WO2005021943A1

WO2005021943A1 - Flow distributor with screens to optimize the gas flow in catalyzers and use

Info

Publication number: WO2005021943A1
Application number: PCT/BR2004/000161
Authority: WO
Inventors: Marcelo Augusto Faraco De Medeiros; Eduardo Murilo Rosas Arantes
Original assignee: Fundacão De Amparo À Pesquisa Do Estado De Minas Gerais-Fapemig; Pontifícia Universidade Católica De Minas Gerais-Puc
Priority date: 2003-08-29
Filing date: 2004-08-27
Publication date: 2005-03-10
Also published as: BR0303014A

Abstract

This invention refers to a flow distributor using screen(s) to be applied in the catalyzers used in combustion internal motors, automotive or not. This purpose of the distributor is to optimize the flow distribution of the fumes that come from the motor when passing through the catalyzer. The screen(s) are duly determined and carefully placed at the entrance of the catalyzer, thereby reducing the pressure gradient, and avoiding the reverse flow or separation of the limiting layer responsible for the flow non-uniformization inside the catalyzer. This flow distributor with screens is cheap to produce, has a low counter pressure and little heat absorption, besides showing excellent results in improving the uniformity coefficient whenever applied in a catalyzer and compared with other devices currently on the market.

Description

FLOW DISTRIBUTOR WITH SCREENS TO OPTIMIZE THE GAS FLOW IN CATALYZERS AND USE This invention refers to a device installed in catalytic converters—or simply catalyzers—used in the exhaust systems of internal combustion motors for automotive applications, or not. This device is has "screens" whose purpose is to optimize the flow of the combustion gases that pass through the catalyzer, and redistribute it uniformly through all the cells of the "monolithic ceramic" or "monolith", where the due chemical oxy-reduction processes of the most polluting gases eliminated by the motors occur. So, the flow uniformization directly increases the efficiency of the catalyzer. The monolith can also be made with other non-ceramic based materials. The catalyzer was introduced in the Brazilian industry in 1992. According to carmakers, the current useful life of a catalyzer must exceed 80 thousand kilometers, or five years. However, to reach this useful life, some precautions must be taken, such as a perfect maintenance of the motor and of the power supply and ignition system. The function of catalyzers is to catalyze, in other words, to accelerate the changes in the chemical composition of the fumes exhausted by internal combustion motors so as to reduce the main pollutants that are present in gases to acceptable levels in accordance with the legislation in force, on the basis of the cycle of emissions established by the FTP 75 (Federal Test Procedure) , and which specifies the parameters and conditions for emission tests. The competent bodies must conduct the pollutant emission tests, since they license the vehicle to come and go, and notify the owners whenever the catalyzer is damaged, or not functioning properly. The catalyzers have a central core that is impregnated with a catalytic reagent. This reagent is made of substances that have noble metals (like rhodium, platinum, nickel, among others), which, as a result, makes the catalyzer quite expensive. Generally speaking, this core is shaped like a "beehive", and placed along the axis of the car's exhaust pipes or system. The beehive or ceramic monolithic is usually made up by cells that can have a square (the shape mostly used) , rectangular, triangular, or circular transversal section with a density of approximately 100 cells per square centimeter. Other shapes for the transversal section of the beehive cells can also be found. The exhaust fumes, such as, carbon monoxide (CO), nitrogen oxides (NO_x) and nonburned hydrocarbons (HC) are sent to the catalyzer through the exhaust pipes, where they react, oxide, and reduce before turning into non-polluting gases like nitrogen N₂, carbon dioxide C0₂, and steam H₂0. Some of those chemical reactions are shown as follows in TABLE 1.

