effects of porosity and slag former

7/29/2019 Effects of Porosity and Slag Former

http://slidepdf.com/reader/full/effects-of-porosity-and-slag-former 1/10

Instructions for use

TitleEffects of Porosity and Slag Former Amount on Rate of Heating-up Reduction of Self-fluxed Pellet

Author(s) Vahdati khaki, Jalil; Kashiwaya, Yoshiaki; Ishii, Kuniyoshi

Citation北海道大學工學部研究報告 =Bulletin of the Faculty of Engineering, Hokkaido University, 162: 273-281

Issue Date 1993-01-29

Doc URL http://hdl.handle.net/2115/42331

Right

Type bulletin (article)

AdditionalInformation

Hokkaido University Collection of Scholarly and Academic Papers : HUSCAP

http://eprints.lib.hokudai.ac.jp/dspace/about.en.jsp

http://eprints.lib.hokudai.ac.jp/dspace/about.en.jsp



北海道大学工学部研究報岱

第162号（平成4年）

Bulletin of the Factilty of Engineering

I’lokkaido University， No．！62 （1992）

Effects of Porosity aRd Slag Former Amount on Rate of

Heating−up Reduction of Self−fluxed Pellet．

Jalil VAHDATI lg−IAKI， Yoshiaki KAsHlwAyA

and KuniyoshHSHII

（Received September ！8， 1992）

Abstract

Six kinds of self−fluxed pellets were used in heatingunup reduction experiments． lt was

fo岨d that the reducibility of pellet increased wlth the increase of porosity， increase of CaO／

SiO2， and decrease of slag former amount． Starting temperature of slag flowing−out

decreased with increase of slag former amount and sintering temperature and increased with

the increase of porosity． If the original characteristics of pellets （porosity， sintering temper−

ature and slag fonner ainount） are known， then starting temperature of siag flowingunottt can

be calculated under certain experimental conditions． lt was also found that the shrinkage

was controlled by the reduction rate at high temperature over ！00eOC．

i● 至二簸｛⊃rodLucもio簸

In regard with blast furnace operation， the reducibility of iron ore is an important

parameter which influences its efficiency and operating processes． On the other hand， the

reduction of iron ore is affected by a lot of factors related to its chemical composition，

physical properties， and micro structural aspects．

Many investigations have recently been made to reveal the factors influencing the

reduction of iron ore． Some of these factors related to chemical compositioR of iron ore，

which included SiO2 contentirm5）， CaO content”3’ 6＞，／lVlgO content3一“・ ff＞， A120．3 content2・7−8）， and

potassium content3’9一’2＞． Effects of sulphur content in the reducing gasi3m”4）， and particle size

of iron orei5＞ have also been studied．

In this study， heating−up reduction under load experiments were carried out tising six

kinds of se，lf−fluxed pellets and CO／N2 mixture as reducing gas． The effects of porosity，

sintering temperature， slag former amount （sum of weight percent of CaO， SiO2， A1203， and

MgO）， and basicity （CaO／SiO2） on reducibility of pellet and its behaviour in high temperature

region were studied．

2． Method an〔嚢如醗a重eit’ia澄s

A scheine of experimental apparatus used in this study is shown in Fig． 1． The furnace

used was a vertical one equipped with six SiC heating e｝einents and an alumina reaction tube

金属工学科金属工学第工講座



274 ∫alil VAHI）ATI KHAKI， Yoshiakl KASIHWAYA and Kuniyoshi ISHI工

lnfrared Gas 真隠lyS巳r

〈＝

｝ial Gauge

／

T， C，

Pressure

Tra葺s虹er

seR Co葭Puter

）t13’ss Fie’s” Centrellei

Fig．1 A scheme of experimental apparatus．

pressure transducer， wet gas volumeter，

and infrared gas analysers were sampled

by a micro computer．

Fig．2 shows a scheme of graphite cru−

cible in which six pellets of almost same

size （11．2−！1．5 mm in diameter） were placed

on the top of coke layers supported by two

graphite side plates． A coke plate （20×10

mm） was trimmed off and placed on the top

of pellet column for uniform loading．

Thus， the pellets were surrounded by car−

bonaceous materials． The uniform flow

of reducing gas was also made possible

（42 mm 1． D． and 100e mm L．） and controlled by a SCR power regulator． Taking Xmray

pictures and also recording of video tape of samples’ were made in the course of reduction

experiment using X−ray system． The reducing gas was prepared by mixing each component

gas whose flow rate was precisely controlled by the thermal mass flow controller （］v3［FC）．

