This section is from the book "Modern Buildings, Their Planning, Construction And Equipment Vol3", by G. A. T. Middleton. Also available from Amazon: Modern Buildings.
For a correct understanding of the working principle of any form of hot-water apparatus, whether for domestic supply or for warming buildings, it is essential that the cause and action of the circulatory movement of heated water be described and understood. The want of this knowledge is chiefly accountable for the many failures experienced in this work, the average hot-water fitter being, as yet, but little better than a copyist - able to repeat previous successes, but often failing when the work entails a departure from ordinary undertakings.
The circulatory movement that occurs in water when it is heated, on which the hot-water engineer relies, is due to the action of Convection. To describe this action, it has first to be explained that water, mobile as it is, consists of a mass of separate particles- molecules as they are termed, each particle being independent of the other and having the property of mobility in a very pronounced degree. The mobility of the particles, as the term implies, admits of their gliding over, under and around one another in a manner that indicates an entire absence of friction or resistance, some authorities going so far as to say that there is a separating or repelling force existing, causing the molecules to work free from one another instead of rubbing. The physical condition of water, therefore, is that of a mass of extremely minute particles in a state that admits of their moving freely in any natural direction when quite a trifling force is brought into play. It will be seen directly that a motive power of exceeding weakness will set the molecules of water moving with astonishing rapidity.
A very simple experiment will illustrate convection currents, it only being necessary to suspend a glass jar, filled with water, over some simple source of heat, such as a small benzoline or spirit lamp, or over the chimney of a small paraffin lamp. It will be found, however, that water is invisible - certainly the action of convection is ; but this difficulty can be overcome by using a solid visible substance in the water, which will indicate whatever movements may be occurring. Doubtless the best material to use is amber in a powdered form, this having a specific gravity very nearly the same as water, and it will consequently remain suspended in it, stationary when the water is still, and moving in accordance with any motion of the water. A piece of broken pipe stem, easily obtained where pipes are repaired, will serve, if a part of it is filed into powder. There may be disappointment due to there being so many clever imitations of amber, but a piece of true material is not difficult to obtain, and this is worth doing if experiments in hot-water circulation are to be undertaken.
The amber dust is put into a bottle with some water, the two being well shaken together. After allowing the bottle to rest for a little time, it will be found that some of the dust has risen to the top as a scum (which can be removed) while some has sunk to the bottom. There will, however, be a good quantity suspended in the body of the water, and this is the part that should be decanted off for use. On making an experiment it will be found that the movements of the amber particles are quite visible, and they correctly represent the movements due to convection currents. A rather full explanation is devoted to this owing to the great advantages obtained by being able to observe -to see - what is occurring with the water.
The cause of the circulatory movements is explained as follows. When the water in a vessel is cold - the particles being of uniform temperature - the whole mass will be found to be stationary. On applying heat to the bottom of the vessel the particles nearest that point become warmed, and, like almost everything in nature, they expand, becoming larger than their colder fellows. An increase in bulk of each warmed particle occurs, but there is no increase in weight. It is common knowledge that any substance which is lighter, bulk for bulk, than water, cannot remain at the bottom of a vessel of water. It, instead, immediately rises to the top. The rise is not directly due to the lighter quality of the substance, for if a piece of cork, or a hollow rubber ball, be placed on a table it shows no inclination to rise, but if either be placed at are given an increase in bulk, and so made lighter for a given size than their fellows, that movement occurs.
There are several instructive and distinctly interesting results to be noted from even so simple a contrivance as that just explained. There will be seen, for instance, what has been termed the "skin" of water; for however freely the main body of water may be circulating, a thin surface at the top remains quite still and unaffected. It can be seen, too, how an upward movement occurs in the centre with a downward current around all sides, if the lamp is placed centrally; while the up-current can be confined to one side and the descending movement to the other by putting the lamp nearer one side than the other. It can also be seen how easily the circulation can be reversed by shifting the lamp from side to side. Finally, the theoretical explanation just given can be proved by letting the contents of the jar be first heated, and then, after the lamp is removed and the water has become stationary, the circulating movement can be started by placing a cold object in or on the top of the water. In this case the warmed particles are cooled, contracted, and made heavier (bulk foe bulk), and the circulation is then more plainly due to the superior weight of the coolest particles. The foregoing, it will be noted, refers wholly to the circulation of heated water in a vessel, a thing worth studying in the way described, remembering the small trouble and expense involved. For practical hot-water work, however, the vessel, representing a boiler, must have pipes extending from it, and the circulatory movement must not only occur freely through these pipes, but the movement must take place up one pipe and down the other, and not both up and down in each.
It is not a very easy thing to explain why the up and down movements occur separately in the two pipes, in the very positive and reliable way they do. When the pipes are attached to a boiler they are connected one at top and one at a low point (as is fully explained later), and the common explanation is that this fact accounts for the phenomenon just referred to. It is not so, however, for if, for experimental purposes, both pipes are connected quite level with one another at the top of a boiler, the circulation, up one pipe and down the other, occurs just as freely and starts as instantly as usual, without hesitation.
It is to be strongly recommended that a simple apparatus, constructed of glass, be made and experimented with by all interested in this work ; and for initial investigation an easily erected and inexpensive model, such as is shown in Fig. 32, will serve to settle many points. That an effective circulation can be obtained with the pipes both connected at the same level at top can be tried by arranging them as shown in Fig. 33. It will be found, as stated, that the circulation will set in promptly, but there will be no certainty as to which pipe will have the up-current and which the down. This will largely depend on whether the lamp is more under one pipe than the other. The object, therefore, in making boiler connections, one high and one low, is not to produce the requisite circulation, but to ensure its always occurring in one desired direction. With the connections at the high and low points there is always a certainty as to which pipe will be the " flow " and which the " return." The flow-pipe is that which has the ascending, while the return-pipe has the descending current in it.
The small model illustrated in Fig. 32 consists of a glass jar of 2 to 3 inches diameter, having a large neck aperture. It is closed by a tightly fitting cork, which can be waxed over with sealing wax if desired. Two holes are bored in the cork for the pipes to fit tightly into. The pipe is glass tube of about 5/16 or 3/8 inch size, costing about one penny per yard, and this can be easily bent in any gas flame - a bunsen flame for preference, as it does not discolour the tube. The tees can be of copper or zinc tube, soldered up. The joints are made with short pieces of flexible rubber tube (such as can be obtained from a chemist) of rather smaller size and stretched on. Amongst several points to be observed with this model is the rapidity with which the circulation sets in, it commencing almost instantly the lamp is put under, which, remembering the low conductive power of rather thick glass, goes to show that an extremely small rise in temperature will set the water in motion. It will also be seen that, instead of the circulation being-local to the boiler for a little time, as is commonly supposed, it commences at once both in boiler and pipes. Experiment can also be made as to the effects of putting the air pipe in the wrong place, or of omitting it, or of grading the pipes wrongly, etc. Needless to say, a much more extended apparatus can be erected about as easily as this simple one, and much, very much, can be learned, while the solution of problems is made moderately easy.