Casting of Bloom Beam Blank
Casting of Bloom Beam Blank
Casting of Bloom Beam Blank
Algoma's interest in the contin- dia, seamless tube grades, and rails. hearth, and casting buildings
uous casting of blooms and shapes The beam machine would also be were required; in addition, the Isley
arose out of the prospect of attain- convertible to the production of stacks serving No. 25 open hearth
ing substantial gains in yield from 32 x 10 in. slabs for our narrow furnace were relocated (Fig. 4).
liquid steel to product, and also strip mill. The operating floor of the casting
of improving productivity on our Construction of the' plant began in plant occupies as area of 125 x 140
primary mills. We were approach- September, 1966. The casting plant ft, and contains the four-strand
ing the position where ingot pro- is located in a building formerly bloom machine and two-strand
duction would be greater than the occupied by an eight-furnace open blank machine, located on 52-ft
capacity of the two blooming mills; hearth shop (Fig. 2). It is supplied centers (Fig. 5 ) . The operating floor
these mills devoting a large part with 108-ton heats by the three- is divided by the 150-ton crane col-
of their time to rolling blooms and furnace basic oxygen shop, from umn line into a 90 ft wide casting
shaped blanks for the rail mill- which it is separated by an operating aisle and 50-ft tundish aisle. In addi-
wide Aange beam mill complex. open hearth shop. The movement tion to two of the four preheat sta-
Thirteen of the 21 wide Aange of steel to the casting machines thus tions, the tundish aisle provides
sections produced are most econom- involves placing of the ladle on a space for skulling, relining, and set-
ically processed by rolling the ingots pick up stand at the south end of the ting up tundishes, preparation and
to a shaped blank on one of the BOF shop by the south crane in drying of stoppers, and some stor-
blooming mills. Rolling rates on this shop (Fig. 3). The ladle is then age of refractory and other sup-
these blanks are comparatively low, picked up by a high level crane, plies.
with roll changes taking up an im- raised an additional 55 ft, a total Three ladle cars are provided, a
portant part of mill time. of 84 ft above grade, and transported powered car serving each machine,
The requirements of the 18 in. across the open hearth shop and and an idle car which can be coupled
merchant mill would involve a large over one of the 300-ton furnaces, a to either, permitting rapid ladle
' number of billet sections, were bil- distance of some 330 ft. It is then switching for sequential or multiple
let casting to be considered; there- lowered 34 ft into position on the ladle casting.
fore, it seemed apparent that cast- ladle car. This movement, from The emergency ladle is located
ing machines capable of producing furnace to ladle car, requires an between the machines on an inter-
a range of bloom sizes and shaped average of 16 min. Obviously, mediate floor level; an overflow
blanks would best suit our require- extensive modifications to the chute off the end of the tundish
ments. A four-strand bloom ma- roof structure of the BOF, open diverts metal from tundish to
chine would allow casting times of
1 hr or less for 108-ton heats in the
range of sizes considered. Modifica-
tions to mill design and practices
permitted a reduction in the num- I.. .k I
ber of blank sizes required from 12
in conventional practice to five for
the casting machine. They would
have a common web thickness of
4 in. to simplify machine design,
and the smallest, at 131 sq in. area,
would allow a two-strand machine
to cast the oxygen furnace heats
in 1 hr.
It was therefore decided that the
casting plant would consist of a
four-strand bloom machine, produc-
ing blooms up to 10% x 14 in., and
a two-strand beam-blank machine,
which would cast the five blank
sections ranging from 131-224 sq
in, in size (Fig. 1 ) . The capacity of
the shop is estimated at 600,000 net
tons comprising structural grades, I - ::
.I I . I b
n ::x,S
UY 11'4i91.
