Above the mantle only two styles of construction have been used in recent years. For many years nothing but plain firebrick from two to five feet in thickness was used from this point to the top of the furnace, but the tendency for the lining to be eaten away more extensively just above the top of the bosh than anywhere else, which had existed very generally for many years, became worse with the general use of fine ores, and the lining troubles which they brought with them.

As the cooling plate construction had given a good account of itself in the bosh, furnacemen began to put rows of cooling plates above the mantles to prevent this action, and this practice is now quite frequently followed. These were at first simply built in the brickwork and the pipes to them were brought in through small holes cut in the shell for that purpose. This, of course, put these cooling plates entirely beyond the possibility of renewal; if they failed they were done with and must be shut off and were soon burnt up.

To meet this situation some furnacemen have cut holes through the furnace shell large enough to permit the removal of the whole cooling plate exactly as is done on the bosh, but this cuts up the shell very badly and even when this provision is made it is a matter of much difficulty to remove these plates through such a great thickness of brickwork. It is, therefore, doubtful whether the trouble and expense of making these plates removable is worth the small gain to be obtained by that means. It will be seen that in the case of the furnace shown by Fig. 165 the course was adopted of bringing the water pipes for these blocks out through the shell so as to give access to the several connections, but without making any provision for the removal of the cooling plates.

Fig. 168. Cast-iron cooling plate.

In early days of cooling plates they were made of cast iron with coils of pipe cast in them, these in all cases being built solid in the brickwork, without any provision for removal. The pipes were first made in one continuous coil zigzagging back and forth through the cooling plate, but an improvement on this was subsequently introduced which consists in making each pass of the pipe through the block separate and bringing the ends of all the pipes out through the outside surface of the plate, and putting the connections which put them in series on these projecting ends. By this means if the inside pipe failed it could be cut out and the one back of it left in service; if this failed by the gradual burning back of the furnace, it could be cut out and so on back for as many passes as the pipe might make through the plate, this in turn number generally being from three to five. This arrangement is shown in Fig. 168.

For the protection of the bosh, cooling plates made of bronze, described above, were a great improvement. Bronze has a great advantage over iron, in that iron and cinder do not adhere to it, so that the cooling plates stay perfectly clean and as a result of this and of their shape they can be pulled out of the furnace without much difficulty and replaced whenever they fail so that the inside of the furnace is kept practically to the original lines for a long period of time.

On account of this superiority for service below the mantle, the bronze cooling plate has often been used above the mantle also, but after considerable experience with it I prefer for that service the cast-iron plate with pipes cast in, for two reasons.

First, these plates above the mantle being the highest of any of the water-cooled portions of the furnace they are the first to lose the water in case of any partial failure of the water supply which reduces its head. I have known of several cases where this happened without the complete failure of the water supply. This means that these cooling plates above the mantle are the most likely to lose their water, and if they do so the bronze plates are bad for two reasons. First, the melting point of the bronze is many hundred degrees below that of the cast-iron plate and its mass is much smaller, the body of the plate being hollow, so that when the water on these plates has been lost, even though the furnace be stopped at once, the heat of the surrounding materials causes the bronze to melt after a comparatively short period, this leaves a considerable cavity in the brickwork into which the bricks from above are likely to settle, thus causing at least the possibility of a crack in the lining, even if the gas does not blow out through the burnt cooling plate itself, as it often does.

With the cast-iron plate, on the other hand, the conditions are quite different. The melting point of cast iron is not reached by the contents of the furnace at a point much above the mantle, so that these plates are much less likely to be melted out completely than the bronze plates, and even if the inner end next the stock is melted off the outer end is protected by the brickwork for a long time, if not indefinitely, and being practically solid except for the pipes cast through it, a partial melting of the cooling plate does not form a cavity as in the case of the bronze plate.

Secondly, if even a small portion of the nose of the bronze plate be melted off it is destroyed completely, as this opens its water space into the furnace, but with the multiple-pipe system several inches of the plate may be melted away and still, by using some of the rear pipe passes, water may be put through it again, and its usefulness as a cooling member preserved indefinitely.