ceramic parts, two main approaches are present. The first approach is to use a dry powder made from a hard phase and a binder phase that are mechanically mixt together or the binder phase is coated on the hard phase, the later offers the best results regarding the green part density. The second approach is to use a slurry-based SLS process. This process has the advantage of starting from more homogeneous and highly packed powder layers due to the deposition method, the semi-liquid mixture consisting from structural powder plus the polymer binder and the fact that each layer is dried before laser sintered. In the case of ceramic materials, the slurry-based process has good results with reported densities up to 98 % CITATION Hwa11 l 1031 17.
J. P. Kruth et al. investigated the approach using dry powder at the Leuven University in the research papers CITATION Khu13 l 1031 14 CITATION Khu131 l 1031 15 for two ceramic materials, alumina (Al2O3) and zirconia (ZrO2). Both ceramic materials were coated using thermally induced phase separation (TIPS) with polypropylene (PP) as the polymer binder to produce homogenous spherical alumina-polypropylene composite powder. The composite powder was laser sintered in a nitrogen atmosphere using a DTM Sinterstation 2000 machine. The SLS process parameters were tested and adjusted with the purpose of obtaining as high as possible green part density and strength.
After the laser sintering process, the resulted green part densities in the case of zirconia-PP powder was 36 % and in the case of alumina-PP, powder was 34 %. The green parts with the best mechanical properties were obtained when using laser energy densities between 0.1-0.2 J/mm3 and those parts were further used in post-processing operations.
The authors of CITATION Khu13 l 1031 14 CITATION Khu131 l 1031 15 have shown that pressure infiltration (PI) and warm isostatic pressing (WIPing) increase significantly the green part density and this translates to an increased final density after the full sintering process is completed. In the case of zirconia-PP powder, WIPing provided the biggest increase in green part density from 36 % to 90 % translating into a sintered density of 92 %. PI also improved the green part density for the same zirconia-PP powder form 36 % to 45 % resulting into a sintered density of 54 %. When combining both PI and WIPing the green part density increased up to 83% and the sintered density was 85 %. For the alumina-PP powder, similar results were reported with WIPing providing an increase in green part density up to 93 % and a sintered density of 89 % could be achieved. PI increased the green part density up to 51-53 % resulting in a sintered density between 62-64 %. Combining PI and WIPing provides a green part density of 83 % and sintered density 88 %.
In both powder composition cases the author of CITATION Khu13 l 1031 14 CITATION Khu131 l 1031 15 report that the best post-processing method is the combination of PI and WIPing providing a smaller degree of shrinkage compared to only WIPing, resulting in less part distortion and avoiding unwanted cracks after sintering. Although this processing methods increase the density of the green parts to nearly full density, still a good degree of porosity is present in the structure causing a low flexural strength in the case of alumina sintered parts, reported value 148±20 MPa CITATION Khu131 l 1033 15. Examples of such sintered structures can be seen in the SEM figures 21 and 22. Through this processing method of indirect SLS combined with pressure infiltration and warm isostatic pressing, crack-free ceramic parts can be obtained offering an advantage over the direct SLS and SLM techniques.
Isostatic pressing is another viable prost-processing operation for the indirect SLS of ceramic materials studied at the Leuven University by J. P. Kruth et al. in the paper CITATION KSh11 l 1031 16. The purpose is the same, to improve green part density and the final sintered density.
30391100000000SEM micrographs of cross-sectioned sintered alumina samples produced by SLS + PI + WIP CITATION Khu131 l 1033 15303657026098500026098500
SEM micrographs of cross-sectioned sintered zirconia samples produced by SLS + PI + WIP CITATION Khu13 l 1033 14Cold isostatic pressing (CIP) and quasi isostatic pressing (QIPing) are the post processing methods used by the authors of CITATION KSh11 l 1031 16 to improve green part density in the case of components made from an alumina (Al2O3) – polyamide (PA) composite powder. The composite powder is produced by planetary ball milling, resulting in a mixture of 0.3 µm alumina particles and 100 µm polyamide 12 (PA12) particles. A DTM Sinterstation 2000 laser sintering machine was used to generate the green parts with laser energy densities between 0.2 – 1.9 J/mm3. The resulted green part densities were very low, 22 – 24 %TD, due to the SLS working principle, a roller is used to spread the powder layers and the fact that the polymer binder is mixt not coated on the alumina powder.
