Thermometry

Linde's Air Liquefier.—The Linde system depends upon the same principle as the Hampson system, but differs in that the air is not expanded to the atmospheric pressure, but to an inter mediate pressure of about 17 atmospheres. The advantage of doing this is as follows. Suppose that, as is commonly the case, the air is compressed in two stages, the pressure being raised from one atmosphere to p, atmospheres in the low pressure cylinder of the compressor, and then cooled and taken at P.2 atmospheres into the high pressure cylinder, and compressed to 180 atmospheres. Then, in order to balance the plant the work done by each piston must be equal, so that a machine the intermediate pressure would be about 17 atmos pheres, so that by expanding down to 17 atmospheres only, nearly half the work of compression might be saved. Against this there is a loss of cooling effect, but only in approximate proportion of 17 to 180. The Linde principle is the basis of industrial plants for the production of liquid air, and of oxygen from air. 1 he simple plant, making use of free expansion only, as a means of cooling is not, however, used, and industrial plants usually include a pre cooling cycle, in which the air is partly cooled by a carbon dioxide refrigerating cycle, which is relatively much more highly efficient than the free expansion process as a means of cooling through a limited range beiow atmospheric temperature. Since the efficiency of the Hampson Linde process increases as the temperature at the inlet to the regenerator falls, pre-cooling is advantageous. However, it would not suffice merely to cool the gas at the inlet to the regener ator coil, since the expanded gas would then leave the plant at a temperature below that to which the compressed gas was cooled by the carbon dioxide. This difficulty is overcome by utilizing the cold gas to cool the compressed gas before it enters the carbon dioxide refrigerator. The arrangement of the ap paratus is illustrated by the diagram (fig. 12). The air is compressed in two stages by the cylinders a and b, and the re turned air from the liquefier re enters the compressor at the in termediate stage at C. The com pressed air at 18o-20o atmos pheres passes through a water separator d, and a purifier e, to the heat exchanger f, in which it is cooled by the cold compressed air from the liquefier. From f the compressed air passes through g, where it is cooled by the circulation of carbon dioxide, as in the cascade system already described. It then enters the Linde regenerator, a short section of which is represented by h. It consists of a triple coil of copper pipes. Through the inner pipe passes the high-pressure air. Through the middle space, the air at 17 atmospheres pressure, returns to the com pressor, very effective exchange of heat being obtained be tween these streams of compressed air flowing in opposite direc tions. Through the outer space flows a stream of the coldest air at atmospheric pressure, serving as a heat insulating layer.

At the valve m expansion takes place, and here the pressure falls from about to about 17 atmospheres, and cooling with partial liquefaction takes place, the liquid collecting in 1 and the unlique fied air passing by m to the middle pipe of the regenerator coil. The liquid flows through the valve n, and the pressure falling to atmospheric pressure in o, it is partially evaporated, the vapour passing by p to the outer pipe of the regenerator coil, the liquid being drawn off through q. The air returning through the outer pipe of the coil amounts to about one-fifth of the whole : it passes out of the system at r, but actually may be used to cool part of the incoming air by passing it through a subsidiary heat exchanger, similar to f. The air entering the middle section of the regenerator coil by m passes to S, and thence to the heat exchanger f, and to the intermediate pipe of the pump c.

Claude's Air Liquefier.—Another method for increasing the efficiency of the Hampson-Linde system, which was first sug gested by the late Lord Rayleigh, involves a return to the prin ciple of the completely reversible refrigerating engine, by the introduction of an expansion cylinder or turbine into the system in place of the expansion valve, thus making the gas do work on some mechanism outside the liquefier. It would appear at first

sight possible to increase the efficiency of the process enormously, even when working at a low pressure ; but as a matter of fact, the efficiency of the engine is small when the process of expansion results in partial liquefaction. In May 1902 Claude succeeded in constructing such a machine, and in liquefying air by expanding it in an engine which was lubricated with light petroleum, which does not congeal at liquid air temperature. Later he found that liquid air is itself an efficient lubricant. He therefore modified his plant in several ways, passing part of the air only through the expansion cylinder, and expanding it so as to produce a cold exhaust, without considerable liquefaction, the cold gas being used to cool air under pressure in a regenerator coil, this air being partially liquefied by subsequent expansion. Claude's plant ap pears to work very successfully; but though the fact that the air does work on expansion results in increased efficiency of the lique faction process, the energy communicated to the expansion engine is not worth while recovering by coupling this engine to the air compressor, and it is actually wasted.

According to Greenwood, Industrial Gases (1920), the output from various types of liquid air plants is as follows :— Linde of a practical method of fractional distillation of the liquid air for the production of oxygen, which will be described later.

Liquefaction of Hydrogen.

The early attempts to liquefy hydrogen have already been described. The first to obtain liquid hydrogen in quantity for experimental purposes was Dewar at the Royal Institution. His earliest experiments were carried out in 1895, and were described in a paper read before the Chemical Society in 1898. The apparatus used is shown in fig. 13. Hydrogen passed in succession through coils in the vessels B and C, contain ing solid carbon dioxide and ether and liquid air respectively. The gas, which was cooled to the neighbourhood of —15o°, then entered the coil F, which served the purpose of the regenerator coil of the Hampson air liquefier. As the hydrogen was cooled below the inversion point of the Joule-Thomson effect, on ex panding at the valve G, controlled by the spindle H, it became cooled, the cold gas returning through the interstices of the coil cooling the compressed gas regeneratively, and escaping at the outlet E. By means of this apparatus Dewar does not seem to have obtained static liquid, but the temperature at G fell very low. In these experiments he made use of his beautiful invention, the vacuum vessel, which will be described more fully later, and which is an outstanding contribution to the technique of low temperature investigation. Dewar then proceeded to build a large hydrogen liquefier, of which a description was never published. In referring to it, he writes—"the general type of the arrange ment appeared so promising that in the next two years there was laid down in the laboratories of the Royal Institution a large plant (it weighs two tons and contains 3.00o ft. of pipe), which is designed on precisely the same principles, though far more elaborate." With this plant he obtained first small quantities and afterwards large quantities of liquid hydrogen. In 1898 he pub lished the first quantitative observations on the boiling point and density of the liquid. In 1899 he succeeded in manufacturing a litre of the liquid. In the year 190o M. W. Travers published an account of a laboratory apparatus, embodying the same principles, but weighing less than 3o lb., and which could be operated in connection with a compressor, or by means of hydrogen com pressed into cylinders at 125 atmospheres, and which was cap able of producing half a litre of liquid hydrogen after pre-cooling with about five litres of liquid air. Olszewski improved this ap paratus, of which a convenient form is shown in diagram (fig.