NO129320B - - Google Patents

Description

Apparatus for heating a fluid, especially steam

of halides, excluding the fluorides, of titanium,

silicon, aluminum and zirconium.

The invention relates to an apparatus for heating a fluid, in particular vapors of halides, with the exception of the fluorides, of titanium, silicon, aluminum and zirconium.

It is known to heat a fluid by passing it through a heated metal tube, and the invention relates to the problem of reducing the amount of metal required by weight. The problem may arise as a result of requirements that the device must be light, or as a result of economic considerations, especially when it is necessary to use an expensive metal.

In the U.S. patent no. 2^96327 describes an electric gas heating device which comprises a rudder with electrically conductive strips arranged inside the rudder, but electrically isolated from it. An electric current is passed through the conductive strips, and the gas is heated by passing this through the rudder in contact with the electrically conductive strips. The cross-sectional area of the electrically conductive strips increases in the direction of gas flow through the tube in order to keep the temperature of the conductive strips constant during use of the device,

In the U.S. patent no. 32^7888 describes an electrically heated evaporation apparatus which comprises a rudder arranged as a resistance element in an electrical circuit, the solution to be dehydrated or concentrated being led along the rudder. The thickness of the rudder wall varies in order to vary the heat supplied to the seal of the rudder and thereby avoid overheating of the forge. It is therefore proposed for this purpose to increase the thickness of the downstream end of the tube wall by increasing the amount of metal there.

With the help of the invention, an apparatus for heating a fluid is provided, comprising at least one tube through which the fluid can be led and which is provided on the inside with metal fins or other elements which are in contact with the fluid, as the fins or the other elements cross-sectional area increases in the direction of flow of the fluid, and the device is characterized by the fact that the rudder is also made of metal and that the fins or the other elements are in electrical contact with the rudder wall, and the total cross-sectional area of the rudder wall and the fins or the other elements increases along the rudder in the direction of flow of the fluid so that when both the rudder and the fins or the other elements are heated by the application of an electric potential difference between two points at a distance from each other in the longitudinal direction of the rudder by thus producing an electric current both through the rudder wall and through the fins or the other elements, said increase in the overall cross-sectional area will prevent the temperature of the rudder wall inner surface which is in contact with the fluid, in increasing in the direction of flow of the fluid, at least over a certain length of the rudder, while. the surface area of the fins or the other elements is such that the temperature of the surfaces of the fins or the other elements which. is in contact with the fluid, at no point does it exceed its maximum temperature. inner surface of the rudder between the mentioned two points..

The invention makes it possible to reduce the amount of metal compared to the case where the rudders are heated by radiation from the walls. a furnace in'which the rudders are placed-. A 'minimization'

of the metal weight requires that all metal has as high a temperature as possible. As corrosion, requirements for mechanical strength and/or other considerations dictate an upper limit for the metal's maximum temperature, the requirement that the metal's temperature be as high as possible means that the metal's temperature must be as even as possible.

The invention makes it possible to achieve a more uniform temperature distribution in the metal in two ways. Firstly, the use of electrical resistance heating in the metal even at the end of the tube where the fluid has the lowest temperature makes it possible to supply heat quantities that could otherwise only be achieved by radiation, but by using furnace wall temperatures so high that they would not be able to used in practice. Secondly, the increase in the cross-sectional area of the electric current's flow path towards the end of the tube, where the fluid has the highest temperature, leads to a corresponding reduction in the heat generation per unit length of the pipe in this direction, which, by tending to compensate for the reduction in the cooling effect of the fluid as its temperature increases, enables the application of current densities in the upstream end of the pipe which would otherwise lead to overheating of the downstream end of the pipe.

If one chooses a single rudder as the basis of comparison, i.e. a rudder of uniform diameter and wall thickness, which is not equipped with fins or other elements, and which is heated by passing an electric current through it, instead of a single rudder which is heated by radiation , the advantages obtained by means of the invention become even greater. The reason for this is that when the rudder is heated by wood

radiation, the pipe receives a greater heat application in the relatively cold upstream end than in the relatively warm downstream end (which tends to compensate to some extent for the fluid's greater shearing effect in the upstream end), while the heat application becomes uniform over the entire length of the pipe when the pipe is heated by conducting a electric current through it. For a given supply temperature of the fluid and a given maximum temperature of the metal, the length of a single, conventional tube required to raise the temperature of the fluid to a given value (provided that the other variables, mainly the mass flow rate of the fluid and the diameter of the tube are the same in each case), be smaller when the rudder is heated by radiation than when the rudder is heated by passing an electric current through it. It is surprising that by equipping the rudder with internal fins or other elements and heating the rudder and fins or said elements by passing an electric current through them according to the invention, not only can the rudder be made shorter, even

compared to a conventional, single rudder heated by radiation, but may also make it shorter than would be necessary to compensate for the amount of metal used in the fins or other elements.

