" Tatara " Process*

- A Pig Iron- and Steel-Making Process, Transmitted from

Ancient Times in Japan -

By lukichi KOZ lIka**

S y n op sis at " C ha -n o-Yu ") were m a d e from it.

From Gncient times 111 j ajJa n Ihere are excellent swords called In th e fo llowing the a utho r d escribes " T a ta ra"
" NijJjJon- t6" as a symbol of " Samu rai" . The)' are weLL-kn owll and p rocess putting th e stress o n th e direct steelma kin g
sland high in their aesthetic value aLL over Ihe world. The material of m ethod , " Keraoshi " .
" N ijJjJo /l-to" is " Tamahagallc" (c/'ltde steel ) or " H ocho-Ietsu"
(crude wrought iron) p roduced by " T atara " process w hich was devel- II. The Historical Development of "Tatara ))
oped jJeCl.tiiariy ill j ajJaIl alld has bem hallded dow n to us f rom ollr Process
ancestors.
This pajJer briefly describes the method of " Totaro " /J I"Qcess and
From th e a rchaeo logica l viewpoin t, it is r egard ed
several malters related to it. th at the process was brou gh t to j a pan for th e firsL
Lime from th e Asia tic Con tin ent in th e first a nd secon d
I. Introduction centuries . The prim itive m e thod of th e process ex-
R aw m ateria l used in " T atara " process is th c isted in the fa bulo us age of j a pa n, when o ur a ncesto rs .
specia l iro n san d ca lled " M asa " o r "Aka m e" whi ch m an ufac tu red b y a id of n at ura l dra ft a mixture of
is min ed onl y in San-in distri ct, so u th-wes t pa rt of iro n, steel, sponge iro n , slag, a nd cha rcoal in a hole
th e m a in isla nd of j a p a n . C ha rcoa l is used as fu e l. which is dug u nd erground . After d estroyin g the
Th e sha ft of th e furn ace is low a nd rec ta ng ula r, a nd furn ace, thi s mixture was fo rged by ha nd into swo rds,
th e body is m a d e of fire clay. Photo. 1 shows th e weapons, a nd agricul tura l in struments.
furn ace bod y. By th e early 13th cen tur y " T atara " process was
improved to the present ty p e a nd the process beca m e
to be opera ted indoors in stea d of outdoors.
In this way the direct steelma king industry in j a pa n
was establish ed , earlier th an th e estab lishm ent of it
in E urop ean countr ies by on e hundred an d severa l
d ecad es .
Speakin g of th e d evelopment of iron sand sm elting,
the d evelopmen t of bellows must be consider ed. Th e
history of blast in j a pa n began with na tural dra ft ,
a nd t hen leather bag bello w, blow pipe, a nd ha nd-
worked reciproca tin g bellow " Sa shifuigo )) were d e-
ve loped . Th e las t one was a lread y contrived in th e
Ph oto. I. " Tata ra" fu rn ace eig hth century in j a pa n, a nd a ft er tha t several inven- •
tions a nd improvem ents wer e in troduced in it. Th e
By " T atara " p rocess a re produ ced sm a ll lumps o f m otive power of bellow cha nged from ha nd power to
crud e steel " T a m a haga ne ", bliste r stee l " H era " treadl e as we see a n exa mple in tread bellow , " F umi-
(" T a m a hagane" in cl ud ed in it), pi g iron, etc. In fui go )) (see Photo. 4 ) . W a te r mill ha d not b een used
the p rocess there a re two kin ds of methods : one is as th e power until the ea rl y 20 th century. It is be-
direct steelm a ki ng met hod ca ll ed " K eraos hi " by cause t he wa ter stream n ea r t he furn ace was r egarded
which m o re crud e steel is p rod uced th a n b y th e other m oistu rizing th e fu rnace , a nd it was too di ffi cul t at
method , a nd the o th er is pig iro n-ma king one " Zuku- that time to se t bl as t pipes over long dista n ce .
os hi " . Bu t these method s a re so simila r in m a ny Aro und th e la te 17th a nd earl y 18th cen t uri es, wh en
poin ts tha t the a bove-mentio ned produ cts are m ore th e feuda l sys tem in j a pa n was es ta blished , th e p rod-
or less p r odu ced by eith er method . ucts by " T a ta ra " process becam e a n impor ta n t
Crud e steel ob tain ed is formed in to th e shell o f so urce of revenue in som e p rovinces. Thus the p rocess
sword b y r epeated forg ings, while b lis ter steel a nd p ig was d evelo ped m ore a nd mo re, a nd it b ecam e the
iron a re con ver ted to wro ug h t iron " H ocho-tetsu " present accomplished a r t.
which is used as the co re of sword . Th e furnace was opera ted by a chi effor em a n called
H owever, the produ cts were used not on ly a s the raw " O motemura ge" a nd fo ur assista n ts. Bu ilding of
m ateri a ls of the sword bu t a lso as importa n t ma teria ls the furnace, cha rging o f th e m a teria ls, pulling ou t of
o f hi g h q ua lity tool steels. As pig iron has specia l the p roducts, etc. were carri ed out on " Murage " 's
q ua lity, specia l cas tings (fo r exa m p le, teake ttle used own responsibili ty. H e op era ted th e furna ce in ac-

* O r igin a ll y pu bli shed in T etsu-to-H agalli, 52 ( 1966), 1763 in J apanese . E nglish version rece ived A pril 12, 1967 .
** Ya sug i works. Hi tac hi M eta ls, Ltd ., Yasug i.

