Diffusive Combustion of Ethanol in a Dual-Fuel Direct Injection Compression Ignition Engine

(1)

Doctoral Thesis in Machine Design

Diff usive Combustion of Ethanol in a Dual-Fuel Direct Injection

Compression Ignition Engine

NICOLA GIRAMONDI

Stockholm, Sweden 2021 www.kth.se

ISBN 978-91-7873-893-9 TRITA-ITM-AVL 2021:25

KTH ROYAL INSTITUTE OF TECHNOLOGY

NICOLA GIRAMONDI Diff usive Combustion of Ethanol in a Dual-Fuel Direct Injection Compression Ignition EngineKTH 2021

(2)

Diffusive Combustion of Ethanol in a Dual-Fuel Direct Injection

Compression Ignition Engine

NICOLA GIRAMONDI

Doctoral Thesis in Machine Design

with specialization in Internal Combustion Engines KTH Royal Institute of Technology

Stockholm, Sweden 2021

Academic Dissertation which, with due permission of the KTH Royal Institute of Technology, is submitted for public defence for the Degree of Doctor of Philosophy on Friday the 11th June 2021, at 10:00 via Zoom webinar

(3)

Copyright pp 1-93 Nicola Giramondi, 2021 Publication I © 2019 SAE International Publication II © 2020 SAE International Publication III © 2021 SAGE Publications Publication IV © 2021 SAE International ISBN 978-91-7873-893-9

TRITA-ITM-AVL 2021:25

Printed by: Universitetsservice US-AB, Sweden 2021

(4)

Nomenclature

AFR Air-fuel ratio

ARoHR Apparent rate of heat release BDC Bottom dead center

BEV Battery electric vehicle CAD Crank angle degree

CAx Crank angle degree at x% apparent heat released CFD Computational fluid dynamics

CI Compression ignition

CO Carbon monoxide

CO2 Carbon dioxide COV Coefficient of variation ED95 Ethanol based fuel EGR Exhaust gas recirculation EVO Exhaust valve opening FGM Flamelet generated manifold

HC Unburned fuel

HCCI Homogeneous charge compression ignition HCV Heavy commercial vehicles

HRR Heat release rate

ICEV Internal combustion engine vehicle IMEP Indicated mean effective pressure IVC Intake valve closure

KH Kelvin-Helmholtz

LCV Light commercial vehicles LHV Lower heating value

LTC Low temperature combustion MCV Medium commercial vehicles

MM Molar mass

MPR Main-pilot ratio MPS Main-pilot separation NEB Net energy balance NO_x Nitrogen oxides

(7)

Ox Fuel oxygenate fraction PM Particulate matter

PPCI Partially premixed compression ignition PPRR Peak pressure rise rate

PSR Perfectly stirred reactor

RANS Reynolds-averaged Navier-Stokes

RCCI Reactivity controlled compression ignition RF Response factor

RT Rayleigh-Taylor SD Standard deviation SOC Start of combustion SOI Start of injection SR Substitution ratio

TKE Turbulence kinetic energy TDC Top dead center

(8)

Abstract

The impact of climate change due to global warming necessitates rapid and extensive measures to enhance the sustainability of the energy and transport sectors.

In this context, there are large environmental and societal benefits to be gained by replacing diesel with renewable fuels for road freight transport. This solution may facilitate and expedite the transition towards fossil-free, carbon-neutral transport, while the electrification process takes shape. Short-chain alcohol fuels have favorable properties for the enhancement of engine performance and the abatement of pollutant emissions, however, they necessitate ignition aid systems in compression ignition engines. The present research investigates a novel concept of dual-fuel combustion for heavy-duty compression ignition engine applications by means of engine tests and three-dimensional combustion simulations. This concept involves the direct injection of pure ethanol as main fuel through a centrally mounted injector, and minimal quantities of diesel as pilot fuel via a separate injector. The objective is to achieve diffusive combustion of ethanol in a process analogous to conventional diesel combustion throughout the entire engine load range, with a higher thermal efficiency and lower pollutant emissions. Single-cylinder engine tests were carried out to evaluate the influence of combustion characteristics and performance with respect to dual-injection strategy, engine load, ethanol ratio and configuration of the diesel pilot injector. The characteristics and performance of ethanol-diesel direct injection compression ignition (DICI) combustion were compared to two sets of baselines, that are conventional diesel combustion and dual-injections of diesel via the main and pilot injector in the same proportion as in the dual-fuel test points. At low load conditions, increasing the separation between the diesel pilot and ethanol main injection enabled the achievement of diffusive combustion of ethanol, avoiding combustion instability and partial misfire thanks to minimal quantities of diesel injected. At high load conditions, a minimum main-pilot separation was instead required to limit the degree of ethanol premixing at ignition. Using a diesel pilot injector having a lower number of sprays with a wider hole diameter promoted a more robust ignition of ethanol, while also causing a reduction of engine performance. Parallel to the experimental work, three-dimensional combustion simulations were carried out in order to investigate the interaction between diesel and ethanol sprays during ignition at various engine operating conditions, from low to full load. At the operating conditions investigated during engine tests, the