It will be understood that with the device described in US Patent No. 32'+7888 it is not possible to save metal by reducing the length of the rudder as it is necessary for the rudder to have a certain length in order for the desired degree of evaporation to be achieved. Metal is added instead to avoid overheating of the forge.

It will also be understood that because the fins in the present device are in electrical contact with the wall of the rudder or rudders (which itself consists of metal), there is no need to use electrically insulating elements, as in US patent document no. 2596527, and that this is an important advantage when a corrosive liquid or gas is to be heated.

It should also be noted that while the design of the fin in the device according to US patent document no. 2596327 is chosen so that the temperature of the fin itself is essentially uniform, the fins in the present device have been given such a shape that the temperature of the rudder wall is essentially uniform . As explained below, the temperature of the fins in the existing apparatus does not even need to be approximately uniform.

The advantages obtained by means of the present apparatus,

are particularly pronounced when the fluid is the vapor of a halide (other than the fluoride) of one of the elements titanium, silicon, aluminum and zirconium, or a mixture of more than one such halide, and at least the surface of the rudders and of the fins or other elements which comes into contact with the halide vapor, is of platinum or of an alloy of platinum with rhodium, ruthenium or iridium.

The heating of such halide vapors to relatively high temperatures of the order of 700° - 1100°C, e.g. for the purpose of bringing them to a temperature at which they will undergo oxidation reactions with oxidizing gases, presents problems due to the corrosive properties of the halides at such temperatures. The use of platinum or a suitable platinum alloy as construction material for the parts of the apparatus which are exposed to the hot halide vapor makes it possible to reduce the corrosion to an acceptable level, and the saving with respect to platinum or platinum alloys made possible by invention, is very significant from an economic point of view due to the high price of these materials.

The fluid must of course not have such a high electrical conductivity that it has any significant effect on the strength of the electrical current that passes through the rudder wall and through the fins or other elements.

The device can comprise several metal rudders, each of which is supplied

on the inside with fins or other elements, and which are connected in such a way, as explained below, that the fluid flows parallel through them.

The rudder or rudders can advantageously be arranged in a heat-insulated casing, such as e.g. in a container lined with a non-metallic, refractory material, and, in order to reduce heat loss from the rudder or rudders, the enclosure walls can also be heated in a different way than by radiation from the rudder or rudders. Preferably, however, the inner surface of the casing is kept at a temperature lower than the temperature of the outer surface of the rudder or rudders at the downstream point of the two mentioned points, because the resulting net radiation loss from the rudders further increases the maximum current that can flow in the rudder wall without overheating, and this makes it possible to further reduce the amount of metal. The optimal degree of cladding depends on an assessment of the savings in terms of metal weight on the one hand against the increased consumption of electricity on the other.

The importance of maintaining the inner surface of the casing at a temperature below the temperature of the outer surface of the rudder or rudders at the downstream point of the two points mentioned will be more easily understood when considering an apparatus comprising a rudder which is internally provided with fins shaped (as described below ) of lami-'' material of the same thickness as the rudder wall, the thickness of the rudder wall being small compared to the diameter of the rudder. If one then assumes that the rudder and fins are constructed in their entirety from a single material, and disregards the effect of free edges (if any) of the fins and the fact that the exposed area of the inner surface of the rudder is reduced by the fins there where these are attached to the rudder, the surface area of the fins that come into contact with the fluid becomes twice as large per mass unit of the fins as the surface area of the wall exposed to the fluid, per mass unit of the rudder wall. On the other hand, the heat generation per unit mass of metal is the same for the fins as for the rudder wall. Thus, if the inner surface of the casing is kept at the same temperature as the outer surface of the rudder at some point along the rudder length, so that no net heat loss from the rudder occurs at this point by radiation to the zone walls, the rudder wall must emit heat to the fluid at this point in a quantity per surface unit exposed surface that is twice as large as the corresponding amount for the fins at this point. This means that at the point in question the temperature of the rudder wall must be significantly higher than the temperature of the fins, so that the fins have a temperature that is lower than the optimum and therefore do not work at the maximum heat output per mass unit metal. If the point in question is the downstream point of the two previously mentioned points, where the fins are wide (so that the mass of the fins per unit length of the rudder can be greater than the mass of the tube wall per unit length of the rudder), this represents a significant welding with metal. Although the temperature of the fins can be easily increased simply by increasing their thickness without changing their total cross-sectional area, the heat output of the fins per mass unit of metal, and no saving of metal is therefore achieved in this way. On the other hand, it is possible to save metal by arranging for the rudder to emit radiant heat to the walls of the enclosing zone, because fins of smaller cross-sectional area will then be sufficient to prevent the rudder wall from reaching temperatures that exceed the desired maximum temperature.

