T I T A N I U M I N BLAST FURPU’ACES.
BY AUGCSTEJ. ROSSI. The influence of titanium or titanic acid on the fusibility of the slags and upon the regular working of the blast furnace is a question upon which we have had but very few direct investigations in this country and in which, to the enlightened practice of Europe, has been opposed a kind of prejudice, possibly justified in certain circumstances but not based on well defined or precise grounds, and frequently in contradiction with the facts. Phosphorus has certainly an injurious effect on iron and cast iron, but for many purposes to which the latter is applied phosphorus pig is much in demand for fine castings not requiring great strength, and, although in general it is a n element to be avoided in a n iron ore, the question is one of relative quantities admissible for certain purposes. But while as to phosphorus the subject has been exhausted, so that we know with certainty between what limits me can use l>hosphorus ores with :dvantage, the same cannot be sitid of titanium. The question may be stated as folloivs : Is titanium an element so obnoxious that ores, otherwise excellent, easily and cheaply obtained, shovld be rejected on account of its presence, even in smail or limited quantities? Is it the excess of thiselementwhich has caused this general prejudice, and is the latter justifiable in all cases? Are there any limits within which the presence of titanium would pass unobserved and would not influence the working of the furnace? We have endeavored in ‘this paper to throw some light on these points by presenting the facts collected upon the subject in the course of certain researches that we hare been called upon to make. We shall give, in a condensed form, the literature of titan-
92
TIT.\XIL'.\I IN BLAST FTRNACES.
inm, satisfied if we can contribute i i i some measure the elements of a conscientions investigation, independent of any personal or p recon ce i ved opinion T h a t a prejudice may have existed for ,z long time without any apparent reasoii is strikingly illustrated in tlie practice of t h c blast furnace. For many y t w s lime aloiie has been considered n s the only real fluxing clement in a limestoiie. ('alcites, oftcii more diflicult or more expensive to ohtain, have been exclusively used :is fliixes, while dolomites, near a t h a n d and cheaper, have been rejected u n d e r the impression t h a t magnesia had 110 fluxing properties. If ~iic:liail impression arose from the fact that magnesium silicRtc is infusiblc, the same could have been sitid with equal truth of aluniiniuin silicate, and still aluminons ores have been almay considered as very advantageons in the blast furnace. niit the s i m e o1)jection would apply to lime, for of all the compounds that si1ic:a can form with lime nlo??e?only onc is prwtically fnsible iii the blast furnace. Lime in excess in :Lslag will render it so infusible aiid so pasty its to be t h e cause of grave disorders, ii filling up the lreartli with iiifusible blocks, which, in sonic cases, 11i~vcl m i i foiiiid t,o weigh as much as SO fairs. Will it IJC said t1i:it lime 11:~s110 fliLsing properties? The fact is that tlic clonl,le silicatcs of lime an11 magnesia or of alumina and magnesia iire t:iiormousl?- inorc fusible than tlie simplc silicaks of these bases : ~ n dc/s~f'i:sihlao r eveii inore so than tile corresponding silicates o f litnt: ;111(1 ;tluriiiuii whicli for wars have constitutctl the slags of ni:iny furiiiiccs c:xclasi\-ely. 'I'he introtlnctioii of :I. tliircl liasc into ii tlonljle silicate will coiisiderably increase its fusibility, triiile silicates f'iisiiig much niore readily than double silica;ites. Ikside; 15s fttr as saturation of silica is coiiceriiecl, one p01.ind of rnagiiesia. to forin ;L silicate of R certain character, :L neutral silicate. for iiistaiicc. will take 111)1 -pounds of silica, while limc will take u p only 1.0; 1))s. lielice wit'li ;I tlolornite as a flux, with ores containing a certain qnantity of :Lluiiiina, as they almost invariably do, there n-ill bc n o t only n skiving (J11 the first cost,, but the q u m t i t y of liniestoiic necessary will be decreased, the aniount of foreign matters t o be fnsed diminished. and t h e fusibility of the slag increased. It is only at a comparatively recent date, however, ttint itoii masters have recognized these
.
TITANIUY IN RLAST FURNACES.
93
facts, and such an authority in metallurgy as Dr. Percy stated in 1864 t h a t ‘‘ The use of dolomite ns $tix instead of calcite must be avoided; it tends to produce infusibility of the slay” (Percy’s Metallurgy, p. 506). Titanic acid is a substance very tedious and difficult to determine in iron ores and pig irons, as every chemist knows. Sometimes, according to the state in which i t is present in the ores, it may be precipitated with the insoluble residue of silica obtained from t h e fluxing of t h e ores with alkaline carbonates and nitrates and subsequent treatment, and i t may then be mistaken for silica. Although it imparts to the latter a yellowish coloration when ignited, this coloration disappears by cooling and it may be overlooked, or, when titanic acid is present in small quantities, the coloration may not be noticed. It may and will generally pass into the acid solution obtained from the filtration of t h e insoluble residue and there, occasionally, may show its presence by causing the filtrate to run milky when the precipitate on the filter is washed with water. Frequently these indications are lacking and, if not specially determined by approved methods, titanic acid will be overlooked and mistaken for iron, to which it “sticks,” as Rlr. Riley says (Journal of British Iron and Steel Institute). It will vitiate the determination of phosphorus, in many cases absolutely preventing the precipitation of phosphoric acid. (See for determination of titaninni in iron ores, pigs and steel, the excellent papers of Messrs. Shirmer, Drown and others, American Institute of Mining Engineers ; A. A. Blair’s Analysis of Iron, Tenth Census of the U. S., Vol. XV., ISSO.)
