RI 2524 Progress in Blast-Furnace Research

An investigation of the production of iron in the blast furnace is obviously not an easy subject for research . Enough is known of the mechanical , thermal and chemical conditions existing inside the furnace to say certainly that they are complicated. Furnace operators have found that the process is sensitive to a number of known and to many unknown operating conditions . In order to form any conclusions of value , it seems necessary to collect a large amount of data taken with considerable care and accuracy. These facts make a blast furnace research difficult but not impossible . The most serious obstacle in attacking the problem comes from the fact that it is useless in the design and operation of the blast furnace to use the method of " trial and error , which has proved valuable in so many other industrial problems . The variables are so interrelated and are so dependent one on the other that the number of experimental combinations to be tried is enormous . There are certainly not less than twenty factors in furnace operation and design that are known to influence the results . This is apparent if it is remembered that a phrase like " size and shape of the furnace" includes height of the furnace , height , diameter and angle of the bosh , inwall batter , stockline diameter , and furnace volume ; and that a phrase like " chemical and physical properties of the fuels charged " includes size and shape of the lumps , density, porosity and strength of the coke , in addition to its chemical analysis . Let us suppose that we need choose for each of twenty factors one of two values , for example , suppose we must say only that the furnace shall be " tall " or " short , " that the bosh angle shall be " flat " OF "steep , that the ore shall be " coarse " or " fine , " or that the blast has to be "hot" or " cold , " the number of resulting combinations is more than one million . In putting twenty pennies on the table , either heads up or tails up, the number of permutations possible is the twentieth power of two , slightly more than one million. It is obvious , therefore , that some sort of mathematical relationship , a set of rules of thumb , laws , curves or the like , are needed . Whether the relationships , curves or formulas are derived from experiment , from theory, or from mere guess -work, makes no difference provided they agree with the known facts . If they do not agree with known facts , they are worthless no matter what their origin or how weighty the " authority" vouching for them. Few such relationships have appeared in blast- furnace literature , and it may be doubted whether such formulas can be discovered .