Influence of Hot Working on Internal Cracks in Continuously-Cast Steel Billets

E. B. HAWBOLT, F. WEINBERG, AND J . K. BRIMACOMBE

Continuously cast bi l let sections containing halfway cracks and in-plant, hot rol led rod f rom the same heat have been examined for open defects. It was found that signif icant open defects were present in the steel after a reduction of 4:1. A direct correlat ion be- tween init ial crack d imensions and open defects after a given reduct ion could not be made as the init ial crack configuration var ied appreciably along the bi l let length. Labora- tory scale, hot rol l ing of smal l continuously cast bi l let sections containing a known half- way crack configuration showed hot reductions greater than 6:1 are necessary to effec- t ively close halfway cracks. This is applicable for cracks or iented both paral le l and pre- pendicular to the rol l ing plane. Smal l pores and l ines of inc lusions present in mater ia l reduced 10:1 did not affect the low temperature f racture s t ress of the steel.

THE quality of steel produced by continuous casting, as compared to static casting, is a function of the de- fects present in the as -cast steel and the change in the size and distr ibut ion of these defects with subse- quent hot working. Static castings tend to exhibit coarse dendrit ic s t ructures in the center of the ingot with macrosegregat ion and center l ine porosity. Con- t inuously east steel, which solidif ied more rapidly than a stat ic ingot, exhibits a f iner dendr i t ic s t ruc- ture with less macrosegregat ion and center l ine porosity. The static castings general ly are extensively hot worked during roi l ing, which tends to close the pores and reduce the effective maerosegregat ion in the steel. On the other hand the eontinuous casting, having a smal le r cross-sect ion, is general ly subjected to less hot working. Accordingly, to compare the quality of steels produced by continuous casting with the equivalent static casting requ i res some knowledge of the effect of hot working on the reduction and redis- tr ibution of defects in the continuously cast steel.

The effect of hot working (either rol l ing or forging) on the st ructure and defect d istr ibut ion in stat ical ly cast ingots has been examined extensively. These studies have concentrated on establ ishing the mini - mum rol l ing ratio (the c ross -sect iona l a rea of the bi l - let divided by the corresponding area of the rol led product) required to ensure the breakdown of the as- cast st ructure, 1-5 and the attainment of sat isfactory mechanical propert ies. ~'6 Centerl ine porosity is the major defeet considered in these studies. The amount of hot reduct ion required to close center l ine porosity and give sat isfactory mechanical propert ies in the final product var ies among the different studies, within the range 4:1 to 10:1 2,6 In a recent study by Ushi j ima 7 on Si-Mn spr ing steels, a min imum rol l ing reduction of 6:1 to 10:1 was reported as necessary to give sat isfactory mechanical propert ies.

These observat ions cannot be extended to continu- ously cast steel with confidence, without a c lear defini- tion of the type, size and distr ibut ion of the defects in

E. B. HAWBOLT, F. WEINBERG, and J. K. BRIMACOMBE, are Associate Professor, Professor, and Associate Professor respectively, all with University of British Cohtmbia, Department of Metallurgy, Vancouver, B.C. Canada.

Manuscript submitted May 30, 1978.

METALLURGICAL TRANSACTIONS B

both stat ical ly and continuously cast steels before and after hot reduction. This information is general ly not known or is unavai lable.

The present study was undertaken to examine the effect of hot ro l l ing on casting defects in continuously cast steel b i l lets , attention being directed pr imar i l y to halfway cracks as a representat ive defect. This in- vestigation is part of an overal l program to examine and account for defects in continuously cast steels as a function of steel composit ion and thermal history during casting.

CRACKING AND POROSITY IN CY)NTINUOUSLY CAST STEEL

Many types of internal cracks may be found in con- t inuously cast steel bi l lets and slabs, the two most f re- quently observed being halfway and center l ine cracks. An example of the crack configuration for a bi l let with appreciable halfway cracking is shown in the sulfur pr int reproduced in Fig. 1.

The presence of the halfway cracks has been att r i - buted to high casting speeds, s high casting tempera- tures, 8'9'n'12 excessive water spray cooling, s-n mis- a l ignment of support and pinch rol ls , 8'l~ and reheat- ing of the steel below the spray zoneY In addition it has been reported that S and P levels in excess of 0.025 and 0.030 pct respect ively can lead to enhanced cracking.8, ~s

Center l ine cracking and center l ine porosity has been related to thermal s t resses , 12 bulging of the strand due to ferrostat ic p ressure , and volume shr ink- age.

