The invention comprises a method to determine _^ the temperature of a gas and a tool to realize the method.

The temperature of a gas is normally determined by use of a thermometer, i.e. by a measuring system or a body, for example a liquid thermometer (H -thermometer) , a resistance thermometer or by a thermocouple. A common feature for these instruments is that they contain a sub¬ stance that has a measurable property that is dependent on the temperature in a certain relation. By measurement this substance is brought into thermal equilibrium with the gas and the measured quantity, volume of the liquid, the elec¬ tric resistance or the electro otoric force indicates the thermal state of the gas. When the system of the ^' gas also comprises other bodies their surfaces will exchange radiant energy with the thermometer. The thermometer accepts there by an equilibrium temperature difference from the tempera¬ ture of the gas. The difference causes an error in measured temperature of the gas. The error can be considerable in many cases, even if precautions are made by shielding the thermometer. Actual cases for temperature measurement are the gas in the combustion chamber or the gas in the reac¬ tion zone of any electro metallurgical furnace.

A measurement that is being made during a certain time, is limited upwards to- the highest permissible ¹ temperature for the material in the thermometer and in any protecting tube, for example to an order of 1800°C for a thermocouple.

Known measuring technique for the ^' ' temperature of gases is under such conditions not satisfactory. The aim of the present invention is therefore to get a method and a tool that can give more reliable temperature measure¬

In kinetic gas theory it is known for an ideal gas that the temperature of the gas is proportional to the mean kinetic translation-energy of the molecules. Further is the molecular diffusivity of the gas (diffusion coeffi- cient) proportional to the mean translational velocity, from which it follows that the temperature, theoretically is proportional to the diffusivity in square. The diffusivi ty of the gas can therefore be taken as a measure of the temperature. The kinetic gas theory is further developed to comprise real gases. The diffusivity for a binary mixture of gases,e.g. is expressed by

^D ab' ^cons *- ^{Tl>5 (} | -*I _B ³ ' ^P '^m ^Λ where

T - temperature (K)

M ^a molecular weight for gas A, resp. gas B p = pressure of gas mixture (Pa) yf ^s diameter of collision cross section (A)

- ≥ ^» collision integral (-) The diffusivity is defined phenomenologically from the Fick's law

J _Λ ^β D _Δ . c <^A dz where

2 _A ^» diffusive flux of gascomp. A, (mole/m s) c = mole concentration of the gas components

(mole/m.-") γ. - mole fractipn of gas compenents A (mole/mole) z = length coordinate .( )

D.= diffusivity (m /s)

The diffusivity is determined experimentally by measuring a flow and the relevant concentration gradient of the equalson above.

The invention is based on the recognition of the relation between temperature and diffusivity. According to the invention it has been found that the temperature of a gas can be measured by preceeding as given by the patent claim 1. A device or a "thermometer" for realization of the procedure is given in patent claim 4. Further preferab-l-e features of the invention is given in the subclaims.

Under certain prerequisites to the design and the dimentions o-f the device, the measurements will give a base for determination of the diffusivity. By known rela¬ tion- between diffusivity and temperature, the temperature can be evaluated. The device is in the following called "diffusion thermometer".

The invention will in the following be described in more detail by a examplified design given in the draw¬ ing. . This shows in principle a sketch of the main compon¬ ents of a diffusion thermometer for determination of the thermometer in a hot gas, according to the invention.

The thermometer as examplified consists of two concentric tubes 1 and 2, an outer tube 1 for supply of an inert gas to a so called turn-chamber 3, and an inner tube - 2 for an exit of the gas out of the turn-chamber 3. The turn-chamber 3 has connection to the gas 5, through a de¬ vice for diffusion, a so called diffusion body 4 which may consist of a conveniant porous material. The diffusion body 4 may also be made as an open ended tube. The concen- tic tubes 1 and 2 are surrounded by a concentric jacket 6 for cooling, e.g. by water.

