Seetharaman S. Fundamentals of metallurgy

Подождите немного. Документ загружается.

process development. Thus, property and process are closely interlinked for

mutual progress. The integrated manufacturing system of BH-IF SEDDQ steel

sheet consists of unit processes listed in Table 13.1.

13.2.2 Processes driven by properties, environment and

energy

Piping for gas transmission is usually produced in large tonnage, sinc e pipelines

need to transmit a large volume of gas over a long distance from well site to

shipping port. High strength low alloy steel (HSLA) plate has been utilized for

making the pipe. HSLA contains C, Mn, Cu, Ni and Cr each in a small amount

depending on the mechanical and corrosion properties required. Also, some

microalloying elements, Ti, Nb, V and/or Mo are commonly utilized. The plate

is press formed into a U-channel, further pressed into an O-tube, and the seam is

submerged arc welded in length direction, and finished into pipe. This is called

UO pipe.

UO pipeline passing through Northern areas (e.g. the Trans-Alaska Pipeline

System, TAPS) demands superior low temperature ductility for security reasons.

Such ductility can be achieved by fine grain matrix of ferrite, , that is obtained

by recrystallizing fine grains of austenite, . Fine  grains of diamet er d increase

TS or decrease the fracture appearance transitio n temper ature (FATT),

according to the Hall±Petch relation or Cottrell±Petch relation, each predicting

that YS/TS or FATT is positively or negatively proportional to (d)

ÿ0.5

Formerly, the grain refining was carried out by normalizing, i.e. heating hot

rolled HSLA plate just above Ar

transformation temperature to form fine 

grains, and air cooling the plate to let the fine  grains be transformed into fine

ferrite and pearlite. The normalizing process is, however, costly and low in

productivity. Elimination of the normalizing was trialed, and the thermo-

mechanical control rolling process (TMCP), later followed by accelerated

cooling, AcC, has been successfully industr ialized.

The success was due to

lasting development efforts, progress of most modern plate rolling mills, online

cooling facilities and advanced rolling and cooling schedules, combined with the

use of the microalloying elements in HSLA steels.

TMCP utilizes low temperature hot rolling of  grains near Ar

temperature

to form fine recrystallized  grains and non-recrystallized deformed  grains as

shown in Fig. 13.7.

The deformed  grains contribute more to refining  grains

by providing a greater specific boundary area for the nucleation of an increased

number of  grains. In addition, the deformed  grains contain deformation

bands within a grain that also serve as nucleation sites for  grains. Intensive

AcC after rolling generates intra-grain transformation nuclei that contribute to

refine  grains as well. The AcC converts pearlite (forming on air-cooling) into

finely dispersed bainite that increases strength without impairing duct ility (see

AcC1 in Fig. 13.7). When the AcC is interrupted and taken over by air cooling

516 Fundamentals of metallurgy

13.7 Thermomechanical control process with accelerated

cooling for microstructure control of steel during hot rolling (



: austenite, F:

ferrite, P: pearlite, B: bainite, M: martensite), (Oosawa).

(interrupted accelerated cooling, IAC), the bainite becomes tempered bainite

(see AcC2 in Fig. 13.7). Strengthening by IAC includes refining of  grains and

increase in precipitation hardening that make YS higher, and increase in volume

fraction of bainite that increases TS.

Microalloying of Nb, Ti and V forms very fine precipitates of Nb (C N), TiN

and VN that keep  grains fine on reheating owing to their grain boundary

pinning effect. Grain growth of recrystallized  is inhibited particularly by the

fine precipitates of TiN. Also, Nb and Ti dissolved in the  matrix suppress

recrystallization on deformation, and increase Ar

in excess of 100K. This

makes the advantages of TMCP available at higher temperatures. Also, Nb, Ti

and V dissolved in  matrix precipitate as fine carbonitride, carbide and nitride,

respectively, during and/or after  !  transformation, and increase the strength

of  matrix. In IAC, addition of Nb, Ti and V increases the quench sensitivity

and volume fraction of bainite, and hence strength.

