WO2012131374A1

WO2012131374A1 - Welding of titanium and other group iv metals

Info

Publication number: WO2012131374A1
Application number: PCT/GB2012/050704
Authority: WO
Inventors: Derek John Fray; Carsten Schwandt
Original assignee: Green Metals Limited
Priority date: 2011-03-31
Filing date: 2012-03-29
Publication date: 2012-10-04
Also published as: GB201105398D0

Abstract

A method of limiting oxidation and/or ingress of oxygen when welding, comprises: providing a molten salt flux (22) to cover a pool of molten metal (18) generated when welding a first metal component (16) to a second metal component; and applying an electric potential using power source (24) to the molten salt flux (22) whilst covering the pool of molten metal (18), using an electrode (26) in contact with the molten salt flux (22) as an anode and at least one of the first and second metal components as a cathode.

Description

TITLE: WELDING OF TITANIUM AND OTHER GROUP IV METALS

DESCRIPTION

Field of Invention

The present invention relates to a method of limiting oxidation of, and/or ingress of oxygen in, welded metal when welding together metal components, and particularly but not exclusively to such a method when welding together components of titanium or other Group IV metals (such as zirconium and halfnium)

Technical Background

The welding of titanium, especially in thick sections, is a difficult process because the metal readily picks up oxygen from the surrounding atmosphere when exposed to the elevated temperatures that occur under typical welding conditions. The reason is that titanium and other Group IV elements have the unusual property that these metallic elements can dissolve considerable quantities of oxygen in both the liquid and the solid states. Thus, on solidification of liquid titanium in the welding zone, this oxygen remains dissolved in the solid. In addition to the actual weld metal, the heat-affected zone of solid titanium on either side of the weld and at its back also dissolves oxygen. Overall, the total volume of titanium exhibiting increased oxygen contents after a welding process can be substantial. A critical issue in the technological application of titanium is that an increased oxygen content in the metal leads to a decline in its mechanical and fatigue properties. This means that, after welding, a substantial volume of the titanium component will have inferior properties which will then affect its performance in further service

Prior Art

Problems concerning titanium welding have existed since titanium and its alloys became commercially available. Several methods to prevent oxygen ingress have been proposed, which include the welding of the component under a protective inert gas shield in an open atmosphere or in a flushed cubicle, and the welding of the component with a liquid flux layer around the heated zone. However, even with gas shielding there is frequently sufficient oxygen present to alter the properties of the titanium. Therefore, the development of a suitable protective flux has long been a major goal but, while various flux systems have been tried, no satisfactory solution has as yet been reached.

As an example, Eutectic Canada Inc. has developed different types of flux such as XuperFlo Flux and Wonder Flux. These are thought to be a mixture of potassium- bearing compounds such as potassium fluoroborate with boric acid and silica. However, the fundamental problem with many flux systems is that they contain oxides which can be reduced by titanium. As a consequence, this oxygen is incorporated and dissolved into the heat-affected zone of the titanium component and raises the oxygen content. Pure fluoride-based fluxes should be more appropriate, but frequently these contain significant amounts of calcium oxide, which again leads to enhanced oxygen contents. Overall, there is at present no titanium welding technique that precludes oxygen ingress to an extent that would be sufficient to preserve the properties of the component.

It is known that it is possible to remove dissolved oxygen from titanium by making the titanium cathodic in a bath of a halide of a very electropositive element such as calcium, ie either calcium chloride or calcium fluoride. However, the removal depends upon the diffusion of oxygen in the solid state which is a very slow process and would not be applicable to thick sections of contaminated metal. What is needed is a method of preventing the oxygen reaching the titanium in the first place. Statement of Invention

In accordance with one aspect of the present invention, there is provided a method of limiting oxidation and/or ingress of oxygen when welding metal, comprising: providing a molten salt flux to cover a pool of molten metal generated when welding a first metal component to a second metal component; and applying an electric potential to the molten salt flux whilst covering the pool of molten metal, using an electrode in contact with the molten salt flux as an anode and at least one of the first and second metal components as a cathode.

At least one metal component may comprise a Group IV metal, such as titanium. The Group IV metal may constitute at least 50% by weight, and possibly at least 75 % by weight, and even at least 90% by weight, of the at least one metal component.

The electrode configured as the anode may comprise or consist of carbon, iridium or calcium ruthenate. For example, the electrode may be a ceramic comprising calcium ruthernate.

The molten salt flux may comprise a chloride and/or flouride of an alkali or alkali-earth metal. For example, the molten salt flux may comprise calcium chloride, or calcium fluoride or mixtures thereof. A mixture of high melting point salts such as fluoride and chlorides of calcium and their mixtures may be desirable.

In the method, the electric potential may be applied with a potential difference of up to about 3V between the anode and the cathode. In any event, it is desirable that the applied potential is below the decomposition potential of the molten salt flux.

The molten salt flux may be provided to cover not only the pool of molten metal generated during welding, but also part of at least one of the metal components surrounding the pool of molten metal. In this way, the method may prevent oxidation of, and/or ingress of oxygen in, the fusion zone and at least part of the heat-affected zone of the weld.

Taking a titanium component as an example, the novel technical concept of the present invention is the application of cathodic protection to the titanium component during the welding process. More specifically, the titanium is maintained as the cathode with respect to an anode in a high-temperature molten salt flux, such that any oxygen, either in the form of molecules or ions, will be prevented from entering the titanium. This approach is similar to that of the well-established cathodic protection in aqueous media, where metals are protected from corrosion at ambient temperature through the application of a cathodic potential. In order to provide the desired electrochemical protection, the molten salt flux must be thermally stable at typical welding temperatures, chemically stable in the presence of titanium metal, possess a wide electrochemical stability range of several volts, and be conductive to oxide ions, O^2".

The electrochemical basis of the new approach is that any oxygen that enters the titanium will be transformed into oxide ions and expelled into the flux. This may be expressed through the following cathodic half-cell reaction:

O (in Ti) + 2 e^"→0^2" (in flux) (1)

Similarly, any oxygen that enters the flux will be maintained therein as oxide ions and not dissolve into the titanium, as follows:

0₂ (in gas) + 4 e^"→2 O^2" (in flux) (2)

The oxide ions in the flux transport to the anode, where they are discharged and removed from the electrolytic cell in the form of gaseous species. At a carbon-based anode, the gases released are carbon monoxide (CO) and carbon dioxide (C0₂); at an inert anode, the gas is molecular oxygen (02). This may be expressed by the following half-cell reactions: xO² + C→C0_X + 2xe^" , with x = 1,2 (at carbon anode) (3) 20²→0₂ + 4 e^" (at inert anode) (4)

In accordance with another aspect of the present invention, there is provided a means of preventing ingress of oxygen into weld metal by applying a cathodic potential to the weld metal via a molten flux.

In accordance with another aspect of the present invention, there is provided a method for protecting a metal from oxidation by applying a cathodic potential, via a molten salt (flux) phase to the metal.

In accordance with another aspect of the present invention, there is provided a method for protecting Group IV metals from oxidation by applying a cathodic potential, via a molten salt (flux) phase to the metal.

In accordance with another aspect of the present invention, there is provided a method for protecting titanium from oxidation by applying a cathodic potential, via a molten salt (flux) phase to the metal.

In accordance with another aspect of the present invention, there is provided apparatus comprising a powerpack, anode, leads and molten flux to cathodically protect a metal.

In any of the preceding aspects of the invention, the flux may be a mixture of high melting point salts such as fluoride and chlorides of calcium and their mixtures. The anode may be carbon, iridium or ceramics containing calcium ruthenate. The voltage applied between the cathodic metal to be protected and the anode may be of the order of 3 V which is below the decomposition potential of the flux.

It will be appreciated that many different energy sources may be used for welding the metal parts together, potentially including a gas flame, an electric arc, a laser, an electron beam, friction and ultrasound.

Brief Description of the Drawings

An embodiment fo the invention, together with examples illustrating the principle of the present invention, will now be described with reference to the accompanying Figures, in which:

Figure 1 is a schematic illustration of a welding operation embodying the the present invention;

Figure 2 is a schematic illustration of apparatus for cathodic protection of titanium strip; Figure 3 is a photgraph of titanium strip after processing in the apparatus of Figure 2; Figure 4 is a photgraph of titanium crucible after processing in in a manner similar to the titanium strip of Figure 3 ;

Figure 5 is a schematic illustration of an experimental arrangement for cathodic protection of a molten titanium/copper alloy; and

Figure 6 is a photograph of an iridium metal sheet after electrochemical processing, using the experimental arrangement of Figure 5, in an air environment and after removal of solidified salt.

Detailed Description of the Drawings

Figure 1 is a schematic illustration of an electric arc welding operation embodying the present invention. As is well known in the art, electric arc welding uses a welding power supply 10 to create and maintain an electric arc 12 between an electrode 14 (which may be consumable) and titanium components 16 to melt metal and form a pool of molten metal 18 at the welding point. The pool of molten metal 18 is covered by a molten salt flux 22, with both forming part of an electrolytic cell with a second power supply 24 during the welding operation. An electrode 26, which is in contact with the molten salt flux 22, is coupled to the second power supply 24 and configured as an anode. At least one of the titanium components 16 is also coupled to the second power supply 24 and configured as a cathode. In this way, oxidation and/or ingress of oxygen in and around the resulting weld may be reduced or even substantially eliminated.