The flow rate of the fluid through the tube or through each tube, the amount of electrical potential applied between the two mentioned points and the configuration of the fins or other elements can be such that above at least 50$ (preferably at least 80$)

of the surface area of the rudder or rudders between the two points, the temperature of the inner surface of the rudder or rudders exposed to the fluid is less than 200°C (advantageously less than 100°C and preferably less than 50°C) from the highest temperature of the inner surface exposed to the fluid. Likewise, the flow rate of the fluid through the pipe, the strength of the electric potential applied between the two mentioned points, and the configuration of the fins or other elements are advantageously such that the temperature of the inner surface of the pipe or of the pipes exposed to the fluid, everywhere between the said two points, apart from the immediate vicinity of the point of origin, is within a range of 300°C (preferably within a range of 150°C). When

the cross-sectional area of the wall of the rudder or of each rudder is constant between the two mentioned points, the flow rate of the fluid through the rudder, the magnitude of the electric potential applied between the two mentioned points and the configuration of the fins or other elements can preferably be such that the temperature distribution over the the inner surface of the rudder or of each rudder exposed to the fluid is substantially level between the two points mentioned, except in the immediate vicinity of the updraft point.

The configuration of the fins or said other elements can vary widely. This is largely due to the fact that the rate at which heat is generated in the wall of a given short tube length is the same for all configurations of the fins or other elements that give the same total cross-sectional area of the electric current's flow path through the considered tube piece. If one thus ignores any effects that variation in the configuration of the fins or other elements may have on the heat transfer between the fins or other elements and the tube wall or on the flow of fluid through the rudder, any change in the configuration of the fins or other elements that does not change the total cross-sectional area, nor change the temperature of the inner surface of the rudder wall. In contrast, any change to the configuration of the fins or other elements that results in a bend in the cross-sectional area per length unit of the rudder, result in a reduction in the temperature of the fins themselves or of the other elements themselves. In order to thus ensure that the fins or the like do not themselves reach too high a temperature, it is sufficient to ensure that they have a sufficiently large surface area in relation to their cross-sectional area.

The fins and the like can advantageously be made of laminar material. The thickness of the laminar material can conveniently be substantially the same as the wall thickness of the rudder or rudders. When the Finns and others is formed from laminar material, the surface area of the fins etc. which is exposed to the fluid per length unit of the rudder, directly proportional to the cross-sectional area of the electric current flow path provided by the fins, etc. If, under these circumstances, the temperature of the fins, etc. at the downstream point of the two mentioned points is mainly equal to the temperature of the rudder wall, it will be found that the temperature of the fins etc. decreases along the rudder in the downstream direction if they are dimensioned with a view to keeping the temperature of the rudder wall essentially constant in the longitudinal direction of the rudder. Although this variation in the temperature of the fins etc. in the longitudinal direction of the rudder can be avoided or reduced by increasing the thickness of the fins etc. and at the same time their surface area decreases, so that the surface area of the electric current flow path remains unchanged, the improved evenness in the temperature of the fins etc. not result in any saving of metal.

Finns and others can advantageously be an integral part of the rudder or rudders, but this can lead to construction difficulties. is called, and the Finns, etc. can therefore be produced separately and attached to the rudder or rudders, e.g. when welding,,

Finns and others can be straight fins sora in the longitudinal direction extends parallel to the axis of the rudder and in the transverse direction extends inwards (e.g. radially inwards) from the rudder wall, as the width of the fins increases through the rudder in the direction of the flow of the fluid. Instead, the rudder or each of the rudders can contain a rudder-shaped element which is arranged coaxially in the rudder, and which is attached to the rudder wall by means of flat fins which in the longitudinal direction extend parallel to the axis of the rudder, and in the transverse direction are arranged radially, as the number finds ocher step by step through the rudder in the fluid's direction of flow. Instead of extending through the rudder parallel to the axis, the fins etc. can for example be spiral-shaped, with the axis of the spiral coinciding with the axis of the rudder.