( 36 ) Report
 Transactions ISIJ, Vol. 8, 1968 [ 37)

Section aa

G.L. G.L.

"u
'"E

•
~

Section bb ....
""

G.L.

Charcoal
4352
Fur. bottom, "Hondoko"

Air chamber
----t ""~

---------l
Dry clay ""

-----+ '"'
~

•
Shingles

Round stone, "Bozuishi "

Ditch '"'
""
'"' Fig. I.
"Tatara" furnace, (mm),
a scale 1/40

cordance with the art succeeded from his forefathers, adhered on its surface are chipped off. Then it is
as well as with his own several experiences. cut off by machine or hammer to select the parts which
The operation method was kept secret and handed contain a lmost constant carbon. They are crude
down to only his heir, consequently if " Murage " steel called " Tamahagane".
died out, the method was not transmitted to the The remainder of the product is used as the raw
younger generation. materials of wrought iron, crucible steel, and tool
steel.
III. Products and Their Treatments
The amount of the pig iron produced in the process
By " Keraoshi ", direct steelmaking method in is about 20% of the total products.
"Tatara" process, an iron block of about 1.1 x2 .9x
0.27 m (in thickness ) is produced in the furnace. The IV. Construction oj Furnace
block called "Kera" (blister steel) is a mixture of The furnace is constructed at the weathered and
steel, iron, sponge iron, pig iron, slag, and a few pieces perfectly dried granite zone in San-in district.
of charcoal. As Fig. I shows, digging about 3.5 m underground
After the furnace is destroyed , the block is pulled is made in the centre of a ditch that is an outlet of
out a nd cooled in air, and then foreign substances moisture. On the ditch , round stones (abo ut 40 cm

Report
( 38 ) Transactions lSI], Vol. 8,1968

130-+- - -430 - -- ' - --';:-Td 320 110 130

"--- -445 - - ! - --jl/

View a Section b

r -- - 375 - '----! I"

View e Section d

~mw~
10?

•
- ::T~ tu r:re 145 ISO
.. ~ .. -<t---j{- .. -R- ~~~H-r.: Fig. 2.
38 ~ Tapping hol e
"Tatara " furnace body, (mm).
Scale, upper 1/ 10, below 1/20.

in dia. ) are tightly spread covering the area of 6.5 x dust, charcoa l powder, clay, mud , etc.
5.5 m , then it is paved 30 cm thick with shingles, Above the ground the furnace body IS erected,
and moreover dry clay is put on it in a thickness of closely connecting with the furnace bed.
60 cm. Thus the furnace foundation is completed. Fig. I shows "Tatara" furnace entirely, while
Upon the furna ce foundation the furnace bottom called Fig. 2 the part of the furnace body in detail (see
" Hondoko " is set up in the centre, and on both the Photo. I ).
left and right sides of it are installed the air c hambers, The furnace body is built with spccial fire clay (see
" Kobun e ", which are insulations from heat and mois- Table I) by foremen.
ture. In the lower part of the furn ace bottom there There are three tapping holes, (for pig iron and slag)
is another air chamber which is used as a fuel hole on the right and left sides and at the centre of the
for charcoal making. furnace foot. The tapping holes on both sides and
The furnace bottom is filled up with firewoods , at the furnace foot are partitioned by wall. But, as
which are burnt into charcoals just as in a charcoal kiln. operation goes on, the furnace wall is eroded by slag
The upper 1/5 ofthe charcoals is crushed a nd squeezed and the partition is automatically cleared off. It
by mallet. Thus the furna ce bed is completed. means that the fire clay used for furnace building be-
The walls of the furna ce bottom and air chambers comes slag making material, too.
are built of stones or gravels with clay as a binding There are 19 tuyeres horizon tally lined on both
agent. Th e crevices in them are filled up w ith dry sides of the furnace. These tuyeres are connected

Report
 Transactions ISH, Vol. 8 , 1968 ( 39 )

