- Email: [email protected]
Effect of vacuum arc remelting and processing parameters on structure and properties of high purity niobium
Int. Journal of Refractory Metals and Hard Materials 50 (2015) 120–125
Contents lists available at ScienceDirect
Int. Journal of Refractory Metals and Hard Materials journal homepage: www.elsevier.com/locate/IJRMHM
Effect of vacuum arc remelting and processing parameters on structure and properties of high purity niobium M. Sankar ⁎, V.V. Satya Prasad, R.G. Baligidad, A.A. Gokhale Defence Metallurgical Research Laboratory, Hyderabad 500058, Telangana, India
a r t i c l e
i n f o
Article history: Received 12 August 2014 Received in revised form 25 November 2014 Accepted 6 December 2014 Available online 9 December 2014 Keywords: Niobium Vacuum arc remelting Arc length Thermomechanical processing
a b s t r a c t Niobium and niobium alloys are normally produced by vacuum arc remelting (VAR) process. In the present work, the effect of various process parameters such as arc voltage, arc current (melt rate) and fill ratio on arc stability during VAR process and solidification structure of remelted niobium ingot is investigated. The effect of thermomechanical processing of VAR ingot on microstructure and mechanical properties is also investigated. It has been established that arc voltage is a function of arc length. For melting of niobium in 110 and 150 mm diameter crucibles, the optimum arc voltage is in the range of 34–37 V. It has been observed that the arc voltage needed for melting niobium is much higher than that used for melting stainless steels, reactive metals and superalloys. It has also been observed that for a given arc voltage, fill ratio has no significant influence on the melt rate of consumable electrode. However, arc current has significant influence on the melt rate and consequently on the solidification structure of the ingot. Thermomechanical processing of the VAR ingot is shown to result in significant increase in room temperature strength and ductility. However, the processing temperatures do not have any significant influence on the tensile properties. © 2014 Elsevier Ltd. All rights reserved.
1. Introduction Niobium exhibits a fascinating combination of attractive properties such as high melting point, low density, good fabricability, low ductile to brittle transition temperature and low thermal- neutron absorption cross section [1]. These properties make niobium a potential candidate material for applications in atomic power reactors, reentry vehicles, space power equipment etc. [2]. However, niobium suffers from poor resistance to high temperature oxidation, and active research is underway to overcome this limitation [3]. Historically, powder metallurgy methods involving high temperature vacuum sintering and carbon reduction have been extensively used to produce niobium metal [4]. However, these methods yielded niobium with higher oxygen contents which adversely affect its room temperature ductility, weldability and ductile-brittle transition temperature. Alumino-thermic reduction (ATR) and electron beam melting became the standard practice of producing high purity niobium since the early 1960s. Currently, industrial scale electron beam melting furnaces are available to produce niobium ingots of 300-500 mm diameter and over 2 m in length [5]. However, an electron beam melting process is mainly used only to refine impure niobium and has not been a very successful process for producing alloys based on niobium. This is because of the very high vacuum levels employed in the EB melting
⁎ Corresponding author. E-mail address: [email protected] (M. Sankar).
http://dx.doi.org/10.1016/j.ijrmhm.2014.12.001 0263-4368/© 2014 Elsevier Ltd. All rights reserved.
process which, while refining Nb may result in undue losses of alloying elements. Vacuum arc remelting (VAR) process enables production of alloys based on reactive and refractory metals. Many of the commercial alloys based on niobium have been produced by vacuum arc remelting. The high temperature encountered in electric arc melting, the refining associated with protective vacuum cover, use of refractory free water cooled crucible and solidification in a water cooled mould enable production of high purity ingots with a desirable cast structure [6]. The success of the vacuum arc remelting process is essentially decided by the proper choice of the process parameters. The parameters which are crucial to maintain the melting process stable and to obtain good solidification structure are arc length, fill ratio and melt rate. However, the information available in literature regarding vacuum arc remelting of niobium and the effects of various process parameters on the quality of niobium ingot is inadequate. In general, niobium is subjected to thermomechanical processing by hot forging followed by cold working. Hot forging temperature of commercial niobium is in the range of 650–950 °C. However, heating to such a high temperature leads to excessive oxidation of niobium. Further, oxygen, nitrogen and hydrogen can be picked up during heating and hot forging which make niobium brittle at room temperature. Therefore, niobium is protected against oxidation during heating, hot forging or intermediate annealing by employing protective coatings or inert gas atmosphere or by providing stainless steel jacketing. There is no information available in the literature on the influence of (i) different protection methods during heating and (ii) different thermomechanical
M. Sankar et al. / Int. Journal of Refractory Metals and Hard Materials 50 (2015) 120–125
40
Ingot diameter (mm)
Average input power (kW)
Average melt rate (kg/min)
110
115 165 210 175 220 260
1.0 1.5 2.0 1.5 2.0 3.0
processing conditions on the structure and properties of VAR melted Nb ingot. In the present paper, an attempt has been made to bring out the effect of various process parameters such as arc input power and voltage on melt stability and solidification structure of niobium ingots. The effect of different thermomechanical processing conditions on the structure and mechanical properties of VAR melted high purity niobium is also discussed in this paper.
Arc voltage (volts)
Table 1 Niobium ingots produced under different melt rates.
150
121
35
30
25 5
10
15
20
Arc length (mm) Fig. 2. Plot showing variation of arc length with arc voltage during VAR of Nb under vacuum.
2. Experimental procedure A few VAR experiments were initially conducted in water cooled copper moulds of 60 to 150 mm diameter. Electron beam melted pure niobium rods were used as consumable electrodes. These experiments were conducted with different arc voltages with the following objectives: (i) To establish the relation between arc length and arc voltage and (ii) To optimize the arc length to produce and maintain stable arc.