.
p ,,I..''
l, l . , C
3/16-1% in. Oscillator speed is ref-
erenced from t h e straightener
drives to give a 66% negative strip
p , , Y. 0 '
,piI, **a .' irrespective of stroke or withdrawal
speed. Mold lubrication is presently
gb 68 .Iso'
rapeseed oil, the rate varying from
,,,a, .I* c 20 to 30 cc per ton depending on
,!!,n .. .--
7--
CRANE GALLERV I
**'.'J US.- ..- II~-~'I?cI ) ?lPLI.%lll
I 1
I I
Fig. &Sectional ~rrongenientthrough plant showing bloom machine.
co:mmodate the different blank represent yields from raw ladle steel ladle, the other 80% pouredto ingots
profiles. to good product of 94-95%. The net at the emergency pouring stand. The
The first cast was made on the yield includes only those heats production per turn has been influ-
bloom machine on September 5, which were completely cast on the enced particularly in more recent
1967. Prior to start up, supervisory machine, while the gross yield months by priorities for other uses
personnel trained for a 4-week refers to all ladles sent to the casting of raw steel, and by limitation of
period on multi-strand bloom ma- plant, including those which we grades and proven products from
chines producing a range of prod- were unable to cast because of tem- continuously cast blooms. To date,
uct.~ similar to those we proposed perature, ladle stopper failure, or the bulk of production has been in
to cast. This training period, to- mechanical problems, and those in- structura1,and in high carbon grind-
g'et.her with the several weeks dur- completely cast for similar reasons. ing media grades. Trails are being
ing which the supervisors partici- Of these, about 20% were returned conducted regularly on various
pated in the final assembly, testing to the furnace via the emergency seamless tube grades, but final prac-
and "dry-running" of' the machine,
was of major value in accomplish-
ing a reasonably satisfactory start
up. Supervisors manned the machine Toble I. Production and yield statistics.
on a one-turn basis, for the first (from Sept. 1967 to Feb. 1968)
week during which time the first
operating crew was training. For
the remainder of the first month, this Sept. Oct. Nov. Dec. Jan. Feb.
first crew operated on the same five
turns per week basis, with the Casting Turns 21 36 45 41 43 38
Casts Attempted 17 55 95 91 116 106
second crew training. One or two Product Tons 1420 5036 8889 8155 10901 10048
casts per turn were made while Tons per Turn 68 140 197 199 253 266
Net Metallic Yield 82.9 83.1 84.9 83.6 83.5 82.8
machine adjustments were made and Gross Metallic Yield 68.7 75.7 77.2 73.8 77.8 78.1
operating techniques developed. Fol-
lowing the 24th cast, a two-turn
operation was begun, and remained
at t.his level until start of operation Toble II. Costing machine performance.
of the two-strand machine on March (from Sept. 1967 to Feb. 1968)
12, 1968.
Table I details some of our pro- Sept. Oct. Nov. Dee. Jan. Feb.
duction data during the first six
- -
months of operation. The yield
figures quoted are based upon total Casts Attemoted 17 55 95 91 116 106
% ~omplete'~&ts 76.5 80.0 80.0 80.2 87.9 90.5
metallic charge to the basic oxygen % Complete 4 Strands 41.2 40.0 34.8 48.3 48.3 50.9
furnaces since we have no means of :; Strand Completions 66.2 71.9 75.9 76.7 85.7 84.3
Breakouts % Strand Starts 9.2 4.4 3.6 3.0 2.4 2.O
measuring ladle weight. Based upon Ladle Stopper Failure ';; 11.8 9.1 9.5 6.6 1.7 2.8
typical furnaces yields, these would
Continuous Casting 137
tices for these have not yet been flow in the area of the nozzles. heating, is expected to be of benefit
established. Other more common causes of in a number of areas. There is evi-
Table I1 deals with some elements strand losses have been loss of noz- dence that tapping temperatures
of casting machine performance zle control, particularly in the latter may be considerably lower if tem-
during the start up period. The per- stages of multiple casts, equipment perature stratification in the ladle
centage of complete casts has grad- failures i n , the oscillator-stqaight- can be eliminated; this would im-
ually improved to a level of 90% ener-dummy bar disconnect and prove furnace operating conditions
or greater, the principal factors af- torch cut functions, and breakouts. by lowering refractory wear, tap
fecting this performance being ladle Nozzles used at present are con- t o ' t a p times, and oxygen consump-
stopper failures, unsatisfactory tem- ventional casting pit quality fire- tion. Uniform temperature distribu-
peratures, and mechanical problems clay with mullite inserts. Bore varies tion in the ladle should greatly re-
involving the entire machine. from 3/4 in. to 1 in., depending upon duce the frequency of "cold starts,"
A considerable amount of devel- steel grade and strand size to pro- one of the major sources of strand
opment work has been required to vide a casting speed of 40 to 45 in. losses prior to, and during cast, and
reduce the rate of ladle stopper in the 10% x 9 in. section, 30-35 in. also eliminate the tundish tempera;
failures from around 10% in early in the larger sections. Performance ture "peaks" to which many break-
months to 1-2% presently. A 7-in. of these nozzles has been satisfactory outs are attributed.