In order for the construction to be as simple as possible, the wall thickness of the rudder or of each rudder is advantageously made practically constant between the two points, as the variation in the total cross-sectional area of the electric current's flow path is then a result only of a variation in the cross-sectional area, vertically on the rookery, of the fins, etc. When the temperature to which the rudder wall can be heated is limited as a result of the requirements for mechanical strength, it can however be advantageous to vary the wall thickness in the longitudinal direction of the rudder. Because the smallest thickness of the rudder wall that will provide the necessary strength is lower where the wall's temperature is lower, it is in principle possible to achieve a saving of metal in cases where the temperature of the wall is not constant in the longitudinal direction of the rudder, by reducing the thickness of the wall where the temperature is lower. Furthermore, it is in principle possible to achieve a saving of metal, compared with the case where a rudder of constant thickness is used, whose inner surface temperature is kept essentially constant in the longitudinal direction of the rudder between the two points, by choosing the configuration of the fins etc. so that the temperature of the inner surface of the rudder wall is lower over a part of the rudder immediately downstream of the point of origin' of the two mentioned points than anywhere else between the two mentioned points, pg by reducing the wall thickness of this part of the rudder accordingly. The reason for this is that when the temperature difference between the rudder wall

and the fluid is large, a small decrease in the wall temperature will only slightly reduce the transfer of heat from the tube wall to the fluid, while at higher temperatures even a small decrease in the wall temperature will lead to a significant increase in the material's mechanical strength and a significant decrease in the required minimum wall thickness.

The manner in which the total cross-sectional area of the electric current flow path must vary in the longitudinal direction of the rudder or in each rudder to achieve a given temperature distribution over the inner surface of the rudder exposed to the fluid depends on a number of factors including the mass density of the fluid through the rudder, the original temperature of the fluid, the maximum temperature desired by the surfaces exposed to the fluid, and the physical properties of the fluid. It will also depend on the length of the intervals, if any, in the longitudinal direction of the rudder or rudders between the points where the fins etc. has electrical contact with the rudder. Provided that the changes in the total cross-sectional area remain small in each such interval, the effect of the changes in the length of the intervals will be small.

Although an essentially uniform temperature distribution requires that

the total cross-sectional area changes continuously in the longitudinal direction of the rudder or each rudder, a satisfactory approximation to this can be obtained by changing the total cross-sectional area step by step, the cross-sectional area remaining constant between these successive steps. If this latter arrangement is used, and if the fins or similar are not in electrical contact with the rudder or rudders in which they are arranged continuously over the entire length of the rudder, electrical contact between the fins and the like can and the rudder is suitably provided in the areas where the total cross-sectional area changes. In addition, a mainly even temperature distribution requires that fins etc. are arranged there. of very small cross-sectional area towards the upstream point of the two mentioned points, and the construction difficulties this entails may outweigh the relatively small advantage

which is achieved. An acceptable compromise is a la finnen et al. continue a certain distance in the upstream direction beyond the point where, in order to achieve a uniform temperature distribution along the tube wall, they reach the smallest appropriate dimensions, but maintain the cross-sectional area of the fins etc. constant over said distance at a value determined by said smallest appropriate dimensions, and then terminate the fins etc. completely at a point on the downstream side of the upstream point of the two points between which the electric potential drop is applied.

The configuration of the rudder or rudders and the way in which they are supported must of course be subject to consideration of the thermal expansion and contraction that takes place in the rudders during operation. One can advantageously arrange the rudders in pairs, as the rudders in each pair

at least over the length where the fins etc. are arranged, are straight or coaxial, and the pairs of tubes are parallel at least over said length, with the inner ends of each pair of tubes being connected to one or more outlet branch tubes which are common to several pairs of tubes, and the outer ends of the tubes being connected with supply branching rudder through flexible connections, - e.g. through pods. This arrangement has the advantage that you avoid using flexible connections at the point where the pipes have a very high temperature. When the fluid is a halide vapor and the surfaces of the ribs and fins etc. which is exposed to the halide vapor, is of platinum or a platinum alloy, the outlet manifold or each outlet manifold must of course also be constructed of platinum or a platinum alloy.