When inert gas is flowing into the turn-chamber 3 from the supply line 1 and from there through the diffu¬ sion body 4 into the hot gas 5, the diffusion body 4 will behave as a sink, toward which the gas will diffuse, due to drop in its concentration. By controlling the flow pr unit of time of inert gas that .flows through the diffu- sion body 4, a stationary situasion will establish where the amount of the gas 5 that diffuses through the diffu¬ sion body 4 toward the turn-chamber 3 equals the amount

The concentration of gas 5 in the turn-chamber 3 can be calibrated as a measure of the diffusivity of the gas 5 in hot condition. This concentration can be determined by use of a detector 7 located in the tube 2. The measurement by the thermometer can according to the invention be done in several manners. In one manner an excess of inert gas can be s.upplied, that is the flow of inert gas, e.g. nitrogen, pr unit time, to the turn- chamber 3 is kept larger that the flow sucked back through the tube 2 and the detector 7. Inert gas will thereby flow out through the diffusion body 4 and countered by diffusion of gas 5, e.g. CO. The efflux of gas is determined as the difference between the flows to and from the turn-chamber. It can be varied while keeping constant the gas sucked out of the chamber. Thereby it is possible to determine the efflux that by a. differential reduction gives a differantial, yet measurable concentration of the gas in the turn- chamber. The amount of the efflux can be expressed m thematically in dependency of the diffusivity of the. gas/inert gas, and thereby of the temperature of the gas.

By another operation manner the amount of gas is supplied to the turn-chamber -equal to the amount of gas sucked from the turn-chamber. Under equilibrium condi¬ tions inert gas will diffuse out through the diffusion body and an approximately equi olar amount of gas will diffuse into the turn-chamber. Hereby is two measuring principles possible: a) stationary diffusion through the diffusion body, with quantitative determination of the concentration of gas in the sucked inert gas from turn-chamber, and b) transient diffusion through the diffusion body. It is thereby assumed that the diffusion body is washed sufficienτ- ly by inert gas before any diffusion process is recognized. The time lasting from a given moment until a measureable gas concentration is recognized by the detector, is thereby a characteristic measure for the diffusion and it can serve as a measure for the diffusivity and further for the temp-

The diffusion body must be designed proportion¬ ately that the concentration of the gas 5 in the turn- chamber 3 is sensitive for the diffusivity of the gas 5 in its state. This prescription puts restrictions upon the relation between the outer diffusion, i.e. the diffusion in the hot gas toward the diffusion body, and the internal diffusion of cooled gas through the body. The reaction is expressed by a dimensionless gτo*up, the Biot number Bi- ^h c . L

where h - mass transfer coefficient from hot gas to¬ ward the diffusion body

D ^» diffusivity of the gas through the diffusion, body

L - characteristic length of the diffusion body.

The potensial diffusion of the gas toward the turn-chamber puts further restriction upon the flow of inert gas through the diffusion body. The restrictions are characterized by the dimension!ess group, Peclets nuiber that expresses the relation between comvection and diffusion in the body, namely

Pe= u . L D where u ^a mean velocity of inert gas through the diffu¬ sion body. These numbers constitute a base for the design of the thermometer in stationary operation..

The determination of the mass transfer coeffi¬ cient _c is for the present uncertain. The coefficient represents mass transfer by diffusion as well as by super¬ imposed convection. It is reasonable to assume that the convection is laminar, since the gas is assumed to have high temperature and thus to have high viocosity. The mass transfer will therefore predominantly be governed by the DRE

If a non-stationary diffusion is assumed then the time can be used as the independent variable. The time • it takes for the gas to diffuse into the turn-chamber can be expressed by the dimens ionless group, the Fourier number

Fo ^■ D - t

L ² where t » characteristic time.

Biot number and Fourier number constitute the base for the design of the thermometer for transient condi¬ tion.

The diffusion thermometer reacts, in principle,, and according to the invention upon the highest absolute temperature of the gas at the location of measurement. The temperature that is determined is not influenced by radia- ton between surfaces in the gas volume. The measured quantity is simple, it is not influenced by electric fields. The thermometer can in principle be calibrated in a gas of room temperature with known composition, when the assump- tion can be made that the dependency ^' of the diffusivity upon temperature is known.

The diffusion thermometer can by preference he applied by measurement in a gas that consists mainly of one component of known concentration. In ferrosilicium electrode-furnaces CO can, for instance, apply for tempera¬ ture measurement, with nitrogen as the inert gas.