The authors of CITATION KSh11 l 1031 16 report that wet bag CIP at 200 MPa can double the green part density, up to 50 %TD in the case of the sample made with the highest laser energy density 1.9 J/mm3. Further polymer debinding and sintering steps improve the final density of sintered parts in the range of 85 – 92 %TD. The highest density is obtained with the parts made in the SLS process with the highest laser energy density.
3120390371919500left372554500The green part that were post-processed using QIPing for 15 min at 165 °C and 20MPa showed a sintered maximum sintered density of 94 %TD. This is because of the elevated temperature and pressure that enables the PA to deform plastically and reduce some of the porosity. Although both post-processing methods improved significantly the final density of the sintered parts, in both cases closed porosity was obtained. Two types of pores are reported in the paper CITATION KSh11 l 1031 16: long elongated isolated 100 µm pores and substantially smaller 10 µm sized residual pores. To eliminate the unwanted porosity the authors suggest ways to homogenize the packing density of the powder by increasing the flowability and binder content or by using an alternative composite powder preparation rout, allowing to generate spherical agglomerates. Furthermore, during the polymer debinding and furnace sintering steps cracks are present in the center of almost all test parts. The cracks are attributed to the heating rates during the polymer debinding and furnace sintering steps, but this is still under investigation. The figures 23 and 24 represent the sintered structures obtained in the paper CITATION KSh11 l 1031 16 with both post-processing methods, CIP and QIPing at different zooms.
SEM images of the sintered structured with CIP post-processing CITATION KSh11 l 1031 16right21797500left21956800
SEM images of the sintered structure with QIPing post-processing CITATION KSh11 l 1031 16The approach using slurry-base SLS to fabricate ceramic parts is studied by Hwa-Hsing Tang et al. in the research paper CITATION Hwa11 l 1031 17. The aim of this research paper is to fabricate high strength alumina parts through SLS by creating a new slurry based process, using as polymer binder polyvinyl alcohol (PVAl). The process of slurry-based SLS includes an additional step to the normal procedure. The step involves drying the slurry layer with a heater to form a fresh green layer between the previous step of casting a thin slurry layer and the next step of laser scanning.
The authors of CITATION Hwa11 l 1031 17 report that high strength alumina parts with the value for flexural strength of 363.5 MPa and a value of 98 % from the theoretical density can be fabricated through this new slurry-based approach. This is possible by first coating the alumina (Al2O3) particles with fully hydrolysed PVAl (degree of hydrolysis of 98.5?99.2 mol %) to form spherical agglomerates, one micrometre in size. Secondly by mixing this coated PVAl- Al2O3 powder with deionized water (solvent), ammonium polymethacrylate (dispersant) and a 6% aqueous solution of sub-partially hydrolysed PVAl (degree of hydrolysis of 72?76 mol%) to form the slurry. The purpose of the sub-partially hydrolysed PVAl is to allow the removal of the green part after the SLS process is completed due to its solubility in water at room temperature. The laser scanned PVAl will be insoluble in water due to recrystallization.
The relative density of the green parts reported in the paper CITATION Hwa11 l 1031 17 after the completion of the laser sintering process is 56.7 %. This density value is much larger than that of the green parts obtained by dry powders, 22-36 %, used in the conventional SLS process. This high green part density value translates, as I mentioned before, to a final sintered density of 98 % with parts free of delamination and cracks and a flexural strength of about 363.5MPa. The authors of CITATION Hwa11 l 1031 17 attribute these good results to the use of the dispersant. This forms an absorptive layer on the ceramic particle surface to obtain a strong steric effect. Furthermore, by controlling the pH level and after each layer was laid and dried the next deposited layer, because of the 70 vol% water content in the slurry, had the chance to be absorbed in pores of the previous layer. The total amount of the PVAl binder used in CITATION Hwa11 l 1031 17 was 3.95 wt% of the slurry; the water-insoluble PVAl coated on alumina powder and the water-soluble PVAl binder were 2.55 wt% and 1.4 wt%, respectively.