TABLE 1: Reactions inside the Catalyzer 2CO + O₂ -» 2CO₂ (I) 2C₂H₆ + 7O₂ -» 4CO₂ + 6H₂O (II) 2NO₂ + 4CO -.> N₂ + 4CO₂ (III) Catalytic efficiency is required, as stipulated by the legislation, and defines the useful life of the catalyzer. The efficiency values required are in the region of 99%, but it is difficult to keep them all through the useful life of the vehicle. The most critical phase in terms of fume emissions is just after cold starting motors. In a typical test cycle, motors produce 60 to 85% nonburned hydrocarbons—among which the olefin and the alkene series— during the first 200 seconds of running. This is partly due to the fact that automotive catalyzers cannot convert those hydrocarbons before a monolithic ceramic has reached a temperature above 300° C, after rising through the heat of the gases originating from the combustion, that is to say, from the motor fumes. The cycles of pollutant emissions executed by the car makers follow, for example, the procedures specified in the FTP 75 (Federal Test Procedure) . Generally speaking, the fumes are not distributed uniformly through the different cells of the monolithic ceramic {Angele, B . and Kirchner, K. , 1980, "The Poisoning of Noble Metal Ca talysts by Phosphorus Compounds - III", Chemical Engineering Science, v.35, pp. 2101 -2105) ; in other words, the speed of the gases for each cell of monolithic ceramic is not uniform; this generates a series of performance losses for the catalyzer. The chemical efficiency of the catalyzer, or its capacity to make polluting fumes react and change into non-polluting fumes, drops with the gas flow increase per cell of monolithic ceramic {Angele, B. and Kirchner, K. , 1980, "The Poisoning of Noble Metal Catalysts by Phosphorus Compounds - III", Chemical Engineering Science, v. 35, pp. 2101 -2105) . In other words, the chemical efficiency of the catalyzer is inversely proportional to the speed of the flow in every cell. Therefore, differentiated chemical efficiencies occur between the cells of the catalyzer's monolithic ceramic. For a non-uniform flow distribution in the monolithic ceramic, the cells in the central area go at extremely high speeds compared to the cells located further on the outside. Those speed distribution distortions in every cell of the monolithic ceramic become quite serious with the increase of the number of Reynolds . The number of Reynolds is a non-dimensional parameter, which is calculated for the flow, and represents the ratio of inertial to viscous forces acting. The discharge non-uniformization causes an irregular consumption of the catalytic reagent {Angele, B. and Kirchner, K. , 1980, "The Poisoning of Noble Metal Ca talysts by Phosphorus Compounds - III", Chemical Engineering Science, v. 35, pp. 2101 -2105) . As a matter of fact, at the end of the useful life of a catalyzer, some regions of the monolithic ceramic may have a high concentration of reagent, which has a significant impact on the cost of the equipment and on its useful life. Moreover, because of the flow non-uniformization, the emission of the solid particulates of carbon, or soot, and other contaminants exhausted by the motor are deposited irregularly throughout the cells of the monolithic ceramic transversal section. This irregular distribution also increases the counter pressure added to the exhaust system of the motor {Day, J. P. and Socha , L. S . Jr. , 1991 , "The Design of Automotive for Improved Pressure Drop and Conversion Efficiency", International Congress and Exposi tion Detroi t, SAE Paper 910371 , Detroi t, Michigan) and consequently reduces its effective power (power measured at the end of the motor axis) . It is important to highlight that the exhaust system is made up by all the components through which the fumes pass, such as the pipes, or galleries in the motor, the end collectors of the motor fumes, the exhaust pipes, the catalyzer, the connections, the flanges, the mufflers, the silencers, among others. For ^■ an irregular discharge, the load loss in the monolithic ceramic is significantly higher than the corresponding loss for a uniform or an ideal flow. This results in a higher counter pressure in the exhaust system of the motor, and consequently, reduces the fluid-dynamic efficiency. The heating time of the catalyzer—a predominant factor in reducing the emission of polluting fumes in the atmosphere after an internal-combustion motor is cold started—is directly influenced by the flow distribution in the cells of the monolithic ceramic, since the flow distribution is directly related with the heat flow distribution { Jobson , E. et al . , 1991 , "Spatially Resolved Effects of Deactivation on Field-Aged Automotive Ca talysts", Interna tional Congress and Exposi tion, SAE Paper 910113, Detroit, Michigan) . Besides, the temperature non-uniformity of the monolithic ceramic, as a consequence of the non-uniform distribution of the heat flow, generates significant thermal tensions that very often damage and reduce the useful life of the catalyzer. Those problems could be reduced if the flow of the fumes were distributed uniformly between the cells of the monolithic ceramic. So, various systems have been applied to gain a more uniform flow. This invention presents this type of device. Generally speaking, the flow non-uniformity in the catalyzer comes from the following phenomenon. Normally, the installation of a catalyzer in the exhaust system of a vehicle requires increasing the transversal section of the exhaust pipe because the diameter of the monolithic ceramic is larger than that of the pipe. This section increase can be done using a component known as a diverging diffuser or mouth, as shown in