Preliminary Experiments with Flames of Atomic Hydrogen
To try out the possibility of blowing atomic hydrogen out of an arc, 20-ampere arcs from a constant-current transformer were passed between two tungsten rods 6 mm in diameter mounted transversely in a horizontal alundum tube (10 cm diameter) through which a stream of hydrogen flowed. With voltages from 300 to 800V, arcs could be maintained with electrode separations up to 2 cm. The magnetic field of the arc caused the hydrogen to move transversely so that it became fan-shaped. Iron rods 2 or 3 mm in diameter melted within 1 or 2 seconds when they were held 3 to 5 cm above the arc. By directing a jet of hydrogen from a small tube into the arc, the atomic hydrogen could be blown out of the arc and formed an intensely hot flame. To maintain the arc in a stable condition the electrodes were brought close together (1 to 3 mm), but the arc did not remain entirely between the electrodes, but extended as a fan to a distance of 5 to 8 mm. The flame of atomic hydrogen, however, extended far beyond the arc. At distances of 1 or 2 cm from the arc molybdenum (m.p. 2900oK) melted with ease. Near the end of the arc tungsten rods and even sheet tungsten (m.p. 3660oK) could be melted. The use of hydrogen under these conditions for melting and welding metals proved to have many advantages. Iron can be melted without contamination by carbon, oxygen, or nitrogen. Because of the powerful reducing action of the atomic hydrogen, alloys containing chromium, aluminum, silicon, or manganese can be melted without fluxes and without surface oxidation.Temperature of Atomic Hydrogen Flame Compared with Other Flames Let us suppose we could obtain atomic hydrogen in bulk at atmospheric pressure and room temperature and that we could then let this burn to the molecular form in a flame. What would be the temperature of this flame and how would it compare with that of other flames? Taking the heat of reaction (for 2 grams) to be 98,000 calories and taking the specific heat of molecular hydrogen (for 2 grams) to be 6.5 + 0.0009 T, we find that the heat of the reaction would be sufficient to heat the hydrogen to 9200oK. The dissociation of the hydrogen, however, would prevent the temperature from rising to any such high value. If x is the degree of dissociation at the maximum temperature reached, the available heat of recombination is only (1-x) 98,000. Langmuir plotted two curves based on his accumulated data . These two curves intersected at T = 4030oK and x = 0.642. Thus atomic hydrogen at room temperature and atmospheric pressure would heat itself to 40300K and the degree of dissociation would then be 0.642. There is another factor which tends greatly to increase the temperature of the atomic hydrogen flame even above the calculated value of 4030oK. The atomic hydrogen, instead of being originally at room temperature, is already at a high temperature at the moment of its escape from the arc. The conditions are analogous to those in an oxyhydrogen flame in which both gases are preheated. Thus the upper limit of temperature is fixed only by the degree of dissociation of the hydrogen and the rate at which heat is lost by radiation or contact with bodies of lower temperature. Rate of Surface Heating by Flames Although the high temperature of the atomic hydrogen flame is of great importance when it is desired to melt substances of very high melting point such as tungsten, a factor of even greater importance in for example ordinary welding operations is the speed with which heat can be delivered to a surface per unit area. If a Bunsen burner flame delivers 51 watts per sq. cm to the whole surface of a black body, it would thus heat it to a maximum temperature of 17300K. If heat is applied by the flame to one side of a plate-shaped body and the heat is radiated from both sides, the maximum temperature reached would be 1450oK. The fact that the Bunsen flame does not heat bodies so hot as this indicates that the rate of surface heating decreases as the temperature of the body increases. The heat reaches the surface from such a flame by conduction through a relatively stationary film of gas. The decrease in the temperature gradient when the body becomes hot would explain the lower rate of surface heating. With 1330 watts per sq. cm delivered by the atomic hydrogen flame, the temperature of a black body would rise to 3900oK. The power radiated from tungsten at its melting point is 395 watts per sq. cm, and 1330 watts per sq. cm should heat tungsten to about 5300oK. At such high temperatures, however, the rate of surface heating by an atomic hydrogen flame must decrease because of the fact that the hydrogen remains partly dissociated so that the recombination is not complete. With surface temperatures below 2000oK, however, this factor would be negligible. It is probable that the rate of surface heating would be dependent not so much on the temperature gradient in the surface film of gas as on the rate of diffusion of atomic hydrogen through this film. Thus we may expect the rate of delivery of energy to a metal surface to remain nearly constant until the surface reaches a temperature of at least 2000oK.It now became of interest to determine what fraction of the total energy in an arc or a flame could be delivered to a large flat surface against which the flame was directed. For this purpose a cylinder of copper 10.5 cm in diameter and 9.8 cm long was used, which weighed 7950 grams. The flame was directed against one of the flat polished ends, and the rate of temperature rise was measured. An atomic hydrogen flame was produced by a 60-ampere a.c. arc using a torch like that shown here: The voltage across the electrodes was 70 volts. A wattmeter showed that the power consumption in the arc was 3510 watts, which gives a power factor of 0.84. The electrodes were tungsten rods 3.2 mm in diameter which made an angle of 55 degrees with one another. The rate of flow of hydrogen which bathed the electrodes was 14.2 liters per minute (30 cubic feet per hour). From the temperature rise of the copper block the heat delivered to the surface was found to correspond to 3100 watts