nozzle with 1%-in. bore is used, in for all high carbon grades, even on The original molds on the bloom
A , Cone 32-33 fireclay composition. multiple casts, and for single casts machine were of electrolytic tough
conventional quality 8-in. diameter of lower carbon structural grade pitch copper. No provision was made
sleeves and heads are used, and steels. The problem of nozzle qual- for supporting the bloom from thc
despite ladle temperatures of 100F ity, and of nozzle-stopper configura- bottom of the mould through the
above normal practice, and retention tion to provide fully reliable con- cooling plate area to the top apron
times averaging 80-85 min, we trol for all grades and practices has rolls. We experienced an extremely
haven't found it necessary to use not been fully solved. rapid rate of mould wear in the
air cooled rods. A ladle preheating Equipment malfunctions in the lower part of the plates, in which the
facilities is under construction; if first months of operation accounted original 0.060-in. taper on the
ladles are to be preheated with for about 15% of all strand losses. straight sides was lost within 20
stoppers set, air cooling of the rods In the past two months, however, to 30 casts, and wear continued at
may well become necessary. Since as a result of modifications to both a rate of 2-3/1000's per cast. Ther~e
ladles are used interchangeably be- equipment and techniques, these is little doubt that the resulting
tween; 'conventional and continu- failures have been reduced to less "negative taper" results in erratic
ously.'cast heats, no evaluation is yet than 3% of total strand losses. cooling of the strand, and in irreg:
available of the effect of higher While the trend in breakout fre- ular corner and wall thickness,
t e m ~ e ~ a t u r eand
s extended holding quency has been generally. down- which manifests itself first as rhom-
times on lining performance. Lining ward, breakouts continue to be the boidity, later as longitudinal face or
life on all oxygen furnace ladles single most serious problem in corner cracks, and ultimately as
presently'averages 32 heats. It is also maintaining machine availability breakouts. It has also been our e x -
too early to assess accurately the and controlling maintenance costs. perience that the corner radius pro-
influence of higher tapping tempera- In the first 500 casts, a total of 56 vided on the liner plates makes the
tures on 'oxygen furnace lining life; breakouts have occurred, a fre- machining of the plates to restore
however, at full production a loss of quency based on strand starts of mould dimensions and geometry a
20-25% on conventional lining ma- 3.1%. In the early stages of opera- very difficult undertaking. Mainten-
terials might be expected, and will tion, a large percentage of these ance of properly fitted corners after
have to be compensated for through could be attributed to such operator machining has been almost impossi-
the use of higher quality refrac- errors as packing failures, and ex- ble, and opening up of these cor,ners'
tories. cessive or variable speed, or to such during cast, particularly during ex-
The percentage of casts completed equipment malfunctions as prema- tended multiple casts, also lead's to
on four strands has remained below ture disconnects and roll pressure corner defects, cracks, and break-
50% up to the present. This figure fluctuations. More recently, however, outs.
is not a precise measure of strand the majority of breakouts appear to There are many potential solutioris
losses during casting, since it also be associated with temperature of to the problem of maintaining mould
includes the casts which were begun metal and with mould condition. As geometry and tightness. It should
on fewer than four strands. In the indicated earlier, temperature stra- be possible to reduce the rate of
course of the first 500 casts, for tification in the ladle poses a prob- copper wear through the use of a
example, 152 casts, or 30% were lem in starting the cast without type of copper with greater strength
started on three strands or fewer. nozzle freeze-up. In order to insure at operating temperatures or of a
When a breakout or other strand a reasonable chance of starting higher strength surfacing material.
failure occurs early in an operating without such freeze-ups, a ladle Early results on hard chrome plating
shift, it is our normal practice to temperature greatly in excess of that have been encouraging, the first
cast the remaining heats on that required for castability is necessary. mold installed having some of the
day on three strands, making repairs Typically, on 0.18-0.20 carbon original taper remaining after 150
on the down shift. Three-strand grades, a tap temperature of 3000F casts; several similar trials are in
casts typically required 75-80 min. is required to provide a ladle tem- progress. Trials of silver-bearing
A second major cause of strand perature at the machine of 2925 to copper are planned. The first trial
losses is frozen nozzles, particularly 2935F. At these levels, the tundish of an attached foot-roll arrangement
on outer strands. Ninety-nine such temperature at start of cast will be also showed some improvement,
losses were encountered in the first 2800-282O0F, and in the first 20 to presumably by equalizing air gap
500 casts. This problem appears to 25 min of cast, will peak at 2860- formation on all sides of the mold.