Arc length is found to be an important process parameter during the VAR process. The distance between electrode bottom surface and top surface of the solidified ingot is taken as the arc length. Direct measurement of arc length is not possible during melting in the VAR process. Hence, this was measured during the VAR process by suddenly switching off the power and stopping the electrode feed. Then the electrode was moved downwards until it touched the ingot and the distance travelled by electrode was taken as arc gap. Another set of experiments was carried out in 110 and 150 mm diameter water cooled copper moulds with different fill ratios (the ratio of cross-sectional area of consumable electrode to the crosssectional area of the mould) and melt rates to study the effect on
soundness and solidification structure of the ingots. Another experiment was conducted in 110 mm diameter mould wherein the power was switched off and the arc extinguished twice at different intervals, each time for a period of 4-5 min. The process was restarted using higher arc current initially to study the effect of power interruption on soundness of the ingot. The relevant process parameters (such as voltage, current, and vacuum levels used) were continuously recorded during all the VAR melts. Two trials were performed for each set of experiments. Three pieces of 110 mm diameter and 470 mm length were cut transversely from the VAR ingot (110 mm diameter × 150 mm long) produced using the optimized process parameters. These pieces were subjected to different thermo-mechanical processing conditions as detailed in Table 1. The first ingot piece was coated with commercial grade (Deltaglaze FB 412) oxidation resistant coating and forged at 900 °C to 30 mm thickness. The second and the third ingot pieces were directly warm (250 °C) and cold (25 °C) forged to 30 mm thickness. Finally all the forged pieces were of 30 mm thickness. These forged pieces were stress relieved at 900 °C for 1 h and subsequently cold rolled to sheets of 2.0 mm thickness by using DEEMAG rolling mill. All the cold rolled samples were recrystallised at 1100 °C for 1 h. The as-cast VAR ingots were tested for internal soundness using gamma-ray radiography (5 Curie Co-60 radioactive source) with an
Titanium
Fig. 1. Plot showing variation of arc length with arc voltage during VAR of Ti and steel under vacuum [7].
122
M. Sankar et al. / Int. Journal of Refractory Metals and Hard Materials 50 (2015) 120–125
Fig. 3. Proposed configuration for (a) transient arc due to motion of molten metal and (b) drop short.
voltage (Fig. 1) [7]. In the case of refractory metals the relationship between arc length and arc voltage is not reported in the literature and knowledge of this relation is essential to control the arc during melting. Arc voltage as a function of arc length is plotted in Fig. 2. It is clearly seen that arc voltage has significant effect on arc length. Therefore it may be inferred that arc voltage can be used as a parameter to control arc length during consumable vacuum arc remelting. The arc length increases from 8 to 17 mm with increase in arc voltage from 28 to 37 V. Above 37 V arc gets extinguished. It is observed that arc is not stable even when the arc length is too short. This may be due to transient arc interruption because of multiple and simultaneous contacts between the electrode and the liquid metal pool due to the motion of the liquid metal pool. This phenomenon is schematically described in Fig. 3a. It has been reported that during VAR melting, the molten metal transfer occurs when metal spikes hanging from the electrode (cathode) form a low resistance bridge by touching the molten metal pool (anode)
7,000
a
6,000 5,000 Current, A
exposure factor of about 5 curie- h. The chemical composition (impurity elements) of the Nb electrode and vacuum arc remelted Nb ingots was determined. All the metallic impurities were analyzed by using inductively coupled plasma optical emission spectroscopy (ICPOES) method. The carbon and hydrogen contents were analyzed by CC-441 Leco carbon and sulphur analyser. Oxygen and nitrogen were analyzed by Leco-136 oxygen & nitrogen analyser. The longitudinal section samples of as-cast VAR Nb ingots were cut, machined, ground to 180 grit finish and then etched with an etchant comprising of 3 parts HF + 2 parts H2O2 + 4 parts H2O (by volume) for macrostructure studies. Samples of (a) cast + glass coated + hot forged + cold rolled + annealed, (b) cast + warm rolled + annealed, and (c) cast + cold rolled + annealed were mechanically polished to 0.5 μm finish and etched with an etchant consisting of HCl- 30 ml, HF-30 ml, HNO315 ml and water -10 ml for metallographic examination. Tensile specimens of 4.0 mm gauge diameter and 20 mm gauge length were prepared from as-cast VAR ingots of 60 and 110 mm diameter as well as from as cast VAR ingot of 110 mm diameter produced with arc interruption with gauge portion containing the interrupted region. Tensile specimens of 6 mm gauge width and 32 mm gauge length were also prepared from 2.0 mm thick section of cold rolled and annealed sheet. All the tensile tests were carried out at room temperature by using Instron 5500R Universal Testing Machine. The fracture surfaces of tensile tested samples were examined under a scanning electron microscope.
4,000 3,000
3. Results and discussions
2,000
In the case of Zr and Ti remelting, arc length is a function of arc voltage whereas in the case of steel and superalloys it is not a function of arc
1,000 0 0
100
200
300
400 Time, S
500
600
700
800
2.5 Ingot diameter: 150 mm
45
Melt current: 5.5-6.0KA
40
b
35 1.5
Voltage, V
Melt rate ( kg/min)
2
Ingot diameter: 110 mm Melt current: 4.5-5.0KA
1
30 25 20 15 10
0.5
5 0
0
0 0
0.1
0.2
0.3
0.4
0.5
100
200
300
400 Time, S
500
600
700
800
Fill ratio Fig. 4. Plot showing effect of fill ratio on melt rate.
Fig. 5. Plot showing current (a) and voltage (b) versus time traces during melting of niobium.
M. Sankar et al. / Int. Journal of Refractory Metals and Hard Materials 50 (2015) 120–125
123
Fig. 6. 110 mm diameter VAR Niobium ingot produced using 1.5 kg/h melt rate (a) as-solidified and (b) machined.
leading to its rupture. This phenomenon is called drop short. This is described schematically in Fig. 3b. At slightly higher arc length (12 to 15 mm), it is very difficult to maintain arc stability due to drop short even though the transient arc interruption due to motion of the liquid metal is totally prevented. This may be due to very high frequency of arc interruption because of drop short. At 15 to 17 mm arc length, drop short frequency is at a minimum level and the arc is stable. The variation of typical process parameters (voltage and current) as a function of time is shown in Fig. 4. While melting of 110 mm diameter mould at optimum arc length is 15-19 mm. The figure shows that the arc is burning at a voltage of 34-37 V and current of 4500–5000 A. Both arc voltage and current are stable throughout the melting process. Occurrence of drip-short has a unique electrical signature which allows them to be detected [8]. In the present work the electrical signature
(sudden voltage drop to 2 V and at the same time the current begins to increase up to 10 kA) corresponding to drip short could not be recorded. In the present work, the drip short phenomenon could be noticed during the melting process from the change in the frequency of sound resulting from the drip short process as well as from the change in the process current and voltage values. However, the voltage values could not be recorded and need for process control due to equipment limitation. Though the drip short can also used to control the arc gap, our objective in the present case has been limited to minimize the drip short frequency and maintain a stable arc. Arc voltage was used to control arc length. Based on these experiments it can be said that for melting of pure niobium in 60 to 150 mm diameter crucibles the optimum arc length is about 15 to 17 mm and to achieve this arc length, arc voltage of 34 to 37 V is required to be maintained. These arc voltages are much higher than those used for melting steels, reactive metals and superalloys. The fill ratio is also an important process parameter as it has direct influence on the arc stability. A low fill ratio produces more stable arc than does a high fill ratio because of the formation of a relatively centralized heat source in the former case even when the cathode spot wanders over the electrode tip. Smaller size electrode also facilitates rapid pumping away of gas from the mould. This helps in maintaining the partial pressure above the molten pool below a critical value. With higher fill ratio, the conductance path is reduced which raises the partial pressure above the molten pool. If the fill ratio is too high, the gap between the mould wall and the crucible becomes too small leading to arcing between the electrode and crucible wall resulting in crucible puncture. Fig. 5 shows the experimental results of 110 mm and 150 mm diameter
Table 2 Chemical analyses of Nb before and after VAR.
Fig. 7. Macrostructure showing grain orientation with increase in melt rate (or power) (a) 1.5 kg/h and (b) 2 kg/h.
Elements
Before VAR (EB melted stock) (wt.%)
After VAR (wt.%)
Fe Mo Ta Ni Si Ti W Al Zr Hf C N O H Nb
0.001 0.0035 0.2 0.003 0.002 0.001 0.016 0.005 0.0016 0.002 0.013 b0.002 0.005 0.0004 Bal.
0.001 0.0034 0.2 0.004 0.002 0.001 0.015 0.004 0.0015 0.002 0.012 b0.002 0.005 0.0005 Bal.
124
M. Sankar et al. / Int. Journal of Refractory Metals and Hard Materials 50 (2015) 120–125
Table 3 Optimized VAR process parameter to maintain stable arc and produce sound VAR ingot. Sl. no
Ingot diameter (mm)
Electrode diameter (mm)
Fill ratio
Arc voltage (V)
Arc current (kA)
Melt rate (kg/min)
Vacuum (mbar)
1 2
110 150
60 100
0.3 0.44
34–37 34–37
4.5–5.0 5.5–6.0
1.5 2.0
1 × 10−4 1 × 10−4
Fill ratio = cross sectional area of electrode/cross sectional area of ingot.
ingots melted by VAR using different fill ratios. It was observed that a stable arc without any glow discharge and side arcing could be achieved while melting in moulds of 110 mm and 150 mm diameter with fill ratios of 0.3 and 0.44 respectively. It has also been observed that for a given arc voltage and arc current, fill ratio appears to have no significant influence on the melt rate (Fig. 5). It has also been observed that the fill ratio for niobium (which is refractory metal) is much lower as compared to that used for superalloys and reactive metals. During VAR, directional solidification is an important requirement to produce relatively dense, homogeneous and segregation free ingots. Table 1 shows the experimental results of ingots melted by VAR at different melt rates. Fig. 6 shows typical photographs of 110 mm diameter ingots produced by VAR. The surface quality of the ingots is found to be excellent at the melt rates of 1.5 and 2 kg/h. Fig. 7 shows the effect of melt rate on the solidification structure for the 110 mm diameter ingot. It is clear from the figures that at a melt rate of 1.5 kg/h, columnar grains almost parallel to the ingot axis indicating directional solidification. As melt rate increases from 1.5 to 2 kg/h columnar grains are more inclined to the ingot axis. The direction of the columnar grains corresponds to the direction of the maximum thermal gradient at the solidification front. As the direction of the heat flow is always perpendicular to the solidification front, the direction of the columnar grains depends on the profile of the molten metal pool during the solidification process. The solidification rate remains constant as a result of the