EXPERIMENTAL

The effect of hot rol l ing on halfway cracks in con- t inuously cast steel was examined in two ways:

1. Observat ions were made of cracks in as -cas t bi l - lets and in bars hot rol led from these bi l lets, with reductions of 2:1 to 10:1.

2. Sections of as -cas t bi l lets containing halfway cracks were hot rol led in a smal l laboratory mil l . The cracks in the rol led product were examined and com- pared to the init ial cracks. In addition tensi le tests were carr ied out on samples cut from the rol led steel.

ISSN 0360-2141/79/0611-0229500.75/0 9 1979 AMERICAN SOCIETY FOR METALS AND VOLUME lOB, JUNE 1979-229

THE METALLURGICAL SOCIETY OF AIME

, i , i i i i |

A . . . . . . . .

,ooo

80o \ .

,~ 600

7 o A \ / crocked 400 Q B \ / billet LL

200 A C V

O

holes halfway cracks centre structure

~ io 40 Distance From Outside Surface(ram)

Fig. 2-Fracture stress vs distance from the outside surface of the billet: (a) material with detectable cracks (filled symbols), (b) mate- rial with no detectable cracks (open symbols). A.B.C. repeat tests. Heat 3. Table I.

Fig. 1 -Sulfur print of half a cross-section of a continuously cast billet showing halfway cracks~ magnification 0.65 times. The configuration of tensile test specimens cut from the billet is also shown.

In-P lant Hot Rol l ing

1. As -Cast Bi l lets . Two composit ions of continu- ously cast and ro l led bi l lets were examined, these be- ing heats 1 and 2 in Table I. T ransverse sect ions of the as -cas t bi l lets (152 152 mm) were surface ground and sulfur printed. Cracks in the stee l appeared as dark l ines in the sulfur pr ints . The resolut ion of the sulfur print in del ineating cracks in the stee l cannot be c lear ly defined since the te'chnique requ i res sulfur to be segregated to the crack surface and acid to enter the crack. The halfway c racks which could be reso lved had an average crack width of 15 ~ and ranged f rom 9 to 33 ~ in width. It is possible that smal le r cracks were present which could not be reso lved and which could markedly influence the ducti l ity of the steel .

To invest igate this possibi l i ty, tensi le tests were conducted on samples f rom continuously cast bi l let sect ions both with and without observable cracks (based on sulfur prints). The composit ion of these bi l lets is given in Table I, heat 3. The tensi le test samples were cut f rom the bi l lets With the conf igura- tion shown in Fig. 1. The test samples were cyl indri -

Table I. Composition of Strand Cast Billets

Heat No. C Mn S P Cr Cu Si Ni V

1 0.30 0.62 0.024 0.004 0.078 0,17 0.20 0.07 0.034 2 0.59 0.80 0.044 0.025 0.14 0.23 0.23 0.098 0.003 3 0.34 1.03 0.042 0.032 0.18 0.38 0.30 0.13 0.002 4 0.35 0.98 0.046 0.036 0.20 0.45 0.30 0,14 0,003

cal with a reduced gage length of 30 mm and a gage diam of 3.2 mm. The samples were pulled in an In- stron machine at a strain rate of 3.3 10-4/s at liquid nitrogen temperatures (-196~ This low test tem- perature was se lected to min imize the plast ic defor- mation during tensi le test ing and to maximize the ef- fect of smal l defects on the strength of the steel .

The resul ts of the tensi le tests are shown in Fig. 2 where f racture s t resses are plotted as a function of distance f rom the outside surface. In these tests the f racture s t ress is also the maximum st ress attained. The f racture s t ress is markedly reduced in the region where halfway cracks are observed in the sulfur print (curve A). It a lso progress ive ly decreases f rom the outside to the center of the bi l let. For a bi l let with no observable halfway cracks (curve B), there is no drop in the f racture s t ress in the halfway region, but the progress ive decrease to the center is present . This is a c lear indication that sulfur prints are a sens i t ive means of detect ing cracks present in the steel , in agreement with other repor ts . 12 The factors contribut- ing to the progress ive decrease in f racture s t ress with distance f rom the outside surface were not determined.