vii

(9)

ignition of a subset of ethanol sprays was locally triggered by the contact with the products of diesel combustion. Subsequently, ignition propagated towards the neighboring ethanol sprays, until reaching the furthest ones from the diesel pilot injector. The coupling between experimental and numerical results highlighted the noteworthy predictive capability of the adopted combustion model with respect to the ethanol combustion characteristics. In conclusion, the present research work provides a solid starting point for future studies on diffusive combustion of alcohol fuels in compression ignition engines. The structured knowledge built in the course of the doctoral project lays the foundation for the development of a fuel-flexible engine for heavy-duty applications.

(10)

Sammanfattning

Klimatförändringarna, som en följd av den globala uppvärmningen, gör det nödvändigt med effektiva och omfattande åtgärder för att göra transport- och energisektorn mer hållbara. Både ur miljö och socialt perspektiv, finns det stora fördelar med att ersätta diesel med förnyelsebara bränslen för godstransport på väg. En sådan lösning underlättar och påskyndar övergången till fossilfri, kolneutral transport, under tiden elektrifieringen av transportsektorn utvecklas. Korta alko- holer som biobränsle har fördelaktiga egenskaper som kan öka motorprestandan samt minska utsläppen av hälsovådliga avgaser. Dock är det i kompressionsantända motorer nödvändigt med hjälpmedel för antändning av dessa bränslen. Detta projekt undersöker ett nytt koncept med förbränning av dual-fuel för framtida användning i tunga kompressionsantända motorer. Detta undersöks genom motorexperiment av konceptet, samt tredimensionell simulering av förbränningsförloppet. Koncep- tet innefattar direktinsprutning av etanol, som huvudbränsle, genom en centralt monterad insprutare. I kombination med ytterst små mängder av Dieselbränsle som pilotbränsle genom en separat insprutare. Målet är att uppnå diffusiv förbrän- ning av etanol i en förbränningsprocess, likt den av konventionell diesel, för alla lastpunkter med en ökad termisk verkningsgrad och låga utsläpp av föroreningar.

Motorprov i en en-cylindrig motor genomfördes för att utvärdera förbränningsegen- skaperna och prestanda med avseende på en dual-fuel strategi, last, etanolmängd, och dieselinsprutarens konfigurering. Förbränningsegenskaper samt prestanda av etanol-diesel direktinsprutad kompressionsantänd (DICI) förbränning jämfördes med konventionell dieselförbränning samt dubbel insprutning av diesel i samma konfiguration och mängd som etanol-diesel insprutningen. Vid låga laster gjorde en ökad separation mellan dieselpiloten och den huvudsakliga etanolinsprutningen att diffusiv föorbränning av etanol lättare kunde uppnås. Detta gjorde att förbrän- ningsinstabilitet och missantändning kunde undvikas då ytterst små mängder av diesel injiceras. Vid höga laster behövdes en kort separation mellan de två olika bränsleinsprutningseventen för att begränsa förblandningen av etanol innan an- tändning. En dieselinsprutare med färre strålar och med större håldiameter gav en mer robust förbränning av etanol, men det försämrade också motorns prestanda.

Parallellt med det experimentella arbetet genomfördes tredimensionell simulering av förbränningen där interaktionen mellan diesel- och etanolstrålarna vid antändning undersöktes vid olika lastförhållanden, från låg last till full last. Vid de driftsförhål-

ix

(11)

landen som undersöktes genom motorexperiment visade simuleringen att antändning av etanolen delvis triggades genom lokala interaktioner med förbränningsprodukter från dieselförbränningen. Därefter propagerade förbränningen till de närliggande etanolstrålarna och så vidare tills att förbränningen propagerat även till etanolstrå- larna längst bort från dieselinsprutaren. Kombinationen mellan experimentella och numeriska resultat framhäver den anmärkningsvärda prediktiva kapaciteten av den antagna förbränningsmodellen med hänsyn till förbränningsegenskaperna hos etanol.

Sammanfattningsvis, projektet ger en stabil startpunkt till framtida studier av diffusiv förbränning av alkoholbränslen i kompressionsantända motorer. Kunskapen som utvecklats under doktorandprojektets gång ger en grund för framtida utveckling av tillämpning av flexi-fuel i tunga förbränningsmotorer.

