Alexander GRECHKO, Dr. Sc. (Technology), leading researcher of the State Institute of Non-Ferrous Metallurgy (GINTSVETMET)

Heat exchange is one of the most complex physical phenomena which is hardly amenable to formalization and, therefore, is studied mostly by experimental methods. First, we collect experimental data on various instances of heat exchange, then we generalize these data by the similarity theory and deduce criteria-related dependences in the form of functions and mathematical equations. A large contribution to this field has been made by Academicians Mikhail Mikheyev and Samson Kutateladze.

Heat exchange through contact (contact heat exchange, or transfer) is one of the hardest cases for pyrometallurgy. This phenomenon relates to thermal processes taking place between the contacting surfaces of two machine parts. Such assemblies are quite common in power engineering (generators, engines), machine manufacturing (detachable and permanent joints, clad, or composite metals), in transportation (slide faces and surfaces, fuel and lubricant layers on the surface) and the like. A theory has now been developed for obtaining quite reliable criteria of technical decision making in various industries.

The contact heat-exchange theory involves research findings on the heat conduction (heat resistance) of contact places (joints) of two surfaces having a definite degree of roughness. The heat-transfer (resistance) value comprises the conductivity indicator at points of the actual contact of surfaces (prominent microroughnesses) and the same indicator for the medium (liquid or gas) filling the contact gap. A correlation of these two characteristics is the subject of the theory of the heat-exchange of contacting surfaces. Here such things are considered as the quality of surface working (of particular interest are cases when the precision factor is in the 1 to 10 u, range of roughness height),

page 69

compressive force (in values of thousands and tens of thousands of kilopascals), the thermophysical and mechanical characteristics of contacting materials, among other factors.

Lately attempts have been made to use the contact heat-exchange theory in bubble pyrometallurgy ^* (water-jacketed smelters, fuming furnaces, Vanyukov smelters ^** ). This is very important, for the normal performance of assemblies that have parts without direct cooling can be ensured through contact heat transfer from the intensely cooled elements of a furnace (water jackets, lances, tuyeres and related blast units). Technical solutions suggested by the new theory are attractive for their simplicity: it becomes possible to do without forced cooling of furnace elements which are in direct contact with the melt; this, in turn, guarantees safety should such elements go out of commission-otherwise, if some of the cooling water gets into the melt, a powerful explosion may occur.

Two extreme cases can be specified for contact heat-exchange applications in pyrometallurgy: absolute contact in practical terms, say in plates of composite metal (copper/steel) which are manufactured through welding and by the explosion method, and rather loose joints with a wide clearance in between surfaces. The former case poses no problems, for one can well use the conventional materials; as to the latter, special analysis is needed.

Here we are dealing with large contacting parts. Their surfaces are rough- worked or machined (by rolling, turning, milling, cutting, planing), hence the roughnesses - all the various traces left by cutting tools, buckles, ripple markings and deformations; all that results in wide clearances at contact points, 1-2 mm and more. And yet another important consideration: these assemblies are in a specific medium, the melt, which can streak into the clearances and aggravate the heat-transfer situation for the contacting parts. And the high temperature of the ambient medium (1,200-1,400C) hampers heat-transfer control too.

The problem of melt streak into clearances has been but little studied so far. What we know is this: the lower the stickiness of the melt, and the higher its temperature (and the wider the gap between two surfaces), the more likelihood of the melt filling the corresponding "cracks". Japanese researchers have conducted many experiments on simulated and actual metallurgical melts (for this purpose they drilled holes of different diameter in smelter partitions) to find out that streaking sets in at 2.5-3.0 mm clearances.

The melt medium is determined by the nature and rate of melt motion (flow- about), formation of skull (a lining of hardened melt on a smelter's cooled wall) and the melt characteristics (resistance, stability and the possibility of cracking-depending on the temperature and rates of heat flows); it also depends on the deformation of parts that have long been in operation, and so forth. Here we must point to a complex pattern of heat conditions in bubble metallurgy smelters. The blowing of the melt with gas jets sends up the flow- about rates of circulating currents which can damage respective parts and even put the whole unit out of order. Considering the above features of heat exchange in pyrometallurgy, we have come to the conclusion that conventional computation techniques are hardly suitable in this case, and additional analysis of the processes is needed.

