Porous tungsten with controlled porosity by low temperature sintering

Abstract

Porous tungsten as a high current density cathode is one of the important applications of the metal, which is mostly used in high temperature conditions due to its exceptional resistance to melting (T _m = 3410±20°C). Its porous form has been a crucial component of dispenser cathodes used in electronic valves and high power lamps. Porous tungsten skeleton forms the matrix, which is then impregnated with an electron emissive compound. Upon every emission from the surface, new material has to be fed into the surface pores via the open pore channels. Hence it may be proposed that a uniform porosity is needed for a better performance. However, a controlled porosity has not been achieved yet. Moreover, sintering of tungsten has always been difficult due to the extreme process conditions. A high sintering temperature (T _s≥2000°C) and a strong reductive atmosphere (hydrogen) have been the absolute necessity in making these parts. This study further explores an alternative sintering technique being developed. The idea is based on the reactive sintering concept. The energy output from the exothermic reactive system of tungsten oxide and aluminium has been the heat source for sintering porous tungsten. As a result, sintering temperature and time have been reduced considerably. Higher homogeneity, thus more uniform pore distribution, was observed. A better control of porosity related to the pressing and sintering conditions was achieved by the characterisation method previously developed. Microhardness has been a useful monitor of the scatter in porosity of the parts. Throughout the study, SEM was used to observe the porous structures and powder morphologies. DSC and XRD were useful to follow the microstructural evolution in the reactive system.     

Effect of simulated thermal cycles on the microstructure of the heat-affected zone in HSLA-80 and HSLA-100 steel plates

The influence of weld thermal simulation on the transformation kinetics and heat-affected zone (HAZ) microstructure of two high-strength low-alloy (HSLA) steels, HSLA-80 and HSLA-100, has been investigated. Heat inputs of 10 kJ/cm (fast cooling) and 40 kJ/cm (slow cooling) were used to generate single-pass thermal cycles with peak temperatures in the range of 750 °C to 1400 °C. The prior-austenite grain size is found to grow rapidly beyond 1100 °C in both the steels, primarily with the dissolution of niobium carbonitride (Nb(CN)) precipitates. Dilatation studies on HSLA-80 steel indicate transformation start temperatures (T _s ) of 550 °C to 560 °C while cooling from a peak temperature (T _p ) of 1000 °C. Transmission electron microscopy studies show here the presence of accicular ferrite in the HAZ. The T _s value is lowered to 470 °C and below when cooled from a peak temperature of 1200 °C and beyond, with almost complete transformation to lath martensite. In HSLA-100 steel, the T _s value for accicular ferrite is found to be 470 °C to 490 °C when cooled from a peak temperature of 1000 °C, but is lowered below 450 °C when cooled from 1200 °C and beyond, with correspondingly higher austenite grain sizes. The transformation kinetics appears to be relatively faster in the fine-grained austenite than in the coarse-grained austenite, where the niobium is in complete solid solution. A mixed microstructure consisting of accicular ferrite and lath martensite is observed for practically all HAZ treatments. The coarse-grained HAZ (CGHAZ) of HSLA-80 steel shows a higher volume fraction of lath martensite in the final microstructure and is harder than the CGHAZ of HSLA-100 steel.     

Effect of simulated thermal cycles on the microstructure of the heat-affected zone in HSLA-80 and HSLA-100 steel plates

The influence of weld thermal simulation on the transformation kinetics and heat-affected zone (HAZ) microstructure of two high-strength low-alloy (HSLA) steels, HSLA-80 and HSLA-100, has been investigated. Heat inputs of 10 kJ/cm (fast cooling) and 40 kJ/cm (slow cooling) were used to generate single-pass thermal cycles with peak temperatures in the range of 750 °C to 1400 °C. The prior-austenite grain size is found to grow rapidly beyond 1100 °C in both the steels, primarily with the dissolution of niobium carbonitride (Nb(CN)) precipitates. Dilatation studies on HSLA-80 steel indicate transformation start temperatures (T _s ) of 550 °C to 560 °C while cooling from a peak temperature (T _p ) of 1000 °C. Transmission electron microscopy studies show here the presence of accicular ferrite in the HAZ. The T _s value is lowered to 470 °C and below when cooled from a peak temperature of 1200 °C and beyond, with almost complete transformation to lath martensite. In HSLA-100 steel, the T _s value for accicular ferrite is found to be 470 °C to 490 °C when cooled from a peak temperature of 1000 °C, but is lowered below 450 °C when cooled from 1200 °C and beyond, with correspondingly higher austenite grain sizes. The transformation kinetics appears to be relatively faster in the fine-grained austenite than in the coarse-grained austenite, where the niobium is in complete solid solution. A mixed microstructure consisting of accicular ferrite and lath martensite is observed for practically all HAZ treatments. The coarse-grained HAZ (CGHAZ) of HSLA-80 steel shows a higher volume fraction of lath martensite in the final microstructure and is harder than the CGHAZ of HSLA-100 steel.     