(12)

List of Publications

This thesis is based on the four appended publications in the list below.

Publication I

Giramondi, N., Mihaescu, M., Christiansen Erlandsson, A., Jäger, A., CFD-Driven Preliminary Investigation of Ethanol-Diesel Diffusive Combustion in Heavy-Duty Engines, SAE Technical Paper 2019-01-2192, 2019, doi:10.4271/2019-01-2192 Publication II

Giramondi, N., Jäger, A., Mahendar, S.K., Christiansen Erlandsson, A., Combustion Characteristics, Performance and N O_x Emissions of a Heavy-Duty Ethanol-Diesel Direct Injection Engine, SAE Technical Paper 2020-01-2077, 2020, doi:10.4271/2020- 01-2077

Publication III

Giramondi, N., Jäger, A., Norling, D., Christiansen Erlandsson, A., Influence of the diesel pilot injector configuration on ethanol combustion and performance of a heavy-duty direct injection engine. International Journal of Engine Research, 2021, doi:10.1177/14680874211001260

Publication IV

Giramondi, N., Konstanzer, D., Christiansen Erlandsson, A., Evaluation of the Ethanol-Diesel Spray Interaction during Ignition in a Dual-Fuel DICI Engine Using an Experimentally Validated CFD Model, SAE Technical Paper 2021-01-0521, 2021,

doi:10.4271/2021-01-0521

The publication listed below is not appended to the thesis.

Publication V

Mahendar, S.K., Giramondi, N., Venkataraman, V., Christiansen Erlandsson, A., Numerical Investigation of Increasing Turbulence through Piston Geometries on Knock Reduction in Heavy Duty Spark Ignition Engines, SAE Technical Paper 2019-01-2302, 2019, doi:10.4271/2019-01-2302

xi

(13)

The conference paper listed below was selected for publication in the In- ternational Journal of Engine Research, leading to Publication III after an additional review process. The conference paper is not appended to the thesis.

Giramondi, N., Jäger, A., Norling, D., Christiansen Erlandsson, A., Influence of the Diesel Pilot Injector Configuration on Ethanol Combustion and Performance of a Heavy-Duty Direct Injection Engine, 2020. THIESEL 2020 Conference on Thermo-and Fluid Dynamics Processes in Direct Injection Engines.

(14)

1 Introduction

1.1 Transition towards future transport technologies

This thesis is being written in a time of rapid and unforeseen changes in the world stage. For this reason, the projections that will be referenced hereon will have to be revised based on near-future developments. Nonetheless, the author wishes to provide a view on a sustainable way forward leading to a lower dependence of the transport sector on fossil fuels, as well as to a lower energy and carbon intensity.

Before the events of 2020, a steady economic expansion was predicted to occur until 2040, leading to an increase in the global gross domestic product up to double the level for 2017 [1]. Parallel to the economic expansion, the worldwide energy demand is expected to grow by around 20%. Under this scenario, the subsequent increase in the global energy consumption need to be counterbalanced by strong measures for enhancing energy efficiency, in order to reduce greenhouse gas (GHG) emissions and to limit global warming within 2 degrees compared to pre-industrial levels [2]. This increment was identified as the critical threshold to mitigate the risk of adverse consequences caused by climate change [3]. Considering the worldwide energy matrix for electricity production, renewable sources are expected to double the capacity installed in 2017 by 2040. Nonetheless, the overall penetration of renewable energy sources for electricity production would stop at around one-fourth of the total capacity, given the predicted growth in the global electricity demand [1].

In order to decarbonize power generation, "you have to run very fast just to stand still" [4]. The economic expansion depicted above would be followed by a parallel increase in the world population by around one quarter [1], thereby leading to an average growth in the individual purchasing power. The volume of commercial and private vehicles sold has been historically correlated to economical indicators like the gross domestic product [5] and purchasing power [1], and therefore, it is expected to increase.