The appearance of the f ractured sur faces of tensi le spec imens containing halfway cracks (12 mm from the edge in Fig. 2) are shown in Fig. 3(a) and (b). Most of the f ractured surface is smooth and undulating, in a pattern character i s t i c of an as -cas t dendr i t ic st ruc- ture, (Fig. 3(a)). The remainder of the surface exhibits a faceted topography associated with br i t t le f racture . This is shown more c lear ly at a higher magnif ication in Fig. 3(b) where the smooth.sur faces and the faceted br i t t le fai lure sur faces are evident. The smooth sur- faces often contained a h ighdens i ty of MnS inclusions.

The appearance o f the sur faces of the halfway cracks shows that the cracks formed interdendr i t ica l ly when

230-VOLUME 10B, JUNE 1979 METALLURGICAL TRANSACTIONS B

liquid was st i l l present between the dendr i tes 12 The res idual interdendr i t ic liquid would be r ich in res iduals, and therefore would be the region where MnS inclusions would develop. This would account for the high density of MnS inclusions observed on the crack surface, and the high density of sulfur associated with the cracks in the sulfur pr ints .

2. Hot-Rolled Bars. Rolled bar samples from heat 1 having reductions of 3, 6 and 132 to 1 and from heat 2 with reductions of 4, 28 and 80 to 1 were examined. Sections of al l the bars were surface ground, examined for cracks and open pores, and sulfur pr inted. The

(a)

(a)

(b) Fig. 3-Fracture surface. Test temperature - 196~ Brittle fracture region A, smooth surface B. (a) magnification 34 times, (b) magnifi- cation 400 times.

(b) Fig. 4-(a) Large void and elongated MnS inclusions. Longitudinal section after reduction of 3:1. Heat 1, Table I, magnification 230 times, (b) Small void with cusped outline containing multicomponent inclusions after reduction of 4: 1. Heat 2, Table I, magnification 1450 times.

bars having reduct ions of 16:1 or greater contained signif icant numbers of fine pores less than 4 microns in diam and sulfide and oxide inclusions of the same general s ize. In this case the pores may be associated with inclusions removed from the surface by polishing. Larger pores, c lear ly resul t ing from cracks or pores in the as -cas t st ructure, were observed for the reduc- t ions of 3:1 and 4:t in heats 1 and 2 respect ively. An example is shown inF ig . 4(a), the residue of a half- way crack. A second example is shown in Fig. 4(b), in which the crack containing mult icomponent inclusions, has a cusped outline associated with an interdendr i t ic crack.

Wulpi 6 has suggested that a hot reduction of 4:1 should be suff icient to effectively e l iminate defects which affect the mechanical propert ies . The present resu l ts which Show defects st i l l present after a 4:1 reduct ion contradicts this. However, nei ther the pres - ent invest igat ion or those of Wulpi have made a direct corre lat ion between a specif ic crack or porous region

METALLURGICAL TRANSACTIONS B VOLUME 10B, JUNE 1979-231

before and a f te r hot ro l l ing . In the b i l l e t sect ions ex - amined in th i s invest igat ion , the s i ze and loca l d i s - t r ibut ion of ha l fway c racks var ied apprec iab ly a long the b i l l e t . Thus it was found that random samples of ro l led rod cou ld exh ib i t no res idua l c racks , o r s ign i f i - cant c racks , depend ing on the in i t ia l c rack conf igura - t ion of the par t i cu la r sample . Accord ing ly , i t is d i f f i - cu l t to es tab l i sh d i rec t ly the amount of hot reduct ion requ i red to ensure the c losure of ha l fway c racks of known sever i ty by examin ing random sect ions of ro l led bar , For th i s reason a laboratory hot ro i l ing tes t ing procedure was under taken .