Literatwe of Titanium. Titanium is an element found almost everywhere. It is present in variable quantities in many rocks, ores and minerals. Magnetic ores frequently contain titanic acid, and thus titanium finds its way into many blast furnace slags and into pig irons. It is met with in trap, basalt, mica, gneiss, garnet, amphibole, hornblende, etc., hence in many clays and certain mineral waters. I t has been detected in meteorites and is an important constituent of the solar atmosphere (Roscoe & Schorlemmer’s Chemistry, p. 255), its pres-
ence being indicated by certain dark bands in the ultra violet spectrum (\\-atts’ 1)ictionary of C‘hemistry, ‘13;2, p. 197G) ; it is found in nearly all crystalline rocks, hematites and magnetites (Baucrman’s Metallurgy, 11. 53, etc. j. Titanic acid is found in comparatively large quantities in the ash of coal. Tery many samples of coal from widely separated districts gave it in every instance (Chemical Xews, 1583, p. 157). in some cases as muzh as 0.16;; of the ash. It is also found in limestones in small quantities, and iii fact in all t h e materials t h a t enter :L blast furn:ice. It is fount1 i n many iron ores :uid clays, and generally in siliciites, consequently also in blast fnrnace slags (Fresenius Quantitative i h i l y s i s . p. 1 5 j . Chemically, titaniuni is closely related t o tin : it iised t o be described as a rare element, I ) u t it has been found lately to exist in considerable quantities i i i i i v n ores and ciays (1:losh:im’s Chemistry, Titanium). and may ?,e re,gardecl as one of the usual constituents, having beeii very generally found associated with iron ores (21. A. 13lair. Cliemical An-
-411 of t h e rocks and minerals from Berks, Montgomery and 1’l:iladelphi:i ( ~ ~ tics 1 1 (I’L), syenite, dolerite, gneiss, pyroxene, schist, I i o ~ ~ ~ i l t l e r i i i the e, rock formation of tlic famous Cornwall milles (I’LL.). contain titanic acid i l l ciiiantitiesfrom o . X $ t o 5%and more (Geological S u r v ~ yof l’ennsylvaiiin Cb., 11. (;I). It exists in the clays : ~ n dslate clays of l’ork, L4damsand Lancaster counties (l)~i,)> ill’ whicli it is to ,bc rcgarded :is a11 essential ~oniponeiit, its amount being remurkably coiistaiit (Ani. Inst. of 3Iin. Eng., 6. 1). 1W). Froni rocks it finds its way into ores. 1 n -le70 Jersey t h e very gcncial presence of this eleineiit in the clays and some of the gneiss rocks lias led to the belief t h a t it is rarely absent from irou ores and nearly all of the later aiialyses of ores of this State since 1870 show its presence from traces to 1 1 . G O $ and more of T i 0 2 (Geological Survey of Yew Jersey, 1$;‘3, p. 152), and still some of these ores, consideked aniougst the finest of the country. have been and :we daily smelted in blast furnaces in Pennsylvania and S e w Jersey i i i admixture with other ores. ‘l‘lie Dickerson mine, one of the very best, contains 0.70$ TiO,, the Mount
TITANIUM I N BLAST FURNACES.
95
Pleasant 0.58$, the Teebo 0.59$, the Hibernia mine 0.55% (Arr Inst. Min. Eng., Vol. XIV.). The Ringwood mines (Passaic Co.), worked since the last century, and belonging now to Messrs. Cooper & Hewitt, have supplied for years the ores used ih the Durham and Ringwood furnaces (Am. Inst. Min. Eng., Vol. XIV.). They contain from 0.30% to 2.72$ TiO, (Geological Survey of New Jersey, 18'73, p. 53). Other orescontaining 11.60% TiO, have been smelted in the same furnaces or in others in admixture (Ib,, 1873, p. 55, et seq.). Ores from New Jersey containing F.23$, 5.21$,11.606 TiO, have been smeltedalone in a blast furnace for several months, or nearly one year (personal information from the manager of the furnace). Iron ores of many districts of ATortuay and Sweden, forming enormous, almost unlimited deposits, and containing in some case8 25 to 45%of TiO, are waiting for the development of an industry which only t h e lack of combustibles otner than charcoal impedes (Journal of British Iron and Steel Institute, 1580, p. 132) ; many of them have been smelted in these countries and some have been imported to England for treatment. It IS present in the ores of the district of Ivenkeping (Sweden), to the extent of 6.37% TiO,, yet these ores are smelted alone in the blast furnaces of this region furnishing one of the well known brands of Swedigh iron (Jordan, Revue de 1'Exposition Universelle, 1867, p. 155,II.); in the ores of Newbottom district 9.108 TiO,, of the Norland province, Ulfo, 9.50$ TiO,, of Kronaberg district 8.50% TiO, (Report of the U. S. Commissioner to the Vienna Exhibition, 1873, p. 182), and in many other ores of that country and Norway ?.lo$ to 15.10%and 40.80% TiO, (Prof. Forbes, Jour. of Br. Ir. and St. Inst., lS80). It is found in Russian ores smelted in blast furnaces in the U r d district (4.06% to 4.86% TiO,) ; this district manufactures lS,OOO tons of pig metal per year (Revue de 1'Exposition Universelle de 1867, p. 548, III., Jordan) ; in ores smelted in Bavaria in the blast furnaces of Hochstein and Eisenberg (Munzerheim ores) 1.396% TiO, (Ib., p. 594, 111.). A slag of the Concordia Iron Works (Coblenz), contained 6. '70$ TiO, (Percy's Metallurgy, p. 515) ; that of a Styrian charcoal furnace 6.71%TiO,