FIGURE 1. A "flange" or "glove" is also commonly used as a component to connect the catalyzer to the exhaust system of the vehicle. Those types of components enable an increase in the pressure and a reduction in the speed of the discharge. The gain in pressure can be moderate or blunt and will be directly proportional to the increase of the transversal section with the longitudinal length. Consequently, you can determine how much the pressure increases through the "pressure gradient". This pressure gradient defines the pressure variation rate along the length, that is, flow- wise. The positive pressure gradient is adverse to the discharge, as it forms a "reverse flow". This reverse flow is characterized as part of the discharge, at a specified location inside the component (diffuser, glove, or flange) , in which it goes in the opposite direction of the main discharge. This is represented in the drawing of FIGURE 2, which shows the electric wires describing the flow. The electric wires show the shape and trajectory of the discharge. So, whirls or twirls appear and limits the main discharge, thereby reducing the useful transversal section. Those twirls can be stationary or move periodically in the same direction as the main flow, thereby provoking new twirls and, very often, larger limitations of the main-flow useful transversal section. It is as if various "bubbles" formed and were released in the course of time, making the electric wires of the flow distort and, consequently, making the flow non-uniform. The diverging diffusers or mouths applied in catalyzers are considered as "wide angle" diffusers and have a high area ratio, that is, the ratio between the transversal section areas at the outlet and the inlet of the diffuser. Those diffusers have a high total aperture angle or divergence angle. The "wide angle" diffusers are those whose aperture angle is above 12°. To uniformize the flow, those diffusers need control devices or flow distributors. The feature normally encountered in the diffusers of catalyzers is the displacement of the inlet exhaust pipe towards the inner part of the diffuser. Diffusers presenting this feature sometimes have a more uniform distribution in the speed profile at the outlet compared with diffusers without such a constructive feature. Apparently, it is due to the formation of the twirl at the outlet of the pipe which "drags" the fluid towards the inner walls of the diffuser {Kachara , N. L . , 1973, "Wide- Angle Axisymmetric Diffusers Employing a Specified Steep Pressure Rise on the Boundary Walls", Internal Report, FM/14/73, Universi ty of Salford) . The optimization of the tube length towards the inside of the diffuser can improve the pressure recovery coefficient (C_rp) by over 10%. The pressure recovery coefficient (C_rp) is defined as the ratio between the difference in static pressures between the outlet and the inlet of the diffuser and the difference between the stagnation and static pressures at the diffuser inlet. This determines its efficiency and the flow uniformization at the outlet. We confirmed that for diffusers with inlet tubes going inwards and area ratios below 1.4, the pressure recovery changes in the range of 5% maximum (Farell , C and Xia , L . , 1990, "The Note on the Design of Screen-Filled Wide-Angle Diffusers", Journal of Wind Engineering and Industrial Aerodynamics, v. 33, pp. 479-486) . However, those studies were conducted for diffusers with relatively small angles, which is not the case of the diffusers normally used in catalyzers. With wide-angle diffusers, this effect is almost insignificant, and does not present any substantial advantages in terms of flow uniformization. Experiments conducted with automotive catalyzers improved the speed distribution, or made the flow more uniform, and increased the distance of the diffuser inlet tube up to the front of the beehive of the catalyzer (Lemme, C. D. and Gives, W. R. , 1974 , "Flow Through Catalytic Converters - An Analytical and Experimental Trea tment", Automotive Engineering Congress, SAE Paper 740243, Detroi t, Michigan) . Those solutions, in most cases, are inadequate because of the high production cost involved using exhaust pipes of a large diameters and, mainly, because of the lack of space and the heavy weight of the set. The researchers also claim that the geometry at the start of the monolithic ceramic, in other words, at the outlet of the catalyzer, has little influence on the flow distribution or uniformization. The flow deflector or flow distributing vane is a device very much used to uniformize the distribution of speed in catalyzers. Those devices come in various shapes and sizes, and are normally inserted at the diffuser inlet to improve the flow uniformization in the beehive. Those devices increase the counter pressure and the heating time of the catalyzer after the motor is cold started {Howi tt, J. S . and Sekella , T. C. , 1974 , "Flow Effects in Monoli thic Honeycomb Automotive Ca talytic Converters", Automotive Engineering Congress, SAE Paper 740244 , Detroi t, Michigan) . They used a high counter pressure while absorbing part of the heat discharged by the gases, and produced "shadow" areas on the front of the beehive inlet, where there is practically no flow. The devices analyzed in the literature on the subject are conical deflectors, disk deflectors and deflectors of the vane or aerofoil type with pitch angles of 60°, 45°, and 30°. The highest counter pressure was produced by the aerofoil with vanes at 30°, followed by those at 45°, and 60° for the disk and conical deflectors. It is important to note that those devices do not manage any proper flow uniformization because of the

