Note: Descriptions are shown in the official language in which they were submitted.
~'7~
~ his invention i~ concer~ed with a device for the
continuous measurement of -temperature in a metallurgical furnace,
with particular reference to the temperature at the ~urface of
the refractory which lines the walls o~ an electric arc furnace.
TlleDe wal]~ can be exposed, from tlme to tlme, both to ex~remely
rapid increase~ in temperature and to excessively hlgh temper-
atures. ~he former cau~es cracking and ~palling, while the
latter causes melting which results in erosion. ~oth are dele-
teri~us to the ~urnace lining and result ln the need for frequent
re lining.
It ha~ been found that both problems can be avoided
by controlling the power input to the electrodes, in other
words, by reducing the power input when the rate of increase
in temperature at the face of the refractory lining become~
excessive or when the te~perature becomes too high. To do this,
the temperature mu~t be scanned continuously, even durin~ the
~palli~g and melting.
The junction of a conventional thermocouple is, of
course, destroyed at the ~ame time as the surface of the lining
a~ ~oon as the attack on the walls begins. However, i~ ~uch a
thermocouple i~ adequately protected against th~ rigors to
which the walls are exposed, the response time is 90 lengthened
as to make the thermocouple u~eles~ ~or the purpose envisaged.
If the protection i9 in the form of a graphite block, the
re~ponse time i9 good, but the block mu~t be water cooled ~or
its own protection. Hence, such a devlce detects only change
in temperature; it does not ~easure the true temperature.
Radiation pyrometry can be used, but the many problems
arising from the conditions under which the measurements are made
lead to uncertain temperature value~. ~he problems include poor
knowledge of the emis~i~ity of the refractories, interference by
the hot furnace atmosphere; and difficulty in locating a
. ' ~ .
-- 1 --
~7~26~
si~hting hole.
The device most sui-ted to -the need is a self-healing
thermocouple, namely a device in which a bridge immediately
forms between the ends of the two thermocouple wires at the hot
face of the wall lining as soon as the original junction or
subsequently formed junc-tion is destroyed either by spalling
of the wall in which it is embedded, or by melting, or by any
inadvertent mechanical means such as being struck by the charge,
as, for example when the scrap piled around the sides of the
furnace slips downwards. This lmplies -that the brid~e must be formed
as soon as possible after the newly exposed cross-sectionof the
thermocouple is exposed to the high tempera-ture radiation from the arc.
British patent Nos. 1,230,633 to G.D. Spenceley and
1,370,465 -to P. Collins published ~ay 5, 1971 and October 16, 1974,
respectively, propose several embodimen-ts of essentially similar
thermocouples in which an insulating powder is packed between
the two limbs of the thermocouple. 'rhis packing becomes
conducting, forming a junction between the conductor ends, when
conductive material resul-ting from the furnace environment is
deposited on it, this deposit being a mixture of me-tals, metal
oxides and slag. However, because -this new junction originates
from ~he furnace ~unction, there coulcl be considerable delay
between the destruction of an existing junction and the formation
of a new function between -the newly exposecl ends of the conducting
limbs. In other words, there could be considerable periods of
time when the thermocouple is not producing a signal.
U.S. patent No. 3,845,706 of November 5, 1974 to
Strimple et al. descrlbcs a slightly different approach.
In this case, the two conducting limbs of the thermocouple
are insulated from each other by means of a sleeving of a
glassy material on at least one of the wires and are packed
in a powdered refractory material that is electrically
conductive at high -temperatures. In the description of
~,
~ '
these ther~ocouples, it is pointed out that although the refrac--
tory material~ used, namely ma-terials taken from the group chrome
ore and chro~e oxide, are initially non-conductive at temperatures
below about 2000F, after a slngle exposure to normal ~urnace
operating condition~, a conducting hridge i~ formed b~tw~en th~
conducting limbs, this bridge conducting at temperatures lower
than 2000F and even as low a~ a~bient temperature, as evidenced
by the fact that signals are received from the device ~he
reason for this is not ~ully understood. However, according to
the patentee; it i~ believed that when the de~ice is exposed to
relatively high temperatures, the gla99 insulating sleeve
surrounding at least one of the thermocouple wire~, when heated
in the presence of -the chrome oxide or chrome ore, reacts with
or i9 caused to ~orm a conducting path that i9 independent of
temperature.
It is appropriate here to examine more closely the
electromotlve ~orce (e.m~f.) generated in the wires of a thermo-
couple. An electric current will flow in a closed circuit of
two dissimilar,metallic materlals if their junctions are at
different temperature~ If the circuit ia opened at some point,
the e.m.f. caused by the temperature difference can be measured.