3. Laboratory Hot Ro i l ing . Hot ro l l ing exper iments were car r ied out on samples hav ing a 48 mm square c ross sect ion ; these spec imens were cut f rom the as - cas t b i l l e ts l i s ted in Tab le I, heat 4. Four of the sec - t ions ro l led (4A to D) conta ined ha l fway c racks , two

9 (4E and F) had no apparent c racks . The conf igurat ion of the c rack p lane wi th respect to the ro l l ing d i rec t ion was var ied , as shown schemat ica l ly in Tab le II I . For sect ions 4A and B, the c rack p lane was perpend icu la r to the ro l l ing d i rec t ion ; fo r sect ion 4C the c rack p lane was para l le l ; and fo r sect ion 4D the c rack p lane a l te r - nated between perpend icu la r and para l le l to the ro l l ing d i rec t ion by ro ta t ing the sample 90 deg a f te r each pass .

A l l samples were hot ro l led at 1200~ us ing a 101 mm d lam, two h igh laboratory S tanat ro l l ing mi l l . The in i t ia l a rea reduct ion of 2:1 was car r ied out in 4 to 6 rap id , sequent ia l passes of 1.5 mm reduct ion per pass . The samples were then reheated to 1200~ and the same procedure repeated unt i l the requ i red reduct ion was reached. The sample was then coo led to room temperature , a 25 mm sect ion was removed f rom the end fo r examinat ion , and the remainder reheated and ro l led fu r ther to the next s tage of hot reduct ion . Us ing th i s p rocedure samples hav ing reduct ions of 2:1, 4:1, 6:1 and 10:1 f rom the same s tar t ing mater ia l were ex - amined. In one case , a sample hav ing a reduct ion of 18:1 was examined. The laboratory ro l l ing program is compared to the in -p lant ro l l ing pract i ce fo r the in i t ia l s tages of reduct ion in Tab le I I .

For the tes ts in wh ich the d i rec t ion of the app l ied load was a l te rnated , a s ing le pass of 0.76 mm reduc- t ion was made, the sample ro ta ted 90 deg, and another pass taken at the same ro l l se t t ing . The sample was then reheated to 1200~ and the above sequence re - peated unt i l the des i red reduct ion in a rea was ach ieved . When marked rhomboida l d i s to r t ion occur red , the sam- ple was coo led to room temperature and mach ined to a square sect ion .

Table II. Comparison of Laboratory Rolling with Plant Rolling for the Initial Stages of" Billet Reduction

Laboratory Plant

Initial billet size, mm Roll speed, mm/s Roiling temp, ~ Roll reductions for initial breakdown to I0:1

48 x 48 152 1200 1.5 mm reduction per pass. Results in 3 pet red/pass initially 6 pet red/pass at 5:1 30 pct red/pass 10:1

Roll Diameter, mm 50.8

152 x 152 1520 1200 approximately 25 pct reduction in thickness per pass. Total reduction 8:1 with 7 passes

approximately 530

The 25 mm sect ions cut f rom the ro l led sample were sur face ground, su l fu r p r in ted , po l i shed and etched fo r meta l lu rg ica l examinat ion by opt ica l and scann ing e lec t ron microscopy . An es t imate of the average s i ze of the open defects on the po l i shed sur - face perpend icu la r to the ro l l ing p lane was made f rom measurements of 10 to 20 pores on the sur face .

Pa i rs of tens i le tes t samples hav ing a gage length of 12.7 mm and th ickness of 2.5 mm were mach ined f rom samples cut f rom b i l l e t sect ions 4A and 4B a f te r var i - ous s tages of ro l l ing . S imi la r re ference samples were taken f rom the c rack f ree sect ion 4F . The long i tud ina l ax i s of the tens i le spec imens was in the ro l l ing p lane and perpend icu la r to the ro l l ing d i rec t ion . The sam- p les were s t ra ined to fa i lu re in an Ins t ron at a s t ra in ra te of 4.2 10"4/s a t - 196~ The f rac ture s t ress was ca lcu la ted us ing the in i t ia l c ross -sect iona l a rea of the tes t sample . A f te r tes t ing , the f rac tured sur - faces were examined opt ica l ly and wi th the SEM. Any

Table III. Defect Size Related to Degree of Hot Work

Degree of Di rect ion Average Hot of Pore S ize , Appearance of Defects

Reduction Roll Force Microns

As Cast 300 x 17

2:1 1000 x 750 Large open defects visible on sulfur prints and on polished surface

4:1 270 x 113 Large open defects visible on sulfur prints and on polished surface

6:1 30 x 15 Line of small open defects mostly perpendicular to the applied stress