"shadow" areas (parts without any flow) , and because they increase the weight of the set and the heating time of the catalyzer. The deflectors with vanes work as aerofoils in the shape of a propeller, and have the effect of reducing the pressure gradient on the diffuser walls; nevertheless, the same problems were detected when using the deflectors {Howitt, J. S. and Sekella , T. C. , 1974 , "Flow Effects in Monoli thic -Honeycomb Automotive Ca talytic Converters", Automotive Engineering Congress, SAE Paper 740244 , Detroi t, Michigan) . Besides those deflectors, the methods mostly used to control the separation of the limiting layer, that is to say, the uniformization flow, applied to diffusers are, more generally speaking: whirl generators and the installation of screens transversally to the flow. Experiments were carried on diffusers with curved walls {Mehta , R. D. , 1977, " The Aerodynamic Design of Blower Tunnels wi th Wide Angle Diffusers", Prog. Aerospace Science, v.18, pp. 59-120) . This shape of wall construction can be considered as an ancillary method to control the separation. Diffusers with curved walls in the shape of a "bell" also were noticed to have a higher performance compared to similar diffusers with straight walls. The US patent 4,797,263 refers to the application of a flow distributor installed before the beehive or monolithic ceramic of the catalyzer. This deflector is a plate that has apertures at various points and differentiated dimensions that limit the flow in the central part of the monolithic ceramic, thereby redistributing the flow outwardly. This type of deflector presents the same problems as those already mentioned, such as a high counter pressure, a high heat absorption and "shadow parts of flow", when compared with the "screen" proposed in this invention as flow distributor. The US patent 4,730,786 is on a secondary device of tap faucet, which uses screens and perforated disks as a means of flow uniformization, thereby changing the features of the flow more particularly to promote a laminate flow, that is, one without any turbulence. This way, the level of noise generated by the flow is reduced. This invention applies to washing basin taps and aims at reducing the noise. The other solution studied recently has to do with the change in the format of the monolithic ceramic. One alternative is to increase the length of the central channels of the monolithic ceramic compared with the peripheral channels (Matthew, J. , 1999, "The New Study Looks at the Impact Off Cell Structure on Advanced Converter Performance", Auto Emissions, n.12, pp. 3-4). This solution, besides the fact of it being a difficult construction and likely to generate a waste of ceramic material, requires a far greater length for the central channels than that for the peripheral channels, which jeopardizes the size of the system as a whole. It seems, the problem is barely reduced with this system. Another alternative is to adopt variable-section cells, that is, cells with a smaller transversal area at the center of the monolithic ceramic and cells of a larger transversal area at the outside (Kim & Son, 1999, Improved flow efficiency of the catalytic converter using the concept of radially variable cell density - Part I. SAE Technical paper in the 1999-01 - 0769) . This construction solution is also difficult, more particularly, because of the ceramic material, which usually is very fragile. Additionally, there seems to be a problem in the fact that to have an average speed in the various cells, the relative area seems to be larger in the smaller cells, and so, the chemical reaction, as a consequence of the catalytic substance consumption, is not uniform. If, in that case, the thickness of the cell walls were uniform, the small cells would be responsible for a volume of ceramic proportionally larger in comparison with the larger cells, which would increase the heating problems when cold starting. In fact, those types of solution show further more intrinsic problems to the method of control, which are quite difficult to resolve. The method most usually used to prevent the separation in "wide angle diffusers" uses screens that are placed transversally to the flow. The screens provide a more uniform speed profile and avoid the direct effects of the growth of the limiting layer and its separation. To resolve the inconveniences caused by the other currently studied and used methods for the flow uniformization in catalyzers, this invention was developed; it is characteristic for applying screens inside the diffuser or in another part of the catalyzer. The use, inside the diffusers, of screens with the most varied mesh configurations, shapes and distinctive dimensions has been widely studied to control the "reverse flow", in other words, to control the "limiting layer separation" inside the diffusers or diverging mouths of wind tunnels for the aerodynamic analyses and other applications (Mehta , R. D. and Bradshaw, P. , 1979, "Design Rules for Small Low Speed Wind Tunnels", Aeronautical Journal , November 1979, Paper 718, pp. 443-449; Mehta , R. D. , 1976, "The Design of Wide-Angle Diffusers", I. C. Aero Report, v. 76-03, pp. 1 -14 ; Mehta , R. D. , 1977, " The Aerodynamic Design of Blower Tunnels wi th Wide Angle Diffusers", Prog. Aerospace Science, v. 18, pp. 59-120) . Therefore, this invention is based on this theory and applies it to optimize the flow discharge inside the catalyzers. In this invention, those screens were tested to try and optimize the speed profile by taking into consideration aspects such as the type of screen (flat or curved) , the ideal position (location), the installation (perpendicular or transversal to the flow) and the number of screens. It is important to highlight that wherever the "screens" or similar devices are installed, the distribution of the speed profile or the flow uniformization improves significantly {Mehta , R. D. , 1977 , "The Aerodynamic Design of Blower Tunnels wi th Wide Angle Diffusers", Prog. Aerospace Science, v. 18, pp. 59-120) . The diffuser of the catalyzer used as a prototype for testing in the laboratory and for the development and analysis of the screens as a flow distributor, the purpose of this invention, has the shape of a conical circular section with an area ratio of approximately 4 and a divergence angle of about 47°. You usually find catalyzers with diffuser angles and area ratios above the values of this laboratory prototype, and with different shapes for the transversal section like an elliptical section, for example. Nevertheless, the screen flow distributor can be applied no matter the shape of the transversal section of the diffuser, flange, or glove used to connect the catalyzer to the exhaust system. This invention can also be applied in exhaust systems that use any type of internal combustion motors for the most varied uses, such as, power stations, and nautical, maritime, or aeronautic units, among others. The screen can be defined, basically, according to its mesh, thread diameter and the material used in its construction. The screen mesh corresponds to the number of apertures per linear inch, measured from the center to the center of the wires. The diameter of the thread is usually determined by a number (BWG) . This number is fixed in a table and related to the diameter of the screen thread. Other non-dimensional parameters must be defined to determine the effect of the screen under normal exposure to the discharge. The pressure drop coefficient of the screen (K) is one of those parameters, whereby the value of K can be calculated using the following expression: ΔF K = ^est ₂ (IV)