CO and CO2 contents of the outlet gas were analysed using the respective infrared gas

analysers． The total flow rate of outlet gas was measured by a wet gas volumeter． A

constant load of 1 kgf／cm2 which adjusted by steel balls was applied on the samples in all

experiments． The expansion and shrinkage of the samples was detected by an electrical dial

gauge as the displacement of a silica rod setting to the top of sample bed． The pressure drop

between the inlet and outlet gas was also continuously measured using a pressure transducer．

All data from thermocouples， dial gauge，． te diiatomterT

雛．撫

．懸難轡

Side plate

A−A＋一e．

ge

A

r

．it （）u；ll−lz i一｛．Hl

一．

S．．t一

．／一一？lunslor foi’ i｛）：id

sgd一一 Coke

繋

繊

難

／／／／11

一’

S3， mpie

Coke

Crucible（Graphite）

O．D．＝36mm

互．D．＝23㎜

H ＝？Gmm

｝｛oider

Themocouple

Fig． 2 Graphite crucible assembly．



Effects of Porosity and Slag Former Amount on Rate of

Heating−up Reduction of Self”fluxed Pellet． 275

through 72 holes of e．8 mm in diameter at the bottom of the graphite crucible．

In all heatingmtip reduction experiments under load， the samples were heated in a

nitrogen atmosphere up to 20eeC from which the nitrogen was changed to reducing gas of a

given flow rate and compositioR． Heating−up was started with the rate of 10eC／min and

continued up to meltingmdown of the pellets． Total flowrate of reaction gas was 200ecrr｝3／

miR （STP） and the composition was 30volO／eCO and 70volO／oN2 （table 1）．

Table 1 Experimental conditions，

Cas composition

Gas flow rate

Heating rate

Load

Temperature range

30 volO／． CO−70 vol％o N2

2eOO Ncc／min

lo eC／min

1 1〈gf／cm2

200 OC−meltingrrdown

The rate of reduction （RDR） was calculated from the mass balance of oxygen of the inlet

and outlet gas according to Eq．（1）．

RDR（％ min−i）＝［（CO十2CO，），．，一（CO÷2CO2）i．］．160e／22414／［O］（1）

Where ［O］ is the amount of reducible oxygen （g） in six pellets． CO and CO2 represent

the mass flow rate of respective gas， which were analysed by infrared gas analysers． The

reduction degree （TRD） at each time was also calculated by summation of RDR from the

beginning of reduction expressed as Eq．（2）．

TRD＝＝2RDR． At （2）The ores used in this study were six kinds of self−fluxed pellets． The chemical composi−

tion and physical properties of pellets are given in table 2． ln order to have specified

porosities， each type of pellets was sintered at a different temperature． The samples B−a，

B−b， and Brmc had the same chemical compositions and slag fomaer amounts and different

porosities． The pellets A， B−b， and D were at the almost same porosities and basicities but

different slag former amounts． The porosity and the slag former amount ef types C and D

were almost equal and basicity was different．

Table 2 Chemical composition and physical properties of selfrmfluxed pellets．

Chemical composition（wむ％）Sample

T．Fe FeO Sio2 A1203 CaO MgOS｝ag fom三er

≠高盾浮獅煤iwt，％）

CaO／Sio，

@ （一）

Sintering

狽?ｍＰ．（℃）

Porosity@ （％）

A 63．99 0．33 2．38 0．86 3．36 1．06 7．66 1．41 1220 24．30

a 1300 20．92

B b 59．60 2．17 4．21 1．76 5．85 2．13 ！3．95 L39 1265 25．60

C 1250 26．02

C 55．86 2．40 6．26 2．44 7．97 3．23 ！9．90 1．27 1265 24．96

D 54．89 0．54 6．25 2．34 8．66 2．67 19．92 1．39 1294 25．28

X・NSIag former amount ： SiO2十A1203−YCaOH−MgO



276 JaliユVAHDATI KHAK至， Yoshiaki KAsHIWAYA and Kuniyoshi IsHII

3．聡esu夏ts＆翻discscss璽OR

Reduction rate （RDR） of pellet was calculated in the course of heating−up reduction by

oxygen balance based on flow rates of CO and CO2 of inlet and outlet． The reduction degree

（TRD） was also calculated at each time by summation of RDR fronn’the beginning of

reduction． The curves of reduction degree （TRD） of all six kinds of pellets are plotted

agaiiist temperature and shown in Fig． 3． Generally， whole reduction process may roughly

be classified into two regions of

a） gasrmsolid reaction period in low temperatures and

b） gasmsolidrmliquid reaction period in high temperature．

10e

§一 8e

の￡

留6008

；3409

8ar 20

A

e20e

一一@B−a

一一一一@B−b

一一@B−c

ee−ee b

D

づ．老≦s

．ノ1

／／一一t’／’