10:1 Sections 4A, B 25 x 2 Lines of partly open, very small, elongated defects

18:1 5 z 2 Lines of closed, very small, high density regions of non- metallic inclusions

2:1 . [~- . I 60 x 25 Large open defects visible on sulphur prints. Phosphorus en- riched region adjacent to cracks

4: I 20 x 5 Cracks still open but reduced in size

Section 4C 6:1 15 x 6 Very small open lines of defects

associated with groups of non- metallic inclusions

10:1 12 x 5 Few very small holes with groups of nonmetallic inclusions

2:1 ~3___3 880 x 140 Large open defects visible on _ sulfur prints and polished

surface 4:1 300 x 90 Open lines defects visible

Section 4D on sulfur prints and polished surface

6:1 35 x 30 Lines of smaller open defects on polished surface

10:1 10 x 5 Few very small open defects on polished surface

2:1 ff~--q Very few macroscopic defects visible on sulfur print or polished surface

4:1 No evidence of open defects Sections 4E,

4F 6:1 No evidence of open defects

10:1 No evidence of open defects

232:-VOLUME 10B, JUNE 1979 METALLURGICAL TRANSACTIONS B

(a)

(b) Fig. 5-Interdendritic halfway cracks: (a) etched surface of transverse section of billet, (b) parallel to the billet axis, magnification 17.6 times.

observat ions of Smooth, undulating areas on the f rac - tured sur face were taken to be vest iges of the or ig ina l halfway c racks in the samples .

The large halfway cracks through the 48 mm square sect ions could be identif ied f rom sulfur pr ints taken f rom the t ransverse sur faces at e i ther end of the 100 mm long sect ions . However, the halfway c racks are not continuous, nor p lanar, through the sect ion. A typi- cal c rack conf igurat ion is shown in F ig. 5(a), for a sec- tion perpendicu la r to the b i l let axis which has been pol ished and etched. The c rack is c lear ly interden- dr i t i c and as a resu l t is not p lanar , d iscont inuously shift ing when the dendr i t ic s t ruc ture changes. A typi - cal sect ion para l le l to the bi l let axis is shown in F ig . 5(b), again i l lus t rat ing the i r regu lar in terdendr i t i c f racture morphology. Because the i r regu lar i t ies a re smal l , an individual c rack or group of c racks could be compared before and af ter ro l l ing. There was no evidence that c racks had oxidized dur ing ro l l ing.

The appearance of the defects, and the pore s ize, in the hot ro l led s tee ls a re l is ted in Table IH. Sulfur pr ints of t ransverse sect ions of samples 4A and B in which the appl ied load was para l le l to the c rack plane, a re shown in F ig. 6. Large open defects a re c lear ly v is ib le af ter a 2:1 reduct ion, and to a lesser extent af ter 4:1 and 6:1 reduct ions.

The defects a re shown at h igher magni f icat ions in F igs . 7 and 8. Open defects a re c lear ly evident a f ter

4:1 and 6:1 reduct ions. On c loser examinat ion of the effect of applying the ro l l force para l le l to the axis of the halfway c racks (Fig. 6) the c racks are seen to have been forced open af ter a 2:1 deformat ion. (This as - sumes that the c rack character i s t i cs a re comparab le along the length of the sample) . However, s igni f icant c rack c losure has occur red af ter a reduct ion of 4:1. Smal l openings with thei r major axis perpend icu lar to the appl ied ro l l force are evident a f ter 6:1 reduct ion. Etching with Oberhof fer ' s reagent ver i f ied that phos- phorous was present in the defect a reas conf i rming that the defects a re vest iges of the or ig ina l halfway c racks .

Rol l ing with the appl ied load perpend icu lar to the c rack plane (sample 4C, Table III) resu l t s in a prog- ress ive closing of the ha l fway c racks as ant ic ipated. However, smal l holes are s t i l l evident in the s tee l at high degrees of reduct ion, usual ly assoc ia ted with a high density of nonmetal l ic inc lus ions. Defects present a f ter a 10:1 reduct ion are shown in F ig . 9. Roi l ing a l - te rnate ly para l le l and perpend icu lar to the c rack plane (sect ion 4D, Table III) prov ides resu l t s s imi la r to those obtained by ro l l ing para l le l to the c rack plane (sect ions 4AB). Cracks f i r s t open and increase in s ize, then rapid ly decrease with fur ther ro l l ing.