Table I. Composition and size distribution of fire clay (% ) with air chambers through the boxes "Tsuburi"
whi ch distribute air to each tuyere.
~I
Composition
2 3 The tuyere is bored through the furnace wall by
wooden auger when the furna ce shape is completed.
Si0 2 69.24 68.54 65.59
As shown in Fig. 2, the tuyere is enlarged at the middle
A I20 s 12.24 13.12 18.63
point of its length, because in this process more amount
Fe 2 0 s 5.24 4.44 4.82 of air is required, as the operation goes on and the
CaO 0 . 15 0.25 0.23 block of blister steel grows. That is one of the c har-
MgO 0.30 0.26 tr. acteristics of direct steelmaking process "Keraosh i ".
Si0 2 /Remaind ers 169. 24JI 7 . 93 68. 54JI8. 07 65.59/23 .68 The distributing air box and the tuyere are joined
by a bamboo pipe " Kiro " (3 to 4 cm in dia. ) which
Mesh I up5 20
up
32
up
60
up
100 · 100
up down
I Total has an iron cap at apex.
After the construction the furnace is dried by fire-
Size distribution 110 .2 45 16.4 17 .5 8.7 2.2 I----WO- woods for about 15 hr, and then the operation of t he
The corpuscles of the sa mple are el imin ated furnace is to be started.
__ ~ decantation. The house in which the furnace is installed is called
Notes
2 There is no iron sand in the residue. " Tatara " house or "Takadono ".
- '-----'-- ---- Fig. 3 and Photo. 2 show a " Tatara " house which

• Fig. 3. " Tatara" hOllse, (mm), scale, 1/ 100

en
S

Stock room of charcoal.

I .
Stock room of Iron sands.
I
Stock room of charcoal . ~ci
00
,.,.
<'>

"Sumimachi " "Kogan emach i" I "Sumimachi"

U')
': 3
:0 0
U')
'"30> -~.
I '"
U')
<'> 0
..,

•
:0-:0
_.'"
'0>
1f-- - -5455- - -- U-- - - - 7273+ - - - - --III--t-- -5455- ----i1 :0
~
v
-;;
... v
o CIl
" ...
o
... 0
0-"
o I I1 o
o
o
TUyery-~ T
o
--11--- - --+- - - - - - - - -- IL Furnac0l t-::l- - - - ""- - - - - ! - - - - t
--l-I+-
Rest I <-
room '-.... - / g ~ Rest
fur '-....-, room
forem en I I
Air chamber ~

@ ! ~-
f-- 272S- + - 2066 - ~3~40--2067-! en
o
<'>
, :><'
-i",
"g..o..
'"
..,0> 2727
Stock 3" 8..
room
~~
Stock yard of fire cla y, -':(b

<'>
'Tsuchimachi " Door ii>
c=======~oDoor theo=======~~~~~======<r-~~~-O==~====O
of
house I '" ~===:::=I¥
C1

~'-----~-- 84S5 ------~

R eport
( 40 ) Transactions ISIJ, Vol. 8, 1968

remains the same as three hundred years ago. column. (see Table 2)
The bellow was moved by manpower in ancient
V. Another Equipments of " Tatara" Process
times. The followings are typical examples.
1. Bellow (I ) Tread bellow" Fumifuigo " (see Ph oto. 4 and
As shown in Photo. 3, reciprocating wooden bellows Fig. 5)
are used in the process. Four bellows are worked (2) Balance bellow " Tembinfuigo " (see Photo. 5
in a set by water wheel (see Fig. 4) . A set of bellows and Fig. 6 ) (The shape looks like a balance. )
is able to send about I Nm 3 of air in a minute, which However, water wheel power takes the place of
is calculated by the numbers of revolutions of water manpower an d the reciprocating bellow came to be
wheel , and the pressure is 20 to 30 mm by water used in general.

2. Large Drop Hammer, " ado "

The hammer is used to cut roughly the block of
bl ister steel (see Ph oto. 6) into pieces of 500 to 300 kg,
but not to crush but to cut along with the difference

Table 2. Relation between R /min (wa ter whee l)

and amount of air

R /min of wheel 23 24 25 27 30 31

_A
__m_o_u_n_t _o_fa_i_r _(JV in_)~1__2_2 ___2_3____2_4____26____29____3_0_ . . .
__m_8_/m
__

Ph oto. 2. " Tatara" hou se" Sugaya Tatara "

Photo. 4. Tread bellow" Fumifuigo"

•
Photo. 3. Wooden bellow" Sashifuigo " Fig. 5. Tread bellow " Fumifuigo "

Fig. 4.
cp Wooden be ll ow" Sash ifuigo", (mm).
- ti!l'lIl
!!l!!lLf
I.lE
. ++-1
a::::IEtt~~B H BHHEEIlIHl!I- Water wheel Sca le, upper 1/ 20, be low 1/40.
c:Jljb

Report
 Transactions ISIJ, Vol. 8, 1968 ( 41 )

in grain size or glaze and fine cracks, etc. So in this and more, a nd at the same time , to cut off impurities
work a special art and rich experiences are required. ad hered o n the surface of the block.
Weight of the drop ha mmer is 1.2 to 1.5 t and drop Fig. 8 shows the small drop hammer.
distan ce is 5 m in height. Fig. 7 shows the large These small pieces of the block are further cut by
drop hammer. han d into smaller pieces about I kg in weight. In
There is another way of cutting the block by quench- this way crude steel is obtained (see Photo. 7) . Cut-
ing the block in the pond called "Kanaike". But ting at this stage has another mean ing, that is, to
this is not used in these days, because the yield of produce the smaller pieces of crude steel containing
crude stee l is very low. equal amo unt of carbon.