(Osborn’s Metallurgy, p. 12s):t h a t of a Swedish furnace, Ezerholm, It is found in h u t r i m and other Irish hoy orcs, aluminous ores, in quantities varying from 3.51% to 5% TiO, (Percy’s Netallurgy, p. ?07, 2 2 j ) , ores extensively importcd into the United Statcs by furnaces in Pennsylvania, and elsewhere hcrc and i n England smclted i r admixture with otlicr orcs (Kirnball, Am. Inst. 3Iin. Eng., Vol. IS., p. 14). The Irish Hill 3iiniiig Co. orescontain 5.YOto li.20$ TiO, (Chemical Kews, 1881, 1). ?N),(Encyclopedia Britannica, p. 81). It is found in bauxites (Frriiace and elsewhere), mixed as a flus with other ores, 3.30% TiO, (Ib.). Ln Algerin, Oran province ores I.%$ TiO:, in 8wdi?iia, Iglesias ores 1.25% TiO, (Vathaire. les Hauts Fourneaux). I n many iron sands of ,Sicily 9.90% TiO, (Revue de 1’Exposition Universelle, 18G7), (Bauerman’s Metallurgy), and in ~\~eroZenlcbird (Ibidem). It is found in t’eiz~isylvriaici oTes in quantities of 0.30% TiO,, in RlcIlvee ores, York Co. (Geological Survey of Penn. MM., A. E : ) , in Chestnut Hill ore bank, mined by theBerks Go: Mining Company, 3.93$ Y’iO, ( I b . , AIM., p. azo), in Brandywine ores 10.448 T i 0 , smelted in blast furnace without any trouble in admixture with others in the proportion of 10,: in the charge, about 1.5$ TiO,, average ; in the magnetic sands of Adams Co. ?.37$ TiO, (Geolog. Sur.’ of Penn., 11. 3, p. 101, M X , p. 227 e t sty.). It is found in - 1 - e ~ ~Yovk State ores alniost constantly (10th Census of the United States, 1-01. XI-.,p. 5 5 5 , 556, e t c . ) , in Chateaugay ores 0.47$ Ti09 smelted in blast furnaces (Am. Inst. of Min. Eng., p. 81, 1’01. IX.), in ores of the northern part of the State 0.733$ TiO, (Ib., Yo]. SIT’,), in Lake Champlain ores, Westport, TiO, -l.ZS$ (Ib., lro1. S I . , 1). 159), split rock 14.70$ TiO, (Ib., Yol. II., 1973, 1). 13), used in blast furnaces in aclmisture. I n ores from Oneida Co., Kirlilanil ores? 10 t o 40: TiO,, smelted in admixture in the Coleraiiie furnaces. In Westchestei. Co. aluminous ores Z41$ to 4.41$ Y’i09, recommended in adinixture with others (Kemball Ib., T701. IS., p. 14), in limonites of Staten Island (10th Census of the U. S., p. 194) In -JTorth CLiToliiia o m s in proportions varying from 12.08$ to 13.71:: TiO, (10th Census U. S., p. 311), and u p to 39.Glg (Ib., 9.20% TiO, (Ib.) (Vathaire les Hauts Fourneaus).
TITANIUM I N BLAST FURNACES.
97
p. 562), in Cranberry ore banks 0.95 titanjc acid, in Chatham 00. ores 1.04% TiO,, in Mitchell Co. ores 0.95% TiO,, u p to 5.33$, metallic iron 65.44$, Roan Mountain ores (Am. Inst. Min. Eng., Vol. XI., p. 159), in Centre Co. ores 8.65% TiO,, metallic iron 60.88s (10th Census), (Am. Inst. Min. Eng., Vol. XI.). I n Colorado ores, in quantities of 11.99$, 11.61$, 12.92$, 12,735, 13.84$, 14.86$, 13.06% TiO, in different districts (10th Census U. S., p. 476). In Oregon ores. Putnam Co., Oswego Furnace smelts ores containing 0.54%TiO,. I n ores of Clackamas Co. and of other districts in which its presence was ascertained but quantities not determined (10th Census, p. 566). I n Rhode IsZand ores, Iron Hill Mine, Providence Co., 9.35% TiO,. In Virginia ores, in which i t is “ present ” in almost all the ores in quantities not determined, in some TiO,, 0.16% (10th Census, p. 26’1-576) and u p to 6.53% TiO,, metallic iron 52.20%in North Garden ores (Am. Inst. of Min. Eng., 6, p. 159). I n South Carobinu ores, Silver Mountain bank, York Co., TiO, 39.67% (10th Census U. S., p. 267-576, etc.). I n many other ores of different States its “presence” is “quoted” but without any amount being given (10th Census U. S.). I n some snalyses it may figure in the (‘Insoluble Residue aitd Silicious nintters,” but no analysis of the latter is given (Geological Survey of Pennsylvania and of other States). It has been found in the C‘uncidian. ores, 11.2’1$ TiO,. (Revue de 1’Exposition Universelle de 1867, p: 685), in the Bay St. Paul ores, 48.60% TiO, smelted in two blast furnaces in 1873 furnishing an excellent iron (Mr. Riley, Jour. Br. Ir. & St. Inst., 1874, p. 132), they constitute the mineral ilmenite (Am. Inst. Min. Eng., II., p. 13), in Moisie River sands, 4.15% to 28.955 TiO, (Bauerman’s Metallurgy. p. 53, et seep.) (Chemical News). It is n constant elemeut in the English ores smelted in the Cleveland district in the furnaces of Sir Lothian Bell and others. The Cleveland ores contain a good deal of titanium (Mr. Riley, Jour. Rr. Ir. & St. Inst., 1574, p. 132), while in the R’orthampshire it is found most readily, even in the cinders. In fact, titanic acid
9s
TITANICM IN BLAST PVRNACES.
occurs in all clays. It is a very coininou nzaterial, very difficult to find ; it has a great tendency to stick t u t h e iron, nnd h a s been very qfteiz estimated as m i l l e of iroii a n d mistakeit, &foi* it (hlr. Riley, Ib., p. 132, et s q . ) . It has been found frequently in the U. S., i n the Lehigh furna. ces, in the crevices of the hearth after the furnace had been blown out. 11s for instance a t the Crane Iron IVorks, and pig iron from the same works mas colored p w i y d e by the presence of titniaiiiin ; the ores containing titanium were from Morris Go., New Jersey (Osborn’s Metallurgy, p. 4S5j. It would appear from the preceding quotations t h a t titanium is not a rare element, nor one confined to special localities or countries ; in fact, it is met with in a number of iron ores hereaiid in Europe and Africa, etc., i n quantities from traces to 50% of TiO,. T h a t some of these ores, a t least, have been smelted in blast furnaces sufficiently appears from the above, and from the fact t h a t many pig irons both here and in E u r o p contain it as a constant element in very notable quantities, :ind h a v e contained it for years without any mention being made of troubles resulting from the iise of such ores in a blast furnace or from the slags r u n with these pigs, which slags did also coiitain titanium, a s we will see further. When a proportion of titaniferous ore is added to the charge, it increases the strength of the metal, a t the same time giving it a peculiar mottled character (Banerrnun’s Metallurgy, 1SS4, 1’. 53). Titanium may be present in pig iron to the extent of about 15 (1,G4$ TiO,) : its presence was considered as R very favorable factor (Ib.). Pig irons containing 0.47$, 0.;1$, 1.15s titanium, corresponding t o 0. ;:$, l , l G $ , l.SS$ TiO,, have been regularly produced in English furnaces from a mixture of 7+$ ilmenite, containing 38.845 TiO, with red hematites and local ores, which supposes in the mixture a n average of about 2.913$ TiO,, admitting t h a t the local ores did not, contain any (Percy’s Metallurgy, p. 551). Sir Lothian Bell has obtained from regular mixture of English local ores used daily i n the furnaces of the Cleveland district, Cornish ore, red hematite, Irish bog ore, pig irons containing respectively,
+
+
+
99
TITAKIUJI I N BLAST FURNACES.