It ~hould be emphasized that this e m.f. i~ not a property of
the junctions but depends on the bulk propertie~ of the two
materials and the temperatures of the two junctions. In thermo-
couple~, one junction i5 held at a known temperature. The
voltage is then a function of the temperature o~ the other
junction. ~he effect does not depend on the nature of the
~unction. Hence it may be made in any convenient way ~uch as by
soldering, brazing, spot-welding, fusing or even by a length o~
another material forming a conducting bridge between the two
dis~imilar materials. ~he only restriction is that the ~nds of
the two aforementioned dis~imilar materials making up the junction
must be at the same temperature.
In ~he event that the two ends are not at the same
temperature at thei.r point~ of contact with a bridge between
them, inaccurate readings can be obtained. Examining this in
detail, in a normal thermocouple :
12 Tl 12 T2 (1)
where V i9 the net voltage developed,
(E12)~ is the eOm.f. developed at the hot junction
(~t te~perature Tl) between the di~imilar
materials 1 and 2, and
(E12)~ iB the e.m~f. developed at the cold junction
(at temperature T~) between the dissimilar
materials 1 and 2~
If th~ change in e.m.f. in the couple 19 linear with increaqing
temperature~
(E12)~1 = S12Tl (2)
and (E12)T2 = S12~2 (3)
0 where S12 is the thermoelectric power or relat~e Seebock
coefricient of the couple composed of the two di~simi-
lar materials 1 and 2 and is the voltage developed in
the couple for a unit increase in temperature.
Thus : V ~ S12~1 ~ S12T2
It can be re~tated here that in thermocouple applica-
tions, one junction is usually held at a known temperature~ The "
net voltage is then a function of the temperature of the other
junction. ~ost tables of thermocouple data are gi~en for the
couple with one junction at a reference temperature.
Now, i~ there i8 a bridging material between the dis-
similar materials 1 and 2 at the hot end and t~e temperature of
~L~3762~4
the junction between the material 1 and the bridging material is
not at the sa~e temperature as the temperature of the junction
between the ~aterial 2 and the bridging material, the net voltage
V' developed will be given by:
V (E13)Tl (~23)T~ (E12)T2 (5)
where ~1 is the temperature of the junction between material
1 and the bridging material 3,
~3 i9 the temperature o~ the junction between material
2 and the bridging material 3,
( E13 ) ~ i 8 the e.m.f. developed at temperature Tl
between the dissimilar material~ 1 and 3, and
(E23)~ is the e.m.f developed at temperature T3
between the dissirnilar material~ 2 and 3.
In the very likely event that, at lea~t over the temperature
range T~ to ~3, the voltage change is linear with temperature,
the above equation can be re-written
V' = S13~1 ~ S23T3 ~ (E12)~ (6)
where S13 is the relative Seebeck coefficient of the couple
Z composed of the dis~imilar materials 1 and 3, and
S23 is the relative Seebeck coefficient of the couple
oomposed of the di~similar ma~erial~ 2 and~3.
As discu~sed earlier, the e.m.f. produced i9 not a
property of the junction but depend3 on tne bulk properties.of
each of the two di~imilar materials. Thus :
S13 = S1 - S3 (7) ;
~ and 523 = S2 ~ S3 (8)
; where Sl, S2 a~d S3 are the absolute Seebeck coefficients of the
materials 1, 2 and 3,respectively. ~he sign of Sl~ S2 and S3
can be positiYe or negative, depending on the sign of the
majority current carrier i~ the material. Thus the Bign of the
_ 5 _
.. . . ., ,. ~ .. ~ , ,. ... . :
:
i~76Z~
relative Seebeck coefficients can al90 be po~itiYe or negative.
Thus : V' = S~ S3~1 S2T3 ~ S3~3 ~ (E12)~2 (9)
~ he junctions be~ween each of the two ends of the
thermocou~le limbs and the bridgin~ material should there~ore
be a~ n~arly as po~ible at the ~ame t~mperature i~ ~he mea~ur~d
voltage V' i~ to be used as a measure of the -temperature~ In
that case, iOe. if ~ 3~ equation (9~ a~d hence equation (6) i9
reduced to equation (1). This i9 especially important if S3, ~-
the ab~olute Seebeck coefficient o~ the bridging material 3, is
large compared with Sl and S2. In such an instance9 the error
could be as large as or even larger than the true reading.