be peculiar to very large tundishes, 2880F. A large percentage of break- Maintaining the proper setting, and
even when stopper-controlled, and is outs occurs in this peak temperature alignment of a foot-roll is critically
not altogether a result of inadequate period as indicated on the graph important, and difficult to accom-
preheat. A large ladle nozzle should (Fig. 7). The second graph illus- plish on a curved mold machine;
improve end nozzle performance, by trates the variations in tundish tem- however, its potential value in re-
reducing the time necessary to fill perature occurring on a multiple ducing mold wear and providing
the tundish. In addition, a tundish cast involving four high carbon more uniform cooling has led us to
redesign will remove these end noz- heats (Fig. 8). pursue this aspect further with the
zles from the extreme ends of the An argon stirring facility being development of an infinitely adjust-
tundish and provide a better metal installed, together with ladle pre- able assembly which will be much
138 Open Hearth Proceedings, 1968
- - - - - - - - -- - -
there are decided advantages, eco-
Table I l l . Multiple cajts and distribution by grades. nomic and otherwise, in the insert
(from Sept. 1967 to Feb. 1968) rather than the solid type. As noz-
zles of improved design and refrac-
Sept. Oct. Nov. Dec. Jan. Feb. toriness become available, we would
hope to be able to cast all grades
Casts Attempted 17 55 95 91 llG 106
on a multiple basis in the same way
Sb Cast Multiple - - 23.2 50.6 75.0 48.2 that high carbon grades are pres-
Tundishes Used 17 55 84 67 61 75 ently handled. On these latter
Tons per Tundish 83.5 91.6 105.8 121.7 178.7 134.0
:L Heats < 30 C 100.0 100.0 95.8 52.8 33.6 51.9 grades, three or more heats are cast
Heats > 60 C - 2.1 45.1 60.3 40.6 in sequence regularly, and are rarely
terminated as a result of refractory
problems. On one occasion, six heats
totalling 620 net tons were cast on
liable spray patterns. As a result of tundish availability. If a tundish can four strands with a total casting
our experience on the bloom ma- be used for even two casts on a time of 6 hr, 38 min. The longest
chine, secondary cartridge-type fil- regular basis, the tundish aisle bot- casting period was 7 hr, 6 min, when
ters were installed on the blank tleneck can be eliminated, and re- five heats were cast on three strandsl
machine where the much larger fractory and fuel costs sharply re- 3. Operator fatigue. When more
number of very small nozzles, many duced. than two ladles are cast consecu-
of them inaccessible for routine in- The ladle car arrangement de- tively, it is necessary to provide
spection and cleaning, would be an scribed allows changing of ladles to spell operators to relieve the strain
even greater problem. It is planned be achieved rapidly. On many oc- and monotony of extended casting
to provide similar 25-micron filters casions the interval from emptying periods. It is in such circumstances
on the bloom machine as well. one ladle to opening the next has as these that mold level contro,l
Some of the statistics relating to been 40 sec or less, and rarely over equipment becomes attractive on
multiple casting are shown in Table 1% min, while tundish capacity even large strand machines, in reducing
111. It is considered essential that at normal level will allow 3-4 min the degree of concentration required
this procedure be used on a regular of casting. of the operators. At the relatively
basis if the rated capacity of the There are a number of factors in- low casting speeds employed, there
plant is to be realized. Although fluencing the success of multiple is some doubt of its justification for
the average casting time for four- casting. single ladle casting.
strand casts is 57 min, the overall The potential value of argon stir-
average including casts on fewer 1. Careful co-ordination between ring has been discussed earlier. The
than four strands is 67 min. Simi- steelmaking and casting operation elimination of wide variations an?