0.79$, 1.15$, 1.G%9$ titanium, corresponding to 1.29$, l.SS$, 2.SSfR TiO, (Percy’s Metallurgy, p. 552), (Memoir of Lothian Bell to the British Associtttion, 18G3), and the slag run with these pigs contained 0.75$ TiO, (Percy’s Metallurgy, French translation under the auspices of the author, 1865), (Memoir of Lothian Bell, 1863), (Phenomena of Iron Smelting, L. Bell, 1873). Cleveland pigs contain titanium in a regular and constant manner as a noriitnl constituent, as analyses show, in quantities of 0.09$, 0.13$, 0.14$, 0.20$, 0.22%, 0.266, 0.51$, 0.56$, corresponding to 0.15$, O.?l$, 0.23$, 0.33$, 0.36$, 0.42$, 0.84$, 0.92$ TiO,, and u p to 17; titaniu? (1.64 TiO,), and above (Percy’s Metallurgy, French Translation), (MemoirLothian Bell, 1SG3), (J.of B. Ir. &St. Inst.). Mr. Riley says (J.of BY.Ir. & St. Inst., l8S0, p. 190), titanium is alzanys present in fair amounts u p to 1%(1.64 TiO,), but rarely above that in Bessemer or Clevelnd pig ; it is always €ound in Bessemer pig. I n the discussion of the paper which has given rise to these observations, not the slightest mention is made of troubles resulting for the furnace €rom the normal and daily use of ores capable of supplying the above percentages of TiO, t o the pig, and the English furnaces, as is well known, have longer runs than the American, a run of eight years or more not being a rare case with the former. Pig irons smelted in this country have shown titanium also as a pretty constant element, and the examples quoted below could be multiplied. The analyses were made by Messrs. Drown and Shirmer (Am. Inst. Min. Eng., 17, p. 346). Pig Irons.
Richmond, Greenwood, Hecla, Du tc hess, Glendon, Silver Gray, Leesport, Bushong, Unknown, * Perryville,
Titanium.
Titanic Acid.
0.01s 0.05% 0.045 0.055 0.099 0.114 0.115 0.225 0.318
0.030 0.084 0.080 0.090 0.160 0.190 0.191 0.370 0.520 0.066
0.040
*This latter pig was analyzed by the writer.
Ores containing less than SJ$ of titanic acid are not recognizcd any more as titaniferous ores (Dana’s Mine~,alogyj. Ore cont:tining :19.20$ TiO,, ilmenite, have been successfiilly worked for s w era1 years a t S o r t o n (England) Ly the Sorwegii~nTitanic Ii.oii Co. The slag 1’1111 from tlie furnrice contaiiir:il 26.I d < ‘I’i02, : ~ n d2 7 , S3$ silica. The uncertaiii tv of importation, tlic leaiinrss of the ores, which contained only 32 to 3ci.S1:6 iron, tlie amount of coal necessary to run in slag such itn enormous aniount o f silica and titanic acid (64.015) prevented the sciicme f r o m \icing financially profitable after a few years. 1 %t the ~ metallurgy of the treatment was a success, and showecl tliut, with proper management, a furnacc can carry such R Iwrceeritagc of titanic acid in the charges witliout clogging (Win. 11. Homreii, A m . Inst. llin. Eng., 11, p. l59j, (Osborn Netallurgy, 1). 4T-t). The C’andian Titanic Ore Co. smelted in two blast furnaces ores from 13ay St. l’aul quot,ed above, containing -19.GO$ TiO,, withoiit any trouble i n t h e furnaces. Tlic tires \vert’ very lcnii (MY. Riley, J o u r n a l of 13r. Ir. ti St. Inst., lSi4. I). 132 ; 1876. 1). l!)O). A good liquid slag can be obtained IoitJLout citiy t l ( f F ’ C U ~ f ~with J ores containing 3.555 TiOg (Prof. Forbes Am. Inst. of >fin. Eng., L‘ol. Sl., p. 159). With proper chargesand fliixes, with a n amount of titanic acid not above %, it is not diflicult to work an ore clenn2y and pr@tuhZy (Ih.). With a n acid cinder and it slack blast, ores containing from 8 to 13% Ti02 can be readily treated without any trouble (Ibidem). (Ohcmical Sews, J k c . 11, 1SS8.) These are the statements and practice of an eminent chemist and metallurgist who has been for years consulting engineer for blast furnace managers in Swedeii, S o r w a y and England, who has supervised the running of many furnaces in Sweden and Korway, smelting titaniferous ores, and the furnace at S o r t o n (England), treltting the ilmenite from Korway mentioned a’bove. ~ ~ l t h o u gtroublesome h when present in l u g e quantities and having ii teiicfcncy to render the slags pasty, in working the furnace to a certain kind of pig and iiot a t 5 too high tcmperatnre, thc ,p’u.ce.s of the furnace require but z-ery little f o w i y i i matters in working with titaniferous ores (Osborn Aletallnrgy, 1). 331). Scandinavian iron masters have often expressed to Prof. Forbes
TITANIUJI IN BLAST FURKACES.