Reactive sintering of porous tungsten for higher homogeneity

Abstract

One of the most important applications of tungsten is its use as a high current density cathode in the form of porous tungsten impregnated with electron emissive salts. Powder metallurgy is the usual processing route to make these parts since melting and casting of tungsten is very tedious due to its high melting point (3410±20°C). The overall porosity and porosity distribution are crucial parameters for the performance and lifetime of porous tungsten cathodes. Although there is an active debate over what constitutes the optimal porosity distribution, lack of homogeneity is observed in conventional parts. This may contribute to a shorter cathode and hence lamp lifetime. These parts are conventionally sintered at temperatures in excess of 2000°C. Furthermore, due to the high temperatures involved, the conventional sintering is very costly and energy consuming. This paper briefly looks into an alternative sintering process being developed for porous tungsten technology and its outcomes in terms of porosity distribution across the parts. As a result, the sintering temperature is dropped down to 1150°C and more homogeneous porous structures have been obtained. A characterisation method previously developed by the authors based on microhardness measurements is proved to be a good measure of the homogeneity of porous tungsten parts.     

Fabrication of in situ TiB2–TiC reinforced steel matrix composites by spark plasma sintering

Abstract

In situ TiB₂ and TiC particulates reinforced steel matrix composites have been fabricated using cheap ferrotitanium and boron carbide powders by spark plasma sintering (SPS) technique. The sintering behaviour and the formation mechanism of the composite were studied. The results show that when the composite was sintered at 1050°C for 5 min, the maximum relative density and hardness of the composite are 99·2% and 83·8 HRA respectively. The phase evolution of the composite during sintering indicates that the TiB₂ and TiC reinforcements were formed in situ as follows: first, the solid/solid interface reaction between Fe₂Ti and B₄C, resulting in the formation of a small amount of TiB₂ and TiC below 950°C; second, the solid–liquid solution precipitation reaction in the Fe–Ti–B–C system, resulting in the formation of the main TiB₂ and TiC reinforcements at ～1000°C.     

Thermomechanical fatigue of particulate-reinforced aluminum 2xxx-T4

Thermomechanical fatigue (TMF) and isothermal fatigue of unreinforced and SiC_p-reinforced aluminum 2xxx-T4 alloy were examined. Thermomechanical fatigue experiments were conducted underT _min = 100 °C,T _max = 300 °C andT _min = 100 °C,T _max = 200 °C conditions, and isothermal experiments were conducted at 200 °C and 300 °C. Based on stress range, substantial improvements in fatigue life were observed with reinforcement under both isothermal and thermomechanical loading conditions. Based on strain range, the TMF lives of the reinforced material increased in out-of-phase (OP) loading and remained unchanged in in-phase (IP) loading. A decrease in isothermal fatigue lives of the reinforced material compared to those of unreinforced material was observed in both 3 × 10⁻³ s⁻¹ and 3 × 10⁻⁵ s⁻¹ experiments at 200 °C and in 3 × 10⁻³ s⁻¹ experiments at 300 °C. Crack growth mechanism maps were constructed to identify crack growth behavior of the unreinforced and the reinforced materials. The TMF OP conditions were more favorable to transgranular cracking. Mixed (transgranular and intergranular) crack growth occurred in TMF IP experiments. Evidence of void formation at grain boundaries, crack deflection due to particles, and oxide penetration at the crack tips is demonstrated using scanning electron microscopy (SEM) and Auger spectroscopy analysis.     

Kinetics of Densification and Grain Growth of Pure Tungsten During Spark Plasma Sintering

The kinetics of densification and grain growth of tungsten during spark plasma sintering (SPS) was studied under isothermal conditions. The results show that using SPS, high-density (>97?pct) pure tungsten can be produced without the addition of sintering aids. The estimated sintering exponent (m?=?0.4 ± 0.03) suggests that the rate-controlling mechanism of sintering is diffusion along the grain contacts into the interparticles neck region. The activation energy of tungsten self-diffusion was calculated (Q?=?277?±?15?kJ/mol) in the temperature range 1523?K to 1773?K (1250?°C to 1500?°C). The activation energy is smaller than the values in previous studies using conventional sintering. This suggests that there may be some differences in the sintering conditions and mechanisms during SPS processing compared to conventional sintering. Grain-growth kinetics was studied in the range 1873?K to 2073?K (1600?°C to 1800?°C) and classified as normal grain growth according to the estimated grain-growth exponent (m?=?2?±?0.2). The grain-growth activation energy was calculated as 231?±?15?kJ/mol.     