The synergy between transport electrification and the decarbonization of the energy matrix has the potential to enhance transport sustainability [6]. The effec- tiveness of this technology transition may vary between world regions depending on the local energy sources for electricity generation, while its pace is contingent on the diverse requirements of the transport segments. As an example, the electrification of the light-duty vehicle fleet is expected to cause an increase in the global electricity

1

(15)

2 Chapter 1. Introduction

demand – up to 15% under the scenario where all light-duty vehicles were electric by 2040 [2], prompting the need for a reinforcement and expansion of the the electricity grid. The evaluation of the CO₂emission savings enabled by transport electrification necessitates comprehensive life cycle assessments, accounting for the the carbon intensity of battery electric vehicle (BEV) production and the characteristics of the local power matrix. Producing a battery electric car is estimated to generate up to 30%-50% higher CO₂ emissions compared to the production of an internal combustion engine vehicle (ICEV) [7]. With respect to the overall well-to-wheel CO₂ emissions, the relative gain achieved with a battery electric vehicle depends on the power sources composing the local energy matrix. Kawamoto et al. [8] highlighted that a minimum driving distance is required for a BEV to enable an overall CO₂ emission reduction, compensating the higher emission levels connected to the production of a BEV compared to an ICEV. In countries like Japan, China and Australia, the projected CO2 emissions at the BEV end of life would be comparable or potentially higher than an ICEV [8], without a change of the local power matrix.

A similar study was carried out by Casals et al. [9], comparing different driving cycles and accounting for the power matrix of several European countries. The penetration of BEVs in the automotive market is expected to lead to a substantial benefit in terms of global warming mitigation in countries with low levels of equivalent CO₂ emitted per kWh of electricity produced. This is the case for countries like France, Norway and Sweden, with a high penetration of renewable or nuclear power sources in their power matrix. On the other hand, European countries with high GHG emissions per kWh of electricity produced should decarbonize their energy generation mix in order for transport electrification to enable substantial environmental benefits.

A key enabler for transport electrification is represented by the advancement of battery technology and the subsequent impact on the price of the battery electric vehicles. As an example, producing a battery electric car may become less expensive than an ICE car [10] if the price for the battery pack will decrease below 100$ per kWh from the current level of 137$ [11]. The spread of battery electric vehicles will drive an increase in the demand for Li-ion batteries raw materials – that are not widespread: in fact, only 32 countries in the world contribute to their production [12].

Moreover, lithium, cobalt and especially graphite are assessed to have a relevant supply-risk.

With regard to heavy-duty transport, technical limits may hinder or slow down the transition towards BEVs. As exemplified by Kalghatgi [13], a heavy-duty Class 8 truck would require a battery with a capacity of around 1000 kWh in order to cover a range of 500 miles. A battery with such features would weigh more than four times compared to a suitable internal combustion engine, while costing more than the whole ICE truck – assuming a battery pack cost of 125$ per kWh. Hence, it is argued that a different energy storage technology is needed for commercial transport [13].

The penetration of BEVs in the market of light commercial vehicles (LCVs) is expected to occur at a higher pace compared to the other commercial vehicle

(16)

1.2. The role of bioethanol 3

branches. Under a scenario assuming an entirely electric LCV fleet by 2040, with a penetration of alternative fuels in the medium (MCV) and heavy commercial vehicle (HCV) market of 70% and 20%, oil would still account for 60% of the energy consumption by road freight transport [1, 14]. This is because HCVs account for the absolute majority (55% in 2015) of the fuel consumption of the road freight fleet, despite being only a minor fraction (15%) of the total number of commercial vehicles [1].

1.2 The role of bioethanol

As outlined in the previous section, conventional engine technology is likely to play a relevant role during the transition towards electrified powertrains. Increasing the penetration of alternative fuels with high energy capacity, such as liquid biofuels, is a relevant factor for enabling the decrease of fossil fuel consumption by heavy-duty transport [1, 5]. As of 2016, ethanol was the biofuel with the highest production capacity of 100 billion liters per year [5]. When additized with lubricity and ignition improvers (ED95), ethanol becomes suitable for replacing diesel in compression ignition engines. However, ethanol-based fuels are currently used in a limited share of commercial vehicles – mostly LCVs [5]. The most widely used biofuel for road freight transport is biodiesel, accounting for 1.6% of the total fuel consumption, compared to 0.6% of ethanol as of 2017 [5].

In 2015, road freight vehicles were responsible for one-third of CO2emissions from transport, corresponding to 7% of the global CO2 emissions [5]. As previously mentioned, despite accounting for a limiting number of commercial vehicles, the HCV segment accounts for more than half of their total fuel consumption. The benefit in terms of CO₂emission reduction driven by a higher penetration of biofuels among road freight vehicles is dependent on the fuel’s carbon footprint, that should be evaluated based on a life cycle analysis.

The net energy balance (NEB) ratio is a parameter evaluating the efficiency of biofuel production: it is given by the ratio between energy output of the biofuel – including or not useful co-products – and the energy input from conventional sources required for the production process [15]. If the input from conventional energy sources balances the useful energy output – i.e. with a NEB ratio equal to 1 – there is no substantial environmental benefit in transitioning from fossil fuels to biofuels. As a reference, when accounting for the energy output contribution of useful co-products, corn grain ethanol has a NEB ratio of 1.25 compared to 1.93 for soybean biodiesel [15]. Because of the high energy input required in the production of corn grain ethanol, the corresponding well-to-wheel GHG emissions on an energy basis are projected to exceed the ones of diesel and biodiesel [15].