In order to be capable of measuring the true and
absolute temperature at the surface of the refractor~ lining in
a metallurgical furnace, the thermocouple must be designed in
~uch a way that the bridge can form only at the hot face of the
lining 90 that ~ 3 and that there i9 no pos~ibility that
contact can be made between either of the conducting limbs and
a conduc~ing bridge at a significant distance from the hot face,
where Tl ~ ~ . This r~quirement becomes especially important
if the conducting bridge i9 compo~ed of metallic oxides (includ-
ing slag) becau~e the absolute Saebeck coe~ficient of ~uch
material~ i9 large compared with that of metallic conductors
~uch as those used for the two limbs. Whereas the Seebeck
coe~ficient o~ metal~ lies in the range of microvolt~ to ten~
of mlcrovolts per degree Centigrade, ~hat of oxides can amount
to a~ much as milll~olt~ per aegree ~entigrade.
In the thermocouples described in the a~ore~aid two
British patents, the two conductors are separated over thelr
whole length by an i~sulating powder ~uch as A1203 or MgO. ~he
bridging can occur onl~ a~ the end~ exposed to the ~urnace.
~hu~ the two junctions are inevitably at, or es~entially at
the ~ame temperature. Hence, even though the bridging material
.
. _ 6 - ~
. .
26igL
i~ probably predominantly oxide in nature, the~e thermocouple~
should be sati~factory from the viewpoint of validity of ~ny
measurements that are obtained. ~hese thermocouple~ are ~ot
completely satisfactory, however, because reading~ are only
intermittently available.
In all the de~cribed embodiments of the Strimple and
alO thermocouple, only one of the dis~imilar wires i8 enca3ed
in a gla~s insulating ~leeve. Bearing in mind that the two wires
are then packed in oxide powder which i3 ~tated to become con-
ducting at temperatures above 2000F ~nd that the temperatureof the hot face i9 0~ the order of 2,700~ to 2,900~F, there thus
is the po~sibillty of contact between the uninsulated wire and
the insulated wire via a bridge of conducting o~ide between one
junction at 2000F, a little way back from the hot face of the
lining, and the other at at 2700 to 2900F at the hot face.
If thi~ occurs, completely invalid temperature readings are
obtained from the viewpoint of ab301ute measurement of temperature.
AlthGugh U.S. patent N 3,8459706 contemplates the insulation
of both the dissimilar wires, inaccuracy can still arise through
2~ the use of fiberglas~ sleeving a~ in~ulation. Unle~s thi~
sleeving i9 of some ~pecial type of high temperature glasq that
doe~ not melt until clo~e to the temperature of the hot face of
~he lin~ng, whioh i9 not specified in the patent, lt is 11kely
that a bridge will be formed according to the description of
the patent, but significantly distant fr~m the hot face. ~hus
the temperature measured could be clo~e to the temperature at
which the glass sleeving melts rather than that of the hot
face, i.e. up ~o ~everal hundred degre~s too low. However, it
~hould be emphasized that such a thermocouple can still be
useful for detecting chan~es in temperature at the ~urface of
the lining.
.. . . ... . . ..... . ..
. . . . . . . .
, . . . . ~ , .
1~6Z6~
In summary then, each of -the two types of self-healing
thermocouples described above has an inherent drawback, the first
because it might not func-tion continuously and -the second because
it does not necessarily give readings tha-t represent the true
temperature of the ho-t face of the lining.
It is therefore an object of the present invention to
overcome the aforementioned drawbacks and to provide a device
for continuously measuring the absolute temperature at the surface
of a refractory lining in a metallurgical furnace, throughout
the life of the refractory lining.
It is a further object of the inven-tion to provide
such a device which relies upon the formation of a conducting
bridge between two thermocouple wires at the surface of the
refractory lining, to complete -the electric circuit required
to measure the -temperature.
In accordance with the invention, there is thus provided
a device fvr con-tinuously monitoring the internal surface
temperature of a refractory lining in a metallurgical furnace,
which device comprises an outer sheath, a pair of dissimilar,
metallic wires within the outer sheath and a powdered oxide mate-
rial closely packed within the sheath and surrounding both wires.