The pipe is connected by girth welding in the open field without pre- and

post-heat treatment of the weld to save labor and cost. At such a weld,

coarsening of grains in the heat affected zone (HAZ) occurs to impair low

temperature ductility and strength, particularly for higher strength pipe grades.

Usu ally, C i ncreases strength but decreases both weldability and low

temperature ductility. To assure low temperature ductility of the seam- and

girth-weld, B-containing, low alloy steels bearing only low- to extra low-C have

been developed for higher strength variants. Loss of strength by decreasing C

has been compensated for by (a) further refining  grains with fine TiN

precipitates, (b) hardening with Nb(CN) precipitates and (c) making texture

converted from ferrite±pearlite to bainite with the addition of B. If C content is

high, B is consumed as Fe

(CB)

which precipitates both at the boundaries and

in the matrix of  grains, and hence cannot serve to form bainite. Decreasing C

to less than 0.04% by vacuum degassing of HSLA melt and combining excess C

with Nb and Ti as NbC and TiC prevent the formation of Fe

(CB)

API5LX-X60 a nd X65 UO pipe thus produced have found popular

application to gas/oil transmission for their reliability and good balance between

strength, low temperat ure ductility and weldab ility. However, higher tensile

grades, up to API5LX-X80, are required to reduce total cost, and hence have

found increasing use but yet in a limited amount. Recently, even API5LX-X100

(YS700, TS800 MPa) has been commercially utilized to cover some pipeline

in Canada. For such high strength grades, TMCP-AcC, shown in Fig. 13.7, has

been applied. The fine ferrite  bainite structure is preferred for low FATT

values, and minimization of inclusions and precipitates are attempted for

securing ductility. An example of the chem istry for X100 is C 0.06, Si 0.1, Mn

1.8, Nb 0.04, Ti 0.01, Mo 0.2% and some Cu and Ni. To assure ductility of the

coarse grained heat affected zone (HAZ), control of Ti and N for  refinement,

utilization of TiO

for fine nucl eation of , and reduction of Si, Al and C-

equivalent with B addition have been made.

518 Fundamentals of metallurgy

Gas and oil that contain H

S (sour gas and oil) have been increasing in their

production. H

S reacts with moisture in pipes, forms H

which reacts with

the inner surface of the pipe wall to form active hydrogen atoms. When

desulfurization of steel melt is inadequate, MnS precipitates in steel slab during

solidification. Deformability of the MnS precipitates relative to the steel matrix

is high at plate rolling temperatures, and hence during cross-rolling they turn

into thin, two-dimensional shape inclusions with sharp periphery in the plate.

The hydrogen atoms diffuse into the pipe wall, and some of them recombine at

the incoherent interface of MnS precipitates and steel matrix to form H

molecules, causing decohesion of the interface.

Stress arising from the increasing pressure of the recombined H

gas

concentrates at the sharp peripheral edge of the deformed thin MnS where the

steel matrix has been made brittle by accumulated atomic hydrogen. Cracks

initiate there and propagate into the steel matrix around the periphery. If the

number density (number per unit volume) of the elongated MnS precipitates is

high, the crac k connects the prec ipitates stepwise in the thickness direction of the

pipe wall, finally causing a burst of gas/oil through the cracked pipe wall. This is

called hydrogen induced cracking (HIC). It is not only S but also C and P in steel

melt that can cause HIC. Both C and P segregate in the center-thickne ss of the CC

slab and hence the plate during rolling, forming hard and brittle pearlite banding

or martensite, causing brittle fracture there when extra stress is applied by the

recombined hydrogen gas. The through-thickness crack can propa gate in the

length direction of the pipe over 10 m before the gas is decompressed during

bursting. The burst can ignite the gas, causing fire and disaster in the area.

