GAS THAT FLOWS INTO AN ENCLOSED electric arc from the sides and out through an axial perforation in one of the electrodes emerges as a jet of high-temperature plasma, a mixture of ionized particles hotter than the surface of the sun. The hot plasma can be raised to still higher temperature if the jet is directed through a moderately strong magnetic field and a second arc. At this higher temperature substantially all the gas molecules are broken apart and their constituent atoms lose one or more electrons.

Matter in this state opens the way to a number of engrossing experiments [see "The Plasma Jet," by Gabriel M. Giannini; SCIENTIFIC AMERICAN, August, 1957]. Solids exposed to the plasma promptly evaporate-even tungsten, the most refractory substance known. Metals and ceramic compounds that do not mix readily at low temperatures can be evaporated by the plasma and condensed as a composite substance. The plasma itself is highly reactive and on cooling can form interesting compounds In addition, gases of low atomic mass can be accelerated to impressive velocities. Accordingly the apparatus holds promise as an engine for driving spaceships. It was this property that fascinated Jimmy McAleer, a high school student of Mobile, Ala. At 16 he decided to construct a plasma jet apparatus and determine its thrust. The project went so well that he wound up at the top in this year's National Science Fair.

"A plasma jet is not difficult to build," he writes, "after you solve two problems. First, you have to learn the construction details: the size and shape of the parts, the kind of materials to use, the amount of current required by the arcs, the precautions to take and so on. In my case this was not easy. I located several general descriptions of the apparatus in our local library, but the details had to be discovered at the workbench. This required more experience than I had, so I enlisted the help of David Martin, an electrical engineer, as my adviser on the project and conferred with Father L. J. Eisele, S.J., of Spring Hill College in Mobile and F. E. Marsh of Bell Aerosystems in Buffalo, N.Y.

"The second problem is to learn how to adapt ordinary supplies to the special requirements of the apparatus. For example, I found that the main arc, the one that generates the plasma, required 150 amperes of direct current at about 50 volts. At this voltage and current the arc consumed 7.5 kilowatts, a power requirement considerably beyond the capacity of ordinary house wiring. I finally hit on the idea of using a string of automobile storage batteries connected in series. Moreover, the voltage had to vary inversely with the current to maintain the arc in stable operation. To induce this drooping voltage characteristic I inserted a carbon welding rod in the circuit to act as a ballast resistor. During the first experiment the rod became white hot and burned up, so I submerged the next rod in a bucket of water. The heat brought the water to a boil in about three minutes-which did not matter because the arc never operated that long at one stretch.

"After talking with my advisers, I decided to use helium for the gas because it is both inert and of low atomic mass. Hydrogen has lower atomic mass than helium and therefore can be accelerated to a higher velocity, but it can also combine with air to form an explosive mixture. I also thought of using nitrogen, primarily because it is inexpensive and does not explode. But the atomic mass of nitrogen is greater than that of helium and it can react with ionized carbon in the plasma to form the poisonous gas cyanogen. In the course of asking about these and other gases in our local welding shops I learned that helium cylinders returned to the gas suppliers for refilling normally contain a little gas, more than enough to supply the plasma jet for several operating cycles. One shop agreed to let me drain the 'empties' without charge.

"All the other materials used in constructing the jet were purchased at hardware and automobile-accessory stores or were salvaged from scrap. The parts w ere made and assembled with hand tools, with three exceptions: I had access to a lathe at school for machining a steel sleeve that supports the negative electrode of the main arc and for drilling an accurately centered hole in the positive electrode. And a local welding shop helped by brazing some of the parts."

During his summer vacation Jimmy visited New York City and brought his apparatus along so I could see how he put it together. It consisted of two major parts: the plasma assembly, a watercooled arc chamber in which a slender carbon rod is supported in axial alignment with a perforated carbon electrode; and the accelerator assembly, composed of an electromagnet and a second pair of electrodes at right angles to the field of the electromagnet, as shown in the schematic diagram at the top. Gas enters the chamber at slightly more than atmospheric pressure and escapes through the perforated electrode, a short length of thick carbon rod. The arc forms between the inner end of the perforation and the tip of the solid electrode. The magnetic and electric fields of the accelerator lie in two planes, each perpendicular to the jet. The apparatus will operate in any position but is usually run with the jet pointing up so the heat is carried away by rising air. The plasma emits intense ultraviolet radiation that extends into the region of soft X rays and must therefore be shielded by a housing of thick sheet metal equipped with a window of ruby glass, as shown in the drawing in Figure 2.

The apparatus requires three separate sources of direct current; these can be automobile storage batteries or a combination of batteries and an arc welding generator. The principal arc operates at 150 amperes and, with the included ballast resistor, requires a 72-volt source. Jimmy found that a bank of 12 fully charged six-volt batteries connected in series will supply power for eight duty cycles of three minutes each. The electromagnet of the accelerator draws 12 amperes at 12 volts from a second bank of batteries. The accelerator electrodes require about 15 amperes. This current is taken from a string of three batteries and is adjusted to the desired value experimentally by shifting the connection of the positive lead (by means of a battery clip) to one or another of the cells.