Each wire is separately insulated by a sleeving o~ a refractory
material having a melting point higher than the normal working
tcm~erature of the Eurnace at the surface oE the lining. The oxide
material used as a packing between the outer sheath and the wires
is electrically conductive at elevated temperatures and has also a
melting point higher than the temperature normally encountered at
-the surface of the lining. Both the insula-ting refractory material
and conductive oxide material, when exposed to the aforesaid normal
worklng temperature, react with each other to form a conducting
bridge between the dissimilar wires at the sur~ace of
the lining, which bridge is opera-tive to provide a
.' '
`~
~7~2~
~ignal related to the internal surface temperature of the lining,
~he invention requires insulation of both limb~ of the
ther~ocouple, which can be made of a~y of the well known platinum
/x percent rhodium ver~us platinum /y percent rhodium pair~
where, for example x = O or 6 and y ~ 10, 13 or 30. The insula-
tion 1~ high melting and preferably is of silica (quart~) 9 mullite
(alumina-silica sy~tem), high purity alumina, alumina of a lower
grade (usually containi~g ~ilica) / or magne~ium oxide. If it i9
in the form of solid tubing, it ~hould advantageously have -thin
walls and ha~e the smallest po~ible bore consi~tent with the
wire gauge used such that the in~ulation i9 tightly fit over each
wire and the formation of a conducting bridge with the oxide
packing i 9 ensured between both wires at elevated temperatures.
If it is in the form of a woven sleeving, the minimum size ~or
the ~ire diameter should preferably be chosen.
~he invention al90 requires that a powdered oxide
material which i9 conductlng at high temperatures, that i9, at
about l~looocJ be packed around the two insulated limbs of the
thermocouple inside the outer sheath and that this material
have a melting point somewhat higher,~or in~tance 50 to 300C,
than the normal operati~g temperature at the ~urface of the
lining. A prePerred oxide material is of the group nickel
~errite, NiOox Fe203, where x lies between 1 and 1.2. This
~aterial has also -the advantage that its conducting nature
doe~ not deteriorate due to ¢ontaminatio~ by the oxide~ of iron,
ca].cium, mangane~e, aluminum, ~ilicon and any other ~etals
that might occur in the slag or brl~kin the furnace; in addition,
- its conductivity i5 not adversely affected by changes in the
atmosphere of the furnace which might be either oxidizing or
reducin~. The material can b~ prepared in a nu~ber of differ~nt
fashion~/ but it can advantageously be sintered in air at about
900C for at least 12 hours to increa~e its den~ity and to
: .
_ g _
.
~ ~ ~ 6Z 6 ~
reduce ~hrinkage through ~inter.ing during use in the ther~ocouple.
After sintering, the particle size can advantageou~ly be reduced
by any normal ~ethod ~o that mo~t of the particles (about 90%)
fall in the range - 50 to ~ 230 me~h (Tyler ~ieve). ~his gives
fairly good pouring and packing characteri~tics. The balance
of the particles can be both bigger and smaller~ One way of
preparing the ferrite powder i9 by intimately mixing stoichio- -
metric quantities of nickel oxide, ~iO, and ferric oxide, Fe203,
both being reagent grade chemical~ in fine powder form, and
firing the resulting mixture according to the above sintering
~ethod. In this casej the sintering achie~es the nece~sary
~olid-solia chemical reaction a~ well a3 the desired den~ifica-
tion~
Because of the high melting ~0mperature o~ the oxide
material, the particles packed axound the insulated wires do
not melt on exposure to the normal working temperature at the
surface of the lining. It has been found, however, t~at the
particles sinter to each other In addition, they react with
the insulatlon around the dissimilar wires~ sintering to the
insulating material and undergoing some localiz~d melting. ~his
melting occurs because, almost without exception, when two
different oxide material~ are held in cont~ct with each other
at an elevated temperature, some product of the reaction has
a lower melting point than either of the original material~
In the present case, the result is a vi~cid materi~l, akin to a
slag, that may be con~idered as ju~t liquid or just ~olid and
which adheres to the conducting wires to form a conducting bridge
between themc Becau~e of the characteristics of the materials
e~ployed for both the in~ulating ~leeving and the conducting
oxide, thi~ reaction require~ a temperature ~ub~tantially
equi~alent to that normally encountered at the hot face. ~hus,
there is no danger that it will occur at an appreciable di~tance
-- 10 -- .
~7~
from the hot face, so resulting in a conducting bridge distant
from the temperature to be measured, which would give spurious
temperature readings.
As mentioned previously, the conducting wires can be
any o~ the ~air~ of alloys of platinum and rhodium that are
normally used for high temperature measurements. Their diameter
is not important. However, diameters of OoO10 inch and greater
generally make con~truction easier. They can be physically
joined by twisting or by welding in the usual fashion if readings
are required star~ing at at room temperature during the initial
heating up of the furnace, before the conducting bridge of oxides
is formed. They can be positioned with respect to each other
in a number of ways, each within its own insulating sleeving :
side by side and as close as possible, with the powdered oxide
material packed around them, side by side with the oxide powder
packed between them as well as around them, if they are insulated
by means of flexible, woven sleevings, they can be twisted together
over the whole length that they are insulated,with the oxide
powder packed around them. In addition, where they are positioned
side by side in close proximity to one another, use can be made
of a single double bore insulating tube instead of two single
bore sleeves.