when the electrode tips were 3 to 5 mm from the copper surface. This decreased to 2800 watts at 13 mm, 2500W at 25 mm, and 2200W at 35 mm. With the arc turned off but the molecular hydrogen burning in the air, the rate of heating corresponded to 250 watts with the electrodes 6 mm from the surface. Subtracting this energy delivered by the combustion of the hydrogen in the air, we find that the energy carried to the metal by the atomic hydrogen ranged from 2850 to 1950 watts. Thus with the electrodes 3 mm from the metal 82 per cent of the power input into the arc was delivered to the copper surface. This efficiency became 78 per cent at 6 mm, 71 at 13 mm, 65 at 25 mm, and 55 per cent at 35 mm. The power corresponding to the complete combustion of 14.2 liters of hydrogen per minute is 2360 watts. Actually, only 250 watts or 11 per cent of this reaches the copper. The total energy of the arc and the flame of molecular hydrogen is 5870 watts, of which 3100 watts or 53 per cent is delivered to the copper. An oxy-acetylene flame from a standard welding torch consuming 30.6 liters of oxygen per minute (64.8 cubic feet per hour) and 28.6 liters of acetylene per minute (60.6 cubic feet per hour) delivered energy at the rate of 4400 watts to the copper surface. A smaller torch consuming 13.7 liters of oxygen per minute and 13.0 of acetylene (29.0 and 27.5 cubic feet per hour, respectively) gave energy to the copper at the rate of 3900 watts.Application of Atomic Hydrogen Flames to Welding of Metals The high temperature of The Atomic Hydrogen flame, together with its powerful chemical reducing action and the avoidance of gases containing oxygen and nitrogen, render it particularly useful for welding, not only for iron and its alloys, but for such metals and alloys as contain aluminum, magnesium, chromium, manganese, etc. The previous figure illustrates one of the later forms of torch used for welding. Two tungsten rods, as electrodes, are held at a definite angle to one another by easily adjustable clamps, and a jet of hydrogen is directed from a small nozzle along each of these rods near its end. The hydrogen thus bathes the heated parts of the electrodes and forms a gentle blast of gas which passes through the arc between the electrode tips, and blows the atomic hydrogen away from the electrodes so that these are not unduly heated. Other torches were also built suitable for automatic welding using machine feed. The electrodes were ordinarily separated 3 or 4 mm and the arc assumed a fan shape extending 6 to 10 mm from the electrodes. Alternating current was generally used. To utilize the atomic hydrogen flame for the welding of metals it became important to have easy and complete control of the flame. Many different forms of welding torches were constructed and tested. The electrodes between which the arc passed were mounted at a convenient angle to one another and were adjustable so that they could be brought into contact at a point which was exposed to a blast of hydrogen from one or more orifices. Thus the atomic hydrogen was blown out of the arc in a definite direction and formed a flame which could be brought into contact with the metal to be welded. The jet of hydrogen also served to bathe all the heated parts of the electrodes and the work, thus preventing oxidation and the introduction of impurities such as nitrogen into the weld. The hydrogen was supplied by a tube which passed through the handle and then by flexible tubes was delivered to each of the electrode holders and escaped through the annular spaces between the electrodes and the lava insulators. Sufficient hydrogen was used not only to surround each of the electrodes to their tips but to form a blast which blew the atomic hydrogen against the work and bathed it in hydrogen.
Both the striking voltage and the arc voltage were higher for an arc in hydrogen than for the ordinary welding arc since there was no appreciable amount of metallic vapor generated in the arc. The standard arc welding equipment of those days was therefore not suitable as a power source for operating the atomic hydrogen torch. If direct current was used the arc could be stabilized by a series resistance or a specially designed generator of the constant-current type could also be used. With series resistance a line voltage of 250 was found to give good results. Alternating current was more convenient and, since the arc could then be stabilized by reactance instead of resistance, greater efficiency was usually obtained. A line voltage of 350 to 400 gave satisfactory operation. Voltages as high as this were needed solely to give stability and to enable the arc to be started at any time by separating the electrodes even when these were cold. A number of tests were made to determine the voltages required to strike an arc by means of the lever mechanism of the torch. With cold electrodes an open circuit ac voltage of 320 was needed for striking the arc. After the arc had been started and the electrode tips had reached the operating temperature a line voltage of only 150 was sufficient to restart the arc as long as the electrodes remained nearly at the operating temperature. These lower voltages could also be employed to start and maintain the arc if the electrodes were first raised to high temperature by a high current while separating them very slightly so as to have a high contact resistance or by bringing a thin tungsten rod between the separated tips of the electrodes. It was more convenient, however, to use open circuit voltages of approximately 400V rather than to employ these special means of starting the arcs. In normal operation the drop across the arc was in the neighborhood of 80 volts. To avoid danger to the operator the entire arc circuit was preferably insulated from ground. A motor-generator was used to give either direct or alternating current for the arc, but it was usually more satisfactory to use a specially designed transformer. The connections that were used in most of the work to be described are shown in Fig. 3. When the arc was not operating the electrodes were in contact by the action of the spring attached to the control lever, so