／．．≒シ／

，づ”〃八

一ze

L

4eg 6go soo l ooo 12ge 14ee 16soTeraperature （”c｝

Fig． 3 Reduction curves of six kinds of pellets．

The region （a） is less than leeOOC and the slope of curve（or the rate of reduction） increases

continuously． Then the region （b） is a part above 100e“C and the slope of TRD curve

gradually decreases， which is caused from the closing of pore by formation of liquid slag

inside the pellet．

In the first region （a）， the reducibility of pellets can be compared among different kinds

of pellets． ln comparison with Brma， Brmb， and B−c， which are of the same chemical composi−

tion and different porosity， the reducibility of pellet increases with increasing of porosity．

Sample B−a of the lowest porosity has the lowest reducibility in all studied pellets． Compar−

ing pellets A， Brmb， and D， which are almost of the same porosity and basicity and different

slag former amount， sample A of the lowest slag former amount and the highest hematite

content has the highest reducibility among the all six kinds of pellets． Pellets C and D have

the same slag former amouRt and porosity and different basicity （CaO／SiO2）． From TRD

curves of these two kinds of samples， basicity also shows a positive effect on reducibility of

pellet， when the basicity increases frorn 1．27 in pellet C to 1．39 in pellet D．

IR high temperature region 〈b）， there is a reduction retardatien period in all cases at the

beginning of the region， which is related to sintering of metallic shell at the surface and




Heating−up Reduction of Self−fluxed Pellet． 277

formation of melt inside the pellet． The reduction retardation period is followed by consid−

erable rising of reductioR degree and pressure drop due to start of smelting reductien after

flowingrmout of inolten FeO−slag． lt has been reported that the beginniRg of the cohesive

zoRe iR blast furnace wottld be due to this flowing−out phenomenon．

Starting teniperature of reductioR retardation and srneltiiig reduction and also beginning

of melting−dowll of pellet， which deter面ne its high temperature properties， depend on its

original characteristics and will be discussed later．

一10

0

10

2G 設） 30

ωan 40c5

ML

＝Qり

5C

60

79

80

ge

lOO

㎜㎜s㎜㎜T㎜一一一｝一丁一丁一 Fn…T”「一7「一

A

rm’rm@B−a

一一一一一 @B−b

ww@一 B−c

−i−te b

一一一一@D

ヘセへ

、態，

慧．・

＼＼、＼

t’K”こ．、、、＼

＼〉急、

＼黙＼

800 900 1ooo “ee 1200 1300Temperature ｛’C）

1400 1500

Fig． 4 Variations of shrinkage of pellets during heating−up．

Fig．4 shows the diagrams of shrinkage of the bed of pellets in the course of heating−up

reduction at a heating rate of 10“C／Riin for all six kinds of pellets． An initial sHght

expansion of samples was observed， which turned to the shrinkage as the heating−up ancl

reduction proceeded． There was an abrupt shrinkage before 1200 eC due to formation of

molten slag within the pellets． ln the other

words， shrinkage was fast during reduction

retardation period． Flowing−out of molten

primary slag from about 1300 “C， at the end of

recluceion retardation period， was ac−

companied by slow shrinking of the samples．

Finally， another abrupt shrinkage was obser−

ved over 1400 OC because of melting−down of

metallic iron． Comparing the shrinkage dia−

grams of samples B−a， B−b， and Bmc， the

temperature at which coRtraction of the bed

starts decreases by increase of porosity． lt

is noticeable that higher porosity of pellet led

to higher reduction degree． lt is also found

from Fig． 3 that higher porosity results in

翼）

ω

㎝

工⊂

￡ω

一10

o

10

2e

30

40

so

60

70

80

90

1ee

9

TTrr’rT’un一一一丁 T U．門…1

AX一一・m

@B−a ’ w．

一一L一一曽 a−b

一一@B−c。’唇’。

一一一一一一一一 @D

x

ltNl

N

ヤ、

黙

ll・

｝ll

ムい

弐い

ll

・様＼＼も

＼きミ≧＼＼・

＼．一㌔『一蛍＼

＼墨＼

＼1＼・、さ

ご＼、’

s

Fig． ro

20 40 6e so looReduct i on degr’ee （Z）

Shrinl〈age vs reduction degree of pellets．



278 Jalil VAHDATI KHAKI， Yoshiaki KAsHlwAYA and Kuniyoshi ISHII

lower shrinkage starting temperature． From these facts， it is considered that the reduction

