Development of CNG Injection Engine to Meet Future Euro-V Emission Norms for LCV Applications

V S Midhun, S Karthikeyan, Nagarajan and S Krishnan

Nissan-Ashokleyland, India

2011-26-0002 Published on 19th-21st January 2011, SIAT, India

S D Rairikar, K P Kavathekar, S S Thipse and N V Marathe

The Automotive Research Association India, India
Copyright 2011 SAE International and Copyright 2011 SIAT, India

ABSTRACT
Compressed Natural Gas (CNG) is now looked upon as a leading renewable fuel for vehicles in INDIA due to mounting foreign exchange expenditure to import crude petroleum. Impending stringent emissions regulations for diesel engines, specifically exhaust particulate emissions have caused engine manufactures to once again examine the potential of alternative fuels. Much interest has centred on CNG due to its potential for low particulate and hydrocarbon based emissions and adulteration hostile nature. Signicant amount of research and development work is being undertaken in INDIA to investigate various aspects of CNG utilization in different types of engines. This paper discusses the methodology for conversion of a diesel engine to dedicated CNG engine and to make the engine to meet EURO-V norms. The primary modications are made on the piston, cylinder head, intake manifold, throttle body adaptation and exhaust system. Two different congurations like throttle body injection and multi point injection were evaluated. In the initial trials 1.1l intake manifold was used to develop the power greater than 55 kW. When the manifold volume was increased to 4.6l, there was increase in the power and resulted at torque. As engine is having two intake ports, location of injectors plays important role for getting targeted performance. Selection of ECU, development of logics and right calibration methodology resulted in meeting EURO-V norms. The injection timing, volumetric efciency estimation, deciding short term and long term fuel trims with respect to engine operations are the key factors for engine calibration. Further the metallic catalytic converter with two brick substrates of 3.8l volume was used. Consistency tests were carried out for assessing the repeatability of results and it is observed that the results are well within EURO-V emission norms with enhanced margins. 1

Keywords : CNG, EURO-V, CNG Injection, Spark Energy

INTRODUCTION
The increasing cost of petroleum-based fuels and the stringent regulations regarding limits for exhaust emissions in recent years have increased interest in alternative fuels for automotive engines. More importantly, natural gasfuelled engine has the potential for obtaining higher thermal efciency; less knocking tendency and low CO2 exhaust emissions due to its higher octane value allowing higher compression ratio operation, and lower carbon-to-hydrogen ratio. Until recently, most natural gas-fuelled engines are converted from gasoline or diesel engine. Only small fractions are developed for dedicated CNG operation. The fuel intake system has evolved since the conversions started. Initially mixer units were used, which were similar to carburetion. It was followed by injection with throttle body. The latest technology is the multi point injection, where injection happens in individual intake ports, and CNG direct injection into the cylinder. The converted engine uses the intrinsic fuel system (i.e. throttle body or multipoint port injection) to deliver fuel to the cylinder. These result in some drawbacks, mainly reduced power and limited upper speed, which are due to lower charge inhaled energy and slower ame speed respectively. One of the methods to mitigate the problems is by directly injecting natural gas into the combustion chamber. Direct Injection (DI) system can increase the absolute heating value of the cylinder charge and enhance turbulence intensity for better mixing prior to ignition. As a result, it can improve the combustion efciency for better torque and power, reduce pumping and heat losses and control the air fuel ratio of the engine more precisely.

Symposium on International Automotive Technology 2011

However, the development of new direct injection engine is costly and technically difcult to achieve within a short period of time. This is due to the needs for development of new cylinder head to acclimate with direct fuel injector and also involves tedious calibration of the engine control system. Sequential port injection (or multi-point injection) of natural gas can offer an immediate solution for the drawbacks of CNG converted engine. NG is injected by individual injector at each cylinder intake manifold just before the opening of intake valve. Better control of mixture formation and response to changing speed can be achieved. Thus, it provides the opportunity to reduce the negative effects on the performance compared to single injector manifold injection. This paper discusses the results obtained from experimental investigation of the sequential injection natural gas engine with respect to performance and exhaust emissions. Emission limits for CNG engine is shown in the Fig. 1.

Considering BMEP (brake mean effective pressure) of developed CNG engine it was recommended to use the asymmetric barrel type compression ring. The key stone ring used in diesel engine was not recommended considering its scraping operation due to low clearance and high temperature generated in CNG engines.