where AP_esl , in equation IV, is a the static pressure drop through the screen and .5pU² is the dynamic pressure of the discharge that is uniform and perpendicular to the screen plane. The pressure drop coefficient of the screen (K) is the parameter that defines the ratio between a static pressure drop of the discharge through the screen and the approximate dynamic pressure. The open-area ratio of the screen (?) is defined as the relation between the area of the screen apertures and the total area, that is to say, it defines the screen porosity. It can be calculated according to the following expression

where 1 , in equation V, corresponds to the distance between the centers of two parallel and adjacent wires and d to the distance to the diameter of the wire. Therefore, you can check that the value of {β) is a non-dimensional one. As for the flow uniformization at the outlet of the monolithic ceramic, you can define the rate varying between

0 and 1 and known as the flow uniformity coefficient (γ) . {Bressler, H. et al . , 1996, "Experimental and Predictive Investigation of a Close Coupled Catalytic Converter with Pulsating Flow", International Congress and Exposition Detroit, SAE Paper 960564, Detroit, Michigan, ) propose the value 1 for the flow uniformity coefficient (γ) , meaning the ideal flow, with the same axial speed in all cells of the monolithic ceramic. The zero value for the flow uniformity coefficient (γ) indicates that the entire flow occurs through a single cell of the monolithic ceramic, and that the speed in the other cells is equal to zero, in other words, it is as if the entire flow that passes through the catalyzer were forced to pass only through one single cell of the monolithic ceramic. The value normally encountered for automotive catalyzers ranges between 0.70 and 0.98. Based on those experimental results { Bressler, H. et al . , 1996, "Experimental and Predictive Investigation of a Close Coupled Catalytic Converter wi th Pulsating Flow", International Congress and Exposition Detroi t, SAE Paper 960564, Detroi t, Michigan) also claim that an average increase in the chemical or catalytic efficiency of 15% is obtained when γ is increased between 0,75 and 0,96. TABLE 2 presents the screens used in the laboratory tests for the present invention, as well as its main parameters. You should know that a wide range of screens is available on the market to conduct those tests. However, after conducting a thorough and long selection, those were sufficient to demonstrate the efficiency of the method. The production of a screen is similar to the production of a cloth, and non-commercial meshes can easily be produced; all you need to is ask the manufacturer for an amount that corresponds to your installations ' requirements and adjustments. The cost of a screen is very low, especially if you consider that the size required is not significantly bigger than the average dimension of the transversal section of an inlet diffuser, a flange, or a glove in a catalyzer.

TABLE 2: Dimensional Parameters for the Screens

Screen Mesh (BWG) Thread Diameter (mm) Aperture (mm) 18/30 0.30 1.11 18/32 0.23 1.18 26/30 0.30 0.67 28/32 0.23 0.67 30/32 0.23 0.61

(B) Weight (kg/m²) Screen Material

0.62 0.83 Galvanized Steel

0.7 0.43 Galvanized Steel

0.47 1.21 Galvanized Steel

0.56 0.67 Galvanized Steel

0.53 0.72 Stainless Steel

The screen tests like the flow distributor applied in the prototype of catalyzer were executed in the laboratory using a "flow bank". The flow bank is an equipment capable of blowing air through a system of pipes, where the prototype is installed. This way, the air flow going through the catalyzer prototype can be varied through the use of control valves installed for that purpose, thereby enabling to simulate similar conditions to that which occur in the exhaust system of an internal combustion motor, with the due fluid-dynamic corrections. The methodology adopted to control and measure the flow speeds was, basically, through the anemometry of a Pitot tube, and the measurements of the counter pressure in the catalyzer were obtained using static pressure inlets 5 connected to a liquid column differential pressure gage. The Pitot tube is an instrument used to measure the speed of fluids, and that is made up by two concentric tubes of a small diameter with inlet holes to take the static pressure (local flow pressure) and the stagnation pressure (pressure

LO of a fluid particle that has decelerated down to speed zero) . This principle is used to measure the speed of aircrafts, for example. The screen placed transversally anywhere in the diffuser already offers some gain in flow uniformity, but

L5 some positions are more efficient, depending on each specific case. The number of screens is also an important parameter. Generally speaking, a single screen already furthers some improvement in terms of flow uniformization, but for diffusers with large divergence angles ("wide 0 angles"), the use of a higher number of screens may be necessary for a perfect flow uniformization. FIGURE 3 is a drawing example of the application of a screen represented by the line drawn inside a diffuser. The use of one or more "screens", chosen adequately according

,5 to their shape and mesh, and installed and correctly placed within those components, furthers a reduction in the gradient of adverse pressure, thereby stopping or considerably decreasing the formation of reverse flow. So, a non-uniformization of the flow can be considerably reduced, as the screen (s) is (are) the essence of the device "flow distributor" proposed in this invention. To conduct a comparative study of the effects promoted by any device proposed, the outlet discharge of the monolithic ceramic must be known without applying any controlling device. The tests were conducted in a catalyzer of a low-capacity motor, of 1000 cubic centimeters approximately. The diffuser carcass had a more or less circular format and the aperture angle of the inlet diffuser was of approximately 47^°. The speed profiles of the monolithic ceramic outlet were analyzed for flows corresponding to the motor rotations when running on full load (fully accelerated motor), but with the due fluido- dynamic corrections. The rotations analyzed were 2,000, 3,000, 4,000, and 5,000 rpm. We also checked the profile of the flow without any screen for rotations above 5000 rpm. TABLE 3 shows that as the motor rotation increases, the flow is increasingly stronger in the central part of the monolithic ceramic, in other words, the flow uniformity is more and more impaired (lower γ values) . The uniformity coefficient values (γ) for the flow can be checked in TABLE 3 for the flows corresponding to rotations corresponding to 2,000, 3,000, 4,000, 5,000 and above 5,000 rpm, calculated as proposed by (Bressler, H. et al . , 1996, "Experimental and Predictive Investigation of a Close Coupled Catalytic Converter wi th Pulsating Flow", International Congress and Exposition Detroit, SAE Paper 960564, Detroi t, Michigan) . The Equation used to calculate γ can be written as follows:

where v, is the local speed and v^^is the average speed through the cells of the monolithic ceramic. TABLE 3: Uniformity Discharge Coefficients for flows corresponding to rotations of the motor running on full load.