ConclusionsPost-processing operations are absolute necessary to increase the green part density, in the case of indirect SLS method. Each method that is presented above has its own advantages and disadvantages when we refer to the final part properties such as sintered part density, porosity, cracks and mechanical properties. The best results from the methods presented above, regarding the final part properties are provided by the slurry-based SLS process with the highest obtained density and high flexural strength but due to the step of layer drying the part building rate is modest (?0.89 mm3/s) according to CITATION Khu13 l 1031 14. For the other post-processing methods based on dry powders, similar results can be achieved only when combining two post-processing methods. Still a high degree of porosity is present in the structure, the debinding and sintering process must be carefully controlled to avoid cracking and the resulted mechanical properties are low compared to the standard ones. Indirect SLS can fabricate complex ceramic and metallic parts but because of the problems mentioned, their application is limited.
Direct SLS of Hardmetal powdersPrinciple of direct SLSIn the case of direct SLS process, the laser beam is used to directly sinter a wide range of materials like various types of polymers, metal and ceramic powders without the presents of the binding material. The most used materials in direct SLS are thermoplastic polymers, which include amorphous, semicrystalline and reinforced, or filled polymers CITATION 7 l 1031 12. Preferably, the resulted parts have a high density, require minimal or no post-processing and they present good mechanical characteristics for the use as final parts.
According to CITATION Mon02 l 1031 13 the first attempts to directly sinter single-phase metal powders during the late 90s were unsuccessful due to the quick consolidation of the molten powder into a sphere with the diameter approximately equal to the laser beam diameter. This phenomenon is known as balling. To overcome this problem another approach to the sinter process is needed. By using a two-phase metal powder in which one of the components has a low melting point and acts as the binder for the other component CITATION Mon02 l 1031 13. This method has its own disadvantage; the parts exhibit the mechanical proprieties and characteristics of the weaker metal phase, making it hard for the use of these parts in high stress applications.
To obtain parts with better mechanical proprieties that can be used in such high stress applications we have to use high melting point metal powders, this is only possible only via Liquid Phase Sintering (LPS) CITATION Mon02 l 1031 13. The LPS process works by melting of the low melting components, wetting, rearrangement, and densification CITATION Nar08 l 1031 18.
Research for this process has been extended in direct SLS of metals such as bronze-Ni, Fe-Cu, WC-Co, TiC–Ni/Co/Mo, and TiCN–Ni. Applications for the parts made using LPS include stamping dies, deep drawing dies, and cutting tools CITATION Nar08 l 1031 18 but with relative small success.
Investigation in direct SLS of WC-Co powdersDirect selective laser sintering of polymer powders is very well mastered even since the introduction of the first commercial SLS machines. They have a wide range of applications mainly in the rapid prototyping industry. In the case of metal powders, as I mentioned above, the first SLS trials were made using single-component metal powders with no success due the “balling” effect, thus the approach with two-component metal powders and liquid phase sintering was employed. In this approach, the first metal used has a low melting temperature, acting as the binder phase, with the second metal having a high melting temperature and acting as the hard phase. For the tungsten carbide (WC) and cobalt (Co) system, the base of this Ph.D. research, the direct SLS method combined with liquid phase sintering (LPS) seems a promising and viable approach. In this system, WC is the high melting point metal (Tm=2870°C) and Co is the low melting point metal (Tm=1496°C).
J. P. Kruth et al. have conducted researches in direct laser sintering of WC-Co powders since the late 90s through to early 2000s at the Catholic University of Leuven, Belgium. In the research paper CITATION XCW02 l 1031 19 a special SLS machine with an pulsed Nd:YAG laser system was used to direct laser sinter a powder mixture of 91 wt% WC and 9 wt% Co into a 3D part used as a mold insert. The part was sintered with a laser power of 8 W at a 10 mm/s laser scan speed and a layer thickness of 0.2 mm. In this case, surprising is the relative low laser power (8W) used to melt the cobalt powder. The authors of CITATION XCW02 l 1031 19 have attributed this to the high degree of heat absorption of Co compared with other metallic binder, cooper (Cu), used in Fe-Cu powder systems. Also regarding the energy absorbed in the powder bed, it was observed that almost all (?96 %) is concentrated within a depth of only 0.4 mm. The SLS process must be very well controlled to avoid part curling.