Because the Finns and others usually does not have the same temperature as the rudder wall, the configuration of the fins etc. must and the manner in which they are attached to the rudder wall, whether they form an integral part of the rudder wall or not, is usually chosen taking into account the possibility that differential expansion between the rudder wall and the fins etc. can take place. In order to make such a differential expansion possible without too high voltages occurring, the fins and the like can be constructed with small incisions at certain Intervals in their longitudinal direction. Although the small areas with a reduced total cross-sectional area of the electric current's strbmniHgs path that occur as a result of the incisions will usually lead to local overheating, the consequences of this can be reduced or eliminated by appropriate design. Firstly, if the areas with reduced cross-sectional area are sufficiently small, the heat conduction in the fins etc. be sufficient to prevent severe local overheating. Secondly, the temperature of the fins etc. is reduced by increasing their surface area per mass unit, and it will usually be significantly lower than the desired maximum temperature, except in the area at the downstream point of the two mentioned points, and areas with increased temperature will therefore not necessarily result in overheating. Thirdly, when the Finns and others is in electrical contact with the rudder only at certain intervals in the longitudinal direction of the rudder, the strength of the electric current in the rudder wall in these intervals is determined by the stem movement of the total electrical resistance of the fins etc. in the area,

and the effect on the total resistance of a notch in a fin can be compensated for (at the expense of a small increase in the total amount of metal required) by a small increase in the cross-sectional area of the fins, etc. somewhere else in the interval.

When the fluid is a halide vapor as mentioned above, the tube or tubes are preferably constructed of a platinum/rhodium alloy containing between 5 and 25 weight percent, preferably between 10

and 15 percent by weight, rhodium calculated on the weight of the alloy. If alloys of platinum and ruthenium or platinum and iridium are used, the proportion of ruthenium, respectively iridium, is advantageously between 5 and 25 percent by weight (preferably between 10 and 15 percent by weight), calculated on the weight of the alloy. Finns and others may be constructed of the same material as the rudder or ribs, but it is preferred to construct the rudder or ribs of a platinum alloy and to construct the fins etc. of platinum. This has the advantage that if differential expansion between the rudder and the fins etc. takes place, the lesser strength of the platinum will allow the expansion to be accommodated by deformation of the fins, etc., without excessive stresses occurring in the rudder, provided that the fins, etc. has a suitable shape and is attached to the rudder wall in an appropriate manner.

If the electric potential is applied to the rudder or each

rudder through a structural part such as e.g. a branch pipe or another pipe, which has a surface that comes into contact with the fluid, it may be necessary to ensure that said structural part reaches too high a temperature. Preferably, the rudder is connected to the electrical source only at the point of origin of the two mentioned points.

Thus, when several rudders are arranged in pairs as described above and connected to a common outlet, each rudder is preferably connected to the electrical source only at points which are distant from the common outlet, i.e. at the upstream end of the fins etc. or between this point and the flexible connection at the upstream end of each rudder. Preferably, a positive direct current potential is applied to one of the rudders of each rbr pair and a negative direct current potential to

the other ror of each rbr pair in this way, grounding the common utlbps branch rbr. If the cross-sectional area of the electrical flow path provided by the outlet manifold is sufficiently large to prevent overheating of the outlet manifold, the fins and the like can in the ribs of the rib pair or each rib pair is extended into this.

When an electrical connection is provided to the rudder or to each rudder, so that the electrical potential is applied to the point of origin of the two mentioned points, the rudder is provided, immediately on the underside of the electrical connection, preferably with one or more further internal fins, etc. which is in electrical contact with the rudder wall, and which provides a flow path for the electric current which has a sufficiently large cross-sectional area to ensure that the temperature of the rudder wall in the immediate vicinity of the electrical connection becomes relatively low during use of the apparatus.

The electrical potential can be either a direct current potential or an alternating current potential, but if it is an alternating current potential, the frequency must of course not be so high that it causes any significant "skin effect", because the invention rests on the assumption that the current density is mainly constant over the cross-sectional area of the strbmning path at any point along the rudder, except where materials of different electrical resistance are used together.

When several pipes are used which are arranged in parallel with regard to the flow of fluid between the inlet manifold and the outlet manifold, it is desirable to ensure that the flow rate of the titanium tetrachloride vapor is practically the same through each pipe, and if pressure gradients of any importance occur in the branch pipes, even distribution between the pipes requires either that the pressure drop across each pipe is large compared to the pressure drops in the branch pipe or that it is compensated for by giving the pipes different flow impedance, e.g. by providing them with openings of different pipe diameters.