Examples

Example 1

Titanium metal strips were employed as the cathode, molten calcium chloride (CaCl₂) was used as the flux, and graphite served as the anode. The temperature was set at 900 °C and the atmosphere was argon containing approximately 1 % by volume of oxygen. The advantage of using calcium chloride experiments is that it is a well- characterised, non-aggressive and water- soluble electrolyte of a moderately high melting temperature.

Figure 2 shows a schematic of the experimental set-up. An alumina crucible 30 with a height of around 100 mm and a diameter of around 60 mm was used to contain the molten calcium chloride 32. The molten salt level in the crucible was around 40 mm. A titanium metal strip 340was employed as the cathode. This was around 70 mm in length, 10 mm in width and 2 mm in thickness, and was partially immersed in the molten salt 32. A graphite rod 36 served as the anode. This was around 80 mm in diameter and 300 mm long, and was also immersed. Both electrodes 34, 36 were contacted at their upper ends with 3 mm thick nickel wire and connected to a conventional power supply. The electrolytic cell was placed inside an Inconel retort of a programmable electrical furnace. The tube was closed at the top with a stainless steel plate that had feedthroughs for the electrode leads, a thermocouple, and gas inlet and outlet. The gas atmosphere 38 was controlled by flushing the retort with technical argon. A deliberate leak was introduced into the gas inlet, and its size was varied until the desired oxygen content of approximately 1 % by volume was established. The latter was controlled by a conventional lambda probe-type oxygen gas analyser located at the gas outlet. In this way, reasonably stable oxygen contents could be maintained over many hours. In the electrochemical experiments, the titanium strip was polarised at -3 V versus the graphite rod. The typical duration of an experiment was 8 h. At the end of the experiment, the titanium strip was withdrawn, washed with tap water to remove adherent salt, then washed with distilled water, and finally dried at around 70 °C in ambient air in a drying oven.

Figure 3 shows a titanium strip 40 that was processed as described above. It can be seen that there is a clear distinction between an oxidised area (right-hand side of strip 40) and a non-oxidised area (left-hand side of strip 40). Oxidation occurred where the strip was exposed to the oxygen-containing gas phase, whilst oxidation was absent where the strip was covered by the molten salt. Control experiments showed that no such distinction existed in case the titanium strip was merely covered by the flux but not kept under polarisation. This indicates that the presence of the flux by itself does not offer any protection against oxidation. However, if the titanium is subjected to cathodic polarisation in the flux, then complete protection does occur.

The titanium strip used was CP-2 material with an oxygen content of 1800 ppm by mass. It is a well-established fact that the hardness of titanium and its oxygen content are directly correlated. Therefore, subsequent to the recovery and washing procedure, the hardness of the titanium strip surfaces was measured by means of a commercial Vickers hardness tester and the result expressed in terms of oxygen content. Numbers were found to range from 2400 to 3700 ppm by mass, indicating that only a minor increase of the oxygen content at the surface had taken place during the entire experimental procedure. Sectioning the titanium strips and measuring the hardness across the bulk of the samples showed no change in oxygen content.

These results are the proof that the concept of cathodic protection of titanium metal is feasible at elevated temperatures through the utilisation of a suitable electro- active molten salt flux and the application of a modest cathodic potential.

Example 2

Example 3 was carried out in order to demonstrate that larger and thicker titanium components may likewise be protected cathodically under an electro-active high- temperature molten salt flux. This was done by using a titanium metal crucible as the cathode, with calcium chloride as the flux and graphite as the anode. As before, the temperature was 900 °C and the atmosphere was argon with approximately 1 % by volume of oxygen.

A titanium crucible 50 was used that was around 100 mm high and 60 mm wide and had a wall thickness of 3 to 6 mm. This was contacted with nickel wire at its upper rim and employed as the cathode. The molten calcium chloride level in the crucible was around 40 mm. A graphite rod of the type described before was suspended in the melt and served as the anode. The same experimental equipment and procedures were used as before, including a polarisation voltage of -3 V and an experimental duration of 8 h. After the experiment, the solidified salt was rinsed from the crucible with tap water, followed by washing with distilled water and drying in an oven.