101
their surprise a t the want of information possessed by English metallurgists in general upoil the subject of highly titaniferous ores, containing such percentagcs as?&%to 40$ titanic acid. Their ciperience has shown that the only objection to the use of such ores is that they are found to be more or loss refractory, as they con tain a grcater pcrcentage of titanic acid, and if much titanium is present they require a quantity of coal so much larger to smelt them as to render their employment alone not profitable when ores free from titanium can be obtained cheaply in the district. After considerable experience in smcltirig these ores, the Norwegian ilmenite and others which yield a very good iron, it was not found profitable to smelt them alone for the preceding reasons, b u t their use was found beneficial when employed in about equal proportions with the other ores of t h e district free from titanium (Prof. Forbes, J. Br. Ir. CYS St. Inst., 1877, Vol. XI.). Titanium iron is essentially a forge iron. A furnace that can .make iron under an acid cinder and a slack blast and keep the silica out will not be troubled with titniizzini deposits; the details of furnaces for smelting these ores are like those suitable for ordinary hematites or magnetites (Am. Inst. Min. Eng., 1877, 2.) Titanic acid is also found in slags. Slags containing 0.75$ TiO, have been run in the Cleveland district from the smelting of local ores (Percy’s French Translation) (Rewe de Z’Expositioiz Cnivsrselle, 1). 37). The Concordia iron works slag, quoted by Percy (see above), contained 6.70% TiO, ; one quoted by Osborn as run from a styrian charcoal furnace the same ; a slag from Ekersholm, Smoland district (Sweden), contained B.OO$ TiO, (Osborn), (Vathaire, Les Hauts Fourneaux, p. 41). Blast furnace slags contain titanic acid when treating titanic ores. It is combined with the bases and silica. It does not notably change the appearance of the slag and nothing but an analysis can discover its presence. I t gives a violet coloration to the cinder, t h e shade of which is more blue than that due to oxide of manganese, but this blue coloration can be imparted to t h e siliptes by so many different causes, t h a t it cannot be considered as a proof of the prescnce of t i t a n k acid (Rivot’s Docimacy, p. 157), (Revue de 1’Exposition Universelle, 1867, Swedish ores).
TVlicn snieltiiig titatiifcrous ores with white iron most of t h e ti{!o(icctci'rl e i c ! / ~ r s{ l i e d n y , the pig iron is said to yield an excellent wi~ougliti r o n a i i t l &tee],and aslittle tit:iiiinni is found in these p r o ducts. i t innst bc due t o the iiiclirc.ct b ~ ~ ~ l e f icffect ~ i i l l of smeItingoi*t~s coiit;iiiiiiig titaiiic acid C i ~ ~ ) l . ; ~Xtii~pt:ttio~l ~.'s from the Gcrni:in Kdition (if I'rol'. Kc~rlil's Nctailurgy, London, lS69, 1). 316). .ll,out 1: titanium ( 1 . M 'l'iO?) ni:i! 1x1 i ) ~ ~ s ei ni i tthe pig. it incrwses the strength of tlic nietal ( l b . , p. :;I ; ) . Tlic iron obtained a t S o r t o n ( b h g l ; u ~ dfro111 ) S o i v e g i u i i ilnicnitc ~vitsfonnd t o be estrcmcly stroiig a n d \ws ~isecl i n Europe for armor ~)intescomrnqntling three tinivs tlic pric-c! of ordinary iron (Osbor11's Jletallurgy, 1). 4 7 4 ) . 'l'liis Kortoii iron obtaiiicd from ilmeiiit~.corit:iining 4 0 . 9 ~ ; ~ titaiiic ticid, 30.206 iis :tverage of cargo, went to the nkmor plates of Slieffielcl 011 accouiit o f the touglinsss wliioli this iron riot only possessed but imparts t o other iron i n : d m i s t u w in the puddling f n r n a c c ~ I\.lion rolled into plat& and merclian t bars: it possesses such $1 c l c n ~ i ,,s?t'f, toztyii ~ ! i t ~asi ~should e r e i ~ d e rit invalnable for boiler plates, sheets of cold stampifig and kindred uses . . . and ores now useless rould LE r n d c of vast service to the iroii trade and t o the industries (blr. Ileby. Secretary 13r. I r . &'St. Associntion, ?, p. 19, is;; ). Titanium is fouiicl iii very variable proportions in cert:~inpig iroii, to wliich it appears to conirriuiiicatc a great teiiacity (liirot's Docirnacic, 1). IT,(j). I t increases t11e.st~re~igtli of the metal (13auerman, 11. 53j. Ores containing tit;ii~iumniay be considered very fnvorable foi. the manufucture of pig metitl: in rnany respects they are preferable to the .spathic ores. foi-, with very little :ittentioil they will produce white iron ricli i n carbon, tlic \wry material required for German steel (Osborn blctallurgv, p. 431). I n Sweden it is n cornmon pntctice to add lO$ of titaniferous ores to the charges to remove sulphur. IYhatever m;iy bc thc reason of this effect, what is known witli certainty is t h a t titanic ores i n the .I-. S.,Canada, S e w Zealand, Swedcn and Worway, are such t h a t the metal produced from them is wonderfully good (Dr. Forbes, J. of Hr. 11,. Cy; St. Ilist., 1574, p. 191). Titaniferous
‘KITANIUM IN BLAST FURNACES.
103
ores containing 486 TiO, smelted in Canada have furnished an ezcclleizt p i g (Mr. Riley, Ib., p. 132). I n Norway and Sweden the ores of Krager and Eger, containing 15$ and 7.10$ TiO, respectively, are worked in blast furnaces. If ores do not contain more than 8% TiO, their reduction is not difficult and the product is of good quality. An examination of the ores analyzed in this report (Report 1879 Geological Survey of New Jersey) shows the presence of titanic acid in these New Jersey ores from traces to 5$TiO,. These are allused successfully aiid are said t o niake yoodiroit (Prof. Cook, Geology of New Jersey, 1873, p. 152). The New Jersey ores, containing ll.SO$’, 6.23$, S.21$ TiO,, and averaging about 8.50$ TiO, in the charges, have been smelted alone in a blast furnace for nearly one year and have yielded a pig iron of a "remarkable strength,” (‘which could almost be bent in two without breaking.” The cast iron was graphitic and considered an exceptionally good KO. 1grayfouiidry iron. It commltnded, i t is claimed, one dollar more per ton (personal information). Putting aside the questions of the iq~rovemeiatand good p u l i ties ckzinzed for the iron obtained from titaniferous ores, the practice in Europe and even in this country, within certain limits, seems to establish the fact that titanium cannot- be considered as being altutcys and necessarily a cause of trouble in the blast furnace if properly dealt with, even when present in very large quantities, since ores containing as much as 40$, and even 48$,of titanic acid have been smelted, in known cases, for several years in the same furnace as successftiUy, if not as profitably, as others richer in iron and frce from titanium or containing lcsser proportions of it. On the other hand, certain blast furnace managers in this country go as far as to maintain that i?Lapprecinble paiztities, mere traces of titanic acid, are sufficient to interfere with the good working ot a furnace-clogging and stopping it. Others, a little mort? liberal, will not even admit a limit of 0.25% of titanic acid in an iron ore, nor will they consider it practicable to use i t in a blast furnace. F o r them the smaller percentage is as deleterious as the
104
TITdKIUJf IN BL.4ST FCRXACES.