When the fluid is a halide vapor as mentioned above, the inner diameter of the tube or of each tube can be from 12.7 to 50.8 mm. The optimum thickness of the rudder wall depends to some extent on the internal diameter of the rudder, but it is usually preferred that the wall thickness is between 0.25 and 1.27 mm, preferably approx. 0.51 mm.

A particularly important use of the invention is for preheating

of titanium tetrachloride vapor, which may be mixed with aluminum chloride and/or silicon tetrachloride, because it is oxidized in the vapor phase to produce titanium dioxide pigment. At least the surfaces of the rudder or rudders and of the fins etc. which is exposed to the halide vapor,

will then consist of platinum or an alloy of platinum with rhodium, ruthenium or iridium. When the halide vapor contains titanium tetrachloride vapor, the temperature on the surface of the rudder or rudders and fins etc. which is exposed to the halide vapour, preferably not exceeding 1100°C at any point. At temperatures above this tempera-

turn can the rudder or rudders and fins etc. get twisted,

if the device is not designed so that thermal stresses are kept at a very low level. Moreover, when heating titanium tetrachloride for vapor phase oxidation to titanium dioxide, it is usually not necessary to heat the rudder or rudders to temperatures above 1100°C.

The titanium tetrachloride vapor can be supplied to the rudder or rudders by

a temperature of approx. 160°C and then heated in the rudder or rudders to a temperature of approx. 1000°C. To preheat the titanium tetrachloride to a temperature of 1000°C, it is usually necessary to heat the copper to a temperature approaching 1100°Co

A version of the device which is suitable for preheating titanium tetrachloride vapor for this to be oxidized in the vapor phase to produce titanium dioxide, and which is constructed according to the invention,

shall now be described as an example, with reference to the attached drawings, where

fig. 1 is a schematic plan view of the apparatus with the top

wall of the enclosure removed,

fig. 2 is a schematic section through the line A-A in fig. 1,

fig" 3 shows a schematic, vertical longitudinal section, which does not

are in the same scale, through one of the rudders with associated equipment,

fig. h is a plan view, which is also not to the same scale, of

a pair of spread fins,

fig., 5 shows an axial section through a short piece of one of

the rivers on a larger scale, and

fig, 6 shows a. cross-section through the line B-B in fig» 5.

The apparatus shown in fig. 1 and 2 on the attached drawings include i6 heating radiator 1,. as. apart from the ends, are placed in an enclosure 2, which is constructed of a non-metallic refractory material containing a high percentage of aluminum oxide and is arranged to support the ribs 1 at points spaced at intervals along the length of the rods length.

The cores 1 in the enclosure 2 are straight, and the fibers in each pair of fibers are coaxial. The axes of the ribs in each pair of ribs are parallel to each other and lie in the same horizontal plane. At the downstream end, the pipes 1 communicate with a horizontal, transverse outlet pipe line 3 (the downstream end of each pipe, see Fig. 3, has a somewhat greater wall thickness to enable the transfer of electric current to the outlet pipe line 3 without overheating), which is closed in one end, and whose other end leads out of the enclosure 2.

On the outside of the enclosure 2 buoys the upstream end of the cables

1 of vertically downwards.. The upstream end of the ribs 1 is provided with (see fig. 3) an annular flange h which is attached to a corresponding flange 5 on top of a flexible bellows 6. At the lower end of the bellows 6 there is arranged a annular Ilens 7 which is attached to a corresponding flange 8 which is attached to the upper end of the branches of two inlbps branching branches 9, one on each side of the heater.

The bellows 6 are made of an electrical insulating material, and the upper and lower ends of each bellows are connected to an electrically insulating rod 10, which is rotatably arranged at each end of the bellows. Because each of the rods 1 has only a short vertical portion compared to its horizontal length, the relative movement between the upstream end of the pipe and the branch rod 9, due to thermal expansion and contraction, becomes a mainly horizontal movement, and this movement is not prevented by the rod 10. • with just outside the wall of the enclosure 2, the wires 1 are connected to electrical connectors 11, which are water-cooled

Each rudder 1 is internally, and inside the casing 2, provided with an assembly of fins, as indicated generally by reference number 12. Each assembly 12 consists of 8 fins 13 which are constructed as four pairs and are welded at intervals in the longitudinal direction into short rib-shaped pieces l^f. Each pair of fins 13 is formed by bending a flat strip of laminar material, which is indicated generally by reference numeral 15 (see fig. h), along its longitudinal center line, so that the two fins 13 in each pair, i.e. the parts of the strip 15 which lies on either side of the longitudinal center line forms an angle of 1+5° between them. The pairs of fins 13

Xp -— *w ■%— ~ ~ w are welded to the rib pieces lh in the area at the aforementioned center line, and they are oriented so that also the angles formed between neighboring fins belonging to different pairs become hS° (see fig. 6).