The arc chamber may be constructed of heavy copper or bronze pipe of three different diameters, closed by end plates in the form of perforated disks cut from sheet stock of about the same weight as the pipes. Only one dimension is critical: the inside diameter of the smallest pipe should make a snug fit with the perforated carbon rod. The inside diameters of Jimmy's tubes are 1 inch, 2 1/2 inches and 3 1/2 inches. The lengths are 7/8 inch, 1 1/4 inches and 2 inches respectively. His end plates were cut in the proportions shown by the cross-section drawing below and brazed to the tubing. Water enters the cooling chamber tangentially through an inlet of 1/8-inch copper tubing, as shown. Brass studs were threaded into the end plates and brazed to support a disk of Transite, an asbestos composition board, which closes the back of the cooling chamber. These studs were made long enough to engage a pair of strap-iron brackets that support the assembly above the base of the housing. A similar pair of studs extends from the upper end plate for attaching the Transite base of the accelerator assembly. Gas enters the chamber through a 34-inch copper tube brazed to the steel sleeve that supports the solid electrode. The sleeve is drilled as shown to make a running fit with the 1/4-inch welding rod of carbon that acts as the negative electrode.

The positive electrode is a 1 1/2-inch length of heavy-duty carbon welding rod one inch in diameter. The cut ends were dressed to a right angle on a lathe. A h-inch hole, accurately centered, was then drilled axially to a depth of 1/4-inch in one end of the electrode and a 3/4-inch hole was drilled from the opposite end to meet the 1/4-inch hole. The action of the arc causes the positive electrode to erode rapidly and it must be replaced after about three minutes of operation. The spent electrode is driven from the tube by means of a wooden punch and a new electrode is then forced into the tube. The high current drawn by the arc requires that the resistance of the contact between carbon electrode and cooling chamber be made as low as possible. Hence, the fit between the two should be tight and the copper kept clean.

The arc assembly is attached to a base of Transite by two strap-iron brackets, as shown in the drawing in Figure 2. Electrical, water and gas connections enter the housing through holes in the base as shown. The negative electrode is moved up and down by a flexible cable of the type used to actuate carburetor chokes in automobiles. The plunger of the choke cable serves as a control for striking the arc and adjusting its length.

The core and pole faces of the electromagnet were cut from soft steel shafting and assembled to the accelerator base by a pair of angle brackets as shown in the drawing in Figure 4. The coil draws about 150 watts and is designed for a three-minute duty cycle every 15 minutes. The plane that includes the center of the magnet should cut the plasma about 1/2 inch above the end of the positive electrode. The axis of the accelerator arc is located about 1/8 inch higher. The steel core of the electromagnet is fitted with spool ends of Formica and insulated by a single wrapping of plastic tape The outer layer of the winding is similarly taped and, in addition, wrapped with a double layer of asbestos paper fastened in place by heat-resistant cement. The electrical connections to both the coil and accelerator electrodes (1/4-inch welding carbons) are made by heavy flexible leads of the type used for wiring automobile headlights. The tips of the accelerator electrodes are spaced about 3/8 inch apart.

The positive lead of the plasma arc is standard automobile starter cable and the negative lead is flexible copper braid of the kind used for grounding the battery to the frame of automobiles. The saddle clamp for attaching the braid to the negative electrode of the plasma arc is made of l/8-inch strap copper about 1/2 inch wide. The U bends in the halves of the clamp must make a good fit with the carbon. The clamp should be attached to the negative carbon rod so that it is within about 1/8 inch of the steel supporting sleeve when the rod is in contact with the positive electrode. This minimizes the portion of carbon rod that conducts current but allows the electrode enough upward movement so the arc can be struck.

The dimensions, shape and materials used for constructing the housing are not critical. The enclosure should be large enough and sufficiently well ventilated to dispose of the heat, and for safety must be made of materials sufficiently opaque to shield the experimenter from radiation. Doors of the shielding material may be provided for exposing materials to the plasma as required.

For convenience, an instrument panel should be provided that includes at least one voltmeter and one ammeter for monitoring the power drawn by the plasma arc, together with on-off switches for each of the three circuits. The panel can also serve as a mounting for the plunger that actuates the negative electrode. Jimmy's panel was made of plywood. He used separate voltmeters and ammeters to monitor the accelerator circuits, but the panel could be equipped with appropriate shunts and instrument switches for monitoring all circuits sequentially with one voltmeter and one ammeter.