3. Small Drop Hammer, " Kodo" VI. Raw Materials and Products in " Tatara "
Weight of the hammer is 300 to 200 kg a nd thc drop Process
distance is I to 1.5 m in height. Operation of the In " Tatara " process two kinds of iron sands are
hammer depends on manpower, but it can not work used as raw materials: one is " Masa " which is used
so q uickly as general drop hammers. The hammer,
therefore, is rolled up by the wheel which a man turns
as a mouse turns a mouse wheel. The fun ction is to
cut the b lock above-mentioned into small pieces more

• Cast iron 400"' -H--I~~-:!

C.L.

Photo. 5. Balance bellow" Tcmb infui go "

Fig. 7. Large drop hamm e r " Odo ", (mm), sca le IfIOO

• 400
f------ 4400---~-I

Fig. 6. Balance bellow" Tembinfuigo "

Ph oto. 6. Blist er stee l " K e ra ", 2 920 X I I 00 X 250 mm F ig. 8. Small drop hamm er" Kod o ", (mm), sca le 1/ 100

Report
( 42 ) Transactions ISIJ, Vol. 8, 1968

in direct steelmaking process "Keraoshi", and the From the table it is found that "Akame" contains
other is "Akame" used chiefly in pig iron-making more Ti0 2 than " Masa " and that gangue elements
process "Zukuoshi ". These iron sands are pro- (MgO, A1 2 0 3 , etc. ) are not eliminated completely.
duced in San-in distri ct only, and they have dif- But in the process of" Zukuoshi " such a high Ti0 2
ferent properties as compared with iron sands pro- content is out of the question because the products
duced in other places of Japan. When iron sands (pig iron ) are taken out from the furnace in a molten
are repeatedly dressed by magnetic separator after sta teo
crushed under 200 mesh, Ti0 2 content in the former
iron sands is gradually decreased but in the latter the
Ti0 2 content shows a tendency even to increase.

1. "JvJasa"
The iron sand called "Masa" is mined in the
granite zone which is weathered by rain water. But
the iron content is so small as 1 to 2 % . The furnace
is operated with such poor ores and this may be re-
garded to be one of the characteristics of" Tatara "
process.
As shown in Photo. 8, the efflorescent hill is destroyed

•
by picks, and washed away to mountain river by
water jet, then iron sands and another rocks are rubbed
and separated each other. These outflows are taken
into five wooden ducts (each 8 to 28m long) which are
put in series (see Photo. 9) and dressed by hand with
water. Thus it becomes possible to obtain iron sands
containing about 60 % Fe with water.
Table 3 is data of the composition of iron sand
" Masa". From the table we can see that impurities
m the iron sand are very small.

2. "Akame " Photo. 8. Open cut" Kiriha "

This iron sand is used as materials in pig iron-
making process" Zukuoshi " and gained in the weath-
ered diorite zone which included 3 to 5% iron. In-
stead of "Akame", beach iron sands are sometimes
used as raw materials in plants situating in the neigh-
bourhood of the sea. The compositions of these iron
sands are shown in Table 4.

Photo. 7. Crude steel" Tamahagane" Photo. 9. Dressing plant" Toyonagashi "

Table 3. Compositions of iron sands "Masa" after dressing ( % )

Com posi tion l

T.Fe Fe,Os FeO SiO. CaO MgO AI.O s TiO. MnO V.O , P S
Description
"Nakagura" 59.0 64.45 24.27 8.40 2.24 1.54 2.34 I. 27 0.05 0.258 0 .064 0.009
" Hanaidani " 59.98 62.45 20.98 10 . 02 0 .22 1.27 1.64 1.54 0.19 0.240 0.060 0.018
I
Report
 Transactions ISH, Vol. 8, 1968 [ 43 )

Table 4. Compositions of iron sand" Akame" and beach iron sand after magnetic dressing (%)

"::om posi tionl·

T.Fe Fe 2 0 3 FeO Si0 2 CaO MgO AI 2 0 s Ti0 2 MnO V 20, P S
Description
Zakka (" Akame ") 54.56 5 1.08 18.48 14 . 90 1.60 1.74 4.98 6.82 0.05 0.032 0.036
Beach iron sand
57 . 38 50.79 20.4 1 5.42 0.11 0.76 2.55 6.55 0.35 0.22 0.078 0.026
(" Gotsu") I