larly while the average time from to ensure accurate arrival time of excessive peaks in tundish tempera-
finish cast to "machine ready" is ladles without delaying furnace se- tures would increase the reliability
53 min, the actual average from quences; normally it is possible to of multiple casting through reduced
finish cast to start cast is 77 min. cast successive heats from the same nozzle and stopper erosion. It may
The difference of course lies in the furnace though delays in either shop well be that the better starting con-
fact that ladles cannot be scheduled occasionally dictate changing fur- ditions resulting from uniform ladle
to arrive exactly when required; naces. Immediately the cast begins, temperature, together with possible
no steelmaking shop can be expected steelmaking personnel are advised improvements in metal fluidity at-
to make ladle deliveries with such of the anticipated completion time tainable by modifications in deoxi-
precision, when the cycle time and subsequent ladle arrival times dation practice, will permit us to
varies by up to 30 min, without established. eliminate some or all of the stopper
incurring serious production delays. 2. Performance of tundish refrac- rods. Improvements in tundish re-
Thus, the potential of the machine tories, principally nozzles and fractory costs and tundish avail-
is limited to 3.3 casts per turn on stopper rods. The mullite insert noz- ability, as well as in stream char-
single casts, and while some im- zles currently in use have been dis- acteristics, would be possible if
provements in times and sequencing cussed earlier; erosion of these noz- stopper rods were not required.
can be expected, these are likely to zles over extended casting times has The process for the continuous
be more than offset by production limited multiple casting to two heats casting of blanks for symmetric,al'
delays arising out of mechanical in the case of structural grades. H-Section and I-Section beams (dog
failures, tundish delays, breakouts, Throttling of stoppers to control bones) and certain other similar
and the like. It can be seen that if speed when nozzles become enlarged shapes was developed jointly by A1-
it is possible to average three ladles is not a satisfactory solution, since goma and The British Irori & Steel
per tundish, the cycle time reduces sustained throttling causes rapid Research Association (BISRA) . Pat-
to 96 min and the output potential damage to stopper heads and also ents for this process have been ap-
rises to five casts per turn. Cleanout results in undesirable stream char- plied for in Belgium, Canada,
of tundishes, together with setting acteristics. The use of stoppers ne- France, Germany, Italy, Japan, Nor-
of nozzles, preparation and setting cessitates rather large nozzles in way, Sweden, and the United States
of stopper rods, and preheating is order to provide reliable seating of and patents have already been is-
costly and time-consuming; fre- the head and to produce an accept- sued for the process in Belgium,
quently operations are delayed by able stream. In nozzles of this size, France, and Italy.
by G. H. Woodhouse
Continuous casting has altered for ing of steel is to be regarded as a tinuous casting. Algoma Steel is to
many of us the concept that cast- function secondary in importance to be congratulated on their initiative
- - - - the conversion of steel in the fur- foresight in undertaking the devel-
G. H. WOODHOUSE is operations mbnager, nace. The paper this afternoon on opment of procedures for casting
Manitoba Rolling Mills, Selkirk, Manitoba, the casting of blooms and beam- beam-blank shapes. Common to
Canada. blank shapes is another major step most developments, a number of ad-
forward in the development of con- justments to the process remain to
140 Open Hearth Proceedings, 1968
be solved; but the results achieved mately 12%. Gentlemen, this is not peak temperature. Was this latter
to date are most encouraging. a unique experience. Higher tap- factor assessed in arriving at the
In discussing this paper emphasis ping temperatures, prolonged ex- conclusions stated?
or attention is drawn to several posure on occasions, and the neces- Efforts at reducing tapping tem-
factors which we feel are of par- sity to preheat the ladle to higher perature through preheating of the
ticular importance. Questions have temperatures all contribute to this ladle to higher temperatures or re-
been raised as to the ideal design problem. Anyone contemplating moval of temperature stratification
of the mould. The effect of the continuous casting would do well in the ladle through argon stirring
corner radius has been sighted. Ex- to pay careful attention to the give every indication of being very
perience would indicate that large changes that must be made to pit beneficial. The graph showing fluc-
radii have a tendency to fail by procedures. A significant cost in the tuations of steel temperature in the
cracking at the top of the quadrant, start up of casting machines has tundish during multiple casts indi-
whereas small radii result in crack- been the loss of metal to t h e , pit cates a wide variation-in tempera-
ing at the end of the quadrant. Little and in some cases, damage to the ture. Were these multiple casts made
information is available on the use casting machine caused by stopper using argon stirring and is this tem-
of square corners with virtually no failures. Another aspect of the prob- perature variation typical of the
radius, but it is our personal feel- lem concerns the quality of the ma- results obtained with argon stirring?