uniform rate of heat extraction by the cooling water. Therefore increase in melt rate results in deeper molten metal pool which, in turn, results in columnar grains which are more inclined to the ingot axis. Therefore with the limited number of experiments carried out, it appears that a melt rate of 1.5 kg/h is the optimum. The results of chemical analyses (Table 2) of niobium before and after VAR indicate that there is no pick up gases during VAR. All the optimized process parameters to produce niobium ingots of 110 mm and 150 mm diameter are summarized in Table 3. The radiography shows that the ingot of 110 mm diameter produced with extinction of arc twice for a period of 4-5 min is sound and free from internal defects. Macrostructure of the longitudinal section of the ingot (Fig. 8) shows the discontinuity in the columnar grain growth at two locations in the ingot and this corresponds to the arc extinction events. In the present case, it appears reasonable to assume that the metal would have got solidified during the longer period of arc extinction. The tensile properties of the ingot produced with arc extinction are comparable to the one produced by continuous melting (Table 4). This is in contrast to the reported cracking of Mo ingot having discontinuous columnar grains during rolling. The grain structure discontinuities in the present ingot did not result in cracking during rolling and testing mainly because of the higher currents employed during restarting which ensure remelting of the substrate and good bonding with the former ingot deposited. The commercial grade (Deltaglaze FB 412) oxidation resistant coating which is generally used for hot forging of titanium alloys is also seen to protect niobium from oxidation during soaking as well as hot (900 °C) forging. The hot, warm and cold forging response of vacuum arc remelted 110 mm diameter niobium ingot is excellent. Hot, warm and cold forged 30 mm thick sections were cold rolled to 2 mm thickness without any intermediate annealing. All the cold rolled sheets were sound and free from internal and surface cracks. The microstructure of niobium sheets produced by different conditions is shown in Fig. 9. The cold rolled and annealed samples show recrystallised grain structure. As would be expected, after high forging and rolling reductions, the cast columnar structure of VAR ingot is completely transformed to recrystallized equiaxed grain structure. The recrystallized grain structure of cold rolled and annealed samples shows non uniform recrystallised grains which may be due to the non-uniform deformation of as cast niobium ingot during cold rolling. The room temperature tensile properties of as-cast Nb ingot (110 mm diameter) as well as all the cold rolled and annealed sheets are listed in Table 5 along with the data reported in the literature [9]. Thermo-mechanical processing of as-cast VAR ingot results in a significant increase in room temperature strength and ductility. This is attributed to the breakdown of the cast structure. It can be seen from Table 5 that there is not much difference in tensile properties of niobium sheets produced by different processing conditions, indicating that the properties are dependent on total deformation and not other process parameters. The properties obtained by different processing conditions are Table 4 Room temperature tensile properties of VAR Nb ingot processed with and without arc interruption.
Fig. 8. Macrostructure showing discontinuity in columnar grain growth at two locations in the 110 mm ingot.
Condition of VAR ingot
UTS (MPa)
YS (MPa)
El (%)
Ingot produced with twice arc interruption Ingot produced with continuous melting
107 115
90 95
34 32
M. Sankar et al. / Int. Journal of Refractory Metals and Hard Materials 50 (2015) 120–125
a
c
e
b
d
f
125
Fig. 9. Optical microstructure of (a & b) hot forged at 900 °C + cold rolled at RT + annealed, (c & d) warm forged at 250 °C + cold rolled at RT + annealed and (e & f) cold forged at RT + cold rolled at RT + annealed.
Table 5 Room temperature tensile properties of VAR Nb ingot processed under different processing conditions to 2 mm thick sheet. Sl. no
Processing condition
Total reduction
YS (MPa)
UTS (MPa)
El (%)
1 2 3 4
Hot forging with Deltaglaze coating + cold rolling + annealing Warm forging + cold rolling + annealing Cold forging + cold rolling + annealing Commercial grade niobium (ASTM) (90% min recrystallised)
97% 97% 97% –
123 128 126 73 (min)
213 215 213 125 (min)
56 58 60 20 (min)
superior to properties reported in literature and comparable to ASTM standards. 4. Conclusions The relation between arc length and arc voltage during vacuum arc remelting has been established. Further, the arc length was optimized to produce and maintain a stable arc during the VAR process. Other process parameters such as arc current and fill ratio were also optimized to produce sound and directionally solidified ingots of pure niobium. It has also been demonstrated that sound ingot could be produced without discontinuity in spite of interruption in the process using higher process current. The VAR conditions do not have any significant influence on the tensile properties. The effect of thermo-mechanical processing on the structure and mechanical properties of pure VAR melted niobium ingots was investigated. The forging response of VAR melted pure Nb at room temperature, 250 °C and 900 °C was excellent. All the forged sections could be cold rolled from 30 mm to 2 mm without any intermediate annealing. Processing of as-cast VAR ingot has resulted in significant improvement in tensile properties. The processing conditions do not have any significant influence on the tensile properties. The tensile properties of all the niobium sheets produced by different conditions are superior to the data reported in the literature and of the ASTM standard.
Acknowledgements The authors are sincerely thankful to DRDO for financial assistance through project DMR-295 for carrying out these studies. The authors would like to express sincere gratitude to Dr.G. Malakondaiah, Distinguished Scientist & CC R&D (HR) for his constant encouragement to carry out the present work. The authors also would like to thank the staff members of ERG, PMG, ACG and SFAG for their help in experimentation. References [1] M. Sankar, Y. Satish Reddy, R.G. Baligidad, Effect of different thermomechanical processing on structure and mechanical properties of electron beam melted niobium, Trans. Indian Inst. Metals 62 (2) (2009) 135–139. [2] G.S. Bobrovnitchii, J.N.F. Holanda, Cold consolidation of ATR-niobium powder under high pressure, J. Mater. Process. Technol. 170 (2005) 187–191. [3] K. Matsuura, T. Koyanagi, T. Ohmi, M. Kedah, Aluminide coating on niobium by arc surface alloying, Mater. Trans. JIM 44 (5) (2004) 861–865. [4] T. Carnerio, H. Moura, Proceeding of the Conference on Electron Beam Melting and Refining State of Art, 1998, p. 110. [5] A. Koethe, J. Moench, Purification of ultrahigh purity niobium, Mater. Trans. JIM 41 (1) (2000) 7–16. [6] A.R. Moss, D.T.J. Richards, Arc melting processes for refractory metals, J. Less Common Met. 2 (1960) 405–425. [7] Hand Book of Vacuum Science and Technology, Noyes Publications, 1995. 553–590. [8] F.J. Zanner, Metal transfer during vacuum consumable arc remelting, Mater. Trans. B 32 (1979) 133–142. [9] ASTM standards, B392-03, 2009, 314.