Table 5. Quality of charcoal

Quality I
Humidit y (%) Volatile matter (%) Fixed C (% ) Ash (% ) Powe r (cal)
Descri ption
For "Tatara" usc 6.21 to 4 . 27 26.91 to 34.34 51 . 96 to 59. 20 1. 42t02.00 5500 to 6000
On the mark et 6 . 70 to 8 . 10 9 . 90 to 12.29 78.49 to 80.39 2.53 to 1.61 7300 to 7600

3. Fuel (Charcoal ) (3) Two kinds of pig irons are produced by " Ta-
Charcoal used in the process is of low grade as tara" process. They are "Hachime" and "Kori-
shown in Table 5; it contains some, but small wooden me" . The former is the pig iron which flows out
parts in it. from the furnace during operation . The pig iron tem-
" Tatara " process successfully utilizes charcoal perature at its flowing out is very low, so that many
• which can be rapidly burnt to ashes. So if charcoals gas bubbles are occluded in it. The latter is the pig
on the market are charged into the furnace, it is choked iron which remains in the furnace bottom till the
and the operation can not be continued finally. final period of operation and shows a dense structure.
Both are white pig iron s. The chemical compositions
4. Products (in Detail ) are shown in Table 7.
In " Tatara " process three kinds of products a re
Table 7. Compositions of pig iron (%)
obtained: ( I) lump steel" Tamah agane" (see Photo.
7), (2) b lister steel " Kera " (see Photo. 6 ) , and (3) Composit ion
C Si Mn P S
p ig Iron.
(I) "Tamahagane" i a crude steel of I to 60 kg
-------
Pig iron
" Hachime "
--- 3.6 1 0.03 0.01 0.033 0.010
in weight. The carbon content is almost constant as " Korimc" 3.55 0.02 tr. 0 . 043 0.0 10
shown in Table 6.
Table 6. Composition of crud e steel" Tamahagane" (% ) These pig irons were used as raw materials of
" Nippon-to" after repeated heating in the oxidizing
~I C Si Mn P S flame and forging, or of specia l castings.
The properties of the pig Irons are itemized as
" Tamahagane " 11.0 to 1.6 0.04 tr. 0.008 0.006
follows:
(a) They have very good fluid ity.

••
,
As shown in the table, impurities contained are
very small, moreover, such impurities as P and S react
with Si0 2 and are occluded as slag. Therefore, in
making Japanese sword" Nippon -to" from " Tama-
hagane", the slags are squeezed out or broken up into
(b ) Their carbon-absorbing rate is very big when
they are remelted.
(c) Their cementite crystal is very large, com-
pared with ordinary pig iron with the same carbon
content.
microscopica l pieces by repeated forging (several
thousand t imes ). In this way "Nippon-to" is VII. Practice oj " Keraoshi " (Direct Steelmak-
produced with the best cutting quality and with the ing Method in " Tatara" Process)
best resistan ce to breaking and bending. The operation is performed by five workers . Two
In other words, steel from" T a mahagane " has the of them are head foremen ca ll ed " Murage" who
following properties. manage overa ll furnace operation on their own re-
(a) Transition points appear very clearly, that is, sponsibi lity, and their main works are:
the steel is very sensitive against heat treatment . (a) to check the degrees of dryness of the furnace
(b) C harpy value is very high. and to dec ide how to carry the furnace operation
(c) It has good deep-drawabi lity. (b) to control the blast
(d) It can be welded without Aux. (c) to check the chargings
T hese facts prove well how "Tamahagane" IS (d) to operate the furnace, especially to take care
pure. of iron sands
(2) Blister steel " Kera " is a steel block in which (e) to decide how to operate the next process
the carbon content is not constant. "Kera" was according to the conditions of blister steel produced
u sed as a raw material of crucible steel in ancient (f) to take the lead in building the furnace for the
times, and today is a good material of tool steel of next operation and drying the furnace.
the highest grade. Two assistants called" Sumitaki " help the" M ura-