larger, t h e only difference being t h a t the action is smaller a n d t h h it requires a longer time to fill. u p t h e furnace. To these extremists it might !Jeanswered that, unconsciousiy, then, many iron masters have produced in this country, daily and normally, pig irons in which the quantity of titanium lias reached as much as 0.318$, corresponctiiig to 0.3:?$ TiOZ (from 0.03;~ to 0 . 5 2 $ , considering 0n2y t h e analysis given above). d l 1 of these proportioiw, itwording to their view, the sinallest as well as the largest, were suficient t o have ultimately clogged t h e furnaces, and still the latter have had years of excellent rnnnitig as regards economy and quality of product, a t least some of them did hiire, since one of these pig irons, the Busliong,” containing 0.37$ ‘l’iOz, has corumanded one dollar more per ton 011 the market. Admitting t h a t t h e ores smelted contained 50s of iron, t h a t is, t h a t two tons of ore \yere fluxed per toil of pig metal, it quantity of 0 . 5 2 s TiO, in the pig would correspond t o 0.26% in t h e ores, arid this supposes that there was none left to pass into t h e slag ; how then could the latter be rendered pasty or ii!#ksibZe ? Furthermore, if dZ the titanic acid is t h u s accounted for in the gig, nolie has been left in the furnace to clog it and we are in this dilemma : either some titanic acid did pass into the slug a n d t h e ores contained inore than O.?G$ ’Pi O z or ull qf the titauic acid call. bs v i d e t o puss i i d o t h e p i g ( L l i d thzis lsnue the .furuuce. T h a t none was left in these furnaces, not even trnces, if we adopt the views of these extremists, appears to be suficiently proved by the fact t h a t t h e pig irons analyzed were not special specimens, but the regular product of the iron works from which they came. We find in England, in the Cleveland district, in Staffordshire aiid Northampshire, smelted daily from the regular local ores, normal pigswliich, as Mr. Riley says, contain titanium in good quantity, u p to one per cent. or more and occasionally as much as 1.6$, corresponding to 1.64% or even its much as .?.S7$ TiO,, and t h e siime authority adds, ‘‘ Titanic acid is also found ~ u o s fveadily even in the slags” of these districts. We have every reason then to suppose t h a t t h e slags r u n with the American pigs quoted did olso contain titauic acid. \Viih materials SO$ rich in iron t h e ((
TITANIVM IN BLAST FURNACES.
105
preceding figures would correspond to a t least 0.826 to 1.43%TiO, in the English ores : with materials only 4 0 s rich to, a t least, O.G7$ to 1.18%in the charges, and these English furnaces h:ive much longer runs than the American. We know from analysis given by Percy and others that a t least the Irish bog ores " nsrd ;tbout in proportion of one-third with the other local ores, contain (Antrim) 3.51% to 56 TiO,. This would bring the proportion of titanic acid in the mixture to from 1.17 to 1.66$, or 0.70$ to 1.00% titanium, figures which correspond very nearly with those found for titanium in the normal pigs of the district. Even in this country certain iron masters more enterprising and more experienced than others have successfully disposed of several thousands of tons of ores containing from 9% to 15% titanic acid, itverage 12$, by smelting them in a mixture with other ores in such proportion that the average proportion of titanic acid was 0.38$, and with as much as 0.75$ titanic acid in the mixture with proper care they did not experience any trouble. They have gone so far as to state thnt, after many experiments with titanium, the result of their experience has been that a mixture containing a greater perceiatnge of titanic acid thnii 1.25% could not be successfully used. This admits clearly a possihlc maximum limit, and is in complete contradiction with the asscrtions of the extremists. The following is another illustration of Bmerican practice. It covers a lapse'of ten years' running of two blast furnaces. These furnaces used five different kinds of magnetites, all containing titanic acid in quantities varying from 0.33 to as much as 1.34$, as analyses made a t the time and subsequent ones have proved. The ores averaged in admixture about 0.70% titanic acid. They were exclusively smelted for that period without any trouble whatever from deposits, sluggish slags or clogging. The furnaces had runs of threeand three and a half yearsa very good run in t h e u n i t e d States-and they produced all grades of iron from white to gray No. 1, and even to graphitic iron. Titanic acid was detected in the slags. These two examples corroborate the results of the English practice. They are given here because we are particularly acquainted with the circumstances of each case and can substantiate these assertions. '(
If we pass to Sweden and Sorway we find there ores containiiig 5% to 10% TiO, smelted alone clearly and profitably in c1ia1,coal furnaces, others containing 1.55 t o 20s TiO, o r more. smelted in admixture or even alone wit,li a beneficial effect ; m d , lastly, Korway ores containing a s mncli as 40$ TiO, and only only 36$ of iron smelted by a company in EnFland for a few !ears with a perfect success as far, a t least, as tlie metnZJurgy of thci treatment was concerned. S o clogging or stoppage of thc furnace, 110 pasty'slags interfering with its working are mentioned. In Canada we notice results exactly similar with ores of nearly t h e same composition as the preceding Sorway ore. W e find slags i n England and slags run from iiormal local ores containing 0. '75s Ti O,, others quoted ahove containing 6.'70$ and 9$ T i O,, slags r u n from local ores coiitaining enough titanic acid t o supply to them these percentages. W e find pig irons in England a n d i n this country containing this element as a constant factor in quantities from 0.08 to 1.64% titanium and more. Tliere seems to be :t contradiction somewhere. Without pretending to settle t h e question by our own authority, we believe t h a t enough fkcts have been presented to justify, a t least, the following remarks : Titanic acid, though present it may be said in almost all magnetites and hematites of certain districts, in various quantities, generally small, has been mostly i y w T e d until a comparatively recent date. I n tlie analyses given in tho geological survey of S e w Jersey it does not figure until 1S70, from which date it was looked for and found regularly in different quantities, fr6m 0.30$ to l$, in ores in which its presence had not been previously suspected, and so constantly, from traces to Is$, in different ores from widely separated regions, t h a t this fact lias called for tlie statement of thc late Prof. Cook, t h a t it seems to be almost a normal element of the Kew Jersey ores (Geol. Sur. of K. J . , 1879). As Mr. Riley has observed, itsdetermination is very difficult, and it has been very oftep mistaken for o z i d e of i r o ~ I n the reports of the geological surveys of Pennsylvania, u p to l88tj-87 a t least, most of the analyses are tsbulated ': Silica and insoluble residue," n o composition of this residue is given. Titanic acid might and could have
TITANIUM I K BLAST FURNACES.