Contents lists available at ScienceDirect
Int. Journal of Refractory Metals and Hard Materials journal homepage: www.elsevier.com/locate/IJRMHM
Effect of vacuum arc remelting and processing parameters on structure and properties of high purity niobium M. Sankar ⁎, V.V. Satya Prasad, R.G. Baligidad, A.A. Gokhale Defence Metallurgical Research Laboratory, Hyderabad 500058, Telangana, India
a r t i c l e
i n f o
Article history: Received 12 August 2014 Received in revised form 25 November 2014 Accepted 6 December 2014 Available online 9 December 2014 Keywords: Niobium Vacuum arc remelting Arc length Thermomechanical processing
a b s t r a c t Niobium and niobium alloys are normally produced by vacuum arc remelting (VAR) process. In the present work, the effect of various process parameters such as arc voltage, arc current (melt rate) and fill ratio on arc stability during VAR process and solidification structure of remelted niobium ingot is investigated. The effect of thermomechanical processing of VAR ingot on microstructure and mechanical properties is also investigated. It has been established that arc voltage is a function of arc length. For melting of niobium in 110 and 150 mm diameter crucibles, the optimum arc voltage is in the range of 34–37 V. It has been observed that the arc voltage needed for melting niobium is much higher than that used for melting stainless steels, reactive metals and superalloys. It has also been observed that for a given arc voltage, fill ratio has no significant influence on the melt rate of consumable electrode. However, arc current has significant influence on the melt rate and consequently on the solidification structure of the ingot. Thermomechanical processing of the VAR ingot is shown to result in significant increase in room temperature strength and ductility. However, the processing temperatures do not have any significant influence on the tensile properties. © 2014 Elsevier Ltd. All rights reserved.
1. Introduction Niobium exhibits a fascinating combination of attractive properties such as high melting point, low density, good fabricability, low ductile to brittle transition temperature and low thermal- neutron absorption cross section [1]. These properties make niobium a potential candidate material for applications in atomic power reactors, reentry vehicles, space power equipment etc. [2]. However, niobium suffers from poor resistance to high temperature oxidation, and active research is underway to overcome this limitation [3]. Historically, powder metallurgy methods involving high temperature vacuum sintering and carbon reduction have been extensively used to produce niobium metal [4]. However, these methods yielded niobium with higher oxygen contents which adversely affect its room temperature ductility, weldability and ductile-brittle transition temperature. Alumino-thermic reduction (ATR) and electron beam melting became the standard practice of producing high purity niobium since the early 1960s. Currently, industrial scale electron beam melting furnaces are available to produce niobium ingots of 300-500 mm diameter and over 2 m in length [5]. However, an electron beam melting process is mainly used only to refine impure niobium and has not been a very successful process for producing alloys based on niobium. This is because of the very high vacuum levels employed in the EB melting
⁎ Corresponding author. E-mail address: [email protected] (M. Sankar).
http://dx.doi.org/10.1016/j.ijrmhm.2014.12.001 0263-4368/© 2014 Elsevier Ltd. All rights reserved.
process which, while refining Nb may result in undue losses of alloying elements. Vacuum arc remelting (VAR) process enables production of alloys based on reactive and refractory metals. Many of the commercial alloys based on niobium have been produced by vacuum arc remelting. The high temperature encountered in electric arc melting, the refining associated with protective vacuum cover, use of refractory free water cooled crucible and solidification in a water cooled mould enable production of high purity ingots with a desirable cast structure [6]. The success of the vacuum arc remelting process is essentially decided by the proper choice of the process parameters. The parameters which are crucial to maintain the melting process stable and to obtain good solidification structure are arc length, fill ratio and melt rate. However, the information available in literature regarding vacuum arc remelting of niobium and the effects of various process parameters on the quality of niobium ingot is inadequate. In general, niobium is subjected to thermomechanical processing by hot forging followed by cold working. Hot forging temperature of commercial niobium is in the range of 650–950 °C. However, heating to such a high temperature leads to excessive oxidation of niobium. Further, oxygen, nitrogen and hydrogen can be picked up during heating and hot forging which make niobium brittle at room temperature. Therefore, niobium is protected against oxidation during heating, hot forging or intermediate annealing by employing protective coatings or inert gas atmosphere or by providing stainless steel jacketing. There is no information available in the literature on the influence of (i) different protection methods during heating and (ii) different thermomechanical
M. Sankar et al. / Int. Journal of Refractory Metals and Hard Materials 50 (2015) 120–125
40
Ingot diameter (mm)
Average input power (kW)
Average melt rate (kg/min)
110
115 165 210 175 220 260
1.0 1.5 2.0 1.5 2.0 3.0
processing conditions on the structure and properties of VAR melted Nb ingot. In the present paper, an attempt has been made to bring out the effect of various process parameters such as arc input power and voltage on melt stability and solidification structure of niobium ingots. The effect of different thermomechanical processing conditions on the structure and mechanical properties of VAR melted high purity niobium is also discussed in this paper.
Arc voltage (volts)
Table 1 Niobium ingots produced under different melt rates.
150
121
35
30
25 5
10
15
20
Arc length (mm) Fig. 2. Plot showing variation of arc length with arc voltage during VAR of Nb under vacuum.
2. Experimental procedure A few VAR experiments were initially conducted in water cooled copper moulds of 60 to 150 mm diameter. Electron beam melted pure niobium rods were used as consumable electrodes. These experiments were conducted with different arc voltages with the following objectives: (i) To establish the relation between arc length and arc voltage and (ii) To optimize the arc length to produce and maintain stable arc.