Report
:;0
(I)
'd
o *'"""
~
>-i
>1
po
P
(Jl
po
~
1st dJ,\' 2nd da\ 2nl dOl\' 4th dav 5th day o ·
p
Da te. IS No\'. 1935 a.m. 5 6 8 10 12 14 16 18 20 22 24 2 5 8 10 12 14 16 18 20 22 24 2 5 8 10 12 14 16 18 20 22 2·\ 2 4 6 8 10 12 14 rr 6 8 10 .
1
1500 -0)
1 1 ..-(J) 'TI 1 ~ -<D---. ~
--elY Temp,of the surface of " Kera"~
1400 ~ i) <:
Pig iron and s l'lf(' in furna ce Bottom of * The mot ion of ..be II ow5 is rec ipro · ~
1300 o 0 furnace S\ag t~~~~nace S lag t emp. eati ng and th ere for e it goes up and <D
rD rf{---u.." &SI.~_'~n:l~--- u-----o down in the manometer. In ·'Tatar;.1 I--'
-\..l... 0 -ffi. outside 0 __ -0
'" 1200 2 2 - 2 --- -.o---"J2' ill
3 process blowers as turbo or root s to
E -W- "'QY (J)
11 00 2
aTap
'~ blowcr <Ire not used. <D
Co ~ ffi ~
hote
'"E
~ 1000 ~ ** Blast \'olume can be ca lcu la ted by
2
[Qg I------
'"" ~ the number of re\'o lut ion o f water
.0)... whee l. The relation be tw een IU min of
"E ~ ~4jl
~
l!..
i~ i25 'C*'\!360'C""
800 after 4hr after I6hr
ffi
V ~ ? ---- ~ ~
--- ------ ~::::sR3 ~
~\D water wheel and bla s t vo lum e is shown
in Table 2.
W 4
700 ~ ~
4"@ -!. *** Iron sa nds and charcoa l arc
600 charged in layer. It is under s tood from
the table that the opera ti on is contro l·
Pr essu r e of blasf 40
by water column [ ---- - -T-- --- ]--J ------J--- ---- --- ---- - -- ]C----J--:.--JC~: ::: -- --r---1 led by th e amount of iron sands and
mOl 30 L==-_-_l ==-_-~~ J[ ~~~--U':---':-::_-If_-_-.:-::C =::::-- ------- ------ -- ------ -------- - - - -- - ---- ----- that the charging: amount of c harcoals
20
,Jr--- - --0-- -- - - - - -- - - --0----0------0-- --0- - - -- - - - -- -- --0 is constant thr ough out all th e pe ri ods.
Wa, er ",hee l*~., 30
_--0-- - -- -<r"
R min 25 0- ___ - - - -0- - -0- - - - 0- - - - - 0- - - - - - - - - - 0-----

20
Operation period .! "Komori " 7.5 ill' ·· Komorit s ugi " 7.5~~· \obori ·· 1 8h r ··Kudari " 36 hI' _ (ooliilg Build the fur Dr~ the fur. Ii hr L
I-- 3hr -r 4h
J- hr
Charging amou nt 100 ~I Next
cperatJon
of i ron sand kg 50
~"paro,,~ 11111111111111 IIIIUIHI IIUIIIII I I 111II 1111IIII I111111111111111111111111111111111111111111111 1IIIIII IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIjlllJIlllllll1
0
Interva ls 30 min- - l nt('rv~i1s 24 min Int ervals 24 min ----L-- Intervals 30 min
Chargi n~ int erval 100
Charging amount 50
rn 11111111111111111111 111111111111111111111111111111 IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIJIIIIIII 1111111111 11111111111111111111111 111111111111111111111111111111111II
of cha r coa l kg 0

Lapse
I
time fr om Ih e
01 Bellow
o I 2 3 4
5 1a, lop
6 7 8 10 12 141 5 16 18 2021 22
I
24 26 28 30
I l'i' tap 80 , , 1, 1',.200,.
32333435 36 38 40 42 44 46
Brilliaot ,.el low [b,m,'
48 50
Via""
52
h'i,ht 1.5m
54 56 58 60 62
Th",,,,, 10mm
64 6667686970
I
72
!
74 76 78
IL
r Dry th{' fur . with
96 98

hef{lnning of operatuln s tart I Pi, tap 1301.., I Th· k 8 fire \\ood s

Cewe to. hoi,· ",I ..,od a"d hob '''' h"ti< ",\eo ace ",,,,,01 , • .. . • ,[,hid,w," ,,[ [o,·.w.11 12mm " '" ' ' mm
IhJckne s~ of fur . \\all 210mm. Lea\·e s la~ to fl o w OU I Beginning of
Obsen allon o f
fur .condition
Charj.!t' s tart. Flamt' chan~es fr om dar" n'd to II~ht \('110\\ . Flamf' shines all around fur . \\all eroded The end of ope ration 4th day a.m.5
J next ope ration
Bli Slt' r s t(,el ht'gln s tn be product'll In tht' fur . boltom as far as 10 the mid points of tu~·t'rE'. Pull o ut bli s ter s tE'e l. - Kera
CompletIOn of fur.
Build the fur . for next operation
-- .. __ ._-
--

Fig. 9. Data of" Tatara " furnace opera tion (by direct steel making method)