In figure 21 is represented the mold insert WC-Co part made through direct selective laser sintering. Although the authors of CITATION XCW02 l 1031 19 present the successful use of this WC-Co part as a mold insert in the injection molding of several plastic parts, the resulted density was only between 37% and 40%. Corresponding to the density of the loose powder. The motivation for the low density obtained, presented in the paper CITATION XCW02 l 1031 19, was that although during the SLS process the liquid phase sintering (LPS) mechanism was applied not all its stages are present. Only the first stage called rearrangement stage is present during the SLS process, were the liquid formed due to the melting of the binder is filling up the existing pores in the loose powder CITATION XCW02 l 1031 19. The second stage, solution precipitation and the third stage, solid state sintering are based on the migration of atoms and those densification steps require longer sintering times CITATION XCW02 l 1031 19. Typical sintering time during the SLS process is from 1 ms to 0.1 s CITATION XCW02 l 1031 19. We can conclude that the SLS process is to short and the LPS stages are not fulfilled, thus a proper densification is not possible.
Due to the low density, further post-processing stages are necessary. The first post-processing method tried in the paper CITATION XCW02 l 1031 19 was to heat the SLS parts up to 1496 °C, the melting point of Co, to allow the liquid phase mechanism to complete its phases. This did not lead to satisfactory results; neither the density nor the strength of the part was improved. The second method was to infiltrate the part with cooper. Full density was obtained but the mechanical properties and the hardness of this cooper infiltrated part are considerably inferior to that of classic sintered cemented carbide.
Selective laser sintered WC-Co mold insert part CITATION XCW02 l 1031 19The same method of direct laser sintering was used by Subrata Kumar Ghosh et al. in the paper CITATION Sub15 l 1031 20 to sinter two different powder mixtures, first having 85 wt% WC + 15 wt% Co and the second one with 80 wt% WC + 20 wt% Co, using also as J. P. Kruth CITATION XCW02 l 1031 19 a pulsed Nd:YAG laser.
The focus of this research was in studying the influence of SLS process parameters such as layer thickness, hatching distance, pulse energy, pulse width, distance focal plane and powder composition on the resulting part properties such as density, microhardness and porosity. To isolate the influence of each parameter on the part properties the Taguchi method was used. The laser sintering tests were made on a special built SLS machine composed from a laser source, an inert gas chamber, filled with Argon, and a numerical controlled machine table.
In total 18 tests were made on the two powder mixture with three different values for each SLS process parameter mentioned before. After analyzing the results, the authors of CITATION Sub15 l 1031 20 presented the following conclusions:
Density. The parameter with the biggest influence (50%) on the part density is the hatch distance. A small hatch distance provides best results. Other parameters that have an influence on the part density are the layer thickness and the powder composition. Using a layer thickness as small as possible, not smaller than the powder grain size, and a relative high volume of metal binder, in this case Co, with very fine powder grains will provide the best results regarding the part density. A part density of 12.42 g/cc is predicted by the authors of CITATION Sub15 l 1031 20.
Microhardness. Similar as in the case of cemented carbide obtained through the classic sintering process, the hardness of the resulted part is direct proportional with the percentage of the hard phase (WC) present in the composition. A high concentration of hard phase will result in a high part hardness.
Porosity. It is influenced by the same process parameters as the density, because porosity and density are linked with one and other. Hatching distance having the biggest influence, 50%, from all parameters with the powder composition in the second place with 15.35%. A good representation of the porosity obtained during the test in the paper CITATION Sub15 l 1031 20 can be seen in the figure 22 along with some structural micro cracks.
SEM imagines with the pores and micro cracks present in the SLS parts CITATION Sub15 l 1031 20ConclusionsAlthough the direct laser sintering of two component metal powders provides better results compared to the single component powders, the resulting parts do not achieve full density, offer inferior mechanical properties and their application range is limited. Most of the research made on this laser sintering method was made during the early 2000s and since then a decrease can be observed due to the introduction of selective laser melting machines (SLM). This method is studied in the next pages.
SLM of Hardmetal powdersPrinciple of direct SLMSelective laser melting (SLM) or short laser melting (LM) was developed with two main objectives in mind: to produce fully dense components with mechanical properties comparable to those of bulk materials, and by the desire to avoid time-consuming postprocessing cycles CITATION 10 l 1031 21. Selective laser melting shares the same working principle with the laser sintering process with the only difference being that in the case of metallic powders the process is based on complete melting/solidification mechanism CITATION 10 l 1031 21.