Ethanol production from sugarcane has a NEB ratio of 7.9 [16], outperforming corn ethanol with respect to land use efficiency and yield. The amount of ethanol produced per hectare of sugarcane cultivated is almost double the yield of corn ethanol production [17]. Sugarcane ethanol has also a lower energy intensity

(17)

compared to corn ethanol, since the conversion of starch into sugar is avoided.

Moreover, the well-to-wheel CO2 emission reduction per vehicle-kilometer achieved with sugarcane ethanol is estimated to exceed the GHG emission savings enabled by the use of corn ethanol [16].

Relevant drawbacks of crop-based biofuels are represented by (i) increased water consumption, (ii) augmented land use potentially intensifying deforestation and (iii) food price increase [17]. As a reference, 9% of the worldwide agricultural land would be needed to replace 10% of transport fuel using biofuels with crop-based feedstock [17]. Considering the rather poor performance of crop-based biofuels from an energy, economic and environmental standpoint, it is argued that efforts should shift towards ethanol production from non-food feedstock [15] like cellulosic ethanol [17]. As an example, Von Blottnitz and Ann [16] highlighted the potential benefits of exploiting cellulosic biomass waste like bagasse for ethanol production, with a projected NEB ratio of 32.

Fuel price represents a driver (or stopper) of biofuels penetration in transport.

As an example, the production cost of ED95 from corn was around 0.7-0.8$ per liter of equivalent diesel in 2016 – exceeding the tax-exempted price of diesel, amounting to less than 0.5$ in the United States market –, with almost 80% of cost share due to the feedstock price [5]. Oil prices fluctuations may also have a strong impact on biofuel demand, as observed in the light-duty sector in Sweden – a country with an established infrastructure for E85 distribution – towards the end of the 2000’s [18].

Regulations promoted the purchase of flex-fuel vehicles that could run alternatively on E85 or gasoline depending on the vehicle owner’s choice. In this context, the gasoline price drop that occurred at the end of 2008 caused E85 consumption to plummet [18].

The European infrastructure for biodiesel and bioethanol production is a result of policies promoted by the European Union from the 1990’s [19]. The initial driver was to support the agriculture sector in a time of crisis, followed by the environmental push towards greenhouse gas emission abatement. In the 2000’s, an additional boost to biofuel production came with the introduction of regulation setting a minimum share of bioethanol and biodiesel in gasoline and diesel, respectively, as well as of biofuel tax exemptions. The weight of these incentives on countries budget, raising concerns over the impact on land use and doubts on the extent of GHG emission reduction enabled by first generation biofuels led to the introduction of new policies in 2008. The target was shifted from a minimum share of biofuels in the transport energy matrix to a minimum share of renewable energy, including biofuels, hydrogen and electricity. Overall, the adoption of biofuels with a lower environmental impact and cost competitiveness should be achieved for the boosting their market size.

1.3 Advanced fuels

As previously discussed, feedstock availability, land use, and the energy intensity of the production process are bottlenecks hindering the large-scale use of bioethanol

(18)

1.3. Advanced fuels 5

as fuel. Methanol is a chemical species with comparable properties to ethanol [20], that can therefore be used for similar engine applications. A strong advantage favoring the large-scale use of methanol is represented by the diversified feedstock:

in fact, the most established methanol production process relies on the conversion of syngas [20] – i.e. a mixture of CO, CO₂ and H₂. The latter can be produced from natural gas reforming and coal gasification. However, these production pathways still rely on the use of fossil fuels, and do not comply with the search for a propulsion energy source with a low carbon footprint. As an alternative, biomass and waste can be also used as feedstock for the production of methanol.

As a candidate for the major energy carrier in the future, increasing attention is put on the use of hydrogen as fuel for internal combustion engines. It is mainly produced from gasification of coal or biomass, reforming of natural gas, and via water electrolysis [21]. Previous studies [21, 22] reported that hydrogen has a vast flammability limit favoring lean operation in premixed combustion concepts, as well as a high diffusivity and flame speed promoting homogeneous mixture formation and short combustion duration. Hydrogen’s autoignition temperature is significantly higher than diesel, necessitating an ignition source in compression ignition engines. Naber and Siebers [23] investigated the autoignition characteristics of pressurized hydrogen under conditions typical of diesel combustion, highlighting a strong temperature dependence. Temperatures higher than 1120 K were found to be necessary to reduce the ignition delay of hydrogen to levels acceptable for direct injection compression ignition combustion (DICI). A similar requirement was found for ethanol in the study of Siebers and Edwards [24], and will be discussed in the following chapter of this thesis.

