US2695946A - Inert gas-shielded arc welding electrode - Google Patents

Description

United States Patent Ofiice 2,695,946 Patented Nov. 30, 1954 INERT GAS-SHIELDED ARC WELDING ELECTRODE Louis A. Conant, Tonawanda, N. Y., assignor to Union Carbide and Carbon Corporation, a corporation of New York No Drawing. Application November 4, 1952, Serial No. 318,750

3 Claims. (Cl. 2198) This invention relates to inert gas-shielded arc welding of the kind disclosed by Meredith Patent No. 2,342,086, and more particularly to an improved non-consumable, refractory metal electrode for use in such welding in a stream of gas composed of argon, or helium, or any suitable mixture of argon and helium.

Heretofore, commercially pure tungsten, or thoriated tungsten, in the form of a diameter rods have been used as electrodes in this type of welding; thoriated tungsten being preferred over tungsten, because of its better characteristics with respect to contamination losses, open circuit starting voltage, are starting losses, and are stabilization. However, it has recently been suspected that in such use thoriated tungsten electrodes may be toxic due to the volatized thoria, and also, unless some additional protection is provided, may be dangerous to the health of people in the vicinity due to ingested thorium with subsequent harmful internal radiation of alpha, beta and gamma rays therefrom. Thoriated trungsten electrodes also are somewhat more expensive than tungsten electrodes.

There exists a need, therefore, for a refractory metal electrode for use in inert gas-shielded arc welding, which is safer, non-toxic, cheaper, and also better even than thoriated tungsten with respect to contamination losses, open circuit starting voltage and operating voltage.

The main object of this invention is to provide a novel electrode which satisfies such need, and which is safe and inexpensive to manufacture and easier to fabricate. Another object is to provide an electrode for such use that is lighter in weight than either tungsten or thoriated tungsten electrodes.

According to the invention such objects are accomplished by providing-a non-consumable inert gas-shielded arc welding electrode which is composed of tungsten having dispersed therein a minor amount (0.5-10%) of calcium oxide, 2% of the latter being satisfactory.

Mixtures of tungsten and the appropriate chemical component, such as calcium acetate powder, to yield the metal oxide are used to prepare ingots approximately x A x 18 inches by conventional powder metallurgical techniques (pressing). These ingots are sintered at 2000 C. or above by passing an electric current through them. The ingots are then drawn or swaged to the final dimensions. Considerably more detail can be obtained from standard text books on the subject, such as Tungsten by Smithells, or Tungsten by Li and Wang.

Starting voltage The starting voltage characterizing an electrode under given starting conditions is an important quantity as it determines the open circuit voltage requirement on the welding source. Electrodes which exhibit a relatively low open circuit voltage requirement for starting will permit a stable arc to be established almost instantaneously, thus minimizing labor and material expenses particularly in machine welding. Comparison was made with commercially pure tungsten and thoriated tungsten electrodes. At the time of these tests only inch diameter commercial thoriated tungsten electrodes were on hand. These were mounted in a torch of the type disclosed by Breymeir Patent No. 2,468,804. A inch diameter commercial tungsten rod was similarly tested.

Exploratory tests disclosed conditions which not only indicated starting voltage differences, but also permitted results to be obtained with satisfactory reproducibility.

The test conditions were (1) a hemispherical electrode tip, (2) a cold arc length of 0.050 inch, (3) an established are of 80 amperes, DCSP, (4) an argon flow rate of 15 cfh, and a fairly massive stainless steel anode or workpiece. A motor-generator was used as the power source; external field excitation was used to extend the open circuit voltage range below 40 volts. Gap breakdown was initiated by the output from a high-frequency stabilizer. The starting voltage under these conditions was taken to be the minimum open-circuit voltage for which reliable and instantaneous starting resulted in at least nine of ten consecutive trials. Since no attempt was made to control precisely such factors as surface conditions except for initial grinding, and particularly since the instantaneity was a matter of the observers judgment, precise reproducibility of the data was not always observed. Thus, an estimated limit of error of about $5 volts is to be associated with these data. This limit is set largely by experienced intuition with weight placed on the observed ranges of open-circuit voltage for transition from percent starting to a condition of no starting.

Electrode starting losses Electrode weight loss during starting with high-frequency voltage is significant in connection with spot weld ing and with repetitive machine welding of small parts or short lengths of scam. Starting loss characteristics were established by determining the weight change produced by successive are strikings. Equipment was constructed in which motor-driven cams actuated switches permitting automatic execution of a 16.5 second operation cycle as follows: (1) line power was supplied to a rectified welding current source having an open-circuit voltage of about 70 volts, (2) line power was supplied to a highfrequency stabilizer for about 0.5 second, the output breaking down the arc gap and establishing a ampere DCSP argon-shielded arc, and (3) following operation of the are for approximately 1.5 seconds, power to the Welding current source was interrupted. A short arcing" time was selected to reduce contributions to weight changes from arcing alone and a cycle period large with respect to the time of arcing was used to provide a reasonable interval for cooling the Work metal to avoid puddle buildup problems. Following original adjustment of the arc length to 0.050 inch, no further manual adjustment in arc length was made throughout the test. A stainless steel anode (workpiece) was utilized with an arc current of 110 amperes DCSP. The argon flow rate was 20 cfh.