Thermodynamic Properties of Dilute Aluminum in Liquid Zinc

Abstract

The thermodynamic properties of the liquid zinc-aluminum system in the range of 0 to 0.36 mole fraction of aluminum and over a temperature range of 430 to 530 °C were investigated by using an electromotive force (EMF) cell of the type

Al(s)|LiCl–KCl–AlCl₃|Zn–Al(l),Mo

The activity of aluminum was found to show a significant positive deviation from Raoultian behaviour. The activity coefficient of aluminum in the dilute solution range (γ°) and the free energy of solution of aluminum based on the one weight percent (1 wt %) standard state were determined as a function of temperature and found to follow the relationships:

logγ°=(1195/T)?0.874

ΔG_s=5457?11.4T (cal/mole)

Over the temperature range investigated, the self-interaction coefficient of aluminum in liquid zinc (e^Al_Al) was found to be approximately constant at ?0.01.

On a étudié les propriétés thermodynamiques du système liquide zinc-aluminium, entre 0 et 0.36 fraction molaire d'aluminium, à des températures de 430 à 530 °C, en utilisant une cellule à force électromotrice (FEM) du type:

Al(s)|LiCl–KCl–AlCl₃|Zn–Al(l),Mo

On a trouve que l'factivite de l'faluminium montrait une deviation positive importante du comportement Raoultien. On a determine le coefficient d'factivite, γ°, de l'faluminium pour une solution diluee ainsi que l'fenergie libre de solution de l'faluminium, base sur l'fetat standard de un pour-cent en poids (1%), en fonction de la temperature. On a trouve que ceux-ci suivaient les relations suivantes:

logγ°=(1195/T)?0.874

ΔG_s=5457?11.4T (cal/mole)

Dans le domaine de température étudié, on a trouvé que le coefficient d'auto-interaction de l'aluminium dans le zinc liquide (e_Al^Al) était approximativement constant à ?0.01.     

The solubility of tungsten in molten iron

The solubility of tungsten in iron has been determined at 1560–1620°C by successively saturating molten iron with tungsten while ensuring a uniform distribution of the content over the height. The temperature dependence of the solubility over that temperature range is described by a Schröder equation: C_s=(3.94±0.11)×10³e-(84.4±0.4)/RT.     

A PRELIMINARY STUDY OF THE REACTIONS BETWEEN TUNGSTEN CARBIDE AND TUNGSTEN CARBIDE - 6% COBALT POWDERS AND WET AND DRY HYDROGEN

Abstract

The apparent activation energy for the following reactions has been determined by measuring the amount of carbon removed from samples of powder heated to various temperatures in an atmosphere of either dry or wet hydrogen:

Dry Hydrogen

“As-carburized” tungsten carbide: no reaction up to 900°C.

Water–milled tungsten carbide: 8·38 kcal/mole over the range 600–900°C.

Water–milled tungsten carbide+ 6% cobalt: complex reaction over the range 600–900°C.

Wet Hydrogen

“As-carburized” tungsten carbide: 58·1 kcal/mole over the range 800–900°C.

Water–milled tungsten carbide: 34·8 kcal/mole over the range 700–900°C.

Water–milled tungsten carbide+ 6% cobalt: 38·4 kcal/mole over the range 825–900°C; 7·38 kcal/mole over the range 600–800°C.

It has been shown that ball-milling in water changes both the physical and chemical properties of tungsten carbide powder. These changes are discussed in relation to the marked differences in reaction rates. A number of possible reasons for these differences are given.     

Hot forging of Ti–6Al–4V alloy preforms produced by spark plasma sintering of powders

Abstract

Powder preform forging is a technology that comprises the preparation of near net shape preforms through powder metallurgy and a subsequent hot forging in order to obtain the desired final shape. In this work, two Ti–6Al–4V powder preforms were sintered through spark plasma sintering (SPS) and then hot compressed in a horizontal dilatometer. Varying the temperature of the process, two full density preforms having different microstructures were produced: sintering at 950°C, a plate-like α was obtained, whereas sintering at 1050°C, an acicular α was obtained. The behaviour of the preforms under hot forging has been studied through hot compression tests carried out in a quenching and deformation dilatometer in a range of temperature and strain rates typically used in hot forging this alloy (850–1050°C, 0·01–1 s^?1). Hot workability has been evaluated by measuring the stresses required for deformation and by analysing both the stress–strain curves recorded during testing and the microstructures after deformation. The main microstructural phenomena occurring during hot compression were individuated. The best conditions for the hot forging operation of SPS preform are temperatures above β transus, where the materials are deformed in a regime of dynamic recrystallisation, at every strain rate.     