Figure 4 shows the titanium crucible 50 after processing in the described way. Again there is a clear distinction between oxidised (inner part of crucible 50) and non-oxidised (bottom part of crucible 50)areas. The upper part was not covered by the flux and underwent massive oxidation, as is obvious from the relatively thick and uneven surface layer of various titanium oxides, while the lower part was covered and thus protected from oxidation, as is evident from the metallic grey appearance.

These results are in agreement with those obtained for the titanium strip and moreover indicate that bigger components and larger surfaces can readily be protected.

Example 3

In the examples reported so far, the oxygen content of the gas phase was relatively low at only 1 % by volume. As this is not representative of the conditions that prevail during an actual welding process, it was therefore necessary to extend the experimental conditions toward the use of ambient air as the gas atmosphere.

The utilisation of carbon as an anode material in high-temperature molten salt electrochemistry is commonplace, which is because of its conductive properties, compatibility with most salt systems, ease of handling, and low cost. The anode gas that is released when using a carbon anode consists of carbon monoxide and carbon dioxide as expressed by Equation (3). It is evident that the reactivity of carbon will set limits to its use at elevated temperatures in air. The first goal of the third set of experiments therefore was to assess the limitations that arise from the use of a graphite anode in air.

An improved approach would be based on the utilisation of inert non-carbon materials as the anode. These materials would evolve molecular oxygen as written in Equation (4) and thus not be consumed by the anode reaction, they would not react with oxygen from the ambient atmosphere, and they would not cause carbon contamination of the titanium component which may sometimes pose a problem for carbon-based materials. Based on information from the literature, two ceramic materials and one metallic material were considered in the present study, and these were doped tin dioxide (Sn0₂), pure calcium ruthenate (CaRu0₃), and pure iridium metal (Ir). These materials are electronic conductors with refractory properties and have been applied with some success under anodic polarisation in different types of molten salt. The second goal of Example 3 therefore was to assess the performance of these materials as an anode in an air atmosphere.

A titanium crucible of the type contained the molten calcium chloride and was employed as the cathode. The various materials mentioned, graphite, tin oxide, calcium ruthenate and iridium, were contacted with nickel wire, suspended in the melt, and served as the anode. As before, the graphite was used in the form of an 8 mm thick rod. The tin oxide was doped with approximately 2% by mass of antimony oxide (Sb₂0₃) and 1 % by mass of copper oxide (CuO), which impart high electronic conductivity and good mechanical strength on the oxide, and used in the form of a commercially available 10 mm wide rod. The calcium ruthenate was synthesized by thoroughly mixing and calcining equimolar amounts of calcium carbonate (CaC0₃) and ruthenium dioxide (Ru0₂), and the calcined material was then sintered into dense discs of diameter 25 mm and thickness 5 mm. The iridium was in the form of commercially available sheet of 50 mm in length, 25 mm in width and 0.5 mm in thickness. The experiments were performed at a temperature of 900 °C in an atmosphere of ambient air which was achieved by leaving the reactor open. The polarisation voltage was -3 V and the experimental duration was 5 h. The recovery procedures were as described before.

In the experiments in which the graphite anode was used in an air environment, the anode remained functional for around 3 to 4 hours. Thereafter, the upper part of the graphite anode, which was outside the molten salt, was fully burnt by the oxygen from the ambient air so that the electric circuit was interrupted. Accordingly, there was cathodic protection during the initial part of the experiment but not during the later part and the cooling phase. The titanium crucible, after the experiment, showed there is a distinction between areas of different extents of oxidation. The upper part was not covered by the flux and show massive oxidation, as is obvious from the presence of various titanium oxides. The lower part was covered with flux and was cathodically protected for most of the experimental duration and shows less oxidation, as was evident from the relatively thin film of yellow and purple titanium suboxides at the bottom of the crucible. As expected, the experiments show that the usability of carbon as an anode in air at 900°C is rather limited. It is clear that a further increase in temperature will shorten the lifetime of the carbon even more, and it is anticipated that its use at temperatures as high as 1700°C, as they would be required in the welding of titanium, will not be possible even for short periods.

It is noted that in the above experiments, the graphite anode was employed with direct exposure to ambient air. An advantageous modification might be to use a carbon-based anode of which the segment exposed to air is shrouded with a refractory material in such a way that contact with the molten salt is made but exposure to the gas atmosphere avoided. It may then even become possible to use carbon at the envisaged temperatures of around 1700°C. This would have a beneficial influence on the electrochemistry of the cell in that at those temperatures carbon monoxide would be the thermodynamically stable gas product, which is much less oxidising than molecular oxygen and also does not dissolve in the molten salt. It should be clarified that in all the experiments described in this section, the extent of oxidation was indeed excessive in the unprotected parts of the titanium crucibles.