Arc length is found to be an important process parameter during the VAR process. The distance between electrode bottom surface and top surface of the solidified ingot is taken as the arc length. Direct measurement of arc length is not possible during melting in the VAR process. Hence, this was measured during the VAR process by suddenly switching off the power and stopping the electrode feed. Then the electrode was moved downwards until it touched the ingot and the distance travelled by electrode was taken as arc gap. Another set of experiments was carried out in 110 and 150 mm diameter water cooled copper moulds with different fill ratios (the ratio of cross-sectional area of consumable electrode to the crosssectional area of the mould) and melt rates to study the effect on
soundness and solidification structure of the ingots. Another experiment was conducted in 110 mm diameter mould wherein the power was switched off and the arc extinguished twice at different intervals, each time for a period of 4-5 min. The process was restarted using higher arc current initially to study the effect of power interruption on soundness of the ingot. The relevant process parameters (such as voltage, current, and vacuum levels used) were continuously recorded during all the VAR melts. Two trials were performed for each set of experiments. Three pieces of 110 mm diameter and 470 mm length were cut transversely from the VAR ingot (110 mm diameter × 150 mm long) produced using the optimized process parameters. These pieces were subjected to different thermo-mechanical processing conditions as detailed in Table 1. The first ingot piece was coated with commercial grade (Deltaglaze FB 412) oxidation resistant coating and forged at 900 °C to 30 mm thickness. The second and the third ingot pieces were directly warm (250 °C) and cold (25 °C) forged to 30 mm thickness. Finally all the forged pieces were of 30 mm thickness. These forged pieces were stress relieved at 900 °C for 1 h and subsequently cold rolled to sheets of 2.0 mm thickness by using DEEMAG rolling mill. All the cold rolled samples were recrystallised at 1100 °C for 1 h. The as-cast VAR ingots were tested for internal soundness using gamma-ray radiography (5 Curie Co-60 radioactive source) with an
Titanium
Fig. 1. Plot showing variation of arc length with arc voltage during VAR of Ti and steel under vacuum [7].
122
M. Sankar et al. / Int. Journal of Refractory Metals and Hard Materials 50 (2015) 120–125
Fig. 3. Proposed configuration for (a) transient arc due to motion of molten metal and (b) drop short.
voltage (Fig. 1) [7]. In the case of refractory metals the relationship between arc length and arc voltage is not reported in the literature and knowledge of this relation is essential to control the arc during melting. Arc voltage as a function of arc length is plotted in Fig. 2. It is clearly seen that arc voltage has significant effect on arc length. Therefore it may be inferred that arc voltage can be used as a parameter to control arc length during consumable vacuum arc remelting. The arc length increases from 8 to 17 mm with increase in arc voltage from 28 to 37 V. Above 37 V arc gets extinguished. It is observed that arc is not stable even when the arc length is too short. This may be due to transient arc interruption because of multiple and simultaneous contacts between the electrode and the liquid metal pool due to the motion of the liquid metal pool. This phenomenon is schematically described in Fig. 3a. It has been reported that during VAR melting, the molten metal transfer occurs when metal spikes hanging from the electrode (cathode) form a low resistance bridge by touching the molten metal pool (anode)
7,000
a
6,000 5,000 Current, A
exposure factor of about 5 curie- h. The chemical composition (impurity elements) of the Nb electrode and vacuum arc remelted Nb ingots was determined. All the metallic impurities were analyzed by using inductively coupled plasma optical emission spectroscopy (ICPOES) method. The carbon and hydrogen contents were analyzed by CC-441 Leco carbon and sulphur analyser. Oxygen and nitrogen were analyzed by Leco-136 oxygen & nitrogen analyser. The longitudinal section samples of as-cast VAR Nb ingots were cut, machined, ground to 180 grit finish and then etched with an etchant comprising of 3 parts HF + 2 parts H2O2 + 4 parts H2O (by volume) for macrostructure studies. Samples of (a) cast + glass coated + hot forged + cold rolled + annealed, (b) cast + warm rolled + annealed, and (c) cast + cold rolled + annealed were mechanically polished to 0.5 μm finish and etched with an etchant consisting of HCl- 30 ml, HF-30 ml, HNO315 ml and water -10 ml for metallographic examination. Tensile specimens of 4.0 mm gauge diameter and 20 mm gauge length were prepared from as-cast VAR ingots of 60 and 110 mm diameter as well as from as cast VAR ingot of 110 mm diameter produced with arc interruption with gauge portion containing the interrupted region. Tensile specimens of 6 mm gauge width and 32 mm gauge length were also prepared from 2.0 mm thick section of cold rolled and annealed sheet. All the tensile tests were carried out at room temperature by using Instron 5500R Universal Testing Machine. The fracture surfaces of tensile tested samples were examined under a scanning electron microscope.
4,000 3,000
3. Results and discussions
2,000
In the case of Zr and Ti remelting, arc length is a function of arc voltage whereas in the case of steel and superalloys it is not a function of arc
1,000 0 0
100
200
300
400 Time, S
500
600
700
800
2.5 Ingot diameter: 150 mm
45
Melt current: 5.5-6.0KA
40
b
35 1.5
Voltage, V
Melt rate ( kg/min)
2
Ingot diameter: 110 mm Melt current: 4.5-5.0KA
1
30 25 20 15 10
0.5
5 0
0
0 0
0.1
0.2
0.3
0.4
0.5
100
200
300
400 Time, S
500
600
700
800
Fill ratio Fig. 4. Plot showing effect of fill ratio on melt rate.
Fig. 5. Plot showing current (a) and voltage (b) versus time traces during melting of niobium.