Samples with no in i t ia l c racks (sect ions 4E, F) show only a few smal l pores af ter reduct ions by hot ro l l ing of 2:1 and no s igni f icant poros i ty a f ter 4:1 reduct ion.

METALLURGICAL TRANSACTIONS B VOLUME 10B, JUNE 1979-233

os cost t

Fig. 8-Voids on transverse section after a reduction of 6:1. Section 4A. Table III, magnification 400 times.

t reduct ion 2 : I

1

84 1

l 6"1 Fig. 6-Sulfur prints of transverse section of sample 4A, Table III, as- cast, and after hot reductions as indicated, same size magnification.

Fig. 7-Voids on transverse section after reduction of 4:1. Section 4A, Table III. Magnification 1 O0 times.

The in f luence of the res idua l de fects on the tens i le f rac ture s t ress a t - 196~ is shown in F ig . 10. The f rac ture s t ress in these tes ts i s a l so the max imum st ress observed . For the sect ions conta in ing ha l fway c racks (4A, B) w i th the load app l ied para l le l to the c rack p lane , the f rac ture s t ress c l imbs rap id ly as the reduct ion in a reas is inc reased f rom 2:1 to a l i t t le above 6:1; the f rac ture s t ress remains essent ia l l y constant a f te r a 6:1 reduct ion is obta ined . A t the low f rac ture s t resses , f rac ture o f ten occur red in reg ions co inc ident w i th the la rgest de fect detectab le in a su l - fu r p r in t of the tes t spec imen. The c rack f ree mate - r ia l (4E) in F ig . 10 mainta ined a h igh constant f rac ture s t ress fo r a l l reduct ions .

Table IV. The Mean Thickness to Length Ratio of the Plastic Zone for the Laboratory Rolling and for Industrial

Billet Rolling

Stage of Reduction ho, h* A t mm mm = ho

~Rr [2- r]

Laboratory Rolling (1.5 mm reduction/pass)

1st Pass 48.3 46.8 5.45 Final pass to give a:

2:1 reduction 25.7 24.2 2.87 4:1 reduction 12.1 10.6 1.30 6:1 reduction 9.6 8.1 1.01

10:1 reduction 6:3 4.8 0.64 Industrial Billet Rolling

1 st pass (25.8 pct) 152.4 1 t 3.0 1.29 2nd pass (23.9 pct) 113.0 86.0 1.17

*All reductions were based on assuming no lateral expansion during rolling.

twhere h o = thickness of the sample on entering the roll h i = thickness of the sample on leaving the roll R = roll radius r = h o -h i

ho

234-VOLUME 10B, JUNE 1979 METALLURGICAL TRANSACTIONS B

Fig. 9-Defects on transverse plane of section 4C, Table lIl reduced 10:1, magnification 1200 times.

DISCUSSION

The observat ion that the f racture s t ress of steel containing halfway cracks increases progress ive ly for rol l ing reductions from 2:1 to 6:1 and then remains essent ia l ly constant with further rol l ing, c lear ly in- dicates that in the laboratory roi l ing program a re- duction of at least 6:1 is necessary to e l iminate the ef- fects of the cracks.

To relate these laboratory scale observat ions to in- dustr ia l ro l l ing procedures the data l isted in Table II has been used to calculate the mean thickness to length ratio, A, of the plast ic zone. 14. In part icu lar , LX has been calculated for each stage of the laboratory ro l l - ing; and a comparable zx has been determined for the init ial stages of industr ia l breakdown of as -cas t bi l - lets. The detai ls of the ro l l ing reductions and the cor- responding A values are shown in Table IV.

*In calculating the thickness to length ratio of the plastic zone, A, plane- strain rolling conditions are assumed.

1200 r , , ,

~ooo 9 . . . . . ~ . . . . . _o 9

o4E

g 8oo

600

?, G,- 400

20o

o 21, 41, el," ,01, ,81, Hot Reduction

Fig. 10-Fracture stress vs hot reduction for sections with and without halfway cracks as shown. Test temperature -196~

For the init ia l stages of reduct ion the re lat ive ly smal l radius of the laboratory ro l ls has resul ted in large ~ values, equivalent to those real ized dur ing large ingot breakdown. These large zX values indicate that inhomogeneous s t ra ins would be generated dur- ing the init ial stages of laboratory reduction with most of the s t ra in being confined to the surface of the sam- ple. The ear ly stages of laboratory reduction would thus be less effective in improving center l ine defects.