ing that such a design would op- terials used in the production of Argon stirring is also effective in
erate very satisfactorily. We appre- casting ladle stoppers. The need ex: inclusion removal as is evidenced
ciate, however, that mill men nor- ists to develop better quality' con- by measuring the total oxygen con-
mally prefer a radius on the corner trol procedures and, if possible, a tent in the steel. Unfortunately, at
but there are a number of desirable practical method in inspection of the present time, we do not have
features from the point of view of the stoppers to determine those that a practical method for continuously
casting. With a corner joint, some are faulty. measuring free oxygen in the steel,
of the difficulties experienced in the Excessive temperature gives rise but it is hoped that developments
joints opening up during casting to a delay in the solidification and in this area may give us such a tool
would be eliminated. In this latter a rise in the temperature of the in the near future. Small volumes
regard, we experienced the same solidifying shell. Normally, in solidi- of argon help in the removal of the
problem of the corner opening up fying the shell, the mould will ex- large inclusions, which can become
and appear to have achieved some tract around 10% of the total heat trapped on the mould wall thus
s,uccess by machining a taper on the content in the steel. A 50 degree reducing the strength of the shell.
joint to key it in. rise in temperature increases the One of the most significant costs
( Recognizing the fact that the ideal total heat content of the steel by in continuous casting is the cost
mould is one with taper; we would an amount that represents approxi- of lining and repairing the tundish.
question the advisability of only mately 20% of the total heat ex- Some of the wroblems associated
tapering the straight sides. It would tracted by the mould. On the other with the cost of the tundish lining
be our feeling that this would rise hand, this 50 degree rise in tempera- lie in the severe erosion of the re-
to uneven cooling. Was it difficulty ture of the metal only gives rise fractories at the point of entrance
:in manufacturing that resulted in to an increase heat extraction by of the steel from the ladle. A hard
the decision to taper only the the mould of approximately 3%. wearing refractory in this area is
straight sides? Do you feel the taper Under the ferrostatic pressure of essential. Tundish design including
on the straight sides is essential? the liquid steel, the stress in the such factors as shape, capacity, and
'Was wear on the outside radius of shell is increasing while the shell metal head have a very significant
the mould also a problem? strength is decreasing. The tendency bearing on tundish costs and tun-
The results indicated with hard to longitudinal surface cracking dish stream control.
chrome plating are significant. These giving rise to breakouts increases. The question of ideal steel tem-
results are similar to our experience. Careful control of those factors perature, cooling procedure as re-
Chrome plating ' increases 'the re- which influence the temperature of lated to grade of steel being cast,
sistance to wear; but we have also the metal in the tundish is essential. is a problem which most steel plants
seen some indication that chrome Referring to the graph which have learned through trial and er-
plating permits better penetration shows the relationships of break- ror. This area has been indicated
of the lubricating oil to the bottom outs to time and temperature, the as a critical one in this paper. The
a.reas of the mould. conclusion is drawn that the peak fact that piggy-backing of heats has
The difficulties experienced in the incidence of breakouts occurs at the been difficult on all grades of steel
use of the cooling ,plates is noted.
' time the temperature in the tundish is an indication of this fact. Mea-
There is no question that cooling reaches a maximum. The graph surements of the liquid core by ul-
plates give rise to more rapid cool- clearly indicates that the frequency trasonics, beta gage, or other
ing and, hence, the possibility of of breakouts is less as the tempera- methods, coupled with heat balances
increased casting speed. The prob- ture decreases. The critical period arrived a t by measurement of heat
lems associated with the accurate in continuous casting is, however, extraction of the cooling media,
alignment of the plates, and the during the beginning of the cast, should provide data that, when pro-
nlould wear due to lack of support, as a number of conditions are un- cessed in a computer, will give a
would indicate that at the present stable at this period. A contributing better indication of ideal casting
time, cooling plates cannot be con- factor to breakouts is also cold conditions. Knowledge and the ap-
sidered an economic operational starts. This situation may also be plication of continuous casting in
component for casting billets and a contributing factor to the shape the steel industry is progressing
blooms. of the curve obtained. There is some rapidly. The paper presented here
Stopper rod failures at the start indication that the peak incidence today has made a major contribu-
up ran at a frequency of approxi- of breakouts occurs ahead of the tion to this field.