107
been present there and entirely overlooked. (See Analyses of Geolo ical Survey of Pennsylvenia.) Suc% a supposition is not a gratuitous one. I n the analyses of iron ore given in the 10th Census of the U. S., we find, amongst other examples, p. 311, under the denomination of North Carolina ores, ‘(Silica and insoluble residue, 28$,” and in the subsequent analysis of these 28% of ‘‘ residue,” titanic acid appears for 11.824&, nearly half ! Such a proportion would have been completelyignored except for special analysis. I n many other ores titanic acid figures only as “ present.” Consequently, iron masters in Pennsylvania, New Jersey, and elsewhere must have been smelting for years, and are even smelting to the present day, in admixture with other ores or alone, local ores or magnetites and hematites from New Jersey, for whicn they have been satisfied to pay the best prices on account of their richness in iron, and in which titanic acid had not been suspected or its proportion determined u p to 187’0, and for others much later still. Yet, there has not been any complaint made of these ores on this score ; on the contrary, they have commanded a better market ; trouble has been taken to have them shipped to distant furnaces. We could quote many such ores, but limit ourselves to the Dickerson mine, O.S9$ Ti 0, ; Mt. Pleasant Mine, 0.58% Ti 0, ; Teebo bfine, 0.59%Ti 0,, all in Morris Co.; the Ringwood Mines (Pausaic Go.), 0.30 to 2.72$ T i 0, ; the Hibernia Mine, 0.55$ Ti 0,. (See Geolog. Survey of N. J.) (Am. Inst. 11. Eng.) In the analyses of the Dickerson Mine, Mt. Pleasant, Teebo, Hibernia, given in 1866 in the Geological Survey of New Jersey, titanic acid does ?sot appear, not hnvitrg been sought foi..
.-
~
~
....
_
~
-
_
............
. . .
. . . . . . . . . .
......
...
s. s;
*.s;
1.15 3.13 80.59 YO.5 ! ) 3.4'4
1.31 ( i . 35 ........
.____ ~
, I. . .
..........................
..
-...- -
~
.~ . .
mitli.
Is titanic acid, then, ~ Z m y and s i l l ~ 1 1 quc?ntitiex ' iln objcctionable element in ai1 iron o r e ? Is it, n o t more correct t o say that, within certain limits, wliich we niay perhaps fix a t 1,OO$ or more, it lia3 proved little objectionable : on thr. contr:iry. that it lias 1mssed for years iiiinoticed ? 3-0 disturbances have occurrtd in the running of the furriitccs iiiikiiowingly snielting such titaniferous ores t,o call attei! t.ioii to :I cIctrinic~nt:il clement. So much seeins to LC ass~ircil, n i i t l coilfirmet1 1)). t h e Eiiglisli practice. Above these quantities what is t h e limit :it which its presence in an iron ore ~ v o n l i lbegin t o be felt by its effect upon the riiiiiiiiig of the furnace i n i.es!)ect to econoniy o r a t wliicli tlie judicious treatment, such :is :i clinnge i i i tlicx isoutirie of the charges aiid fluxes adopted for non-titrtuifcrous ores, might become
TITANIUM IN BLAST F U R X A C E S .
109
necessary to insure both economy and success? American furnaces may have unconsciously worked ores containing much more titanic acid than 1 or 1.25$, and some have done SO to our positive knowledge (8% to 10% titanic acid in the charges, as derived, a t least, from verbal assurances of certain iron masters), without experiencing any particular troubles, but in the absence of direct and well authenticated practice, and especially of 7-eguZar analyses of the clinryes in the case alluded to, it may not be fair to draw a conclusion.
If, as stated by Dana, an ore ceases to be recognized as titaniferous when i t contains less than 31% TiO,, this figure might be called t h e limit, and the personal observations of Prof. Forbes affords a corroboration of this statement. Beyond that, if we take to guide us the European practice in Norway and Sweden, the examples quoted above would provo that, with proper fluxes, even regular titanic acid minerals, such as Ilmenite, containing from 40 to 48% of this substance, can be and have been cleanly and successfully smelted in furnaces without any disorders resulting in the latter from their use. The question of economy in the particular case mentioned depended on many factors and, a t any rate, has-nothing to do with the metallurgy of the treatment. If dealt with as corresponding proportions of silica would be when present in ores in very notable quantities, titanium has certainly a tendency to produce “deposits” and to render the slags less fusible. All depends on the judicious and rational choice and proportion of the fluxes to obtain a desired result. ru’othing economical or skillful can be accomplished when the charges are badly engineered (Osborn’s Metallurgy, p. 478) ; the skill of a blast furnace manager as a metallurgist consists mainly in forming a fluid slag with the materials a t his command (J. of Br. Ir. and St. Inst., 1874, p. 7 5 ) . Certain natural compounds of titanium, such as sphene, calcium silico-titanate and keilhsuite, are perfectly fnsible in a blast furnace. ( I h , p. 43.) Their composition is as follows : ‘(
110
TITAYIKX I N m A s r FURBACES.
