The principle of this blower is the same as that of the Rateau, but the number of impellers used is smaller, and the whole of the blower is built in one body, without the feature of balancing by having the flow from the two half sections of the impeller in opposite directions. This produces a shorter and more compact machine. This blower is shown by Figs. 108, 109, 110, 111 and 112, taken from a paper on the "Turbo Blower" presented by Mr. Richard H. Rice, of the General Electric Company, before the American Institute of Mining Engineers, and published in their Bulletin for May, 1914.

Fig. 107. Method of controlling capacity of Rateau turbo blowers.

The governing device of this machine is on a somwhat different principle from that of the Rateau; it is shown by Fig. 108. The inlet pipe ends in a large inverted cone in which a circular diaphragm is suspended from a scale beam overhead. It is obvious that the area of the annular space between the diaphragm and the walls of the cone will vary with the location vertically of the diaphragm, and that the flow of the incoming air up past this diaphragm will have a tendency varying with the velocity of the air through the annular space, to lift the latter diaphragm and the governing is done by this varying lift. Thus while a different mechanism is used, it produces about the same results as the Venturi tube.

If the pressure required by the furnace increases so as to decrease the flow of air under the speed at which the turbine is running, the diaphragm promptly drops down, carrying with it the beam which operates on the steam valves, and though a relay this admits more steam and increases the speed of the turbine until the flow of air is restored to the normal. In this turbine also a centrifugal governor is provided to prevent excessive increase in speed.

The advantages of the turbo blower are its small size, its entire freedom from reciprocating parts, and consequent ability to dispense with any but a light foundation, the small number of moving parts, there being in effect but one, and its ability to dispense with lubricating oil both in the steam and air ends. Great advantages are claimed for it also in the absolute steadiness of the blast which it produces, and the freedom from pulsation which results from a continuous discharge. In the paper by Mr. Rice above mentioned he describes researches which prove the absence of vibrations from blast, compressed with the turbo blower, and their presence with reciprocating blowing engines, but whether this absence may be translated into dollars and cents of advantages in operation is a matter concerning which no proof has as yet been offered and concerning which I confess to a considerable degree of skepticism.

Fig. 108. Constant volume governor for centrifugal compressors.

It is claimed without hesitation that these blowers on account of the steadiness of their action, etc., enable a furnace to produce less flue dust and to make iron with less fuel than can be done with any other type of blowing machinery, but no definite data on this subject have been offered to support this contention, which also seems to me doubtful.

Fine points of this kind are always exceedingly difficult to prove in an operation of such magnitude as the blast-furnace, and the situation is complicated by the fact that the turbines have in the nature of things, in many places at least, replaced obsolete blowing engine installed under bad conditions and therefore likely to be drawing steam into their inlet and discharging it into the furnace along with the air, with consequent detriment to fuel economy. Moreover, these engines being old and uneconomical, and being presumably of the low-speed variety, they gave greater pulsation to the blast than more modern and high speed engines. It is therefore extremely doubtful whether the mere fact that certain advantages are found where the turbo blower is used, which were not found before its introduction, is due solely to the type of blower introduced, or whether they would not have been obtained equally well with any other type of good modern blower.

Fig. 109. Three-stage General Electric turbo blower partly in section.

Fig. 110. Centrifugal compressor rated Type T - 3 - 25,000 - 15/30 -3000/3800 driven by a 2000/3400 condensing Curtis steam turbine.

Fig. 111. T - 3 - 35, 000-15/30 - 2500/3250 F centrifugal compressor direct connected to 2900 H.P. Curtis steam turbine.

Much is claimed also for the accuracy of calibration of the air-measuring apparatus of these machines, and in a paper by Mr. S. G. Valentine, read at the February, 1914, meeting of the American Institute of Mining Engineers, and published in the Transactions for that year, he gives the results of his experience in the use of one of these blowers on a furnace which bear upon this point. Among other data he states that only 41 cu. ft. of air were required to burn a pound of coke based on the calibration of the air-measuring apparatus of the turbo blower. We shall subsequently see that about 63 ft. are required theoretically for this purpose without any loss by leakage, not only in the engine but around the stoves, the tuyere and elsewhere, and if only 41 ft. are used then one-third of all the fuel charged is being dissolved by direct deduction in the upper regions of the furnace and never reaches the tuyeres at all.

Fig. 112. Three type T - 3 - 40,000 - 15/30 - 2500/3250 R. P. M. air compressors driven by 2900/5200 H.P. Curtis steam turbines installed at Iroquois Iron Company, Chicago, Itl., for blowing 500-ton blast furnaces.

This hypothesis does not agree with the relative low fuel consumption reported by Mr. Valentine of about 2150 lb. of coke on an ore mixture consisting largely of magnetite which required a large amount of heat in the hearth, and, in fact, Mr. Valentine admits that his figures, based on the analysis of the top gases and carbon burned, differ from the calibration figures of the turbo blower by about 10 per cent.

I have information concerning practice in which the solution loss is likely to be much higher than it is in Mr. Valentine's practice, and in spite of this the actual quantity of wind per pound of coke figured from the analysis of the top gases and the actual coke burned at a number of furnaces for a long period of time was over 51 cu. ft., or more than 25 per cent. in excess of that shown by Mr. Valentine's figures.

I have recently been advised by Prof. J. W. Richards, of Lehigh University, that he has gone over the data given by Mr. Valentine by two different methods and that both show clearly that the actual air delivered by this turbo blower must have been 15 per cent. greater than the calibration of the machine indicated, and making this correction it would seem that the correct displacement of the turbo only corresponds to about 3 per cent. more air delivered than the piston displacement of the antiquated engines it replaced. This is probably not altogether correct, but it does seem to be established by these data that the calibration of this turbo is decidedly incorrect and that any claims made for superior economy based on these results are a structure on a foundation of sand which will not fare well at the hands of discriminating criticism.

Considering the extremely elusive nature of air and the possibility of error in measurements of its velocity, due to the fact that this velocity at points quite close together may vary very widely, I am led to the conclusion that the calibration of these machines is not yet on a basis of such complete reliability as their manufacturers seem to believe.

The subject as a whole is a new one, and the number of those who have had an opportunity to handle turbo blowers is very limited, the writer not being among the number. It is therefore not possible at this time to say the last word on this or many other details of this subject.