Corresponding Rotation (rpm) Coefficient (?) 2,000 0.840 3,000 0.780 4,000 0.760 5,000 0.740 > 5,000 0.660

A large variety of tests were conducted using various combinations of screens in different positions. The best result in terms of flow uniformization was obtained through the use of 3 screens, whereby two were made of a 28/32 mesh and one of a 18/32 mesh; those screens placed transversally and at 5mm (position P05), 15mm (position P15) and 25mm (position P25) respectively from the diffuser inlet of the catalyzer. Those positions can present deviations of ± 1mm. FIGURE 4 shows the result obtained applying those three screens inside the diffuser of the catalyzer for a flow corresponding to the rotation of a l,000-cm3 motor above 5,000 rpm. In that figure, the vertical axis represents the speed of the flow in meters per second (m/s) and the horizontal axis, the normalized radius of the monolithic ceramic, that is to say, the ratio between the local radius anywhere inside the monolithic ceramic and its maximum radius. The unbroken line, in FIGURE 4, shows the flow speed profile along the diameter of the catalyzer at the outlet of the monolithic ceramic when the present invention is not used, neither any other controlling device. The line drawn shows the speed profile when the invention is used. Both cases have approximately the same volumetric airflow. An almost perfect uniformization flow was checked using the screens. TABLE 4 presents the uniformity coefficient values { γ) , the coefficients of any screen pressure (K) for the different flows and the average speeds inciding on the screen.

TABLE 4: Application of a 26/30 screen in the position P05 of the diffuser.

Rotation Coefficient Coefficient Gain Coefficient Average (rpm) (?) with (j with a (%) (K) Speed (m/s) no screen screen 2,000 0.840 0.860 2.4 2.6 5.1 3,000 0.780 0.900 15.4 2.2 8.6 4,000 0.760 0.880 15.8 2.1 9.6 5,000 0.740 0.864 16.8 2.3 11.8

The pressure drop coefficient of the screen (K) varies with the speed on the screen. However, the level of uniformization reached is not significantly affected. This fact can be observed by analyzing the uniformity coefficients { γ) with and without any screen. TABLE 5 shows the results in terms of uniformity coefficients { γ) for the various positions of the screen inside the diffuser. The test flow was constant (corresponding to 5,000 rpm in full load) in all screen positions and the average speed at the inlet of the diffuser was above lOm/s. To calculate the value of K, the speed on the screen was considered to be the speed at the entrance of the diffuser. Therefore, according to this hypothesis, the pressure drop coefficient (K) of the screen does not change. It is also important to mention that another position (P35) located 35 (±1) mm away from the entrance of the diffuser was made available for this test, and was opened to install the screen. It is important to stress that the positions are only pre-defined to make the laboratory tests easier, since the screens could be installed anywhere along the entire diffuser.

TABLE 5: Application of a 26/30 screen in different positions inside the diffuser of the catalyzer for a flow corresponding to a rotation of 5,000 rpm.

Rotation (rpm) Gain (%) Coefficient (K)

5,000 16.76 2.3

5,000 17.03 2.3

5,000 15.68 2.3

5,000 14.05 2.3

Screen Coefficient (j Coefficient (j

Position with no screen with a screen

P05 0.740 0.864

P15 0.740 0.866

P25 0.740 0.856

P35 0.740 0.844

We confirmed that the best uniformity coefficient is obtained when the screen is installed in position P15 { γ = 0.866). However, the difference presented in relation to position P05 is not substantial (0.002).

So, the area between positions P05 and P15 presents the best results in terms of uniformity coefficient { γ) for the use of a simple screen installed in the diffuser of the catalyzer under analysis. FIGURE 5 shows the result obtained when using only one 26/30-mesh screen, installed transversally (position P05) 5mm (± 1 mm) away from the entrance of the diffuser of the catalyzer. Again, the unbroken line shows the speed profile along the diameter of the catalyzer, at the outlet of the monolithic ceramic, when the screen is not used, while the line drawn shows the profile when using the screen. Both cases have approximately the same volumetric air Flow. The use of one single screen was proven to further a high uniformization to the flow. The combination of two screens was also tested in fixed positions inside the diffuser of the catalyzer for diverse flows. 30/32 and 18/32 screens were installed in positions P05 and P15, respectively. TABLE 6 presents the uniformity coefficients { γ) for the various flows. The sum of pressure drop coefficients of the screens (Ki + K₂) can also be checked. TABLE 6: Application of two screens inside the diffuser in positions P05 and P15, for diverse flows and corresponding to rotations of the motor at full load.

Rotation Coefficient Coefficient Gain (rpm) (y) with (fl with a (%) no screen screen 2,000 0.840 0.882 5.0 3,000 0.780 0.884 13.3 4,000 0.760 0.878 15.5 5,000 0.740 0.870 17.6

Coefficient (Ki) (P05) Coefficient (K₂) (P15) ΣK (Ki + K₂) 2.1 1.1 3.2 1.7 0.9 2.6 1.7 0.9 2.6 1.6 0.8 2.4

We can come to the conclusion that the degree of uniformity furthered through the use of two screens in the diffuser is usually higher than the results presented in TABLE 4 for a screen, as we had expected. In another phase of the tests, we confirmed the influence of the use of a combination of screens inside the diffuser for a flow corresponding to a rotation above 5,000 rpm. One to three screens were used inside the diffuser with a sum of K varying between 1.3 and 3.6. TABLE 7 shows the screens used and their respective flow uniformity coefficient at the outlet of the monolithic ceramic. TABLE 7: Application of combinations of screens inside the diffuser for a corresponding flow above 5,000 rpm.