Large-scale use of hydrogen as fuel for internal combustion engines poses challenges for storage and transport given its extreme flammability [25], diffusion through metals, and significantly lower energy density compared to liquid fuels [26]. Hydro- gen has an extremely low density at ambient conditions and, even when pressurized up to 700 bar, its energy content on a volume basis is one order of magnitude lower compared to conventional liquid fuels like gasoline and diesel [20]. It should be noted that even methanol has a higher volumetric energy density – and a superior hydrogen content on a volume basis – compared to liquid hydrogen, that has twice the volumetric energy density of pressurized hydrogen at 700 bar [20].

Electrofuels may contribute to overcome the fossil fuel dependence of the transport sector, while also facilitating the use electricity produced from intermittent renewable energy sources at times of low power grid demand. They comprise a variety of fuels – including methane, short-chain alcohols such as methanol and ethanol and heavier hydrocarbons – having in common the use of electricity as main energy source for hydrogen production from water electrolysis, and of carbon dioxide either from biogenic, atmospheric and flue gas sources for the fuel synthesis process [27]. Production processes of electrofuels have a well-to-tank efficiency estimated in the range between 30% to 70%. Since their energy source is electricity and considering that a subsequent energy loss occurs in the engine combustion process, the use of electrofuels should be considered only in applications where the

(19)

adoption of electric batteries is technically more challenging and expensive [27]

– e.g. heavy-duty vehicles. Estimating the price of electrofuels has a high degree of uncertainty given their novelty, however, previous studies suggest that electrofuels are currently more expensive than biofuels [27, 28].

1.4 Thesis outline

The general introduction outlined in this chapter aims to give the reader an overview of the technological context of the present research work. In the next chapter, a review of the state of the art of compression ignition combustion in heavy-duty engines is presented, with a focus on the use of ethanol fuel. Chapter 3 outlines the objective and scope of this work, and includes the addressed scientific questions. Subsequently, the experimental methodology followed during an extensive test campaign on a single-cylinder engine is described in Chapter 4. For readers interested in combustion modelling aspects, the CFD model implemented to perform three-dimensional dual-fuel combustion simulations is described in detail in Chapter 5.

Chapters 6 and 7 provide answers to the scientific questions listed in Chapter 3, while presenting the key findings of the doctoral project based on the experimental and simulation results. Eventually, conclusions and recommendations for future research studies are outlined in Chapter 8 and Chapter 9. This thesis is a compilation of four publications appended at its end, preceded by a summary specifying the contributions of main and co-authors.

(20)

2 State of the art

2.1 Combustion concepts for heavy-duty CI engine applications

The present research work aims at investigating a novel combustion system enabling diffusive combustion of pure ethanol in a process analogous to diesel combustion in a heavy-duty compression ignition engine. A characterization of the most relevant combustion concepts adopted in compression ignition engines is hereby presented to contextualize this work.

2.1.1 Diesel combustion

Diesel combustion in heavy-duty compression ignition engines represents the baseline and starting point of this investigation. The conceptual model of diesel combustion outlined by Dec [29] is adopted as reference. Diesel burns following a two-stage ignition process described in detail by Musculus et al. [30]. The high-pressure fuel spray sustains a premixed fuel-rich flame front subsequent to diesel ignition. Polycyclic aromatic hydrocarbons included in the combustion products are precursors of soot – forming downstream the flame front [30]. The products of fuel-rich combustion subsequently undergo diffusive combustion in the outer flame region [29], where soot is oxidized. The high temperatures reached in the diffusion flame region drive NO_x formation from molecular nitrogen and oxygen in air, according to the Zeldovich mechanism [31].

2.1.2 Low temperature combustion

In an effort to decrease soot and NO_x emissions, low temperature combustion (LTC) concepts relying on a higher degree of premixing compared to diesel combustion have been the object of various studies. Homogeneous charge compression ignition (HCCI) [32] addresses a combustion process driven by compression ignition of homogeneously premixed fuel. Both gasoline- and diesel-like fuels can be used in this concept by adjusting the compression ratio according to the cetane number. Benefits with low soot and NOx emissions are counterbalanced by challenges with low combustion efficiency [32] and lack of direct control of ignition timing and

7

(21)

8 Chapter 2. State of the art

combustion phasing [33]. Turbulent flame propagation and fuel injection rate allow control of the heat release rate in conventional gasoline and diesel combustion, respectively. On the other hand, the constant volume combustion event characterizing the HCCI concept leads to high rates of heat release, as well as excessive in-cylinder pressure rise rates [33]. Moreover, the use of fuels with a high boiling point like diesel hinders fuel vaporization [33] with a negative impact on mixture homogeneity, HC, CO and soot emissions [32].

Partially premixed compression ignition (PPCI) combustion represents an intermediate solution between HCCI and conventional diesel combustion. It is char- acterized by an heterogeneous fuel charge at ignition – i.e. partly rich and partly lean. To this purpose, the fuel injection event is separated from the start of combustion [30] using either early fuel injections, low compression ratios, exhaust gas recirculation (EGR) [34] and/or fuels with low cetane numbers [35]. The use of fuels with high cetane number like diesel necessitate extremely high EGR levels posing challenges to the air handling system [33]. Low engine-out emissions of soot and NOx are retained as in the HCCI combustion concept, while combustion phasing is controlled via the local stratification of the equivalence ratio. However, the PPCI concept also suffers from excessive pressure rise rates at high loads, limiting combustion phasing and having a negative impact on indicated efficiency [30].

Reactivity controlled compression ignition (RCCI) is a dual-fuel combustion concept described by Kokjohn et al. [36] relying on the use of a premixed fuel with low reactivity and of a fuel with high reactivity direct injected at subsequent instances during the compression stroke. The use of two fuels with different cetane numbers enables the local stratification not only of the equivalence ratio – as in the PPCI combustion concept – but also of the in-cylinder reactivity. In this way, a spatial gradient of the ignition delay is established, thereby limiting pressure rise rates while retaining the benefits of partial fuel premixing on engine-out emissions [37].

The direct controllability of ignition timing and combustion phasing is a key factor especially during transient engine operation. The comparative study carried out by Dempsey et al. [33] outlined that LTC concepts are sensitive to variations of the engine intake conditions, since chemical kinetics prevails as combustion controlling parameter with increasing degrees of fuel premixing. Consistent with the above overview, Saccullo [38] identified (i) increased CO and unburned fuel emissions, (ii) knocking tendency limiting the engine load, (iii) reduced combustion controllability, (iv) combustion instability and noise as the main challenges of low temperature combustion.

Dual-fuel combustion concepts like RCCI enable reduced NO_x and soot emissions while allowing the use of sustainable fuels alternative to diesel in heavy-duty engines.

These benefits prompt the need for further investigations aimed at retaining the advantages of LTC while mitigating challenges with combustion control.

(22)

2.2. Ethanol fuel 9

2.2 Ethanol fuel

The use of renewable alternative fuels in internal combustion engines has been encompassed since their invention in the nineteenth century [39]. Rudolf Diesel acknowledged that vegetable oils could be used in diesel engines with efficiencies comparable to the ones obtained with mineral oils [40]. The Otto engine in Henry Ford’s first vehicle – the “Quadricycle” – was fueled by pure ethanol, and Ford Model T could run on various blends of ethanol and gasoline [39, 41].

Diesel engines require fuels with low octane numbers and high cetane numbers, hence prone to autoignition by compression, whereas Otto engines necessitate fuels with high volatility promoting premixing and high octane numbers [41] to avoid pre-ignition.

Table 2.1: Physical properties of ethanol and diesel fuel

Fuel Ethanol Diesel Source

Cetane number [-] 12.0 57.6 Measured [42]

Research octane number [-] 109 <0 [43]

Lower heating value [MJ/kg] 25.58 42.76 Measured [42]

Density[kg/m³] at 15 °C 794.0 805.0 Measured [42]

Vapor pressure [kPa] 16.2 2.3 Measured [42]

Kinematic viscosity[mm²/s]at 40 °C 1.082 2.242 Measured [42]

Latent heat of vaporization [kJ/kg] 846 254 [44]

Boiling point [°C] 78.8 192.5 - 358 [44]

Adiabatic flame temperature [K] 2193^a 2300-2350^b [45]^a [46]^b

2.2.1 Physical properties

Ethanol is a short-chain alcohol with formula C2H5OH [43]. The physical properties of ethanol are listed in Table 2.1, together with the ones of diesel for comparison. Some of the properties in the list were referenced from previous studies, while the remaining ones were measured for the ethanol and diesel fuel used in this work.

Owing to the low carbon and hydrogen content and to the presence of the hydroxyl group (OH), the LHV of ethanol is lower by around 40% compared to diesel on a mass basis. This difference prompts the need for injectors with higher flow capacities and for bigger fuel storage systems to cover the same mileage [20].

