Tweet

---

August, 1929

---

By the courtesy of the Washington Chemical Co. Ltd., who are associated with Newall's Insulation Co. Ltd., we have been recently privileged to pay a most interesting visit to their laboratories and to inspect the special apparatus which they have elaborated for testing the efficiency of various heat insulators. Two methods are employed, and while both are applicable to a very wide range of materials they are roughly most suitable to steam and refrigerating problems respectively.

By this time it is well appreciated that the insulation value of any material depends primarily upon the subdivision due to the minute air spaces formed within it. The size of these chambers and the degree with which they communicate largely determine the extent to which resistance is offered to the flow of heat through the material. For high temperatures such as those involved in steam work the material must also be such that it is not affected in its physical nature, and, of course, it is necessary for all purposes that a certain mechanical strength should be present. The choice of materials is, therefore, somewhat restricted according to the use to which they are to be put, but considerable progress has been made in recent years in the preparation of special brands possessing all the necessary qualities in a high degree to meet any given conditions.

Determinations of the heat resistance of materials for insulating steam pipes have been made for some years by means of the apparatus illustrated. This consists of a 3-ft. length of 6-in, pipe, fitted with hemispherical ends, which is slung in free air and covered with the material under test. An electrical resistance heater on a porcelain core is supported exactly in the centre of the pipe and extends its full length. Thermocouples are placed at intervals in the lagging material and on the surface as well as on the pipe surface under the cover. The electrical energy dissipated in the heater is measured by a standardised wattmeter and can be controlled by the use of external rheostats. By means of previous calibration it is known approximately what current is necessary for a given pipe temperature, and when taking a test this current is applied until the internal temperature becomes constant. In such a condition the amount of heat due to the current (watts x 0.057 B.Th.U. per minute) is obviously equal to the amount of heat lost by radiation, conduction, etc., through the pipe and/or lagging, and it is only a matter of arithmetic to arrive at the loss per sq. ft. of surface per hour. This figure is obtained for a number of different temperatures and the results plotted as a graph. Sources of inaccuracy include "end effect," but this can be allowed for. The same method is in use at the National Physical Laboratory, using a 13 ft. 6 in. length of 4 in. flanged steam pipe with 2 ft. long flanged pipes on either end to counter the "end effect." In this case the internal temperature is measured by ascertaining the electrical resistance of the pipe and checked by means of thermocouples. The results of the smaller apparatus agree very well, however, with the more elaborate determinations of the N.P.L. and the data thus obtained are obviously of the greatest value to the firm in its search for improved materials and in the perfecting of the present. compositions used for various purposes.

The second test apparatus is similar to that described in Special Reports Nos. 6 and 7 of the Building Research Board published by the Department of Scientific and Industrial Research but now unfortunately out of print. An electrical hotplate is used in a horizontal position, with slabs of the insulating material above and beneath. Thermocouples are again used to ascertain the temperature at various points and the test given permits determinations to be made of the rate of heat flow per sq. ft. per hour through any given material at a definite temperature. Water-cooled plates are employed on the external surface of the insulating material to dissipate the heat passing and the rise in temperature of the water gives an accurate check on the heat received. The use of a flat hotplate necessarily introduces very serious difficulties in regard to "end effect," since the edges of the hotplate will obviously be at a lower temperature than the central portion unless special precautions are taken. An ingenious application, of the guard ring is employed to surmount this difficulty, the hotplate being surrounded by a secondary heater supplying the heat for edge radiation.

Drawings and Photographs accompanying the article