Measures that have been established to prevent the occurrence of HIC, and

process developments to execute these measures are given in Table 13.2. Most

Table 13.2 Measures and processes to prevent the occurrence of HIC in API5LX-

X60/65 pipeline for sour gas/oil transmission

Preventive measures Process to execute the measures

1 Make C low or extra low · Decarburization by BOF supplemented by RH

2 Minimize H, P, S (extra

low), oxide inclusions

· HMPT followed by LF + RH refining

3 Convert S dissolved in melt

into CaS

· Ca addition to steel melt in ladle/tundish

4 Minimize center- and grain

boundary-segregation of

C, P and S in slab

· Low superheat casting and elect romagnetic

flow control in mold to increase equiaxed

crystal in pool end of CC-slab.

· Electromagnetic stirring and soft reduction at

pool end to disperse solute e nriched melt in

the boundaries of equiaxed crystals

5 Fine grain bainite structure

with dispersed fine

precipitates

· TMCP and AcC of low C or extra low C HSLA

steel added with microalloying elements (Ti,

Nb, Mo, V, B)

Improving steelmaking and steel properties 519

important amon g them are (a) to desulfurize steel melt to extra low

concentrations (single ppm range) by HMPT and RH/LF (or LF only) with a

flux high in CaO (often CaO+CaF

for prompt melt ing) and (b) add Ca (often in

Fe-clad wire) to the desulfurized melt to convert remaining dissolved S into

solid CaS particles that do not deform in subsequent hot rolling. An example of

HMPT to desulfurize hot metal to 10 ppm by mixing hot metal with CaO+CaF

flux with a mechanical stirrer in the hot metal transfer ladle is shown in Fig.

13.8.

Another example of top injection of similar flux onto melt in RH vessel

to desulfurize steel to < 5 ppm is shown in Fig. 13.9.

CaS particles formed are

largely removed during the HMPT or RH processing. Center segregation of C

and P that occurs during CC is reduced in two ways. One is to disperse solute

enriched melt into the boundary area of sediment equiaxed dendrite crystals

13.8 Desulfurization of hot metal with CaO ± CaF

flux by impel ler stirring

(Ogawa et al.).

520 Fundamentals of metallurgy

where the solutes are less segregated. The other is to prevent inflow of solute

enriched interdendritic melt into the thin melt pool end of CC slab. The former

can be done by a combination of low superheat casting and electromagnetic melt

flow control in the mold that increases nucleation, growth and sedimentation of

equiaxed crystals. The latter is practiced by soft reduction of the pool end with

supporting roll pairs to prevent shrinkage-driven flow that is caused by

solidification. Also, the decreased C-equivalent mentioned above effectively

reduced the formation of the hard and brittle segregate zone.

In the pipeline case, the required properties demanded process developments

in steelmaking, casting, rolling and heat treatment that in turn have made the

production of HIC resistant pipe interactively possible up to X65 grade (X100 in

the near future).

13.2.3 Properties driven by processes

The ball bearing is one of the mos t important parts for rotating machines

including hard disc drives of PCs and engine and traction systems of

automobiles. Resistance to wear and fatigue cracking on rolling contact must

13.9 Desulfurization of steel in RH vessel with CaO±CaF

flux blasting through

top lance (Okada et al.).

Improving steelmaking and steel properties 521

be very high to assure durability of the machines, and hence surface hardness of

the bearing must be high with ductile interiors. Accordingly, high carbon

(~1%C) Al deoxidized steel with Cr has been used. As shown in Fig. 13.10,

fatigue limit expressed in terms of 10% failure life, B

, of typical ball bearings

was of the order of 10

rotation under a given test condition. Here, the bearing

was made of steel containing about 30 ppm of O that was mostly in the form of

oxide inclusions. The root cause of the failure was identified to be large

indigenous and exogenous oxide inclusions occurring during deoxidation and

casting of steel melt. The critical size of the large inclusions was roughly 30 m

but it has becom e sma ller, <15 m, as demands on fatigue life get more

stringent. It may be smaller in future.