M. Sankar et al. / Int. Journal of Refractory Metals and Hard Materials 50 (2015) 120–125
123
Fig. 6. 110 mm diameter VAR Niobium ingot produced using 1.5 kg/h melt rate (a) as-solidified and (b) machined.
leading to its rupture. This phenomenon is called drop short. This is described schematically in Fig. 3b. At slightly higher arc length (12 to 15 mm), it is very difficult to maintain arc stability due to drop short even though the transient arc interruption due to motion of the liquid metal is totally prevented. This may be due to very high frequency of arc interruption because of drop short. At 15 to 17 mm arc length, drop short frequency is at a minimum level and the arc is stable. The variation of typical process parameters (voltage and current) as a function of time is shown in Fig. 4. While melting of 110 mm diameter mould at optimum arc length is 15-19 mm. The figure shows that the arc is burning at a voltage of 34-37 V and current of 4500–5000 A. Both arc voltage and current are stable throughout the melting process. Occurrence of drip-short has a unique electrical signature which allows them to be detected [8]. In the present work the electrical signature
(sudden voltage drop to 2 V and at the same time the current begins to increase up to 10 kA) corresponding to drip short could not be recorded. In the present work, the drip short phenomenon could be noticed during the melting process from the change in the frequency of sound resulting from the drip short process as well as from the change in the process current and voltage values. However, the voltage values could not be recorded and need for process control due to equipment limitation. Though the drip short can also used to control the arc gap, our objective in the present case has been limited to minimize the drip short frequency and maintain a stable arc. Arc voltage was used to control arc length. Based on these experiments it can be said that for melting of pure niobium in 60 to 150 mm diameter crucibles the optimum arc length is about 15 to 17 mm and to achieve this arc length, arc voltage of 34 to 37 V is required to be maintained. These arc voltages are much higher than those used for melting steels, reactive metals and superalloys. The fill ratio is also an important process parameter as it has direct influence on the arc stability. A low fill ratio produces more stable arc than does a high fill ratio because of the formation of a relatively centralized heat source in the former case even when the cathode spot wanders over the electrode tip. Smaller size electrode also facilitates rapid pumping away of gas from the mould. This helps in maintaining the partial pressure above the molten pool below a critical value. With higher fill ratio, the conductance path is reduced which raises the partial pressure above the molten pool. If the fill ratio is too high, the gap between the mould wall and the crucible becomes too small leading to arcing between the electrode and crucible wall resulting in crucible puncture. Fig. 5 shows the experimental results of 110 mm and 150 mm diameter
Table 2 Chemical analyses of Nb before and after VAR.
Fig. 7. Macrostructure showing grain orientation with increase in melt rate (or power) (a) 1.5 kg/h and (b) 2 kg/h.
Elements
Before VAR (EB melted stock) (wt.%)
After VAR (wt.%)
Fe Mo Ta Ni Si Ti W Al Zr Hf C N O H Nb
0.001 0.0035 0.2 0.003 0.002 0.001 0.016 0.005 0.0016 0.002 0.013 b0.002 0.005 0.0004 Bal.
0.001 0.0034 0.2 0.004 0.002 0.001 0.015 0.004 0.0015 0.002 0.012 b0.002 0.005 0.0005 Bal.
124
M. Sankar et al. / Int. Journal of Refractory Metals and Hard Materials 50 (2015) 120–125
Table 3 Optimized VAR process parameter to maintain stable arc and produce sound VAR ingot. Sl. no
Ingot diameter (mm)
Electrode diameter (mm)
Fill ratio
Arc voltage (V)
Arc current (kA)
Melt rate (kg/min)
Vacuum (mbar)
1 2
110 150
60 100
0.3 0.44
34–37 34–37
4.5–5.0 5.5–6.0
1.5 2.0
1 × 10−4 1 × 10−4
Fill ratio = cross sectional area of electrode/cross sectional area of ingot.
ingots melted by VAR using different fill ratios. It was observed that a stable arc without any glow discharge and side arcing could be achieved while melting in moulds of 110 mm and 150 mm diameter with fill ratios of 0.3 and 0.44 respectively. It has also been observed that for a given arc voltage and arc current, fill ratio appears to have no significant influence on the melt rate (Fig. 5). It has also been observed that the fill ratio for niobium (which is refractory metal) is much lower as compared to that used for superalloys and reactive metals. During VAR, directional solidification is an important requirement to produce relatively dense, homogeneous and segregation free ingots. Table 1 shows the experimental results of ingots melted by VAR at different melt rates. Fig. 6 shows typical photographs of 110 mm diameter ingots produced by VAR. The surface quality of the ingots is found to be excellent at the melt rates of 1.5 and 2 kg/h. Fig. 7 shows the effect of melt rate on the solidification structure for the 110 mm diameter ingot. It is clear from the figures that at a melt rate of 1.5 kg/h, columnar grains almost parallel to the ingot axis indicating directional solidification. As melt rate increases from 1.5 to 2 kg/h columnar grains are more inclined to the ingot axis. The direction of the columnar grains corresponds to the direction of the maximum thermal gradient at the solidification front. As the direction of the heat flow is always perpendicular to the solidification front, the direction of the columnar grains depends on the profile of the molten metal pool during the solidification process. The solidification rate remains constant as a result of the