In the experiments in which the tin oxide anode was used in an air environment, the anode remained functional throughout the entire duration of 5 h and the subsequent cooling process. Accordingly, there was full cathodic protection throughout the complete run. Again, the crucible showed a very clear distinction between the oxidised and the non-oxidised areas. The upper part of the crucible was not covered with flux and underwent massive oxidation, while the lower part was covered and was not affected by oxidation. However, post-mortem examination of the tin oxide anodes revealed noticeable signs of erosion such as surface roughening and some mass loss. In line with this, at the top of the titanium crucibles some tin-containing deposits were found. Overall, the results indicate that tin oxide is able to operate as an anode in air an environment for short durations but that it is not fully inert in the long term.

The experiments in which calcium ruthenate and iridium were used as the anodes were successful in that full cathodic protection was achieved in the flux-covered areas of the titanium crucibles, whilst massive oxidation occurred in the nonprotected parts. The calcium ruthenate remained mechanically stable throughout the experiment and underwent no noticeable changes in geometry or mass. In case of iridium a small degree of thinning was noticed in the immersed part. This was expected as iridium forms a film of iridium oxide when oxidised in air at 900 °C and since this oxide is somewhat volatile, slow metal loss will occur.

Example 4

This experiment was operated at 1100°C using a molten 50%Ti-50%Cu alloy and the following salt mixtures - equimolar binary mixture of calcium chloride and calcium fluoride, an equimolar mixture of calcium fluoride and sodium fluoride and an equimolar ternary mixture of calcium fluoride, sodium fluoride and potassium fluoride.

Figure 5 presents the experimental set-up that was used for the investigations with the liquid metal pool. The ceramic crucible 60 contains the titanium/copper alloy 62 which re-melted during heating to the operating temperature of 1100°C. As is indicated in the figure, the molten metal cathode 62 rests at the bottom of the crucible 60 and the molten salt flux 64 floats on its surface. Iridium and calcium ruthenate were used as anodes 66 and were contacted with nickel wire and suspended in the molten salt 64. A molybdenum rod 68 of 6 mm in thickness was used as an electrical contact to the molten metal cathode. The experiments were performed at a temperature of 1100°C in an atmosphere 70 of ambient air. The polarisation voltage was -2.5 V and the experimental duration was 5h. After the experiment, the salt was washed away with water. This procedure extended over several days as calcium fluoride and sodium fluoride are both only very sparingly soluble in water. When using calcium ruthenate as the anode, the ceramic material had a strong tendency to crack and fracture during the experiment.

The use of iridium metal as the anode material proved to be highly successful. The material remained fully operational and ensured perfect cathodic protection of the liquid titanium/copper pool throughout the duration of the experiments. Figure 6 displays the iridium metal sheet 80 after complete removal of the solidified salt. The immersed part of the iridium was also mechanically completely intact and only had undergone some discolouration.

Small pieces were cut off the solidified titanium/copper alloy body and subjected to a quantitative oxygen analysis using a LECO-type analyser. Five samples were taken from the metal and the measured oxygen contents were found to range from 1400 to 2800 ppm by mass. While there does not seem to be any literature on the solubility of oxygen in liquid or solid titanium/copper alloys, it is certainly reasonable to assume that the oxygen solubility is somewhat lower than that of pure titanium but still of a substantial magnitude. Consequently, the low oxygen contents measured may be regarded as the proof of successful cathodic protection under the harsh experimental conditions applied.

Claims

CLAIMS:

1. A method of limiting oxidation and/or ingress of oxygen when welding, comprising:

providing a molten salt flux to cover a pool of molten metal generated when welding a first metal component to a second metal component; and

applying an electric potential to the molten salt flux whilst covering the pool of molten metal, using an electrode in contact with the molten salt flux as an anode and at least one of the first and second metal components as a cathode.

2. A method according to claim 1, in which at least one metal part comprises a Group IV metal.

3. A method according to claim 2, in which the Group IV metal is titanium.

4. A method according to any one of the preceding claims, in which the anode comprises a material selected from the group consisting of carbon, iridium and calcium ruthenate.

5. A method according to any one of the preceding claims, in which the molten salt comprises calcium chloride, calcium fluoride or mixtures thereof.

6. A method according to any one of the preceding claims, in which electrolysing the molten salt flux comprises applying a potential difference of up to about 3V between the anode and the cathode.

7. A method according to any preceding claim, in which the molten metal flux covers part of at least one of the metal components surrounding the pool of molten metal as well as the pool of molten metal generated during welding.