To reduce the harm of the inclusions, ladle degassing was introduced which

effectively reduced total oxygen content, [O]

(sum of dissolved [O] in the melt

and O in the deoxidation product, Al

, suspending in the melt), in the high

carbon melt to 15 ppm. Later, RH process for more effective deoxidation under

vacuum has been implemented to make [O]

a low 8.3 ppm. The decrease of [O]

by ladle degassing and RH processing contributed greatly to reduce macro

inclusions. They were formed by the collision and agglomeration of fine Al

inclusions suspended in the melt during melt transfer, but not removed from the

13.10 Improvement of fatigue life of ball bearing with decreasing oxide

inclusion content by process development of steelmaking and casting

(Uesugi).

522 Fundamentals of metallurgy

melt. In addition, variation in the amount of exogenous macro inclusions

occurring from reoxidation by air and slag entrainment, and formation of macro

inclusion during solidification were much reduced by implementing continuous

casting (CC). As a resul t, [O]

decreased to less than 6 ppm. Further

implementation of bottom tapping (EBT) of EAF steel melt has effectively

reduced macro inclusions arising from the reoxidation of the melt by oxidizing

EAF slag and emulsification of the slag into the melt in the ladle. It is to be

noted that the macro inclusions are detrimental to fatigue life but their number

density is so small, and hence cannot be detected by total O analysis.

Later investigation on the critica l size of inclusions that influences fatigue life

has revealed that not only oxide inclusions but also Ti(C, N) inclusions, both in

excess of 15 m in size as mentioned before, are harmfu l. Accordingly, thorough

measures have been taken to minimize the numb er density of the two inclusions

larger than the critical size by attending to every possible source of O and Ti. The

sources of O were the reoxidation of Al-deoxidized melt by ladle refractory, ladle

glaze, air and oxidizing slag and entrainment of slag during melt transfer into the

CC-mold. The sources of Ti were inputs from ferroalloys added to the melt (>

15 ppm). These measures reduced O and Ti to 3.8 ppm and 5 ppm by 1996.

It must be noted that such a low total O only does not guarantee the absence

of inclusions larger than the critical size, since they occur by chance. However,

measures that consistently achieve total O of less than 4 ppm are more likely to

assure fewer num ber densities of such macro oxide inclusions. As a

consequence, the fatigue life has now exceeded 10

rotations, almost two

orders of magnitude of fatigue life prolongation. A recent concern is to reduce

the formation of large, hard octagonal spinel inclusions (MgOAl

). They

probably arise from the reaction between the suspending deoxidation product,

, and Mg produced by the reduction of MgO in slag and refractory by Al in

the melt during intensive refining by stirring of the melt to remove the Al

inclusions.

13.3 Optimization of processes to meet properties and

productivity

To make steel, either the BF±BOF or scrap±EAF route combined with secondary

refining and casting processes is utilized. Function and characteristics of these

processes are summarized in Table 13.3 (see also Fig. 13.5). Depending on the

requirements specific to each grade of steel, one or two-in-series of the

secondary refining processes are chosen to bridge the primary steelmaking

process, BOF or EAF to CC. Investment, productivity, availability of iron

sources, cost of operation and versatility of operation to cover various steel

grades, all determine the choice. Matching of the production rate and cycle time

of processing among the BOF, EAF, secon dary refining process and CC should

be made consistent.

Improving steelmaking and steel properties 523

13.3.1 Processes to optimize impurities and alloying elements

Typical impurities found in steel are H, N, O (including oxide inclusions) P, S,

non-volatile tramp elements (Cu, Cr, Ni and Sn), and in some cases volatile tramp

elements (Zn, Pb, Sb and As). The impurities occur from hot metal (P, S), scrap

(P, S and tramp elements), ambient air (H, N and O) and refractory (O), and

adversely influence steel properties and may cause process upsets. The impurities

need to be removed below a critical concentration which is specific to a ste el

grade as roughly shown in Table 13.4. It depends often on subsequent processing

to semis or final products. To minimize the cost of removing the impurities, the

concentration of the impurities should be set somewhat lower (with tolerance

given for concentration variation occurring from process fluctuation and

segregation) but not excessively lower than the critica l one for cost reasons.

The tolerance depends much on the process itself and operational skill.