It is necessary to hold the complete assembly in an
outer sheath. This sheath can be open at both ends, or it can be
closed at the end that will be placed at the hot face o~ the
lining. The former alternative is possible because, once the
oxide powder is firmly packed, it does not pour from the sheath
unless it is exposed to vigorous agitation. The sheath can ~e of
.i -. ...
any of the normal high temperature materials, or example, alumina,
alundum, mullite, zirconia or magnesia. Alternatively, it can ~e
of a plastic material that holds the assembly initially, so that
it can be handled and inserted in a brick of the furnace wall by
~7~Zf~
whatever means is selected. This plastic ma-terial burns away
during use. Another alternative is that the thermocouple can be
constructed inside a brick, using the brick itself as outer
sheath. This can be done by drilling a hole through a brick
and then conducting the thermocouple in the hole in the manner
to be described below. It can also be done by molding a hole in a
brick of castable refractory and building the thermocouple
directly therein.
The thermocouple is constructed in the sheath, whatever
its nature, by placing the sheath in a vertical posi-tion with its
closed end downwards and so holding it. If an open ended sheath
is used, it should be placed with one end on a flak surface.
Next, the two insulated wires, in any of the alternative arrange-
I ments described above,are inserted so that they approximatel~
reach the bottom. The powdered oxide material is then poured,
a little at a time, into the sheathing, tapping and tamping after
each addition to ensure maximum packing, until the sheath is fili-
ed. If desired, this end of the sheath can be blocked by any
desired means such as by a plug of glass or quartz wool, by a
plug of castable ceramic or by a terminal block.
The completed thermocouple can be inskalled in the
furnace wall in any convenient way. I~ it has been constructed
in a brick, the brick is built into the furnace wall during a
relining operation. If it is in a sheath, one convenient method
of installation is to build it into a brick of castable refractory
and then to build the whole assembly into the furnace wall during
relining. After installation, the wires themselves or other
wires connected to them are led to the outside of the furnace
shell and to a measuring or recording instrument.
In the appended drawings which illustrake preferred
embodiments of the invention -
Figure 1 represents a longitudinal view in cross-section
- 12 -
.
1~76Z~
of a self-healing thermocouple according to the .invention,
Figure 2 is the same configuration as Fig. 1 with the
exception that the normally closed end of the device, to be
positioned at the hot face of the refractory lining, is left
` open, and
Figure 3 is the same configuration as Fig~ 2 except
that the device is shown constructed directly in a brick as outer
sheath.
The device represented in Fig. 1 includes an outer
10 sheath 10, a powdered oxide material 12 closely packed within
th~ outer sheath 10 and a pair of thermocouple wires 14 and 16
located inside the sheath 10 and separated therefrom by the oxide
packing 12. Wires 14 and 16 are in spaced relationship, with
; oxide powder 12 therebetween, and are each encased in insulating
sleevings 18 and 20, respectively~ ~he outer sheath 10 is shown
closed at its end 22 to be placed at the hot face of the refractory
lining in a metallurgical furnace. At this end, the wires 14
and 16 terminate in a junction 24 and are so physically joined
together in order to allow initial temperature detection starting
20 from room temperature. The free ends 14' and 161 oE the thermo-
couple wires are adapted to be connected with a measuring
apparatus to record the internal surface temperature of the
lining in the furnace.
The junctlon 24 of the thermocouple ensures that the
initial temperatures of the lining can be monitered from ambient
temperature all the way up to operating furnace temperature.
Once the furnace has reached normal working temperature, the
original junction 24 is destroyed as a result of erosion of the
refractory lining. However, at this temperature, which is in
the order of 1,400 - 1,600 C at the surface of the lining, the
oxide powder 12 which has become electrically conductive react
with the insulating sleevings 18 and 20 to form at all times a
- 13 --
... , .,, .. . - . - - . , - . - , . - .. .. .. . - .. . - ,. . .. . ..
64
new junction between the thenmocouple wires 14 and 16 at the hot
face of the lining, thus ensuring a continuous operation of the
device and hence a continuous recording of the internal surface
temperature of the refractory lining.
In Fig. 2, the dev.ice i~ shown with the end 2~ of the
outer sheath 10 left open. As mentioned previously, this
embodiment is possible since, once the oxide powder 12 is firmly
packed, it does not normally pour from the sheath 10.
The thenmocouple of Fig. 3 is shown constructed inside
a brick, using the brick itself as outer sheath 10. The whole
assembly can be built into the furnace wall during a relining
operation.