Rotation (rpm) Coef. (j with Coef. (^■ with Gain (%) no screen screens

>5.000 0.660 0.820 24.2

>5.000 0.660 0.840 27.3

>5.000 0.660 0.860 30.3

>5.000 0.660 0.865 31.1

Ki P05 (screen) K₂ P15 (screen) K₃ P25 (screen) ΣK

1.3 (28/32) 0 0 1.3

2.3 (26/30) 0 0 2.3

1.3 (28/32) 1.5 (28/32) 0 2.8

1.3 (28/32) 1.5 (28/32) 0.8 (18/32) 3.6

The ideal uniformity rate { γ = 1) cannot be reached because of the limiting layer in the walls. The flow obtained with the three screens can be considered as ideal in terms of flow uniformization for an automotive catalyzer. The simple screen installed inside the diffuser with a pressure drop coefficient K relatively low (K = 1.3) provides a 24.2% gain in flow uniformization for the flow in question. The counter pressure measurements of the catalyzer were conducted using a liquid column differential pressure gage connected to static pressure inlets at the outlet of the catalyzer. The flows applied varied between 7.5xl0^~3 and 17.6xl0^~3m³/s, and covered the entire flow range corresponding to rotations between 2,000 and 5,000 rpm at full load. FIGURE 6 shows a variation of the counter pressure with a flow for some of the screen combinations used in the tests. The horizontal axis corresponds to a flow in liters per second (1/s), while the vertical axis corresponds to the counter pressure in Pascal (Pa) . The curves A, B, C, D and E are, respectively, the counter pressure without any screen, with 30/32 screens in P05 and 18/32 screens in P15, with a 28/32 screen in P05, with a 26/30 screen in P05, and finally, the counter pressure with a combination of three screens, that is, the 28/32 screen in P05, the 28/32 screen in P15, and the 18/32 screen in P25. We observed that the counter pressure does not varies much with the different screen combinations. Therefore, for flows below 9.2xl0^~3 m³/s and corresponding to rotations below more or less 2,500 rpm at full load, or even a wide range of rotations with the motor at partial load (with an accelerator aperture below 70%), the application of screens inside the diffuser of the catalyzer does not have any substantial influence on the counter pressure, nor does it hinder the motor load. This fact is probably due to the reduction in load loss of the monolithic ceramic thanks to the gain in flow uniformity, which compensates for the counter pressure imposed by the screens. From the combinations evaluated, the combination with two screens, 30/32 in P05 and 18/32 in P15, presented the lowest counter pressure values for the catalyzer, with not very substantial increases when compared with the screenless flow. This combination shows a sum of pressure drop coefficients (∑K) above that of applications with one screen, and furthers a higher gain in uniformity. The counter pressure in the catalyzer partly results from the turbulent discharge through the inlet and outlet diffusers of the catalyzer and through the screens installed. Summed to the counter pressure introduced by the turbulent discharge, the counter pressure is caused by the discharge through the channels of the monolithic ceramic. This flow is predominantly laminate, with a number of Reynolds in the range of 460 maximum at the outlet of the monolithic ceramic cells. The tests show that the non-uniformity of the flow increases with the flow increase through the catalyzer. For the flow corresponding to a 2,000-rpm rotation, the value of the uniformity coefficient { γ) is 0.840, while that for a 5,000-rpm flow has a coefficient γ of 0.740. The results obtained show that the application of screens inside the "wide angle" diffuser of a catalyzer is an efficient method to control the separation of the limiting layer. The screens, when correctly dimensioned, can completely eliminate the separation of the limiting layer inside the diffuser. This way, they uniformize the speed distribution on the monolithic ceramic (or other material) to very satisfactory rates, even in critical situations of non-uniformity, that is to say, whenever applied with a discharge of low uniformity coefficient values ( γ) . On the other hand, the use of a simple screen with a pressure drop coefficient (K) lower that the one calculated to avoid the separation in the walls of the diffuser improves the uniformity coefficient { γ) significantly in a very critical situation. Let us take for example, γ .= 0.660 in flows corresponding to rotations above 5,000 rpm, the use of a screen results in γ = 0.820. The position of the screen inside the diffuser has a certain influence on the gain in uniformity. In the diffuser tested, we obtained the best results with a screen placed between positions P05 and P15. The gain in uniformity reached the value of 17% (0.740 to 0.866) for a screen with K equal to 2.3 in position P15. Nevertheless, this fact must be analyzed with each different catalyzer, since various factors can have influence in determining the best position, such as, the dimensions of the monolithic ceramic, the dimensions of the inlet diffuser and even of the inlet and outlet tubes for the gases in the catalyzer. The combination of three screens in the diffuser presented the most significant gain in terms of uniformity with 31% (γ goes from 0.660 to 0.865) and ∑K is equal to 3.6 for flows corresponding to rotations above 5,000 rpm, the most critical flow analyzed. This condition was the closest of that calculated {Mehta , R . D. , 1977, "The Aerodynamic Design of Blower Tunnels with Wide Angle Diffusers" , Prog. Aerospace Science, v. 18, pp. 59-120) . According to is shown in FIGURE 6, the counter pressure inserted through the screens is not a significant factor, considering that the motors during most of its running time, operate at partial load regimes below 70%. In those regimes, the flows are lower than if compared with the full-load regimes for the same rotations. The results suggest that the counter pressure introduced with the screens is compensated by the flow uniformization and by the subsequent reduction in counter pressure of the monolithic ceramic. The combination of two screens in the diffuser obtained the best results, considering the balance between the counter pressure and the flow uniformity of the catalyzer analyzed. The gain in uniformity reached a value in the range of 18% (going from 0.740 to 0.870) with the screens 30/32 and 18/32, and ∑K equal to 2.4 for a flow corresponding to a rotation of 5,000 rpm. FIGURE 6 also shows the counter pressure obtained through the use of two screens, 30/32 in P05, and 18/32 in P15 (line B) and the counter pressure in the catalyzer without any screens (line A) , which indicates that there are not any substantial changes in the counter pressure. For flows corresponding to rotations above 5,000 rpm, the most significant gain obtained with two screens was of 30% with the application of a pair of screens 28/32, with ∑K equal to 2.8 and installed in positions P05, and P15, which were also tested. The improvement of the flow uniformity coefficient { γ) of the catalyzer must also be analyzed in function of the cost of the material for production and the counter pressure of the catalyzer. As a matter of fact, we strongly recommend you to consider reducing the dimensions of monolithic ceramic in function of the improvement obtained with the use of screens. The combination of three screens, when tested, obtained a higher counter pressure. This increase in counter pressure can probably be compensated by the reduction of the monolithic ceramic. We did not use more than three screens in the tests carried out, however the result in the increase in uniformity can be even greater. It is important to stress that in the "flow distributor per screens", the shape of the screen can vary, in other words, it can be flat, curved, or with have the shape of a grid (thicker) , a small "beehive" or punched plate resembling a screen, and, it can also be positioned as shown in the drawing presented in FIGURE 7, where the screen is represented by the line drawn and installed in a right angle to the line of center of the diffuser. FIGURE 8 represents a type of curved screen installed inside the diffuser. The screen can have a non-uniform mesh or one that varies over its surface, with a variable pressure drop coefficient, that can be distorted or not. The screen can also have the shape of a "casing" or "kit" for an installation inside the inlet tube of the catalyzer or even inside the diffuser, flange, or glove (provided its function is to fix the catalyzer to the exhaust system) , as shown in the drawing presented in FIGURE 9 by the drawn line. As to the material used to make the screen, it can be metallic, synthetic, or ceramic. The important thing is that this material be capable of withstanding the temperatures inside the catalyzers. So, when using screens to optimize the gas flow through the catalyzer, the following advantages are expected: a reduction in the size of the monolithic ceramic, a reduction in the cost of the catalyzer, an increase in its useful life, a reduction of the catalyzer's counter pressure, a reduction of the catalyzer's heating time, and a thermal fatigue reduction of the monolithic ceramic.