'I
KEILHAU-
SPHESE. .
-
III.
~
.- Avrvage.
~
31.20 40.92 22.25 5.06 5.06
0
0
_~
.______-
-
I 1
31.7s 40.00 24.59 2.00 '2.00
1
P8.04 19.56 6.57 8.03
- .___ -
So eminent a chemist and metallurgist as Prof. Forbes has certainly cleared this question from all possible doubts by his practice. He has smelted, without admixture, ores containing 40% titanic acid without any trouble in the furnace. We refer to t h e excellent papers of Mr. W. M. Bowren (Am, Inst. Min. Eng., 9, 159), and to that of Mr. Deby (J. of Ur. Ir. and St. Inst., 2, 1877, p. 13) for the details of the operation. The ores treated were the Norway ilmenite, of which an average analysis of a cargo gave (see papers quoted) : r 1 litanic acid - - - - - - - - - - - - -39.20 1 The coinposition of the slag actually r n n from the furFerric oxide _ _ _ _ - - __ _ _ _ -18.,59 I,ace, as given ill the paper Ferrous oxide- - - - - - - - - - - 30.00 alluded to, approximated, as as can be obtained in Alumina _ _ _ _ _ _ _ 2.89 _ ) - -closely - . - . _ pactice, the coinposition qf Manganous Oxide- - - - - - - - 0* q i h e n e , a n n t w n l ?ni~ern2of' Silica _ _ - - _ _ - - - - 5.70 titniz1'uvz, fusible in the blast furnace. This is what good ._ _.- - _ _ _ _ - - - - - 0.22 Loss _ _ J management can do. Iron= 36.3%.
j ~
i
____
~
111
TITANIUM IN BLAST FL‘RNACES.
Average
Composition of Slag.
Sphe‘ne.
27.SY 31.78 Silica - - Titanic acid 36.1s 40.00 Lime - - 84.36 24.59 Oxideof iron - 1.SG 2.00 Alumina - - 9.1s 66 Magnesia - - . 0. GO As i t can be calculated from the preceding analyses and from the charges of the furnace as given in Mr. Bowren’s paper, all tbe titanic acid of the ore is accounted for and has mostly been made to pass into the slag, excepting the small percentage left in the pig metal, perhaps 1 to 2$, and, of course, titanium having left the furnace i t could not in any manner be the cause of obstruction by hangings or deposits. If these results were not as economically obtained as with other ores free from titanium, it may be a question if the higher prices realized for the products were not a compensation for the extra expense of combustible, or would not be if the ore, though containing as much titanic acid, had been richer in iron; but it is useless to discuss this side of the question ; the possibility of smelting successfully and cleanly such highly titaniferous ores is all that we intended to establish beyond dispute. The objectiont to the presence of titanic acid in an iron ore would thus be entirely modified if we adopt the conclusions of Prof. Forbes and the results of his practice and of that of the Swedish metallurgists. It toould be no Zoiager a question of clogging o f the fiirnace ” but sinzply one of ecouonty in ercch case. Reduced to this i t would seem that such small quantities as 1 to 1.25$, or even more, of titanic acid could not be objectionable on this score, especially if these titaniferous ores present, as they do, such characters as render them very valuable otherwise. Prof. Forbes (J. of Br. Ir. and St. Inst., 1877, p. 132) has called attention to the fact that ores of this kind, in Norway, Sweden, Canada and New Zealand, are remarkably -free f r o m phosp?Lorus and very frequently also from sulphur; though not absolutely and invariably so, still, as a rule, real titanic ores are extremely free from these elements, and when they are not, the question may
-
i(
((
be raised, docs not the phosphoric acid come from sdheriiig barrcn gangue of apatite and n o t from the ore itself? Jlr. Iiileg' (Ib.) commenting 011 these reniarks of Dr. Forbes lias fully confirmed his assertions. It is quite csceptional (says hc) to fiilrl titanic ores containing any p1iox~)horns. ?'lie s:irtio obscrvatioris have been made iiiclepc~iideutly.in 1 ~ 7 5 :1)y I ' d . C'ook f o r the titaniferous ores of New Jersct> (Geol. S i i ~ v c yof S . ,J.. IS'75, 11, 3 5 ) . Analyses (110 sag's) indicate: this curious relation between titanium and T,liosphorns, that a considerable percentagc of t h e former dwag's correspoiicls to :i very s m n ~ lamount of the latter. Thc beliavior of sncli ores in t h e b h s t fnriiace and the character of the product might be :til interestiiig point in the metallurgy of iron, and the chemical examination of iro;i o u i s to tcst this hypothesis f o r t h c r or substailtiate tlicse facts ought t o be coiltin 11ed. It was true with the ores smelted i n the two furnaces wc liavc spoken of as having r u n for ten years, that t,liose containing least phosphorns contained most titanic acid, its analyses showed: I.
11.
ill.
11..
Average Phosphorus - - 0.OG.l 0.160 0.270 0.730 no. Titanic acid - - 0.9s 0. 90 0.5s 0.33 W e liave collected ii iinniber of aiialyscs of ores coming from mines separated by the Atlantic or by distances on land still greatcr, whicli we give below. A n y idea of fortuitous coincidence has to be excluded, a n d still t h e results seem to corroborate, i n :t remarkable manner, t lic above observations. E s r r E D STATE CC. Peitwsylvn~zin. 1. McElvee ore - - - TiO, 0.33$ 1'. 0.013 Met. Iron 6S.50$ (jS.75 2. Chestrlut Hill ore - '( 3.93% traces *. u , . A , traces __ 3. Cornwall ore - - (Geolog. Survey of Pa., AI3.) -1-07.1JL Cu7.0 I inn. J. of Iron (st St. Inst., 1. RockinghamCo. TiO, 13.71% I-'. .O52 2, p. 2. 36.61 traces 10th Census C.S., 1). 3. " O.!X 0.007 311-326-561.