•
 Transactions lSI], Vol. 8 , 1968 ( 45 J

ge ". They are engaged in charging of charcoa l in to the end of the period the pig iron and slag occa-
the furna ce. sionally run over the furn ace bottom and fl ow o ut fro m
The rest of the workers is " Komawari ", who is the tapping holes.
in c harge of treatment of fire clay" Kamatsu chi " for In the centre of the furnace the nuclei of blister
construction of the furn ace. These five men con- steel are form ed , which promote the blister steel to
tinue to work in a " T a tara" house for four days and grow in the fin a l period. The furnace wa ll is eroded
nights (which are the standard p eriod of one cycle of and th e furn ace area is enlarged at the tuyere line to
the operation ). the mid point of the tuyeres.
In a ncient times, more foremen were needed under " Kudari " (The Final P e riod of the Operation)
the necessity of moving the bellows by ma npower, The charging intervals are shortened from 30 to
but the adoption of wa ter whee l power made it pos- 24 min so that blister steel " Kera " g rows up, but
sible to minimi ze the number o f them. charging a mount is the same as in the preceding
Fig. 9 shows the data of " T atara " furna ce oper- period. As time goes the furn ace wa ll becomes thinner
ation which was performed on Nov. 18 to 20, 1935. and thinner, and after a bout 70 hr from the begin-
The operation is divided into four periods as shown ning, the operation ca n nol be continu ed. Then th e
in th e figure. furnace is destroyed a nd the blister stee l is pulled out
" Komori " (The Beginning Peri od of the Operation) into the air. After that, the furna ce bOltom is set,
In th is period as the furna ce tempera lure can not be a nd construction of th e furnace for the next opera tio n
raised fully, iron sands charged react with the furna ce is to be started (see Fig. 10).
wall a nd become slag, which are collected in the fur- As seen in T a ble 8, the total a mo unl of iron sa nd
~ ace bottom. The heat content of the slag helps the and charcoal charged is about 15 t respectively.
~urnace temperature rise for the next period of the The lota l products obtained are about 3 t, of whi ch
operation. 600 to 700 kg is pig iron a nd 2.3 to 2.4 t bli ster sleel ,
" Komorits ugi " (The Second P eriod of the Operation)
The furn ace temperature gradually goes up but is Table 8. Charge amounts of material s 111 eac h pe riod
n ot high eno ug h to reduce iron sands charged , and Iro n sand (t)
P er iod (% ) C harcoal (t) (% )
th erefore the greater part of lhem becomes slag. But
at the end of this period , pig iron begins to be pro- First pe ri od 1.60 II I. 26 8
duced little by li tt le in the furnace bottom. Second period 1. 50 10 I. 35 9
" Nobori " (The T hird Peri od of the Opera tion) Third period 3 .60 24 3.38 23
As the heat con tent of th e slag a nd pig iron increases Final p e ri od 8.20 55 8.92 60
enough to redu ce iron sands, c harging amount of iron
sands is increased from 90 to 11 0 kg each time. At
Total
1 14. 90 100 14.8 1 100

1
] s t per iod after 2 t05 hr from th e begi nning of opera tion 2nd period afte r 10 to 15 hi' f,"om the beginning o f o pe rati on
I

•
3 rd period afte r 20 to 30 hr fr om th e beg inning of opera ti on Final peri od after 68 hI' from the beginning of opera ti on

Pig iron . sponge i ron

Fig. 10. Surmise sketches of th e centre secti on of th e furna ce in eac h pe ri od of operat ion

Re por t
( 4 6 ) Transactions ISlJ, Vol. 8, 1968

Table 9. Compositions of iron sand c harged in each period (%)

composi ~ 1 T.Fe FeO Fe 2 0 a Si0 2 MnO AI 2 0 a CaO P S Ti0 2 V 20

---------------
Per iod
S

First period 56.55 21.52 66.05 7.90 0.57 5.50 0.38 0 .056 0.032 2 . 15 0.30
Seco nd period 56.96 20.33 60.50 8.18 0 .57 5.90 0.58 0 .077 0 . 027 2.46 0.27
Third period 58.13 21.20 60.46 7.90 0.48 3.86 0 .70 0.077 0.020 2.24 0.29
Fina l pe riod 59.89 22 .85 62.45 7. 45 0 .32 2.87 0.42 0.062 0 . 018 1. 67 0 .27

Table 10. Compositi ons of slag in each pe riod (%)