Due to advances in the laser processing conditions, since 2000, higher laser power smaller focused spot size, smaller layer thickness lead to significant improvements in microstructural and mechanical properties related to earlier laser sintering processed parts. High speed steal and nonferrous metals (Ti, Al, Cu) show very good results regarding the densification rate, surface smoothness and microstructural homogeneity compared with the laser sintering processed parts. The laser melting of pure metal is highly controllable and offers densities up to 99.5 % through the full melting mechanism CITATION 10 l 1031 21.
The disadvantages of this method are that liquid melting requires high laser energy level, normally realized by applying good beam quality, high laser power, and thin powder layer thickness translating in long building time and increased risk of instability of the molten pool. Another one is the large degree of shrinkage that tends to occur during liquid/solid transformation, accumulating considerable stresses in LM-processed parts. The residual stresses arising during cooling are regarded as key factors responsible for the distortion and even delamination of the final products CITATION 10 l 1031 21. Reasonable selection of both laser processing and powder depositing parameters to determine a suitable process window, in order to yield a moderate temperature field to avoid the overheating of the LM system.
Investigation in SLM of WC-Co powdersSelective laser melting (SLM) is the technology that continues the research in additive manufacturing of metal powders from the unsuccessful method of direct laser sintering in manufacturing metal parts or components applicable in the industry. It works using the same principle as of direct laser sintering, building a part from a 3D CAD model layer by layer with the help of a laser that melts a selected path in the powder bed. SLM employs a high power laser focused in a small spot, combined with a small layer thickness and a relative low scanning speed it can achieve the complete powder melting and after cooling, the solidification of the processed material. In the case of metal powders, this mechanism of melting/solidification can theoretically provide full density and mechanical properties comparable with that of bulk materials.
In the case of tungsten carbide-cobalt (WC-Co) powder researches have been made regarding the possibility of manufacturing cemented carbide through the SLM method. Eckart Uhlmann et al. made such a research in the paper CITATION Eck15 l 1031 22 at the Fraunhofer Institute for Production Systems and Design Technology. The motivation for this investigation is to provide an alternative manufacturing method for cemented carbide that is more economical and more flexible when fabricating complex geometrical shapes, especially in the case of cutting tools with complex cooling channels.
The research CITATION Eck15 l 1031 22 was conducted using a SLM 250 machine, on a single powder composition of 83 wt% tungsten carbide (WC) and 17 wt% cobalt (Co). The aim is to determine the influence of the SLM process parameters such as focus position, laser power, scan speed, scan line spacing and layer thickness, on the relative density and chemical composition, mainly Co content of the cemented carbide parts. The high Co content in the composition is to maximize the density and minimize the cracking but this translates in limiting the application range of this method. The cubic shaped test parts were made using a high laser energy density in the internal, core area of the parts and with a low laser density in the external, contour area of the parts.
The areas fabricated using the high laser energy density (Ev = 1667 J/mm3) have attained relative densities larger than 95 %, but looking closely at the material structure some unwanted phenomenons are observed.
In the area were the laser acts directly on the powder bed molten material pools are formed, very benefic for high density, but due to the high temperature 2900 °C, the cobalt starts to evaporate from the composition. In this areas a content as low as 1 % of Cobalt is observed and combined with the high cooling rate of the molten pools a fine grained acicular solidified structure is formed resulting in a strong embrittlement of the material. The areas not directly affected by the laser beam present a similar cobalt content with the starting value of 17 % and an unchanged grain structure is present. The resulted composite material presents an inhomogeneous layered structure.
Representation of inhomogeneous laser molten carbide structure CITATION Eck15 l 1031 22The areas fabricated using the low laser energy density (Ev = 185 J/mm3) attained the original cobalt content, the grain structure is unchanged and homogenous, the cracks are present mostly between the pores but the resulted residual porosity is high with density around 50, 60 %.
Carbide structure fabricated with low energy density CITATION Eck15 l 1031 22The authors of CITATION Eck15 l 1031 22 have observed a clear incompatibility between the relative density and Co content of the laser melting manufactured parts. Their optimization directions are opposed to each other, meaning that either a high relative density can be obtained using a large laser energy input or a high cobalt content can be maintained in the structure by a low laser energy input.