Effect of cobalt addition on the liquid-phase sintering of W-Cu prepared by the fluidized bed reduction method

A new process, fluidized bed reduction (FBR) method, was applied for fabrication of uniform W-Cu sintered material. Liquid-phase sintering was carried out to obtain fully densified W-Cu composite, and the effect of cobalt addition on the sintering behavior was investigated. It was found that fully densified material could not be obtained even after sintering at 1200 °C for 4 hours in the case of 75W-25Cu, while more than 96 pct density could be obtained as soon as the sintering temperature reached 1200 °C when 0.5 wt pct cobalt was added prior to the sintering. It has been found that the wetting angle of the liquid copper is reduced significantly by the addition of cobalt, and the formation reaction of Co₇W₆ intermetallic compound at the surface of the tungsten powder is mainly responsible for the enhancement of the densification process.     

Fabrication of a multi-phase porous high-temperature Mo–Si–B alloy by in situ reaction synthesis

ABSTRACT

The pore formation process has been studied for a high-temperature porous Mo–Si–B alloy produced by in situ reaction synthesis from elemental starting powders. The expansions, phase transformations, microstructure and pore parameters were investigated systematically as a function of the sintering temperature. For sintering temperatures of 1200°C and above, a multiphase mixture of the oxidation-resistant intermetallic compounds Mo₃Si, Mo₅Si₃ and Mo₅SiB₂ was obtained. It was found that the porous alloys exhibit uniform skeletons with fine grains, high porosities and centred pore size distributions. Characterisation data obtained from samples sintered at temperatures from 1200 to 1600°C are used to deduce the processes which occur at each stage in the reaction synthesis. This helps to explain the formation mechanism for the pores and skeletons in this novel material.     

Production of NiAl–matrix composites by reactive sintering

Abstract

This work was devoted to the development of NiAl–matrix composite and its production by reactive sintering powder metallurgy. Various types of reinforcement (aluminium oxide, silicon and tungsten carbides, titanium silicide) were tested. The best chemical compatibility and the highest hardness and wear resistance were achieved by Al₂O₃ fibres. Electroless nickel plating pretreatment of Al₂O₃ fibres improves both distribution of fibres and hardness of the composite. However, it strongly reduces the wear resistance, probably due to phosphorus content in the nickel coating. In situ formation of NiAl–Al₂O₃ composites by reactive sintering of a pressed powder mixture of Ni, Al and NiO was unsuccessful. Only a small amount of cubic γ-Al₂O₃ was detected after reactive sintering and hence no significant hardness increase was observed.     

Sintering a disperse mixture of W-Sc2O3 powders

The effect of additions of scandium oxide (1, 5, and 10 vol. %) on compaction during sintering of a disperse mixture of W-Sc₂O₃ powder at 2000°C has been studied. It has been found that the scandium oxide particles activate compaction during heating and conversely retard shrinkage of tungsten in the stage of isothermal soaking. The kinetics of compaction during sintering is determined by the geometry of the heterophase system. Depending on how much the tungsten and scandium oxide particles increase in size, the nature of thestructure changes markedly from matrix-statistical to statistical.     

Activated sintering of W-HfC composite materials

The features of consolidation of the particles during the activated sintering of tungsten powders with different values of dispersity (d _av = 2–3 and 0.8–1.0 μm) are investigated. Sintering was activated by introducing nickel additives (up to 0.5 wt %), tungsten nanoparticles (up to 30 wt %), and finely dispersed hafnium carbide (5–30 vol %) with subsequent milling in a vibrating mill. The uniaxial compaction of the samples has been performed under pressures from 50 to 1000 MPa, and sintering was performed in vacuum at 1850°C with holding for 1 h. It is shown that the additives of tungsten carbide increase the density of sintered billets and, in combination with dispersed hafnium carbide, tungsten-based composite materials with a grain size up to 2 μm can be obtained.     