--------------- Com posi ti 0 n I

Period ~
T.Fe FeO Fe 2 0 a Si0 2 MnO AI 2 0 a CaO P S Ti0 2

Second period 49.52 58.85 5.40 22.52 I. 23 5.40 0. 18 0.02 5.10

Third period 34.40 39 .47 5.32 30.16 2.28 10.8 1 1.16 0.03 9 . 24
Final period 27.20 30.76 4.62 41.30 I. 16 9 . 21 1.49 0.03 9.51

and the latter contains 700 to 800 kg of crude steel the furnace is not so simple.
" Tama hagane ". As seen in Table 10, iron content in slag is consider-
A typi ca l composition s of iron sand charged in ably high , and moreover, iron sands (not fully reduced ).
each period are shown in Table 9. may sometimes come down to the surface of blister
Slag in each period is various in chemical composi- steel on account of low shaft. From these facts it is
tion, and it is not possible to present its typical or considered that slags or hot iron sands are reacted with
mean values. the blister steel directly and bessemerizing reaction
Therefore, in Table lOa tendency of slag composi- takes place.
tions is g iven for reference. But such reactions cause a decrease in the amo unt
of products, especially" Tamahagane " . According-
VIII. Consideration of R eaction in the Furnace ly the chief foreman " Murage " endeavours to keep
Carbon content in blister steel" Kera " is not uni- that operation condition as possible as norma l so that
form , but crude steel "Tamahagane" obtained by iron sands charged may be reduced to pig iron. In
cutting the blister steel into pieces has nearly constant this sense steelmaking by "Tatara" process much
carbon contents, through considerably wide range. owes to " Murage " 's own specia l arts.
Phosphorous amounts in "Tamahagane " are less
than those in pig iron. IX. "Zukuoshi " (Pig Iron-Making in " Tatara "
These phenomena can be explained in reference to Process )
the Fe-C equilibrium diagram; in the first period of The purpose of" Zukuoshi " is to produce more pig
operation, temperature of the furnace ca n not be iron than in "Keraoshi", direct steelmaking method.
raised enough so that iron sands react with the furnace But both methods are very similar each other in shape
wall and become slag, but in the next period, the fur- of the furnace as well as in the operation method . If
nace temperature sufficiently goes up , and a part of compared, the followings ca n be said. •
iron sands are reduced, absorb carbon, and change In " Zukuoshi " the distance among tuyeres is .
into pig iron. shorter, tuyeres are placed closer to each other,
In the third period, the furnace wall is eroded, and the furna ce is somewhat narrower in width and
tuyere distan ce becomes gradually wide, and blast can higher in height. As far as the furnace operation is
not reach the centre part of the furnace. Then semi- concerned , blasts a re sent more severely, and" Akame"
molten substance (spongelike) is made in the furnace iron sands or bea ch iron sands are used as raw materi-
bottom and grows as the operation proceeds. Mean- als instead of " M asa " iron sands in " Keraoshi ".
whi le in the upper part of the substance (blister steel),
many pieces of molten pig iron, wh ich are reduced by X . Conclusions
CO gas or C, exist among charcoals, and these pig iron In the study made above the followings were made
fa ll down toward the bottom. ' '''hen they come to the clear:
surface of the substance (blister steel ), they hang on it ( I) Blister steel " Kera" is not made in the mol-
and turn to the solid phase because the temperature of ten state but in the semi-melted (at most) state.
the blister steel is lower than that of molten pig iron, (2) There are few impurities in the final product
while the remaining liquid falls down to the bottom. " Tamahagane" but they are contained in it as slag.
Then , while the furnace temperature is co nstant, (3) In the beginning of the operation, slag is basic
carbon content of solid phase is kept constant as a as iron oxide acts as base, but it becomes acidic as the
result. operation proceeds.
By that reason carbon content in " Tamahagane " (4) Accordingly, slags in crude steel can be di-
is constant. But the above reaction occurs only in minished by forging it.
an idealy normal running, and the actual reaction in (5) Operation is not carried out unnecessarily at

Re port
 Transactions ISIJ, Vol. 8, 1968 ( 47 )

high temperature. At present" Murage "'s age is 70 on an average,

(6) Such useless impurities as CaO, Si0 2 , etc. are and the survivors are only two or three.
scarcely charged into the furnace. Now the author and his co-workers are forming a
(7) Products obtained by " Tatara " process have plan of rehabilitation of the process in order to trans-
several specialities as mentioned in section VI.-4. mit a tradition to the younger generation, and make
Most of the steels produced in a "Tatara" furnace a study to adapt the process to the present arts.
had been used as materials of Japanese swords For instance, if we are able to raise the temperature
" Nippon-to" till quite recently. But in accordance of sponge iron by about 200°C in a stationary state in
with a decrease in demand for the swords the produc- the reduction process, we will gain the product as
tion of "Tamahagane" fell away, and for these 20 excellent as " Tamahagane ": the melting yield of it
years the "Tatara" furnace has not been operated, will be higher than that of sponge iron, and it will
consequently "Tamahagane" has not been produced. make by forging superior tool steels which have the same
In short, the process has been forgotten in the world. qualities with" Nippon-to". But here is a problem
Here is another difficulty: the furnace can not be how to pull it out into the air. Probably there is no
operated without the "Murage " 's own art, then if other way except to destroy the furna ce just as our an-
they die out, the process will become extinct, too. cestors thought out this measure some 800 years ago .

Report