uniform rate of heat extraction by the cooling water. Therefore increase in melt rate results in deeper molten metal pool which, in turn, results in columnar grains which are more inclined to the ingot axis. Therefore with the limited number of experiments carried out, it appears that a melt rate of 1.5 kg/h is the optimum. The results of chemical analyses (Table 2) of niobium before and after VAR indicate that there is no pick up gases during VAR. All the optimized process parameters to produce niobium ingots of 110 mm and 150 mm diameter are summarized in Table 3. The radiography shows that the ingot of 110 mm diameter produced with extinction of arc twice for a period of 4-5 min is sound and free from internal defects. Macrostructure of the longitudinal section of the ingot (Fig. 8) shows the discontinuity in the columnar grain growth at two locations in the ingot and this corresponds to the arc extinction events. In the present case, it appears reasonable to assume that the metal would have got solidified during the longer period of arc extinction. The tensile properties of the ingot produced with arc extinction are comparable to the one produced by continuous melting (Table 4). This is in contrast to the reported cracking of Mo ingot having discontinuous columnar grains during rolling. The grain structure discontinuities in the present ingot did not result in cracking during rolling and testing mainly because of the higher currents employed during restarting which ensure remelting of the substrate and good bonding with the former ingot deposited. The commercial grade (Deltaglaze FB 412) oxidation resistant coating which is generally used for hot forging of titanium alloys is also seen to protect niobium from oxidation during soaking as well as hot (900 °C) forging. The hot, warm and cold forging response of vacuum arc remelted 110 mm diameter niobium ingot is excellent. Hot, warm and cold forged 30 mm thick sections were cold rolled to 2 mm thickness without any intermediate annealing. All the cold rolled sheets were sound and free from internal and surface cracks. The microstructure of niobium sheets produced by different conditions is shown in Fig. 9. The cold rolled and annealed samples show recrystallised grain structure. As would be expected, after high forging and rolling reductions, the cast columnar structure of VAR ingot is completely transformed to recrystallized equiaxed grain structure. The recrystallized grain structure of cold rolled and annealed samples shows non uniform recrystallised grains which may be due to the non-uniform deformation of as cast niobium ingot during cold rolling. The room temperature tensile properties of as-cast Nb ingot (110 mm diameter) as well as all the cold rolled and annealed sheets are listed in Table 5 along with the data reported in the literature [9]. Thermo-mechanical processing of as-cast VAR ingot results in a significant increase in room temperature strength and ductility. This is attributed to the breakdown of the cast structure. It can be seen from Table 5 that there is not much difference in tensile properties of niobium sheets produced by different processing conditions, indicating that the properties are dependent on total deformation and not other process parameters. The properties obtained by different processing conditions are Table 4 Room temperature tensile properties of VAR Nb ingot processed with and without arc interruption.
Fig. 8. Macrostructure showing discontinuity in columnar grain growth at two locations in the 110 mm ingot.
Condition of VAR ingot
UTS (MPa)
YS (MPa)
El (%)
Ingot produced with twice arc interruption Ingot produced with continuous melting
107 115
90 95
34 32
M. Sankar et al. / Int. Journal of Refractory Metals and Hard Materials 50 (2015) 120–125
a
c
e
b
d
f
125
Fig. 9. Optical microstructure of (a & b) hot forged at 900 °C + cold rolled at RT + annealed, (c & d) warm forged at 250 °C + cold rolled at RT + annealed and (e & f) cold forged at RT + cold rolled at RT + annealed.
Table 5 Room temperature tensile properties of VAR Nb ingot processed under different processing conditions to 2 mm thick sheet. Sl. no
Processing condition
Total reduction
YS (MPa)
UTS (MPa)
El (%)
1 2 3 4
Hot forging with Deltaglaze coating + cold rolling + annealing Warm forging + cold rolling + annealing Cold forging + cold rolling + annealing Commercial grade niobium (ASTM) (90% min recrystallised)
97% 97% 97% –
123 128 126 73 (min)
213 215 213 125 (min)
56 58 60 20 (min)
superior to properties reported in literature and comparable to ASTM standards. 4. Conclusions The relation between arc length and arc voltage during vacuum arc remelting has been established. Further, the arc length was optimized to produce and maintain a stable arc during the VAR process. Other process parameters such as arc current and fill ratio were also optimized to produce sound and directionally solidified ingots of pure niobium. It has also been demonstrated that sound ingot could be produced without discontinuity in spite of interruption in the process using higher process current. The VAR conditions do not have any significant influence on the tensile properties. The effect of thermo-mechanical processing on the structure and mechanical properties of pure VAR melted niobium ingots was investigated. The forging response of VAR melted pure Nb at room temperature, 250 °C and 900 °C was excellent. All the forged sections could be cold rolled from 30 mm to 2 mm without any intermediate annealing. Processing of as-cast VAR ingot has resulted in significant improvement in tensile properties. The processing conditions do not have any significant influence on the tensile properties. The tensile properties of all the niobium sheets produced by different conditions are superior to the data reported in the literature and of the ASTM standard.
Acknowledgements The authors are sincerely thankful to DRDO for financial assistance through project DMR-295 for carrying out these studies. The authors would like to express sincere gratitude to Dr.G. Malakondaiah, Distinguished Scientist & CC R&D (HR) for his constant encouragement to carry out the present work. The authors also would like to thank the staff members of ERG, PMG, ACG and SFAG for their help in experimentation. References [1] M. Sankar, Y. Satish Reddy, R.G. Baligidad, Effect of different thermomechanical processing on structure and mechanical properties of electron beam melted niobium, Trans. Indian Inst. Metals 62 (2) (2009) 135–139. [2] G.S. Bobrovnitchii, J.N.F. Holanda, Cold consolidation of ATR-niobium powder under high pressure, J. Mater. Process. Technol. 170 (2005) 187–191. [3] K. Matsuura, T. Koyanagi, T. Ohmi, M. Kedah, Aluminide coating on niobium by arc surface alloying, Mater. Trans. JIM 44 (5) (2004) 861–865. [4] T. Carnerio, H. Moura, Proceeding of the Conference on Electron Beam Melting and Refining State of Art, 1998, p. 110. [5] A. Koethe, J. Moench, Purification of ultrahigh purity niobium, Mater. Trans. JIM 41 (1) (2000) 7–16. [6] A.R. Moss, D.T.J. Richards, Arc melting processes for refractory metals, J. Less Common Met. 2 (1960) 405–425. [7] Hand Book of Vacuum Science and Technology, Noyes Publications, 1995. 553–590. [8] F.J. Zanner, Metal transfer during vacuum consumable arc remelting, Mater. Trans. B 32 (1979) 133–142. [9] ASTM standards, B392-03, 2009, 314.