With the purpose of avoiding this, the relative density and cobalt content incompatibility in the laser melting fabrication of WC-Co powder the authors of CITATION RSK16 l 1031 23 propose a different approach with low content of hard phase. R. S. Khmyrov et al. used commercially available micron-size powder (1-2µm) of Co and nano-size (50-80nm) powder of WC to prepare non-standard mixtures of 75 wt% Co – 25 wt% WC and 50 wt% Co – 50 wt% WC. After a uniform distribution obtained in the ball mill, a water suspension of the obtained mixture was prepared. This suspension was deposited on a substrate surface made from 5 mm thick cemented carbide (BK20) plates. The SLM process parameters that provided the best results were laser power 50 W, layer thickness 100 µm, scanning speed of 100 mm/s and a hatching distance of 50 µm. All the samples were analysed on optical and electron microscopy, with the chemical composition determined by energy-dispersive X-ray spectroscopy (EDS).
The authors of his paper CITATION RSK16 l 1031 23 show that the powder mixture with 75 wt% Co and 25 wt% WC present no cracks in the layer structure, having relative low porosity (?10 %) and after the chemical analysis the Co content is similar with the initial powder mixture. Furthermore, the lack of cracks in the structure is credited to the presents of ?-Co-W solid solution. The submicron grain size, the monophase composition and the well separated pores which are smaller than 10 µm in diameter are the advantages of this powder mixture. In the case of 50 wt% Co and 50 wt% powder mixture although the porosity is the same, less than 10 %, the layer structure presents a relatively high degree of cracks. The cracks are attributed to the presents of W3Co3C phase. The ternary carbide W3Co3C known as the ?-phase is very brittle thus making it undesirable in hard alloys. The cracks considerably reduce the mechanical strength of the parts and furthermore the authors do not recommend this mixture for further SLM research.
ConclusionsStudying the above research papers CITATION Eck15 l 1031 22 CITATION RSK16 l 1031 23, that are relatively new 2015-2016 and take into account the work done up until this point by other researchers in the same field of laser melting with the purpose of improving the method, we can conclude that no set of SLM process parameters or powder mixture in the case of WC-Co system can provide adequate results.Regardless of using high or low energy density on the same powder composition a compromise must be made between the relative part density and the cobalt (Co) content present in the structure. In addition, an inhomogeneous layered structure results due to the different way the laser affects the powder bed. The powder composition (25wt%WC-75wt%Co) presented in the paper CITATION RSK16 l 1031 23 has provided good results regarding the relative density, the grain structure, the Co content also with the absents of structural cracks but parts from such composite material provide limited application especially in the case of the cutting tool industry.
3D Printed Hardmetal: Binder JettingThe range of applicability for hardmetals is very wide, starting from micro wear parts in electronic assemblies, complex molds and dies to cutting tools for machining industry, making them suitable for 3D printing. According to CITATION Bro15 l 1031 7, report being made in 2015, the author points out that a large amount of research was done in the flied of 3D printing and the results do not meet the adequate requirements for the industries, just general statements and a few optimistic patents. The author also points out that not one single hardmetal manufacturer has a practical applied solution for 3D printing hardmetal or at least hardmetal powder suitable for 3D printing.
In September 2016 at the World Powder Metallurgy congress in Hamburg the Fraunhofer Institute for Ceramic Technologies and Systems (IKTS), Dresden, Germany reports that it has successfully 3D printed a range of hardmetal tools through the method of binder jetting.
Principle of binder jettingBinder jetting was developed in the early 1990s at the Massachusetts Institute of Technology (MIT) and the initial term was three-dimensional printing (3DP) CITATION Ian15 l 1031 24. Since then the companies that received the license from MIT have commercial application and are using material as polymers, metals and ceramic.
The working principle of binder jetting is similar to the principle of powder bed fusion processes (SLS, SLM). The part is constructed layer by layer from a powder bed that is lowered by a piston with the layer thickness. The difference being that for binder jetting the binder is not mixt or coated with the powder particles, rather is printed into the powder bed through a printing head CITATION Ian15 l 1031 24. The part is finished when all the layers are completed and the piston reaches the lowest position equal with the height of the part.