that there were no voltages on the electrodes and the torch could be laid down on any material without danger of flashing the operator's eyes. To strike the arc the electrodes were merely separated by pressing the lever. Should the open-circuit voltage at any time be impressed across the electrodes when separated, or the operator break the arc by spreading the electrodes too far apart, a relay in the arc circuit (contactor B as shown in Fig. 3) would trip the feeder circuit, in which case it was necessary for the electrodes to be brought in contact again before the main feeder circuit could be restored. The voltage drop across the arc while in operation varied from 60 to 100 volts, depending on the amount of opening between the electrode tips. This voltage was nearly independent of the current, between the limits of 20 and 70 amp, although a slight decrease in voltage was usually noted when the current was increased. Repeated experiments showed that the lower voltage arc (60 to 80 volts) obtained by separating the electrodes only 1/16 or 1/8 of an inch had a more concentrated working zone and was the most efficient arc for most kinds of welding. By bringing the arc closer to the surface of a larger mass of metal it was found that the metal melted very rapidly. For welding, the maximum rate of heating was desired and this was obtained by bringing the torch so close to the metal that the lower portion of the fan shaped arc was just about in contact with the metal and this caused the arc to change its shape somewhat. The tips of the electrodes were then usually about 3/8 to 1/2 in. from the metal. Portions of the arc could at times become short-circuited by the metal so that the tracks of cathode spots on the metal could be seen,but this seemed to play no important part in the welding process.
Conclusions
The above is a general account of the course of events that led to the development of the atomic hydrogen arc welding technique. Although Langmuir himself had many ideas for other applications of atomic hydrogen, such as using it in melting furnaces, and although General Electric themselves were aware of the vast potential of atomic hydrogen (as expressed by the editor of the General Electric Research Laboratory Publication.The point of inception of many important practical processes can be found in researches in pure science. Following Dr. Langmuir's discovery of atomic hydrogen, conclusion was reached that flames of this gas make possible new applications of far-reaching importance), no subsequent effort was made to develop neither the welding method nor any other application using atomic hydrogen. In spite of the obvious value of the process, industry's excuse for laying the process aside was that it had been replaced by better processes such as Heliarc, TIG, and MIG welding, though plasma arc welding is rarely mentioned which has also almost disappeared from the market. Since plasma arc welding is merely an extension of the atomic hydrogen process, the reasons are undoubtedly the same. As a welding process, atomic hydrogen arc welding was obsolesced by MIG and TIG neither of which compare to its welding efficiency and uses. Considering that atomic hydrogen arc welding hardly got of the ground before it was replaced, it is not a far fetched thought to assume that the interests of welding suppliers and electric power companies were being nursed so that more archaic tanks, transformers, gauges, torches, electrodes, gases, fluxes, power etc. could continue to be sold at profit. The reader is reminded of the fact that Langmuir's experiments and findings were taking place almost 100 years ago. Since then, technology has progressed dramatically in many fields. If Langmuir had at his disposal the know-how and technology of today, our world may have looked different. However, we are now in the fortunate situation that we do possess the data of Langmuir's findings, which can be combined with today's technological know-how. The obvious direction in which to look is the same direction that seems to have been the fundamental reason for Langmuirs discoveries having been swept under the carpet: CLEAN AND ABUNDANT ENERGY. If anyone looks up to find any information on atomic hydrogen arc welding, only a few lines will appear, simply informing that the arc is maintained between two metal electrodes in an atmosphere of hydrogen. Shielding is obtained from the hydrogen. Pressure and/or filler metal may or may not be used. Although the process has limited industrial use today, atomic hydrogen welding is used to weld hard-to-weld metals, such as chrome, nickel, molybdenum steels, Inconel, Monel and stainless steel. Its main application is tool and die repair welding and for the manufacture of steel alloy chains. Also used in special military welding requirements. Nothing is mentioned of the extraordinary properties of atomic hydrogen, nor of its potential for the use as energy to drive the wheels of mankind. However, and quite surprisingly, in one edition of Van Nostrand's Encyclopedia of Science it was stated Hydrogen molecules dissociate to atoms endothermically at high temperatures (heat of dissociation about 103 cal/gram mole) in an electric arc, or by irradiation. The hydrogen atoms recombine at the metal surface to provide heat required for welding. What is surprising here is that the actual energy value needed for the dissociation of the hydrogen molecule is given, but the calorific value for the recombination of the atoms into molecules is strangely omitted. From Langmuir's experiments and findings we know that the minimum calorific value for the recombination of atoms was agreed to be in the region 90.000 cal/gram molecule. In other words we have an input energy of 103 cal/gram molecule and an output energy of 90.000 cal/gram molecule. In conventional science this seems to be violating the law of conservation of energy. Langmuir explained this (however, not very convincingly) by the heat being carried forward from the arc to the metal surface. One area which certainly deserves the attention of modern science, is the replication of Langmuirs experiments using high-tech measurement equipment.