Wt. Change after 100 Starts Electrode Trial I Trial II Avg.

M M V W (Commercial) 6.7 4.1 5.4 2% CaO-W O. 6 0. 6 O. 6

Differences due to torch design and/or electrode geometry appear in the data for commercial tungsten electrodes. However, it is apparent that the new electrodes are substantially better than commercial tungsten electrodes with respect to starting losses.

Operating voltage On the other hand in welding thin materials burn-through and other undesirable heating effects are to be avoided. One method of doing this is to decrease the are power by decreasing the arc voltage. The table below shows that electrodes of'the invention produce appreciable lowering of the'arc voltage;

Another method of reducing are power is by reducing the arc current. The extent to which this can be done is determined by the lower current limit for stable operation. This in turn is a function of electrode composition. It has been shown that stable operation can be realized at lower current than that found for standard tungsten electrodes by using thori'ated tungsten electrodes. Such a determination has not been made for the electrodes of the invention. However, since it is believed that the starting voltage requirement is an indication of the lower current limit for stable operation, inasmuch as electrodes having low starting voltage requirements can be expected to operate stably at relatively low arc currents.

The following measurements were taken with a 150 am pere DCSP argon-shielded arc on stainless steel. The argon flow rate was 20 cfh. At least five readings were obtained at each length setting, the averages being shown in the data table. These data include the IR drop along the electrode.

Arc Voltage at- Electrode Length Length W (Commercial) 8.5 10. 1% 'lhOz-W (Commercial) 8. 2 10.5 W (Commercial) 8. 4 10.7 2% CaO-W 8.3 10.5

Electrode performance on stainless steel Weld beads were made to compare the influence of electrode composition on penetration andthe amount of work metal melted. These welds were made at a welding speed of about 40 ipm on 0.055 inch thick stainless steel with 150 ampere DCSP arcs. The cold arc length was 0.050 inch and the argon flow rate cfh. The data presented below wereobtained from single cross sections. The dependence of melted area on are voltage is shown by these data.

ELECTRODE INFLUENCE ON BEAD SHAPE data which at first glance appears to be a large one. It

is felt that this effect is primarily due to the ditference in thermal expansion of the electrode as a result of differences in the electrode geometry and torch cooling etficiency. Comparing the lengths of 7 inch diameter sections in the two cases one expects for the experimental electrode roughly six times the expansion of the standard. Moreover, experience has shown that it is not too far afield to assume the operating length to be about 50% of the cold setting when standard Linde Heliarc equipment is used. Such considerations yield arc voltage results shown in column four as calculated from the data of the previous section. In most cases the agreement is satisfactory.

The last column has been calculated to give a rough metal melting efficiency comparison.

Contamination losses Losses of electrode by low melting alloy formation following accidental contact with molten metal are likely to be one of the most important factors in determining.-

electrode life, particularly in hand welding applications. For this reason, measurements were made to determine the ability of various electrodes to withstand contamination by.

molten stainless steel. A 200 ampere arc was maintained on a stainless steel anode for two minutes; the electrode was then forced into the molten pool of the anode four times in succession (the arc re-establishing itself between:

immersions) and finally operated for another two-minute penod'to allow impurities on the electrode to burn-01f. During the entire proccssthe argon flow rate was'25 cfh;

Electrode tggg g g W (Commercial) 300-460 2% ThOz-W (Lintle)... 106 1% ThOz-W (Mallory). 92 2% CaO-W sten and calcium oxide in the ratio of about 2% by weight of calcium oxide, the balance being commercially pure tungsten.

3. An electrode for gas-shielded metal fusing arc processes, composed of refractory metal comprising tungsten containing a relatively small amount, i. e. 0.52.0%, inclusive, of calcium oxide.

References Cited in the file of this patent UNITED STATES PATENTS Number Name Date 1,792,943 Terry Feb. 17, 1931 2,515,559 Lancaster et al. July 18, 1950

Claims (1)

1. A NON-CONSUMABLE INERT GAS-SHIELDED ARC WELDING AND CUTTING ELECTRODE COMPOSED OF TUNSTEN HAVING DISPERSED THEREIN A MINOR AMOUNT (0.5%-10%) OD CALSIUM OXIDE.