868
STEEL.— Bessemer Process. Gases Evolved.
Charcoal
Bessemer
Pig Iron.
Ingot.
Iron,.
89-830
99-569
Carbon (graphitic).
2-15
—
“ (combined),.
1-95
•20
Silicon,.
1-390
•033
Sulphur,.
trace.
trace.
Phosphorus,.
•015
•020
Manganese,.
4*665
•168
100-000
100-000
Blast Furnace
Bessemer
Slag.
Convertor Slag.
Silica,.
47‘0
42-85
Alumina,.
2*40
3-94
•45
15-62
Manganese,.
4'46
36-83
Magnesia,.
18'10
trace.
Lime,.
26-75
*70
Potash,.
*62
_
Soda,.....
■41
_
Phosphoric acid,.
•Oil
•015
100-001
99-955
A consideration of the above analyses affords fullinformation of the changes that take place in themetal during the blow. They show clearly thatcarbon, silicon, and manganese can be removed bythe action of the blast, while phosphorus, sulphur,and copper are not removed.
The relative rapidity of oxidation of carbon,silicon, and manganese, in the Bessemer and Siemensprocesses, as deduced from the above and followinganalyses, are shown graphically in the accompanying
Fig. 18.
4 2-5 _ 3 2'5 2 1'5 1 0 5
diagram, after Hackney. B,, Ncuberg sample; B 2 ,Atlas sample; B g , Dowlais sample; manganese, whiteline; silicon, dotted line. The carbon per cent, isshown by the top row of figures. This diagramproves that the old notion, that all the silicon wasburnt out before the carbon began to be acted upon,is not true. Manganese and silicon burn out rathermore rapidly than carbon, for when the carbon hasbeen reduced from 4 to 2-5 per cent, the manganesehas been reduced from 3-5 to 1-6 per cent., and the
silicon from 2 to -4 per cent.; and in all these samplesit is seen that, when silicon and manganese havebeen almost removed, there yet remains ^ to 1 percent, carbon. The action is to some extent one ofquantity; for if the silicon be much over 3 per cent,in English pig, free silicon will be left in the steelafter all the carbon has been burnt out. Again,there is sufficient manganese in the Swedish andGerman iron to allow of a small quantity being leftin the steel, to enable it to be worked when thecarbon has been reduced to a low point.
Up to the present time all attempts to use pigiron containing phosphorus and sulphur for theBessemer process have failed, because these elementspersistently remain after the blow has been com-pleted. Snelus believes that this is due to the fact,that in every case a strongly acid slag is formed,which is not favourable to the oxidation of sulphurand phosphorus, inasmuch as there is never any freebase for the oxidised produce of these bodies tocombine with; and hence he has patented the useof lime or limestone for the lining of Bessemer andother furnaces in which iron is treated, so as tobe able to use a strongly basic slag, such as isobtained in the puddling furnace, as well as havingthe most refractory, and at the same time one of thecheapest materials, that can be used for the purpose.Experiments made some years ago have proved thetruth of this hypothesis.
Composition of the Gases evolved from the BessemerConvertor during the Blow. —A study of the nature ofthe gases emitted from the convertor cannot fail tothrow light upon the process going on, and thefollowing paper on the subject, read before the Ironand Steel Institute by G. J. Snelus, August, 1871,contains an exposition of this portion of the subject:—■
The members of the Iron and Steel Institute willremember that Professor Roscoe, in his lectureon Spectrum Analysis, alluded to the difficulty ofdetermining the cause of the greater part of thelines in the Bessemer spectrum, and pointed outthat while the majority of observers referred themto carbon in some form, there were others whobelieved them to be mainly due to manganese.Assuming the lines to be due to carbon, it is clearthat an analysis of the gas producing the lines wouldbe a step towards solving the difficulty, since itwould show us which condition of carbon we mustrefer them to. This is important, because we knowthat carbon, in different gaseous combinations andat different temperatures, produces, as Dr. Wattshas pointed out, several distinct spectra.
Another question which suggests itself on examin-ing the spectrum is that, supposing the characteristiclines are due to carbon, “ How is it that the spectrumof carbon is not seen in the earlier stages of the blow,when an analysis of the metal proves incontestablythat carbon is being burnt in considerable quantitiesfrom the commencement?”
It appeared to me that analyses of the gas atdifferent stages of the blow might help to a solutionof these problems; for though it is generally assumedthat during the process of conversion the carbon of