The aim of such computation is to determine the temperature of contacting surfaces so as to ensure their reliable operation. Above all we must know exactly what the

^* See: A. Grechko, "Bubble Pyrometallurgy: Step into the Future", Science in Russia, No. 5, 1999.- Ed.

^** See: A. Grechko, "Copper Smelter of the Future", Science in Russia, No. 3, 1998.- Ed.

page 70

thickness and heat conductivity of contacting parts are, and what analogous characteristics are of the medium in the clearances (it may be gas or liquid, perhaps even the melt). The heat resistance of this medium, as we have said, is but little studied, it is quite uncertain and unpredictable - a terra incognita indeed.

Although the medium between two contacting parts is of crucial importance for the normal work of this or that assembly, its heat conduction values may differ by a factor of 10 ⁴ . Say, for air the value is equal to 0.04 W/m x deg., while for melted slag and matte the values are 1.3 and 15, respectively. Other factors are also to be taken into account. The temperature of the melt, its circulation and heat fluxes flowing to the partition wall, clearance width and heat conductivity-all that affects the temperature between the contacting parts of a unit and interferes with its performance; as a consequence the unit may even go out of commission.

That is why our GINTSVETMET Institute is so much involved with this research area (with work supervised by the author of the present article). We are taking a close look into heat transfer in contact assemblies and searching for ways and means of their optimization, design including. As we found out, the geometry of contacting parts and their correlation are of major significance for pyrometallurgy Two factors, К ₁ and K ₂ deduced empirically, have proved to be all-important. The first factor (coefficient), k ₁ , describes the thickness correlation of two parts-one that receives heat (the "hot" one facing the melt) and the other that is cooled (the "cold" one outside). The other factor (coefficient), K ₂ , describes the correlation between the heat-receiving (contacting the cooled part) surface. But this point needs some explanation.

The thing is that some pyrometallurgy units are supplied with inner partitions made of two contacting plates; on one side the melt reaches as high as their top, while on the other the level of the melt is much lower. In this case the areas of the heat-receiver and the heat-giver surfaces will always be equal, and K ₂ =1. But the safeguarding structures of furnaces (outer walls) are now and then reinforced with one or several bars fixed on the inner side (where the melt is) along the entire perimeter. And so the bars contact the belt both sideways and above; the K ₂ value will always exceed 1, i.e. the heat-receiving surface has a higher value than the heat-giving one. As shown by experiments, the higher K ₁ and K ₂ values, the higher the temperature difference will be between the contacting surfaces, with ensuing negative consequences.

Computations of thermal resistance in a multilayer wall of the "part-clearance- part" type, with due account of the above, enabled certain conclusions. In assemblies where copper plates contact each other, with clearances above 0.5 mm and filled with gas (air) mixture at a heat flux of 70 kW/m ² and more (that's what actually happens in practice), a part will always fuse unless it is cooled. There are two ways of raising its resistance: either by narrowing the clearance (which should be well below 0.5 mm) or by obtaining conditions for having it filled with melted slag or matte. Besides, copper units have proved to be more resistant than those of steel.

These and many other computations were confirmed in numerous experiments on a Vanyukov smelter for real melts and in a variety of modes.

Proceeding from our computations as well as numerous empirical data, we have designed a new type of lance (tuyere) for a Vanyukov smelter. It comprises a copper tip (head) cum steel band (shroud) which increases considerably the tuyere's resistance and service life. This invention is protected by a patent of the Russian Federation.

We have made but the first small step on the path of developing the contact heat-exchange theory for pyrometallurgy. Let me stress it again: this is a formidable job of work calling for copious experimental evidence on the performance of various units and assemblies in many melts. Empirical data can be collected either in the course of regular operation (including modifications during scheduled repairs) or by aggressive and purposive interference in the process (discontinuing the work of a unit for dismounting and mounting respective assemblies). Although the latter procedure is the most radical and effective, it is hardly ever realized in practice. Unfortunately.

So research scientists have to make do with small "steps" and inch forward in collecting computation and empirical data. First they process previous experimental data based on prior computations, and then get down to designing new assemblies for smelters. Afterwards these assemblies are tested at operating facilities and, depending on the results of such tests, are either adopted or rejected; and computation techniques are modified in the process. The introduction of geometry factors (coefficients) K ₁ and K ₂ is proof positive of that.

We hope that the 21st century will see at long last a rigorous theory of heat exchange in pyrometallurgy.

Permanent link to this publication: