Main SAE Technical Paper Series [SAE International Spring Fuels & Lubricants Meeting & Exhibition -...
Report a problemThis book has a different problem? Report it to us
Check Yes if Check Yes if Check Yes if Check Yes if
you were able to open the file
the file contains a book (comics are also acceptable)
the content of the book is acceptable
Title, Author and Language of the file match the book description. Ignore other fields as they are secondary!
Check No if Check No if Check No if Check No if
- the file is damaged
- the file is DRM protected
- the file is not a book (e.g. executable, xls, html, xml)
- the file is an article
- the file is a book excerpt
- the file is a magazine
- the file is a test blank
- the file is a spam
you believe the content of the book is unacceptable and should be blocked
Title, Author or Language of the file do not match the book description. Ignore other fields.
Change your answer
Downloaded from SAE International by University of Michigan, Sunday, July 29, 2018 SAE TECHNICAL PAPER SERIES 2002-01-1715 The Influence of Lambda Controller Calibration on Removal Efficiency of a Three Way Catalytic Converter Jorge Echauri R. Volkswagen de México, Engine Research and Development Reprinted From: General Emissions Research and Technology (SP–1714) International Spring Fuels & Lubricants Meeting & Exhibition Reno, Nevada May 6-9, 2002 400 Commonwealth Drive, Warrendale, PA 15096-0001 U.S.A. Tel: (724) 776-4841 Fax: (724) 776-5760 Web: www.sae.org Downloaded from SAE International by University of Michigan, Sunday, July 29, 2018 The appearance of this ISSN code at the bottom of this page indicates SAE’s consent that copies of the paper may be made for personal or internal use of specific clients. This consent is given on the condition, however, that the copier pay a per article copy fee through the Copyright Clearance Center, Inc. Operations Center, 222 Rosewood Drive, Danvers, MA 01923 for copying beyond that permitted by Sections 107 or 108 of the U.S. Copyright Law. This consent does not extend to other kinds of copying such as copying for general distribution, for advertising or promotional purposes, for creating new collective works, or for resale. Quantity reprint rates can be obtained from the Customer Sales and Satisfaction Department. To request permission to reprint a technical paper or permission to use copyrighted SAE publications in other works, contact the SAE Publications Group. All SAE papers, standards, and selected books are abstracted and indexed in the Global Mobility Database No part of this publication may be reproduced in any form, in an electronic retrieval system or otherwise, without the prior written permission of the publisher. ISSN 0148-7191 Copyright © 2002 Society of Automotive Engineers, Inc. Positions and opinions advanced in this paper are those of the author(s) and not necessarily those of SAE. The author is solely responsible for the conten; t of the paper. A process is available by which discussions will be printed with the paper if it is published in SAE Transactions. For permission to publish this paper in full or in part, contact the SAE Publications Group. Persons wishing to submit papers to be considered for presentation or publication through SAE should send the manuscript or a 300 word abstract of a proposed manuscript to: Secretary, Engineering Meetings Board, SAE. Printed in USA Downloaded from SAE International by University of Michigan, Sunday, July 29, 2018 2002-01-1715 The Influence of Lambda Controller Calibration on Removal Efficiency of a Three Way Catalytic Converter Jorge Echauri R. Volkswagen de México, Engine Research and Development Copyright © 2002 Society of Automotive Engineers, Inc. ABSTRACT MIXTURE CONTROL SYSTEM The way how the Air/Fuel mixture controller is calibrated impacts directly the performance of catalytic converters. The Air/Fuel mixture is controlled through the use of an Oxygen Sensor placed upstream of the catalyst. The output signal from this sensor is used by the Engine Control Unit to find out whether the gases represent rich or lean mixture. [reference 8] Due to the transport phenomena and feedback time constants involved, Air/Fuel ratio cannot be held static at its value of stoichiometric equilibrium. The frequency and amplitude of the resulting oscillation can be manipulated to achieve the maximum possible steady state catalytic removal efficiency. In this study, specific effects of each Lambda controller parameter are investigated, and used to determine the optimum calibration setting that results in the highest removal efficiency. INTRODUCTION The mixture’s frequency, amplitude of oscillation and the biasing from the stoichiometric point are the main variables that determine how well catalytic activity is induced. As previously empirically demonstrated, the highest frequency (1.5Hz) in combination with the smallest amplitude (0.5 A/F) is generally desireable [reference 1]. Experience suggests that Proportional Gain can be used to manipulate frequency, amplitude of oscillation and biasing. Experiments are conducted to confirm this observation and to find out how exactly the adjustments regarding these variables work. MEASUREMENTS AND CALCULATIONS A three way catalytic converter used in conjunction with a feedback controlled fuel supply system requires that some compensation action is taken over the injection system, in order to maintain the Air/Fuel mixture at a value close to stoichiometry. Otherwise, the oxidation and reduction reactions will be inhibited in the catalyst, diminishing its overall performance. A correction value is calculated by an algorithm where the Proportional Integral Controller is implemented. Such a correction is then incorporated in the basic calculation of fuel supply. The fuel supply and mixture control systems can be idealized as an air mass flow metering device which feeds an algorithm to find a matching mass fuel flow. There is always, ideally, a mass to mass ratio, given by the fixed stoichiometric property of the fuel. It can be assumed too, for modeling purposes, that air mass equals the stoichiometric value of the fuel, and that the ideal matching mass fuel flow MFE is equal to unity all the time. Eq. 1: ma = 14.8 MFE for gasoline The mass flow of fuel is then corrected with a value that represents the controller’s output as a result of the interpretation of exhaust Lambda deviations and the control actions taken by the PI algorithm. Disturbances to both inputs (like vapors from the tank ventilation system) are taken into account, assuming that the mass of air remains unchanged, and any disturbance enters the system in the form of a variation in the mass flow of fuel. So, in the end, the mass flow of fuel entering the cylinder is given by: Eq. 2: mf = MFE + mr + md Where: mf: mass flow of fuel entering the cylinder MFE: mass flow of fuel for stoichiometric equilibrium. Mr: mass flow of fuel calculated by the controller for correction Downloaded from SAE International by University of Michigan, Sunday, July 29, 2018 md: mass flow of fuel from disturbances to the system and from eq 1: ma: mass flow of air Figure 1 shows a schematic [references 2,3] of the control system described above. The simplified model shown above works under two main assumptions: First: If a rich mixture is prepared, then all of the oxygen will be consumed within the combustion chamber. On the other hand, when excess air is mixed with fuel, all of the hydrocarbons present will be burned, while there will remain unreacted oxygen. In other words, from the oxygen concentration in exhaust gases, it can be inferred whether they come from rich or lean Air/Fuel mixture, enabling the control algorithm to make decisions. Second: Due to transport phenomena, there will be a time lag between the moment at which the mixture is formed and that when the feedback shows the exhaust gas composition resulting from this mixture. This fact means that any control action taken will have a delayed impact, and that its natural state of operation is oscillatory. This time constant can be best described in terms of crankshaft degrees, and is associated with the four stroke combustion engine thermodynamic cycle and gas exchange processes within the cylinder boundaries. PI Controller calibration parameters Positive Proportional Gain Pp Negative Proportional Gain Pn Integration delay towards rich Ip Integration delay towards lean In Integration gain (not adjustable) Integration gain Table 1. Lambda controller calibration parameters The available parameters used to calibrate the PI controller are shown in Table 1 and represented in figure 2. The built-in controller consists of positive and negative proportional gains Pp and Pn which react when state transitions in the gases are found. Pp will be activated when a transition from rich to lean is found. Pn will work exactly in the opposite case. Integral gains, in this particular control algorithm, are not adjustable and their magnitude is set to 1 all the time. The integration component can only be adjusted through the calibration of positive and negative integration times named Ip and In. Ip is the time the controller waits between each integration action when the controller is adjusting towards rich because the gases show a lean composition. In works, again, in the exact opposite direction. These components can be combined to adjust the controller for amplitude and frequency of oscillation. They can impact the oscillation of the mixture as well, making it show a tendency to stay rich or lean longer, thus providing a resource that enables one to take into consideration the characteristics of a given catalyst. For descriptive purposes, variables must be created, so that a combination of parameters can be accurately described and the impact of their calibration can be well characterized. These variables are Vp and A. Eq. 3: Vp = Pn / Pp Positive to Negative Proportional Gain Ratio Eq. 4: A = Pn + Pp Sum of Positive and Negative Proportional Gains Both of these can be used to introduce a biasing to the oscillation. As a general practice, Vp has been associated to manipulation of the oscillation’s working cycle. By making Vp<1 the tendecy will be richer, Vp>1 implies a leaner control. This A parameter has been primarily used to influence the amplitude of the deviations. The following figures show the results from studies executed in order to quantify how combinations of Vp and A will impact the catalytic performance of an exhaust treatment device. EXPERIMENTAL RESULTS Cylinders Displacement (cc) Compression ratio Maximum power (kW) Maximum torque (Nm) Multiport non-sequential fuel injection Specific fuel consumption at maximum torque (gr/kW.h) Maximum mean effective pressure (kPa) Mean piston speed at max. power (m/s) Table 2. Engine Configuration 4 Opposed Air cooled 1584 7.75:1 35 @ 4690 1/min 90 @ 3160 1/min Digifant DF1.82 268 713.2 @ 3160 1/min 11.0 OMG Manufacturer MLXK5 Brand Technology Pd, Rh 11:1 ratio Active Metals 80 Loading (gr/cft) 400 Cell Density (cells/ sq in) 0.008 Cell Thickness (in) Table 3. Catalytic Converter Features Tables 2 and 3 refer to the characteristics of the engine and catalyst subject of this work. Experiments were conducted setting the engine to work at a fixed point, where the following conditions were met. Engine speed: nmot=2400 1/min Downloaded from SAE International by University of Michigan, Sunday, July 29, 2018 Intake manifold pressure: ps=780mbar (local atmospheric pressure) This set of conditions was chosen so that good repeatability would be assured between measurements, besides, due to manifold design related phenomena, an engine speed of 2400 1/min allows to have predictible and consistent flow. Improving engine stability with a reduced cycle to cycle variation rate, and a better performance of the fuel supply system. INFLUENCE THAT INTEGRATION TIMES EXERT ON RATE AND AMPLITUDE OF OSCILLATION CATALYTIC ACTIVITY AT FIXED LAMBDA The first step to characterize the performance of the catalytic converter consists of finding out how removal efficiency behaves when air/fuel ratio is kept at a fixed and steady value. From figure 3, it is obvious that the catalyst should always be operating as near as possible to the stoichiometric point where NOx and CO are removed almost completely, and HC is oxidized by more than 50%. Operating conditions other than that of chemical equilibrium will imply a penalty in reaction rates of either oxidation or reduction. [reference 4]. IMPACT OF AMPLITUDE OSCILLATION PROPORTIONAL GAIN AND FREQUENCY improvement is not completely achieved. Advantage can be taken however, considering that there is a point of the A=Pp+Pn value which makes the three lines of efficiency cross. NOx is the only component that will be more effectively removed, while CO remains virtually undisturbed, and HC removal shows a negative influence. The point where the three lines meet, gives the first hint towards successful calibration of the controller. ON OF As observed in references to some other experiments, increases in frequency when amplitude remains steady have been associated with gains in catalytic performance for the three important species, namely HC, CO and NOx. In this particular case, A = Pn + Pp was manipulated in order to discover whether a significant increase of frequency without inducing a wider oscillation threshold could be attained. The results are illustrated in figure 4, where a close relationship can be established between the magnitude of the A parameter and the frequency of the oscillation, while, the oscillation threshold, in terms of deviations from the equilibrium point, remains practically undisturbed. Hence, according to the observations, oscillation frequency could be increased essentially without the need to pay a penalty in larger deviations from stoichiometry. IMPACT OF INCREASES IN FREQUENCY ON CATALYTIC PERFORMANCE Now that it is confirmed that, as predicted, the mixture will change its state increasingly faster as the sum of proportional gains gets bigger, the next step was to confirm if this increase in frequency would have an effect on catalyst performance. [reference 5] Figure 5 shows how the converter behaves when the controller is given a higher gain to its proportional components. It becomes clear that the expected Once that it is proven that proportional gain is related to frequency, but the latter one does not necessarily improve catalytic performance, it is necessary to know if a relationship can be established between the magnitude of the integration time and the same two descriptive variables. Here, A=Pp+Pn was left fixed at a value of 40, and Pp=Pn=20, inducing no deliberate biasing. Figure 6 illustrates that frequency is sensitive to longer integration times only up to some specific point. Simultaneously, the amplitude of the deviations from equilibrium does not show any particular tendency. Based on the findings, integration times must be set to a magnitude of 4 (point where frequency reaches its peak value). This phenomenon could be explained by the changes that Proportional-to-Integral gain ratio assumes when the controller waits longer between every correction action. This peak in frequency could be the consequence of a point, at which integration times are comparable with the natural time constant due to transport effects, explained earlier. It is important to remark that for this particular engine management system, since integration times are expressed in terms of electronic synchronization periods (each one with a value of 180 crank degrees) this optimum value of Ip=In=4 should not be affected by changes in engine speed, nor changes in engine load. These events lead to the conclusion that, for the sake of simplicity, it will be considered as good practice to maintain integration times fixed during calibration, because integration gains are not available for adjustment anyway. This measure simplifies to a great extent the calibration process, because all the remaining resources available are represented by proportional gain and the two descriptive variables associated to it: Vp=Pn/Pp and A=Pp+Pn. CATALYTIC ACTIVITY AS A FUNCTION OF BIASING INDUCED THORUGH PROPORTIONAL GAIN Traditionally, this kind of PI controller has been manipulated in terms of biasing through its proportional Downloaded from SAE International by University of Michigan, Sunday, July 29, 2018 components. The data which had been demonstrated up to this point was: 1. Positive and negative integration times were both set to 4, aiming at the cancellation of any possible contribution they could have on biasing, and taking into account its confirmed effect on frequency. 2. A was set to the value that demonstrated to raise frequency to its maximum: A=Pp+Pn=40. Subsequently Vp=Pn/Pp was swept from values very close to zero up to values well past unity, introducing significant amounts of biasing in both rich and lean directions, in that order. Figure 7 leads to the discovery that, just as empirically known before, Vp=Pn/Pp is closely related to catalyst performance. This relationship lies in the fact that, making one proportional gain bigger than the other signifies the introduction of an offset to the controller, by means of altering the duty cycle of the feedback signal oscillation. This offset acts as a direct biasing of the controller giving what until now is known as a “tendency” to either rich or lean. From looking at the behavior of the efficiency plots, it becomes quite clear that this proportional gain ratio must be set somewhere between 0.8 and 0.99, but it shall never cross the line drawn by the ratio equal to 1. Such a crossing means that excess air is being fed to the catalyst, depleting the catalyst’s ability to perform NOx reduction. This excess air brings a side effect in the form of a steep negative slope in Oxygen Consumption Efficiency named ηO2 which is defined in equation 9 and illustrated in figure 7. ηO2 shows how the catalyst is able to make use of the oxygen it is fed during the lean portion of the oscillation period, in order to execute oxidation of the HC and CO during the rich phase. The second parameter introduced here is the output voltage that a binary oxygen sensor placed upstream of the monolith, shows when the converter is subject to the experiment. That voltage is known as uuhk. It is important to say that this supplementary feedback oxygen sensor is not included in this particular engine management system, and was built in only for experimental purposes. Figure 7 shows as well, how the output from this oxygen metering device can be related to HC and NOx removal. It can show, in the same way, how well oxygen is being used for catalytic elimination of undesired species. Noticing that we have already arrived to the conclusion that Vp must be set to its optimum value of 0.8-0.9, it is remarkable that, for this range of calibration, the best compromise for elimination of the three species corresponds exactly to uuhk values of 0.64 to 0.66 Volts. This is the fundamental principle used in several applications to perform monitoring of the catalyst’s working conditions. [references 6,7] MONITORING OF CATALYTIC ACTIVITY THROUGH OXYGEN CONCENTRATION BEHIND THE CATALYST A vital feature of an automotive catalytic converter is its Oxygen Storage Capacity. The role it plays resides in the need to sustain both oxidation and reduction simultaneously. This is a side effect of the natural oscillatory characteristic of the controller, which maintains the catalyst continously filling and emptying of oxygen. Hence the need for the controller to be adjusted for biasing, because depending of the catalyst’s ability to retain oxygen, a certain balance between rich and lean stages must be accomplished. It can be inferred that, if excess oxygen is found as a component in the gas downstream of the reactor, two things might be happening: either the catalyst is not being fed with enough reductants (HC and CO) and the oxygen is not fully consumed, or, Oxygen Storage Capacity is deteriorated such that HC and CO oxidation can no longer be performed. Figure 8 shows how the oxygen content in the gases behind the catalyst, as measured by an O2 sensor, is related to removal efficiencies of the main three species, and a corresponding mean value of Lambda oscillation. Based on this figure, it is apparent that a lot can be deduced from HC oxidation activity if the voltage level (uuhk) of the secondary feedback device is known. On the other hand CO shows less relationship to uuhk, and NOx seems to bear no relationship to it. The Lambda RMS value is a clear deployment of the biasing that the calibration of the PI Controller can induce in the overall characteristics of oscillation, and therefore, catalytic activity. As