Claims

CLAIMS 1. FLOW DISTRIBUTOR WITH SCREENS TO OPTIMIZE THE FLOW

OF GASES IN CATALYZERS, CHARACTERIZED by the fact that it uses one or more screens in automotive catalyzers, or not.

2. FLOW DISTRIBUTOR WITH SCREENS TO OPTIMIZE THE FLOW

OF GASES IN CATALYZERS CHARACTERIZED by the fact that it uses devices resembling a screen like (a) punched plates;

(b) or a porous surface containing channels, similar to a beehive; (c) or a grid; (d) or a combination of the latter.

3. FLOW DISTRIBUTOR, in accordance with the claims 1 and 2, CHARACTERIZED by the fact that it uses one or more screens either (a) in the inlet diffuser of the catalyzer; or (b) in the flange; or even (c) in the inlet tube or next to it; or even (d) somewhere else in the catalyzer.

4. FLOW DISTRIBUTOR, in accordance with claims 1 and 2, CHARACTERIZED by the fact that it maintains one a or more screens placed anywhere in the inlet diffuser, from right next to the inlet source of the diffuser to the outlet, with or without the use of a "casing" or "kit" type installation.

5. USE of screens to optimize the gas flow in catalyzers, defined in claims 1 and 2, CHARACTERIZED by the fact that the screen can be separated from its fixing device or system.

6. USE of screens to optimize the gas flow in catalyzers, defined in claims 1 and 2, CHARACTERIZED by the fact that the screen does not depend on its geometrical shape.

7. USE of screens to optimize the gas flow in catalyzers, defined in claims 1 and 2, CHARACTERIZED by the fact that the screen does not depend on its physical setup; in other words, the screen can be made of any type of thread or material, that can be distorted or not, adapted to the ambient temperature inside the catalyzer.

8. USE of screens to optimize the gas flow in catalyzers, defined in claims 1 and 2, CHARACTERIZED by the fact that the screen does not depend on its angle of fixation in relation to the discharge flow.

9. USE of screens to optimize the gas flow in catalyzers, defined in claims 1 and 2, CHARACTERIZED by the fact that it is to be used in any type of combustion internal motors such as power stations, nautical, and maritime, and automotive, aeronautic units, among others.