"You must follow a definite sequence of operations when starting up the apparatus," Jimmy writes. "I learned this the hard way. After assembling the parts of my first jet and making sure that the connections were hooked up properly, I struck the main arc. In seconds the arc chamber was reduced to a splatter of molten copper in the bottom of the housing. I had neglected to turn on the cooling water! This must be done first. (I used a small centrifugal pump to circulate water through the chamber from a three-gallon reservoir. A two-minute duty cycle raises the temperature of the water from 60 degrees Fahrenheit to about 160 degrees.) The gas is then turned on and adjusted to a flow- rate of about 20 liters per minute. (I calculated the rate of flow in terms of pressure, measured with a homemade manometer.) The arc is then struck by operating the choke plunger, its length adjusted for stable operation and a reading made of the current. If the current is higher or lower than 150 amperes the apparatus is shut down. Then the effective length of the ballast resistor is adjusted in the appropriate direction by moving one of its contacts, saddle clamps like the one that connects the flexible copper braid to the negative electrode of the plasma arc.

"When it is operating smoothly at 150 amperes, the plasma, as viewed through dark ruby glass, resembles a welding flame about four inches long. According to the articles I have read its temperature is on the order of 15,000 degrees F. The electromagnet of the accelerator is then switched on. This makes the plasma flame spread and bend toward one side, accelerating it. Voltage is then applied to the accelerator electrodes and adjusted until the jet is straight and as long as possible-about eight or 10 inches. The jet becomes quite noisy when power is applied to the accelerator and emits a characteristic hissing sound when the accelerator voltage is properly adjusted.

"I wanted to measure the thrust of the jet directly, and still intend to do so. My first measuring apparatus did not work, however. It consisted of a baffle linked to a spring balance, and I hoped to determine the thrust by measuring the force the jet exerted against the baffle. But the plasma vaporized every baffle that I made, including those of alumina and zirconium oxide. I wanted to enter the apparatus in our local science fair and did not have time to develop a reaction balance. So I decided to measure the temperature and compute the thrust. I borrowed an optical pyrometer, with a maximum scale of 3,700 degrees F., from a local steel firm, but the tungsten filament of the pyrometer did not come even close to matching the intensity of the plasma stream. I then placed a heavy wire of tungsten in the plasma. It vaporized instantly. This meant that the temperature had to be above 6,178 degrees Kelvin, the boiling point of tungsten. So I decided to settle for this value and compute the equivalent thrust.

"Reference texts state that the energy of molecular vibration is equal to half of the product of the mass of the molecule multiplied by the square of its velocity. The energy is also equal to three halves of the product of the absolute temperature multiplied by Boltzmann's constant.

The velocity is therefore equal to the square root of three times the product of Boltzmann's constant and the temperature divided by the mass oú the molecule. Boltzmann's constant is equal to 1.38 times 10^-16, and the mass of the helium atom is 6.68 times 10^-24 grams. At a temperature of 6,173 degrees the velocity of the particles is therefore 6.19 times 10⁵ centimeters per second, or 22,400 feet per second.

"The thrust exerted by a jet is equal to the product of the rate of propellant flow in pounds per second and the velocity in feet per second divided by the acceleration of gravity (32 feet per second per second). In the case of my jet the propellant flow is .00013 pound per second. The force is therefore equal to .00013 times 22,400 divided by 32, which amounts to .09 pound.

"Another criterion frequently used to indicate the effectiveness of rocket fuels is 'specific impulse,' which is the thrust developed by burning one pound of fuel in one second, or the ratio of thrust to the mass of fuel flow. It is expressed in seconds. For the assumed temperature (6,173 degrees Kelvin) the specific impulse of my jet is equal to .09 divided by .00013, which comes out to 700 seconds. Actually 6,173 degrees is a very conservative assumption. At the more likely temperature of 30,000 degrees the figure would be on the order of 1,500 seconds. In contrast, the specific impulse of conventional chemical fuels such as zinc and sulfur is listed at 20 seconds, that of hydrogen peroxide and hydrazine at 240 seconds, and oxygen and hydrogen at 345 seconds.

"I am now constructing another setup for measuring the thrust directly. Turning the jet so that its axis is horizontal, I have suspended the entire assembly as a pendulum bob at the bottom of a four-foot curtain rod. The suspension hinge is a feeler-gauge leaf .004 inch thick. Cooling water is lifted from a reservoir by a siphon attached to the pendulum rod and is discharged into a catch basin. The electrical leads run up the pendulum rod to small containers of mercury that are connected to the power sources. The gas flows to the jet through a length of thin-walled, flexible tubing. During an initial test run the jet deflected the assembly more than an inch out of plumb. Now I plan to run a thread from the bottom of the suspended jet assembly over a small pulley to a weight pan. Then I hope to measure the thrust by loading the pan just enough to pull the assembly back into plumb and weighing the load on an analytical balance."

Bibliography

THE PLASMA JET. Gabriel M. Giannini in Scientific American, Vol. 197, No. 2, pages 80-88; August, 1957.

The Society for Amateur Scientists (SAS) is a nonprofit research and educational organization dedicated to helping people enrich their lives by following their passion to take part in scientific adventures of all kinds.