Combustion Chamber
Combustion chambers for CNG engine is designed based on parameters like turbulence, swirl and squish. As the existing shape of the combustion chamber in the diesel engine is not suitable, some additional changes to be made for the combustion chamber design. For natural gas due to high activation energy, the laminar ame speed is low which results in longer combustion duration. Thus the total combustion period becomes prolonged as compared to diesel and petrol. This leads to loss in efficiency. The longer the total combustion time, higher chance of the remaining unburned mixture to undergo pre-ame reactions and self ignite. Fuel burning before top dead centre increases the work required for compression, while that of burning late in the cycle performs less work on the piston during expansion. This problem of low ame speed can be solved by two ways. The rst is that the ignition timing could be advanced. The other option is the development of combustion chamber specically for natural gas operation by increasing turbulence and squish in the combustion chamber to increase ame speed. A squish motion combustion chamber has its effects on the burning rates. High levels of turbulence generated from the squish cause faster burning rates, which result in improvement in thermal efciencies. In the present engine the combustion chamber design is based on the above principle.

PM (g/kWh)

Figure 1. Emission Potential of Developed CNG Injection Engine

Intake Manifold
High volume intake manifold is introduced for the naturally aspirated CNG version instead of low volume manifold which is used for the turbocharged diesel application. This is used for mounting of the injectors in the MPFi version of CNG injection. Intake manifold is shown in Fig 2.

ENGINE COMPONENTS MODIFICATIONS FOR CNG Piston

Piston bowl is modied from existing diesel CRDi shape to the deep bowl shape with sufcient crown thickness to increase the bowl volume. The compression ratio is decreased to 12:1. Piston and ring pack is one of the major contributing components to achieve desired performance and emission. Piston and ring pack is reviewed from the below following point of views: to minimize dead volumes to minimize friction losses to minimize oil consumption to improve durability

Figure 2. Intake Manifold for Throttle Body Injection

Symposium on International Automotive Technology 2011

Base engine is mapped with existing diesel engine manifold and multipoint injection system. It is observed that due to charge distribution issue, engine was developing much less power than targeted of 57.6kW @ 3800 rpm and torque 192.14 Nm@ 1200 rpm. It was decided to resolve problems of charge distribution by designing a new manifold. Subsequently the new manifold with individual runner is designed considering constraints of given below: Packaging, injector position Uniform distribution of air to all 4 cylinders Minimum possible resistance in the runners Intake geometry suggestions in a such a way to improve induction process Further the elimination of eddies, in the manifold provides effective charge distribution. The newly designed manifold is having individual runners for each cylinder and having total volume increased as compared to diesel asymmetrical manifold. Figure 3. AVL BOOST 1D Schematic Model

INTAKE MANIFOLD COMPUTATIONAL ANALYSIS Steady State Analysis

The main aim of steady state analysis is to nd the pressure drop across individual runners. The pressure drop across runners helps in identifying the ow structure within the intake manifold. This information can be obtained from a steady ow test (ow bench) too, but purpose of this kind of simulation is to be ready for unsteady simulation of Intake Manifold (IM). Steady state study can be fast and can provide the loss coefcients but this information cannot provide any information about an IM performance in the operating situation. The Boundary Conditions (BC) in steady state simulation are constant pressure.

Figure 4. Intake Manifold for Multi Point Fuel Injection

Exhaust System
Exhaust outlet pipe is introduced with a provision to mount the narrow band lambda sensor .Further exible bellow is provided between the outlet pipe and catalytic converter to reduce the stress due to vibrations.

Unsteady State Analysis

Flow through an intake manifold is dependent on the time since crank angle positions vary with respect to time. Unsteady state simulation can predict how an IM work under real conditions. Fig. 3 shows schematic 1D model. Fig. 4 shows intake manifold for multi point injection system. The boundary conditions are not longer constant but time limit variant. These boundary conditions were obtained from 1D analysis using AVL BOOST software. CFD analysis of individual runner type manifold was carried out to assess the CNG mixing. It is observed that CNG injector position directing valve gives better mixing of CNG with air at the upper area. The additional charge ows through intake port causing the engine to give the desired power with use of manifold with increased volume. 3

Cylinder Head and Swirl

The cylinder head is modied to incorporate the spark plug in the same location previously occupied by the diesel CRDi injector. The spark plug used is M12 size due to packaging constraints. One of most critical aspect of the cylinder head which decides the engine performance is the port and in turn the swirl, i.e. both inlet ports consequently generating the swirl. To start with any CNG engine development task, it is necessary to evaluate the status of the performance of both inlet ports. Hence the baseline port performance is carried out to assess the condition. The existing inlet ports (same as Diesel) exhibit a higher mean swirl value resulting in less co-efcient of ow. It is also observed that both the inlet ports are generating swirl. Fig. 5 shows cylinder used for CNG engine development.

Symposium on International Automotive Technology 2011

depression rises until there is a reversal of conditions within the cylinder and manifold, and, in fact, the mean depression in the induction manifold may, at the end of the exhaust stroke, be greater than in the cylinder. Under these part throttle conditions, some of the exhaust gas escaping from the cylinder will not only go out of the exhaust port, but will also be drawn back through the induction port into the induction manifold where it originated before being burnt. Figure 5. Cylinder Head for CNG Engine

CNG Injectors
The CNG injectors are integrated on the runner of the intake manifold. Trials were done to decide on the position of the injector with respect to the ports. It is found that the injection towards valve along with air ow is better as compared to other location. Fig. 6 shows injector position in the manifold.

Cam Shaft and Valve Overlap

The benets of delaying the exhaust valve closing while opening the inlet valve earlier, such that both inlet and exhaust valve opening periods overlap each other, are better cylinder clearing and lling in the mid to upper speed range of the CNG engine. The improvement in cylinder volumetric efciency owing to the extended valve overlap is caused by the high exit velocity of the exhaust gases establishing a depression in the exhaust port and manifold branches. This greatly assists in drawing in fresh air or air-fuel mixture, for diesel and petrol engines respectively, from the induction manifold even before the piston has completed its exhaust stroke. In the ineffective piston-stroke region the piston is not able to perform as a vacuum pump. Unfortunately, the advantage of opening the inlet valve early and closing the exhaust valve late has various detrimental side effects which are not compatible with the minimization of exhaust pollution. Exhaust valve closing lag induces a fresh charge to enter the combustion chamber and cylinder during the ineffective part of the piston stroke at the end of the exhaust stroke and the beginning of the induction period when the engine is running in the higher speed band. However, as the engine speed is reduced, some of the fresh charge will not only enter the cylinder but will actually be carried out with the fast moving burnt gases into the exhaust system. The loss of fresh charge to the exhaust will become more pronounced as the exhaust valve lag is extended and the engine speed is reduced. Consequently, this will show up as an increased amount of unburnt and partially burnt exhaust gases including larger quantities of hydrocarbon and carbon monoxide being present in the exhaust composition. The inlet valve opening lead provides an opportunity for fresh charge to commence entering the cylinder early, provided there is a difference of pressure across the partially opened inlet valve sufcient to force the fresh charge into the combustion chamber and cylinder space. This is not possible with a wide open throttle but it is also an effective way of initiating the beginning of induction. However, as the throttle opening is progressively reduced, the manifold 4

Figure 6. CNG Injector and its Position

CNG Rail/Splitter
The CNG after regulation is passed through the splitter which houses the low pressure lter, the integrated pressure and temperature sensor. Splitter has one input and four outputs (one to each of the injector).

CNG High Pressure Line Components

The high pressure regulator reduces the high pressure CNG to 2.3 bar differential pressure with reference to the engine vacuum pressure. The high pressure lter is of coalescent type is installed in the high pressure line prior to the regulator to remove oil from the compressor which can be mixed with CNG.

Optimization of Spark Energy

The changeover to a resistive-type of spark plug introduces an additional resistance in the ignition circuit, causing a decrease in the spark energy. This tends to increase the mass emissions of HC and CO compared to the values obtained with the non-resistive spark-plugs. In order to maintain the emissions at the target levels, the spark energy is therefore boosted by suitably increasing the coil-on-time (dwell time) setting in the Igniter ECU, thereby increasing the output of the H.T. coils, and ensuring complete combustion of the CNG fuel.

Symposium on International Automotive Technology 2011

Electronic Management System

Fig. 7 shows Electronic control Unit. The ECU takes the input from several sensors and calculates the air, fuel and ignition based on the different control strategies and maps/ tables. The system works on the drive by wire mechanism were the input from the driver based on the accelerator pedal is taken and the value is added to other torque requirements. This is translated to an airow requirement. The electronic throttle is controlled to input the required amount of air and feedback is taken based on the TMAP sensor values.

The ignition timing and dwell angle are also controlled based on the system requirements. The Maximum Brake Torque (MBT) timing is mapped, the timing is modied for optimum timing considering the torque request, catalyst temperature and other requirements.

Catalytic Converter Loading Assessment

Several improvements were concurrently made in the exhaust system like reducing the length of exhaust pipe between mufer and exhaust port .The reduction in the length of the exhaust piping also facilitated catalytic converter light up at low speed and load desirable for HC reduction. This resulted in a reduction of the exhaust back-pressure, thereby boosting the full-throttle torque output of the engine and helping to improve the specific fuel consumption. Two samples of Catalytic Converters of different loading and size were evaluated, however it is observed that 1.9 liter Palladium based Cat Con showed better performance than 1.6 lit palladium based Cat-con.

RESULTS AND DISCUSSIONS

From the Fig. 8 to 13 it is seen that ow in runner outlet at 1, 2, 3 is uniformly distributed across the cross section. However, ow is comparatively less through runner 1, since its location is farthest from the air inlet. Flow in runner outlet 4 is concentrated in the upper half across the cross section. This is attributed to the fact that, there is a sudden diversion of the incoming ow and by geometry hence; it is not reaching the inner corners of the cross section. However ow is good owing to nearest vicinity of runner 4 to the incoming air from air lter.

Figure 7. Electronic Control Unit

Calibration Methodology involves the followings:

Sensor characterization Actuator characterization Volumetric efciency calibration with lambda sensor Relative AMPC calibration for torque output Accelerator pedal mapping Engine speed governing Injection timing calibration Ignition timing calibration Dwell control calibration

Figure 8. 3D CFD Simulation- TBI Conguration

Idling calibration included following:

Drag torque calibration Individual part Alternator, power steering, vacuum pump torque calibration (presently dummy hence zero) Idle PID calibration Based on the air ow, the fuel ow is controlled for the required lambda value for the condition of operation. Based on the same the injection pulse width is calculated. Further the injection timing is also mapped. 5

Figure 9. 3D CFD Simulation- Manifold Inlet

Symposium on International Automotive Technology 2011

From Fig. 15 for the 30 CNG injector position, the mixing of CNG with air is initated at the upper area and expected to ll-up the cross section as it ows through intake port. Engine is giving the desired power.

Figure 10. Flow Distribution at Runner Exit 4

Figure 11. Flow Distribution at Runner Exit 3

Figure 15. 3D CFD Simulation- MPFI Conguration at Runner 4 and 3 From Fig. 16, Flow pattern at each runner outlet is fairly uniform. The wavy nature is due to wave dynamics inside the runner which acts as a pipe. Average deviation in ow at runner no. 1, 2 and 3 with respect to runner no.4 is 5.98%. Average mass ow rate of the mixture for the major portion of positive ow during the cycle is 0.1711 Kg/s. This improvement over the baseline simulation (0.1678 Kg/s) is due to shifting from TBI to MPFI conguration. TBI may reduce the airow due to interaction of incoming air with CNG at the inlet of the intake manifold. Further engine is showing slight improvement in power in MPFI conguration as compared to TBI one.

Figure 12. Flow Distribution at Runner Exit 2

Figure 13. Flow Distribution at Runner Exit 1 From Fig. 14 it is seen that the ow pattern at each runner outlet is fairly uniform. The wavy nature is due to wave dynamics inside the runner which acts as a pipe. Average deviation in ow at runner no. 1, 2 and 3 with respect to runner no.4 is 9.17%. Average mass ow rate of the mixture for the major portion of positive ow during the cycle is 0.1678 Kg/s. Figure 16. Flow Rate at Runner Using AVL BOOST for MPFI Conguration Fig. 17 shows normalized brake torque and brake power of the engine with WOT from 1200 to 3800 rpm. It is observed that low end at torque with modied intake manifold which is preferred in city driving condition. Figs. 18, 19, 20 and 21 show the emission traces during ETC with different pollutants and its contribution in the mass emission result. NOx emission is slightly higher at rated load condition but due to catalyst efciency it is reduced in the remaining transient cycle. HC and CO very well oxidized with the help of palladium catalyst. 6

Figure 14. Flow Rate at Runner Using AVL BOOST for TBI Conguration

Symposium on International Automotive Technology 2011

Figure 17. Normalized Power Torque Curve for MPFI Conguration with Modied Intake Manifold

Figure 21. CO-ETC Traces for MPFI Conguration with Modied Intake Manifold The Fig. 22 shows that mass emission test results for MPFI Conguration with modied Intake system are within the tremendous margin. Table 1 Shows the consistency of emission results and results are with signicant margin for Euro-V.

Figure 18. THC-ETC Traces for MPFI Conguration with Modied Intake Manifold

Figure 22. Emission levels for MPFI conguration with Modied Intake Manifold Figure 19. NOx-ETC Traces for MPFI Conguration with Modied Intake Manifold Table 1. Emission Consistency Results with Modied Manifold for MPFI Conguration

Figure 20. CH4-ETC Traces for MPFI Conguration with Modied Intake Manifold 7

Symposium on International Automotive Technology 2011

CONCLUSION
Engine Meeting Euro-V Emission Norms required the variation in the instantaneous air-fuel ratio to be minimized, such that the air-fuel ratio is maintained at all times within a narrow band centred around stoichimetric condition. Lambda operating range was effectively controlled by using modied intake system. The present air-fuel system achieved this tight control over the instantaneous air-fuel ratio by means of injection fuel system with optimized setting for the injector conguration ,the ECU control Map, and the Ignition timing. The optimized Fuel system enabled compliance with Euro-V emission norms, with comfortable margins. If application is based on future electric based vehicles then electronic throttle is required. But for dedicated CNG trucks, single point injection, manual throttle is sufcient. Further optimization is possible to reduce the catalytic converter loading for cost reduction and performance consistency.

5.

Fino D, Russo N, Saracco G and Spechia V, CNG Engines Exhaust Gas Treatment via Pd- Spinel-TypeOxide Catalysts, Catal. Today, 117: 559-563, 2006

ACKNOWLEDGMENT
The authors would like to thank Mr. Shrikant R Marathe, Director- ARAI for his support and encouragement during experimental trials at ARAI. The authors would like to thank to Mr. P P Chitins, Asst Director and Mr. Amit Tyagi, Project Engineer for their support on simulation during experimental trials at ARAI. The authors would also like to thank colleagues from Nissan Ashok Leyland Tech Ltd. and Mr. Vishal Singhal and Mr Sandeep Suhag from Advantek Fuel System for their co-operation and encouragement.

CONTACT
V. S. Midhun Manager, PD- Engines Nissan Ashokleyland- Technical Center, Chennai TamilNadu, India [email protected] Cell No- 9094028717 S. Karthikeyan Dy-Manager, PD- Engines Ashokleyland- Technical Center, Chennai TamilNadu, India [email protected] Cell No- 9710447235

REFERENCES
1. Sera, M.A., R.A. Bakar and S.K. Leong, CNG engine Performance Improvement Strategy Through Advanced Intake System, SAE Technical Paper No. 2003-01-1937, 2003 Czerwinski J, et al., Sequential Multipoint Trans-ValveInjection for Natural Gas Engines, SAE Paper No. 1999-01-0565, 1999 Aesoy, V and Valland H, Hot Surface Assisted Compression Ignition of Natural Gas in a Direct Injection Diesel Engine, SAE Paper No. 960767, 1996 Kavathekar K P, Rairikar S D and Thipse S S, Development of a CNG Injection Engine Compliant to Euro-IV Norms and Development Strategy for HCNG Operation, SAE Paper No. 2007-26-029, 2007

2.

3.

4.

The Technical Paper Review Committee (TPRC) SIAT 2011 has approved this paper for publication. This paper is reviewed by a minimum of three (3) subject experts and follows SAE guidelines. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form by any means, electronic, mechanical photocopying, recording, or otherwise, without the prior written permission of SIAT 2011.

Positions and opinions advanced in this paper are those of the author(s) and not necessarily those of SIAT 2011. The author is solely responsible for the content of the paper. The Automotive Research Association of India Survey No. 102, Vetal Hill, Off Paud Road, Kothrud, PUNE 411038. Maharashtra, (INDIA) Tel.: +91-20-3023 1111 Fax: +91-20-2543 4190 Email: [email protected] www.araiindia.com Printed in India