Typical binder droplets have 80µm in diameter and form spherical agglomerates of binder liquid and powder particles as well as provide bonding to the previously printed layer CITATION Ian15 l 1031 24. Applications for binder jetting process start with parts made from polymer powders that are typically used for prototypes, visual interpretation and light-duty functional prototypes. Parts made with polymer powders and wax based binders are used as patterns for investment casting 24. Metal powder based parts have a larger degree of applicability starting from prototypes up to production parts. In this case, Fraunhofer IKTS has presented for the first time 3D printed tools, e.g. wire dies, from cemented carbide that have similar mechanical and chemical properties with classic sintering process.
Schematic representation of the binder jetting process CITATION Ian15 l 1031 243D printed hardmetal tools
managed the development of complex hardmetal tools via 3D printing in a quality that is in no way inferior to conventionally produced high-performance tools.
managed the development of complex hardmetal tools via 3D printing in a quality that is in no way inferior to conventionally produced high-performance tools.
The motivation for 3D printing hardmetal tools is obvious when taking a look at the conventional manufacturing methods. Uniaxial or cold isostatic dry pressing, extrusion and injection molding followed by green shaping are all limited and expensive methods when considering making complex geometries, such as helical or meandering cooling channels inside components CITATION Fra17 l 1031 25.
The scientists at the Fraunhofer Institute have succeeded in producing complex hardmetal tools via a binder jetting method. This innovation was presented in September 2016 at the World Powder Metallurgy congress in Hamburg. The powder bed is locally wetted with an organic binder by a print head, bounding together the tungsten carbide and cobalt powder. The resulted 3D printed green bodies are debinded and densified under conventional sintering conditions. The resulted harmetal shows a typical hardmetal structure with a density of one hundred percent and a homogeneous binder distribution. Due to this structure, the mechanical and chemical properties of the 3D printed harmetal are similar compared with the hardmetal obtained through conventional methods. Moreover, the scientists say that is possible to get a homogeneous cobalt distribution, thus achieving a comparable quality to conventionally produced high-performance cemented carbide-based tools CITATION Fra17 l 1031 25. As in the conventional method, the binder jetting method has the possibility of varying the metallic binder. Bending strength, fracture toughness and hardness can be adjusted individually. The prototypes manufactured at Fraunhofer IKTS have a binder content of 12 and 17% by weight CITATION Fra17 l 1031 25.
Schematic representation of hardmetal 3D printing; 1-Structure of WC-12Co; 2-Green body; 3-Sintered wire die U shaped; 4-Sintered wire die helical shaped; 5-Structure of WC-17Co CITATION Fra17 l 1031 25;
ConclusionsFollowing the study made on all the books and research papers in both of the fields involved in this Ph.D. research, cemented carbide powder metallurgy and additive manufacturing, it can be concluded that for successfully 3D printing cemented carbide green bodies all the methods and steps involved in this process have to be specifically designed only for this application. Starting with powder properties used in the cemented carbide composition, the polymer powder that acts as the binder in SLS process, the mixing of the powders to form the SLS powder bed, the SLS process parameters to fabricate the green parts, followed by post-processing operations to increase the green part density and finally the actual sintering process are all important steps in the process of manufacturing 3D printed cemented carbide parts. As analyzed in this report, each step has a big influence of the final result and the problem imposed by one of the steps cannot be solved by the next one.
The process of 3D printing cemented carbide parts starts with the right powder selection. The powder materials that make the composition must have spherical shaped particles, a uniform particle size distribution and the best green part densities obtained in the indirect SLS process are with polymer coated particles. In the indirect SLS, the process parameters must be set in accordance with the polymer properties. The part bed temperature must be set around the initial melting temperature for a semi-crystalline polymer and around the glass transition temperature for an amorphous polymer. In addition, for indirect SLS process, post-processing operations are absolute necessary because of the low density of the resulted green parts. The best results were provided when combining two post-processing operations pressure infiltration and warm isostatic pressing and also when using a slurry based SLS process.
Direct SLS and SLM have the advantage that they do not require post-processing operations but other specific problems for these processes appear. Parts with full density were not achieved, thermal crack are present in the sintered structures, and the porosity is relatively high. In addition, the resulted layered structures are inhomogeneous resulting in parts with inferior mechanical properties and limited application range. The scientists at the Fraunhofer IKTS institute present binder jetting as a successful additive manufacturing technique to produce 3D printed cemented carbide parts, with typical structure, a density of one hundred percent and a homogeneous binder distribution. This offers comparable mechanical and chemical properties to the hardmetal obtained through the conventional manufacturing process.