degree is one of the shrinkage controlling factors at high temperature region． That is more

clear in Fig． 5 in which shrinl〈age of samples is plotted against TRD． ln the first abrupt

shrinkage region， the slope of the curves is almost the same， that is， the ratio of shrinkage

rate to reduction rate is almost same． So in that temperature region， over 100e eC， it is

considered that the shrinkage is controlled by reduction process．

Above mentioned high temperature （over 1000 “C） properties of pellet can be classified

into four characteristic temperatures as follows ：

1） temperature at which shrinl〈age starts （expansion changes to shrinkage）

2） temperature at which formation of melt starts

3） startiltg temperature of flowing−out of molten slag

4） starting temperature of melting−down of metallic iron

Siag for国er ＆皿ounし［％］

ア 11 15 19

i400FA｛ F o．Oo a Lo O ．o

oco

三

日

竃畿

E一

1350

1300

1250

t200

60 70 eD SO IDO

oo

o

o

oo o

og

Primar．v sEag

melting temp．

oo

35 45 55 65 75

1150

鷲oo

｛loc

Oo oO

o

35 4e 45 5g 55

tooo

900

sgo

o

尋

o

o S

Shriiikage start temp．o

e

o

o o

o

o

e

20 22 24 26Porosity ［％］

o

o

o o

o

20 24 2g 32 36 f 290 t 23e 1260 f 290 t 320

TRD ［X］ Sintering temp，［“C］

Fig．6 Effects of slag former amount， sintering temperature， porosity， and TRD on hightemperature properties．

Fig． 6 shows the effects of porosity， slag former amount， sintering temperature， and

reduction degree （TR9） on these high temperature properties of pellet． Based on Fi．cr． 3 and

Fig． 6， the melting temperature of primary slag is affected by reduction degree which is

related to porosity and slag former amount． Samples A and Brma of the highest and lowest

reducibility had the lowest and the highest melting temperature of primary slag， respectively．

The total reduction degrees at this temperature were the highest （52．20／o） and the lowest 〈350／o）

among all six 1〈inds of pellets for A and B−a， respectiveiy． This temperature also increases

by increase of the temperature at which the sample was sintered during preparation process．

The temperature at which flowingmout of slag occurs depends on porosity， slag former




Heating”up Reduction of Selfrmfluxed PeHet． 279

amount， sintering temperature and reducibility of pellet． This temperature increases with

the iRcrease of porosity and decreases by increasing of slag former amount aRd sintering

temperature． Because， the pellet of higher porosity has more capacity for holdiRg of molten

slag and larger pores keeps the molten slag from flowing−out at lower temperature．Moreover， the thicker metallic shell at the surface of pellet， in the case of higher reducibility

due to higher porosity， resists against flowing−out of molten slag and causes to an increase

of this temperature． Therefore at a constant heating rate and composition of reducing gas，

starting temperature of slag flowing−out may be considered as a function of porosity， slag

former amount and sintering temperature． The amount of smelting reduction decreases

with increase of starting temperature of slag flowing−out （Tf） and increase of reducibility of

pellet （Fig．3） which is desirable in blast furnace operation． ln pellet A of the highest

reducibility and the highest Tf， only 380／o of reducible oxygen was removed through smelting

reduction． While， in pellet B−a， of the lowest reducibility and the lowest Tf， 660／o of

reducible oxygen was reduced in smelting reduction process．

From these results， it was found that the starting temperature of slag flowing−out Tf is

a function of slag former amouRt SFA， sintering temperature Ts and porosity E． Using a

multip王e regression analysis method， Eq．3was found for calculation of Tf（starting tempera−

ture of slag flowingrmout）．

T，＝＝f（SFA， Ts， E）

T，＝：i 一！31．3（SFA）一〇．362（Ts）十 287．2（E）十！699．6 （3）

Where：

Variab／es

Tr ： starting temperature of slag flow−out （OC）

SFA： slag former amount in pellet （一）

SFA ＝：：（O／o SiO2 十％CaO十 O／o A1203 十 O／o ］vlgO）／100

Ts ： sintering temperature of pellet （“C）

E ： initial porosity of pellet （一）

Conditions

Reducing gas ： N2−30volO／oCO

CaO／SiO2 of pellet：！．4 （一）

Heating rate ： 10 （OC／min）

Tf is recalculated form Eq． 3 and compared with experimental data in Fig． 7 for all six

kinds of pellet． There is a good agreement betxrvreen calculation and experiment． ln order

to check the accuracy of this model， experimental results of previous studies performed by

other researchers were compared with calculation from present model in table 3． ln those

studies， the experimental conditions and CaO／SiO2 of pellet were the same as present work．