It will be seen from fig. h that the width of each strip 15 had a constant maximum value over short lengths 16 and over a somewhat longer length 17 at one end of the strip, and that the width of the parts 18 of the strip, which lie between the wide parts 16 and 17, increases from one area 18 to the next along the strip towards area 17. It will be seen from fig. 5 that the wide parts 16 are very short (in the longitudinal direction of the fins 13) compared to the length of the parts 18 at gradually increasing width. Correspondingly, the rbr pieces l*f are very short compared to the distances that separate them.

The fin assemblies 12 are placed in the ribs 1 with the wide parts 17 of the strips 15 at the downstream end of the ribs 1, and the outer ends of the wide parts 16 and 17 are welded to the walls of the ribs 1. Thus, each fin assembly 12 is attached to the associated rudder 1

in such a way that the rbr pieces lh lie coaxially with the rudder and the fins 13 come into electrical contact with the rudder wall 1 at each point where they are welded to the rudder.

The fin assemblies 12 terminate at the upstream end on the downstream side of the inner surface of the enclosure 2 wall, but each rudder 1 is internally, immediately on the downstream side of the coupling 11, equipped with another fin assembly 21. Each fin assembly 21 is similar to the fin assemblies 12, in that they are constructed of eight radially splayed fins formed in pairs by bending four strips along the longitudinal centreline, and the fins are welded to a cylindrical piece lh and (at the ends where the fins are widest) to the rudder wall. However, the fin assemblies 21 are much shorter than the fin assemblies 12 and, except for their wider end pieces, the fins in the assemblies 21 have a constant width in the longitudinal direction.

In fig. h, it will further be seen that each strip 15 is provided with small holes 19 along the center line to facilitate the bending of the strip to form the fins 13, and with incisions 20 which are extended to allow differential expansion between the fins 13 and the rudder wall 1 without excessive voltages occur. As described above, such small incisions 20 do not lead to overheating, if they are properly designed. The same applies to the holes 19.

As an example of suitable dimensions, the thickness of the rib walls 1 and the thickness of the laminar material from which the fins are constructed may be 0.51 mm. The diameter of the holes can be 1.59 mm. The inner diameter of the rudders 1 can be 25.W mm and the length of each rudder between 3?66 and H-.57 m. The length of the parts 18 of the strips 15 can vary in a given strip from approx. 0.61 m at the upstream end to approx. 0.23 m at the downstream end, if it is assumed that there are six such parts' The length of each of the wide parts 17 of the strips 15 may be 6.35 mm. The inner diameter of the outlet rudder line 3 can be 76.2 mm and the rudder thickness 0.76 mm.

Although the rudders 1 and the outlet rudder line 3 are preferably constructed of a platinum/rhodium alloy containing 10% rhodium by weight, the fin assemblies 12 and 21 are preferably constructed of platinum to ease the deformation of the fin assemblies when these take up the difference in thermal expansion between the aggregates and the rudder wall 1 (as described above).

During operation, the outlet pipe line 3 and the supply pipe line 9 are electrically grounded, while a positive direct current potential is applied to the coupling 11 at one end of the enclosure 2 and a negative direct current potential is applied to the coupling 11 at the other end of the enclosure 2 (the absolute magnitude of these potentials is the same ).

For each pair of rudders 1, an electric current thus flows from the coupling 11 on one rudder of the rudder pair, through this rudder and the two fin assemblies 21 and 12 in the rudder, through the wall of the outlet rudder line 3 and then through the other rudder 1 of the rudder pair and the two fin assemblies 12 and 21 in this other rudder to the coupling 11 on this rudder.

If one considers a single rudder 1 and disregards the effect of the rbr pieces 1*+, the wide parts 16 of the strips 15 from which the fins are formed, the holes 19 and the incisions 20, the total cross-sectional area of the electric current flow path becomes large immediately on the underside of the coupling 11 on the rudder as a result of the presence of the fin assembly 21. At the downstream end of the fin assembly 21, the cross-sectional area suddenly decreases to its minimum value, which is the cross-sectional area of the rudder wall alone, after which it increases step by step as a result of the presence of the fin assembly 12. The electrical resistance per . unit length of the rudder is thus small immediately downstream of the coupling 11 as a result of the presence of the fin assembly 21, but at the downstream end of this assembly it suddenly increases to a maximum value, after which it gradually decreases in the downstream direction.