Although the laboratory resu l ts may be too conserva- tive in suggesting that 6:1 to 10:1 reduct ions are re - quired to ensure closure of the halfway cracks, the resul ts are in good agreement with those reported by Ushi j ima. 7 The Ushi j ima resu l ts are based on mult i - d i rect ional ro i l ing of 130 mm square spr ing steel bi l - lets using industr ia l ro l l ing faci l i t ies.

The laboratory resu l ts also indicate that after a re - duction of 6:1, a l l cracks and l ines of res idual defects are al igned para l le l to the rol l ing plane. In this or ien- tation, which is also para l le l to the axis of the tensi le specimen, the defects would be expected to exhibit a min imum effect on the tensi le test resu l ts . Through thickness tensi le tests would be required to establ ish if the res idual defects present after a 6:1 reduction would influence the f racture s t ress . This type of in- vestigation would requi re larger samples and indus- t i ra l rol l ing faci l i t ies.

SUMMARY

1. An examinat ion of the as -cas t bi l let and industr i - al ly hot rol led rod f rom the same continuously cast steel revealed that open defects, thought to be vest iges of or ig inal halfway cracks, were present after a 4:1 reduction. However, as the size and sever i ty of defects var ies along the length of a bi l let, it was not possible to make a d i rect corre lat ion between defects remain - ing in random samples of hot rol led rod and those or iginal ly present in the as -cas t bi l let.

2. Laboratory hot ro l l ing tests showed that a reduc- tion in excess of 6:1 was required to effectively close halfway cracks in continuously cast steel b i l lets . This degree of reduction was requi red for halfway cracks or iented either para l le l or perpendicular to the ro l l - ing plane.

3. Due to equipment size l imitat ions, the init ial stages of laboratory rol l ing would be less effective in closing halfway cracks than industr ia l bi l let rol l ing.

ACKNOWLEDGMENTS

We wish to thank the AISI for f inancial suppor t of this project under contract No . 36-337. The ass is tance of N. Wa lker , J. B rezden and I. F rank l in in the exper i - menta l work is acknowledged.

REFERENCES 1. W. B. Pierce: BlastFurn. SteelPlant, 1953, vol. 41, p. 1301. 2. B. H. C. Waters, W. H. Pritchard, A. Braybrook, and G. T. Harris: Z Iron

Steellnst., 1958, vol. 190, p. 233. 3. P. Thomas: a r. Iron SteelInst., 1958, vol. 190, p. 123. 4. I. M. D. Halliday: Iron Steel, 1959, vol. 191, p. 121. 5. G.Fenton and J. Pearson: Z Iron Steellnst., 1958, vol. 189, p. 160.

METALLURGICAL TRANSACTIONS B VOLUME 10B, JUNE t979-235

6. D. J. Wulpi: Met. Progr, 1964, vol. 6, p. 72. 7. K. Ushijima: Trans. Iron SteelInst. Jpn, 1975, vol. 15, p. 380. 8. L. B/icker and P. Gosselin: Open Hearth Proc.~ vol. 53, p. 145, Detroit, 1970. 9. C. I. Miller: Jr.: Open Hearth Proc., vol. 54, p. 316, Pittsburgh, 1971.

10. J. W. Donaldson: J. Metals., 1965, vol. 17, no. 12, p. 1338. 11. K. Ushijima: Continuous Casting of Steel, p. 59, The Iron and Steel Insti-

tute, London, 1964.

t2. G- Van Drunen, J. K. Brimacombe, and F. Weinberg: Ironmaking Steel- making, 1975, vol. 2, p. 125.

13. H. Vom Ende and G. Vogt: Z Iron Steellnst., 1972, vol. 210, p. 889. 14. W. A. Backofen: Deformation Processes, p. 89, Addison-Wesley Pub. Co.,

USA, 1972. 15. W. A. Backofen: ~Ibid, p. 135.

236-VOLUME 10B, JUNE 1979 METALLURGICAL TRANSACTIONS B