In the past, a combination of BOF or EAF with secondary refining was

adequate to remove the impurities to fulfill properties requirements. BOF or

EAF removed most of P and part of S into basic CaO-MgO-SiO

-Fe

O slag as

3CaOP

and CaS, and secondary refining removed the remainder of S. As the

properties requirements increased, P and S had to be further reduced. Process

optimization has been made first by improving secondary refining, and second

by adding HMPT for desulfurization to achieve lower S at a reasonabl e cost. So

Table 13.3 Function of secondary refining processes to reduce impurity elements

and inclusions, add alloys and control temperature

Process H C N O S P Alloy Temp Other features

RH S S S Y High speed mass processing

RH-KTB S SS W S M Y W RH with O

top blow

RH-PB S SS W S M Y W RH with flux side blow

RH-PI S SS W S S Y W RH with flux top blow

ASEA-SKF S S SS Y + Red S Magnetic stirring +

arc heating

LF S S SS Y + Red S Ar stirring + arc heating

AOD S S S S Y + Red S Ar+O

side blow, stainless

VAD S S Y + Red S Ar bottom bubbling in

vacuum

VOD S S M S Y + Red

Sameas above + O

top blow

TN M S Ca injection sulfide control

CAS M M Y Ar bottom bubbling + top

slag

CAS-OB W M M Y W

Same as above + O

top blow

HMPT S S See text

Note: Effectiveness of impurity removal or temperature controlis qualitatively shown by SS (super

strong removal to single ppm range), S (strong), M (modest) orW (weak). For alloying capability,Y

(yes) andY + Red (yes and reduction of alloy metal oxide possible)

524 Fundamentals of metallurgy

far, optimization has been proceeding toward splitting the original refining

function of the BOF into HMPT and secondary refining, leaving the BOF as a

decarburizer.

In Japan, BOFs operate at a higher hot metal ratio (roughly 90%) than in USA

and Europe (75%) owing to sufficient BF capac ity with its cost efficient

operation. HMPT for both dephosphorization and desulfurization has been

extensively developed and industrialized in most of integrated steel plants at

very high availability due to favorable thermal balance with the higher hot metal

operation. In fact, some plants are processing near 100% of hot metal by the

HMPT, making it possible to take advantages listed in Table 13.5. Strong

demands for steels extra low in P and S are behind the choice of this type of

HMPT which could as well be the case in industrially advanced countries in

future if demands for extra low P steels increase.

HMP T in this case is carried out in the following se quenc e: the

desiliconizing fluxes are blasted to hot metal with O

in the BF runner or

tilting tundish which distributes hot metal into transfer ladles. Low Si (0.3%)

operation of the BF, followed by desiliconization to 0.1%Si of hot metal

during tapping into the transfer ladle is preferred to realize the above-mentioned

advantages. The desulfurizing fluxes are either injected with N

into or mixed

with hot metal by mechanical- or gas-stirring in the hot metal transfer/ charging

vessels. The dephosphorizing fluxes are mostly injected into the hot metal

transfer vessels (torpedo- or ladle-type) with O

, with some top addition. Either

pre-mixed or post-mixed fluxes are used, but post-mixing seems to have

advantages in terms of cost and dynamic control of the process. For more

efficient refining with less slag, many efforts have been paid to find fluxes with

Table 13.4 Tolerable maximum inclusion sizes and impurity elements contents in

high performance steels

Application Key property Critical inclusion Critical impurity

size (m) content (ppm)

DI-can sheet Flange crack <40

SEDDQ sheet Average r > 2:0 C < 20, N < 30

Shadow mask Blur in etching <5 Low S

Lead frame Punch crack <5

Sour gas pipe HIC Shape control S<5

LNG plate Embrittlement P < 30, S < 10

Lamellar tear Z-crack Shape control ibid.

Bearing, race Rolling fatigue <10 O < 10, Ti < 15

Case hardening Fatigue crack <15 O < 15, Ti < 50

Tire cord Rupture Shape control < 20 Al < 10

Spring wire Fatigue Shape control < 20 ibid.

Improving steelmaking and steel properties 525