Because the strength of the electric current is constant in each rudder

1 longitudinal direction, is the rate at which heat is generated per length unit of the rudder directly proportional to the electrical resistance per length unito The speed at which heat is generated is thus relatively small immediately on the downstream side of the coupling 11, which (together with the water cooling of the coupling) prevents the coupling 11 from reaching too high a temperature. Immediately downstream of the fin assembly, the heat generation rate per length unit of the rudder 1 marked, after which it gradually decreases in the downstream direction.

The titanium tetrachloride vapour, which may be mixed with smaller amounts of other chlorides, is supplied through the supply branch pipe 9, and its temperature increases progressively as it passes through the pipes 1 to the outlet pipe line 3.

The temperature for each rudder 1 wall increases in the downstream direction over each section of the rudder where the total cross-sectional area of the electric current flow path is constant, and increases more markedly at the downstream end of the fin assembly 21, where the total cross-sectional area decreases" Where the total cross-sectional area increases in the downstream direction as a result of the design of the fin assembly 12, the temperature of the tube walls 1 drops. The temperature profile for the tube wall in the longitudinal direction of each tube 1 has an approximately sawtooth shape over the length that accommodates the fin assembly 12, and it slopes downwards in the upstream direction beyond the upstream end of this unit. The changes in the total cross-sectional area are thus chosen in relation to the mass velocity of the steam and the degree of external coupling of the tubes 1

(as described below) that the temperature profile's peak value is below the maximum temperature that can be accepted.

The temperature for each fin assembly 12 tends to decrease in the upstream direction, but this has (as described above) no particular significance.

The inner surfaces of the walls of the enclosure 2 are kept at a temperature below the temperature of the tubes 1, which (as described above) makes it possible to reduce the weight of platinum or platinum alloy in the fin assemblies 12 which is necessary to prevent overheating of the tube walls 1.

Claims (6)

1. Apparatus for heating a fluid, comprising at least one pipe through which the fluid can be guided and which is provided on the inside with metal fins or other elements that are in contact with the fluid, the cross-sectional area of the fins or the other elements increasing in the direction of flow of the fluid, characterized by that the rudder (1) is also made of metal and that the fins or the other elements (12) are in electrical contact with the rudder wall, and the total cross-sectional area of the rudder wall and the fins or the other elements increases along the rudder in the direction of the fluid flow so that when both the rudder and the fins or the other elements are heated by the application of an electric potential difference between two points at a distance from each other in the longitudinal direction of the rudder by the sun producing an electric current both through the rudder wall and through the fins or the other elements, said increase in the total cross-sectional area will prevent the temperature of the inner surface of the rudder wall which is in contact with the fluid, in increasing in the direction of flow of the fluid, at least over a certain length of the rudder, while the surface area of the fins or the other elements is such that the temperature of the surfaces of the fins or the other elements in contact with the fluid, at no place does the maximum temperature on the inner surface of the rudder exceed between the two points mentioned. 2. Device according to claim 1, characterized in that the rudder or each rudder (1) is arranged in a heat-insulated enclosure (2) whose internal surface during use of the device has a temperature that is significantly lower than the temperature of the external surface of the rudder at the downstream point of the two points mentioned. 3. Apparatus according to claim 1 or 2, characterized in that it comprises one or more additional tubes (1) which are connected to a common outlet (3), and that electrical connections (11) to the tubes are arranged only at points located distance themselves from the common outcome. h. Apparatus according to claim 3, characterized in that one or more additional internal fins or other elements (21) are arranged immediately on the downstream side of the electrical connections (11) to the rudders (1) which are in electrical contact with the rudder and which have a sufficient cross-sectional area to ensure that the temperature of the tube walls in the immediate vicinity of the electrical connections is relatively low during operation. 5. Apparatus according to claim 1-^, characterized in that the cross-sectional area of the fins or the other elements (12) is such that during use of the apparatus the temperature on the inner surface of a part of the rudder or each rudder's (1) wall is immediately on the downstream side of the upstream point of the said two points lower than at any other place between the two points, the said part of the wall having a reduced thickness. 6. Apparatus according to claims<1>1-5, characterized in that at least the surfaces of the rudder or each rudder (1) and of the fins or the other elements (12,21) which are exposed to the fluid are made of platinum or an alloy of platinum with rhodium, ruthenium or iridium.