Sintering of Fe–C–BN steel compacts in different atmospheres studied by dilatometry and DTA/TG

Abstract

In the present work, 2%hBN was admixed with Fe–0·8C, and both dilatometric and differential thermal analysis/thermogravimetry investigations were conducted in Ar and N₂ atmospheres, followed by microstructural studies and mechanical testing. The α–γ phase transformation in both atmospheres was found to occur in the temperature range of plain iron and not, as expected, in that common for Fe–C. In the Ar atmosphere, liquid phase formation is recognised by endothermic differential thermal analysis signal and shrinkage in the dilatometer during the heating stage at ～1275°C. In contrast to the activating effect of the inert Ar for the decomposition of hBN, the deactivating effect of the N₂ atmosphere is visible from the dilatometry results: sintering in N₂ even resulted in slight expansion during the isothermal stage. Therefore, the, at first surprising, conclusion can be drawn that the chemically inert Ar is activating the sintering process while the more reactive N₂ passivates it.     

Thermodynamic properties of Ni-S melts between 700° and 1100°c

The vapor pressure of sulfur over Ni-S melts of various compositions was calculated from the equilibrium weight of the melt in gas streams of known H₂S-H₂ composition. The Gibbs-Duhem equation was used to calculate the activity of nickel and other thermodynamic properties. For the reaction: 3Ni(S) + S₂(g) ⇌ Ni₃S₂(l) the suggested free energy rslationship is: ΔG° = -57,910 + 15.89T (800° to 1100°C). The Calculations were extrapolated to predict that for the reaction: Ni(s) + 1/2S₂(g) ⇌ NiS(l), ΔG° = -26.730 + 10.5T (1000° to 1100°C)     

The relative effects of chromium,molybdenum, and tungsten on the occurrence of σ phase in Ni-Co-Cr alloys

The relative effects of chromium, molybdenum, and tungsten on the occurrence of σ phase have been studied in Ni-Co-Cr alloys. These alloys were designed to simulate the γ matrix in commercial nickel-base superalloys that are strengthened primarily by precipitation of the γ phase, based on Ni₃Al. Three alloy series were studied. The first series comprised four alloys varying in chromium content from 34.63 to 43.65 at. pct. The other two series contained separate molybdenum and tungsten additions of 1, 2, 3, and 4 at. pct at constant chromium contents of 37.5 at. pct. In each of the 12 alloys, the atomic percentages of nickel and cobalt were equal. The alloys were aged in both the annealed and cold-rolled conditions at 1400°F (760°C), 1550°F (845°C), and 1700°F (925°C) for times up to 3000 h. The contributions of the chromium-group elements to σ formation were evaluated both by measuring the volume percentage of σ phase and by determining the final composition of the y matrix after σ precipitation. By these two techniques, critical values of the average electron vacancy number, •N _v , for σ formation at 1550°F (845°C) were found to be 2.518 and 2.512, respectively; σ precipitation was most rapid at 1550°F (845°C). Both techniques in-dicated that under conditions approaching equilibrium, molybdenum and tungsten are equiv-alent in inducing σ formation and about 1.5 to 2 times as potent as chromium. The approxi-mate electron vacancy coefficients(N _v ) for molybdenum and tungsten, as derived from volume-fraction measurements of σ phase, are as follows: 7.35 at 1400°F (760°C) and 1550°F (845°C), and 8.7 at 1700°F (925°C). The values derived from final compositions of the γ matrix after σ precipitation are 7.9 at 1550°F (845°C) and 8.6 at 1700°F (925°C). The bulk diffusion of aluminum into alloys that were otherwise not σ-prone at 1700°F (925°C) caused extensive σ precipitation during aging. This was due to copious precipitation of γ-Ni₃Al and β-NiAl, resulting in enrichment of the matrix in elements of the chromium group. This paper is based on a dissertation submitted by GARY N. KIRBY in partial fulfillment of the requirements for the degree of Doctor of Philosophy, Metallurgical Engineering, The University of Michigan, 1971. The study was conducted in the Ann Arbor Research Labora-tory of the Climax Molybdenum Company of Michigan, a subsidiary of American Metal Climax, Inc.     

High temperature work softening yield point phenomena in polycrystalline aluminum

Work softening yield points have been studied in high purity polycrystalline aluminum tested in the temperature rangeT ₂ = 100°to 450°C at constant strain rate subsequent to prestraining in tension at a lower temperature (T1) = room temperature or 350° C) or higher strain rate. The yield drops increase with prestrain as well as with the difference between the flow stresses at the test and prestrain temperature. Reductions in the yield drop caused by annealing atT ₂ prior to testing indicate that the yield drops are the result of dynamic recovery. It is shown that dynamic recovery is accelerated by concurrent straining.