28e Jalil VAI−IDATI 1〈1−IAKI， Yoshiaki KASHIWAYA and Kuniyoshi ls｝m

o

旺

mppoE

o

鮎anの

Qり

1350

1330

t31e

12SO

1270

o Exper i ment data

A Calculation

愈

125G

12gO 1230 1260 12SO 1320

Sin七eザing 七emρ．（。C）

Fig． 7 Comparison between experiment data and model caicuiation．

Table 3 Comparison between previous studies and model

caiculation of Tr．

Prev｛OUs work Exper｛n｝elltai data CalCulatlO鳶da之a

KOndo e之aい／308DC 1314．2。C

Takahashi＊＊ 1312．C 1302．4’C

＊ S， Kondo， K． lshii， and Y． Kashiwaya

” lxiT． Takahashi ： private communication

4． CO蟄C亘US蓋0互蓋

Reducibility of iron ore and its high temperature behaviour are affected by its original

characteristics and have aR iinportant influence on blast furnace operation． Using six kinds

of self−fluxed pellets， heating−up reduction experiments were performed． The relationships

between reducibility and high temperature properties of pellet were studied． Results

obtained are as follows ：

（1） The reducibility of pellet was affected by its original characteristics and increased with

the increase of porosity and the decrease of slag former amouRt． Also in the case of

constant porosity and slag former amount， reducibility increased as the basicity increased

from 1．27 to 1．39．

（2） Starting temperature of slag flowing−out （Tf）， which was related to capacity of pellet

for holding the melt， decreased with the increase of slag former amount and increased with

the increase of porosity． This temperature also decreased with increasing of sintering

temperature of pellet．

（3） Melting−down temperature of pellet increased with the increase of porosity．

（4） At high temperatures （over ！0eO eC）， shrinkage was controlled by reduction reaction．



Effects of 1）orosity and Slag Former Amount on Rate of

I’leating−up Reduction of Se］．f−fluxed Pellet． 281

（5） Startkig temperature of slag flowingrmout （Tf＞ could 1）e considered as a function of

perosity， slag former amotmt， and sintering teinperature． At certain experii：nental condi−

tions， this temperature for different 1〈inds of pellets ceuld be calculated from Eq． 3．

References

1 ） E． T． Turkdogan ancl J． V， Vinters ： CANAI）IAN METALLURGICAL QUARTERLY， 1．2 （1973） No． 1， p，

9．

2） N．Shigematsti and 1’1．Iwai ：一 Trans， ISIJ， 28 （1988）， p． 206．

3） S． Geva， M． Farren， D． 1’1． St． john， and P． C． 1’layes： Metall． Trans． B， 21B Q990）， p． 743．

4） N．Shigematsu and H．lwai： Trans． ISIJ， 25 （1985）， p． 143．

5） A． A． El−Geassy： Trans． ISIJ， 25 （1985）， p． 1036，

6） T． El Kasabgy and W”1〈． Lu： Metall． Trans． B， 11B （1．980）， p，409．．

7） U． F． Chinje and J， 1’1． E． Jeffes： lronmal〈ing and Steel mal〈ing， 16 （19． 89） Nio． 2， p． 90．

8） U． F． Chinje and J， 1’1． E． Jeffes： lronmaking and Stee｝ mal〈ing， 17 （1990） No． 5， p． 317．

9） M．Gougeon， B． Dupre， ancl C． Gleitzer： Metall， Trans． 13， 17B （1986）， p． 6．57．

10）」． Roederer， B． Bernard， and C， Gleitzer： Steel research， 58 （1987） NTo． 6， p， 247，

！l）」． Roederer， F． Jeannot， B． Dttpre， and C． Gleitzer ： Steel reseai’ch， 58 （1987） No． 6， p． 252．

12） H．， Nakagawa and Y． Ono ： Trans． ISIJ， 25 （！9． 85）， p． 1021．

13） S． 1’layashi， Y． lguchi ancl 3． 1，lirao： ’i’rans． ISIJ， 26 （1986）， p． 528．

14） D． 1’1， St． John， F． Nakiboglu， and P． C， 1’layes： Metall． Trans． B， 17 （1｛86， p． 383’． ’

15） A． A， ElumGeassy and V． Rajakumar ： Trans． ISIJ， 25 （1985）， p． 1202．

effects of porosity and slag former

Documents