important as it is the ability to retain oxygen in the catalyst, it is also important the relative amount of it which is consumed throughout the process. Defining the term Oxygen Consumption Efficiency as the ratio of oxygen concentration up and downstream of the catalyst: eq. 9. ηO2 = 1- ([O2]down / [O2]up) The relationship between Oxygen Consumption Efficiency and the output voltage of the secondary feedback device is demonstrated in figure 9. Since catalyst performance is related to Oxygen Consumption Efficiency, the output voltage of a binary oxygen sensor behind the catalyst can be used as an indication of HC removal efficiency. Downloaded from SAE International by University of Michigan, Sunday, July 29, 2018 FTP75 RESULTS OBTAINED WITH THE CATALYST IN AN ACTUAL VEHICLE Prior to the analysis of the emissions data obtained through the Federal Test Procedure, it is important to highlight the most important features of the Digifant DF1.82 Engine Management System. 1. Multiport non–sequiential fuel injection. 2. Intake manifold pressure is the main engine charge signal. 3. No secondary air injection is provided. 4. No charge prediction during transitories is calculated. 5. There is no catalyst heating subroutine available. 6. Mixture compensation due to load transitions is evaluated as a function of changes in throttle position. The FTP75 test results obtained with the catalyst in an actual VW Sedán are as follows: FTP75 results obtained from the VW Sedan New catalyst, old and new calibration (gr/km) NMHC CO NOx 0.085 1.220 0.121 FTP75 Data (typical) with the previous calibration 0.060 0.750 0.070 FTP75 Data (typical) With the new calibration 0.075 2.110 0.250 TLEV Standard Table 4. FTP75 results obtained before and after recalibration of the engine management system. Special thanks are due to Mark Crocker from OMG who has guided me during the process of writing and correcting the manuscript. I would like to recognize the support given by Dirk Neumann, Javier García and Sergio Villegas from VWMTE. This work is dedicated to the memory of my Father. To my Mother and Brother. To Nayely, her eyes shine brighter than the stars at night. REFERENCES 1,8 2. 3. 4. 5. 6. CONCLUSION The use of a Three Way Catalytic Converter implies that the fuel supply mixture control system must be calibrated to maximize catalytic performance. This calibration, in the case of a proportional integral controller, can be executed by means of manipulation of positive and integral proportional gains. Integration times and gains can be left aside, since their effect on biasing, frequency and amplitude of oscillation can well be taken over by manipulation of the proportional components. Oxygen Consumption Efficiency and the output voltage of a binary O2 sensor downstream of the catalyst can be used as tools to determine catalytic removal efficiency, and have been found as a parallel criteria of calibration, directly associated to the setpoint of the proportional positive and negative gains. Failure to follow the criteria demonstrated during this investigation will lead to deficiencies in the process of optimizing the catalyst performance. ACKNOWLEDGMENTS Thanks are due to Carlos Diaz and Joaquín Hurtado from VWM-TE for the technical and inspirational guidance they have provided me. 7. Heck R., Farrauto R. Catalytic Air Pollution Control. Commercial Technology. Wiley, 1995. Pages 90 and 91. Robert Bosch GmbH. Kraftfahrtechnischestaschenbuch 19. Auflage, 1984. Page 446. Volkswagen AG. Digifant DF1.82 Datenlisting, 1995. Heywood J.B. Internal Combustion Engine Fundamentals. McGraw-Hill, 1998. Pages 301 and 572. Webb C., DiSilverio W., Weber P., Bykowsky B. Phased Air-Fuel Ratio Perturbation – A Technique for Improved Catalyst Efficiency. SAE Technical Paper #2000-01-0891. Page 7. Kumar S., Heck R. A New Approach to OBDII Monitoring of Catalyst Performance Using Dual Oxygen Sensors. SAE Technical Paper #200001-0863. Page 1. Robert Bosch GmbH. Motronic ME7.5 Funktionsrahmen DKATSP 1.130. Page 427. Downloaded from SAE International by University of Michigan, Sunday, July 29, 2018 Figure 1 Block diagram of the mixture control system Figure 2. Relationship between the feedback signal and the response from the PI Lambda Controller Downloaded from SAE International by University of Michigan, Sunday, July 29, 2018 Figure 3. Catalytic Removal Efficiency as a function of normalized Air/Fuel ratio Figure 4. Influence of proportional gain on frequency and amplitude of oscillation Downloaded from SAE International by University of Michigan, Sunday, July 29, 2018 Figure 5. The impact of the increase in frequency of oscillation in removal efficiency Figure 6. The effect of integration time in frequency and amplitude of oscillation Downloaded from SAE International by University of Michigan, Sunday, July 29, 2018 Figure 7. The effect of biasing through proportional gain ratio in removal efficiency Figure 8. Relationship between removal efficiency and the oxygen concentration behind the catalyst Downloaded from SAE International by University of Michigan, Sunday, July 29, 2018 Figure 9. Oxygen consumption efficiency and the output voltage of an O2 sensor behind the catalyst