1 K. M. Andersson, Aqueous Processing of WC-Co Powders, Stockholm: Universitetsservice US AB, 2004.
2 M. P. Groover, FUNDAMENTALS OF MODERN MANUFACTURING: Materials,Processes and Systems, Lehigh University, United States of America: JOHN WILEY & SONS, INC., 2010.
3 V. K. SARIN, D. MARI and L. LLANES, COMPREHENSIVE HARD MATERIALS, Boston: Elsevier UK, 2014.
4 G. S. Upadhyaya, Cemented Tungsten Carbides: Production, Properties and Testing, Kanpur, India: Noyes Publications, 1998.
5 F. Cardarelli, Materials Handbook, A Concise Desktop Reference, 2nd Edition, London: Springer-Verlag, 2008.
6 A. G. a. J.-S. Hötter, Additive Manufacturing, 3D Printing for Prototyping and Manufacturing, Munich: Hanser Publishers, 2016.
7 K. J. Brookes, “3D-printing style additive manufacturing for commercial hardmetals,” Metal Powder Report, May/June 2015.
8 T. A. Osswald and G. Menges, Material Science of Polymers for Engineers, Munich: Carl Hanser Verlag, 2010.
9 M. Schmid and K. Wegener, “Additive Manufacturing: Polymers applicable for Laser Sintering (LS),” Elsevier, Nový Smokovec, 2016.
10 D. L. Bourell, T. J. Watt, D. K. Leigh and B. Fulcher, “Performance limitations in polymer laser sintering,” Elsevier, 2014.
11 C. Yan, Y. Shi and L. Hao, “Investigation into the Differences in the Selective Laser Sintering between Amorphous and Semi-crystalline Polymers,” Carl Hanser Verlag GmbH & Co. KG, 2011.
12 L. Lü, J. Fuh and Y.-s. Wong, Laser-Induced Materials and processes for Rapid Prototyping, Springer Science & Business Media, 2001.
13 M. M. A. Dewidar, Direct and Indirect Laser Sintering of Metals, Leeds: United Kingdom, 2002.
14 K. Shahzad, J. Deckers, Z. Zhang, J.-P. Kruth and J. Vleugels, “Additive manufacturing of zirconia parts by indirect selective laser sintering,” Elsevier, Leuven, 2013.
15 K. Shahzad, J. Deckers, J.-P. Kruth and J. Vleugels, “Additive manufacturing of alumina parts by indirect selective laser sintering and post processing,” Elsevier, Leuven, 2013.
16 K. Shahzad, J. Deckers, J. Vleugels and J. Kruth, “Isostatic pressing assisted indirect selective laser sintering of alumina components,” Emerald , Leuven, 2011.
17 H.-H. Tang, M.-L. Chiu and H.-C. Yen, “Slurry-based selective laser sintering of polymer-coated ceramic powders to fabricate high strength alumina parts,” Elsevier, Taipei, 2011.
18 N. B. Dahotre and S. P. Harimkar, Laser fabrication and machining of Materials, Springer Science & Business Media, 2008.
19 X. C. Wang, T. Laoui, J. Bonse, J. P. Kruth, B. Lauwers and L. Froyen, “Direct Selective Laser Sintering of Hard Metal Powders: Experimental Study and Simulation,” Springer-Verlag, London, 2002.
20 S. K. Ghosh, A. K. Das and P. Saha, “A Case Study of Tungsten Carbide and Cobalt Powder Sintering by Pulsed Nd:YAG Laser,” Springer, India, 2015.
21 D. Gu, Laser Additive Manufacturing of High-Performance Marerials, Berlin: Springer-Verlag, 2015.
22 E. Uhlmann, A. Bergmann and W. Gridin, “Investigation on Additive Manufacturing of tungsten carbide-cobalt by Selective Laser Melting,” Elsevier, Berlin, 2015.
23 R. S. Khmyrov, V. A. Safronov and A. V. Gusarov, “Obtaining crack-free WC-Co alloys by selective laser melting,” Elsevier, Moscow, 2016.
24 I. Gibson, D. Rosen and B. Stucker, Additive Manufacturing Technologies, New York: Springer Science+Business Media, 2015.
25 “Fraunhofer 3D prints hardmetal tools,” Metal Powder report , January/February 2017.