With a lower boiling point and a vapor pressure seven times higher than diesel, ethanol has a high volatility influencing the characteristics of high-pressure fuel sprays [44]. The high latent heat of vaporization of ethanol – three times higher than diesel fuel – poses a challenge for engine cold start [43]. The combined effect of lower adiabatic flame temperature and higher latent heat of vaporization promotes lower in-cylinder temperatures compared to diesel combustion. As a result, a reduction in heat transfer losses and NOx emissions [20] can be achieved by replacing diesel with

(23)

ethanol. However, ethanol’s resistance to autoignition [41] due to its superior octane number – even higher than gasoline – and low cetane number makes it unsuitable as fuel in conventional diesel combustion systems. It should be noted that some of the observations on ethanol in the description above were based on the review of Verhelst et al. [20] on methanol as fuel for internal combustion engines, given the similarity between the properties of the two chemical species.

2.2.2 Combustion chemistry

The chemical kinetics of ethanol combustion was extensively discussed in the review of Sarathy et al. [43]. Ethanol reactivity at temperatures below 1000 K is limited by the absence of hydroxyl radical (OH) chain branching reactions, with a prevailing route of formation of stable acetaldehyde. At high temperatures, hydrogen abstraction followed by radical decomposition to carbon monoxide becomes the leading pathway of acetaldehyde consumption. Moreover, pathways of formation of stable intermediate species like ethylene and formaldehyde are also present at high temperatures.

The limited reactivity of ethanol at low temperatures translates in the absence of low-temperature heat release (LTHR) during combustion, as described by Sjöberg and Dec [47]. LTHR allows to counteract oscillations of the engine inlet temperature: for example, a temperature decrease is balanced by an increase in the low-temperature heat released during combustion. As a result, the single-stage high-temperature ignition characteristics of ethanol makes ignition timing and combustion phasing especially sensitive to changes of the inlet temperature in premixed combustion concepts.

Vuilleumier et al. [48] further investigated low- and intermediate-temperature heat release (ITHR) in a HCCI engine fueled with pure ethanol and various blends of ethanol and n-heptane – i.e. a chemical surrogate for diesel having comparable ignition delays [30]. A n-heptane volume fraction of 25% blended with ethanol was found to trigger ITHR. Increasing the n-heptane volume fraction up to 50% boosted ITHR up to around 5% of the total heat released, while enhancing LTHR. Hydroxyl (OH) radical was identified as the driver of LTHR and ITHR: the radical pool was mostly produced by the chain branching reactions of n-heptane and subsequently consumed in the reactions of hydrogen abstraction of ethanol and n-heptane. Hence, a minimum fraction of n-heptane increased the low-temperature reactivity of the fuel blend. These findings are consistent with other studies on HCCI combustion [49, 50]

highlighting that ethanol acts as an ignition inhibitor of n-heptane, causing an increase in unburned fuel emissions.

2.2.3 Spray characteristics

In the context of compression ignition combustion in heavy-duty engines, it is well established [51–53] that spray characteristics in liquid phase – e.g. the liquid penetration – are correlated to:

(24)

2.2. Ethanol fuel 11

1. In-cylinder conditions, such as the gas temperature and density during fuel injection. The spray liquid penetration decreases with increasing gas temperatures and densities, owing to the enhanced entertainment of gas with a higher energy to vaporize the fuel.

2. The spray hole diameter (do). The distance from the nozzle outlet along the spray centerline (x) within which a sufficient amount of gas is entrained for vaporizing the injected fuel decreases linearly with the spray hole diameter.

This is because the mass flow rate of fuel (m˙_f) increases quadratically with the spray hole diameter, while the amount of entrained gas (m˙g,entrained) increases only linearly, as shown in Equations 2.1 and 2.2 [52]. As a result, the spray liquid penetration has a linear dependence to the spray hole diameter.

˙

m_f ∝ ρ_f· d²_o· U_f (2.1)

˙

mg,entrained∝√

ρ_f· ρg· do· x · U_f· tanθ 2

(2.2) In the above equations, ρ_f is the liquid density of the fuel, ρg the gas density, U_f the velocity of the injected fuel and θ the spray cone angle.

3. Fuel properties like density, volatility and latent heat of vaporization. The engineering correlation developed by Higgins et al. [53] links the spray liquid penetration (L_f) to the aforementioned fuel properties, as outlined in Equations 2.3, 2.4 and 2.5.

L_f

d_o ∝ A^αB^β (2.3)

Where:

• α and β are correlation constants;

• Factor A is the density ratio between fuel and gas shown in Equation 2.4;

A= ^ρ^f

ρ_g (2.4)

• Factor B is the so-called “specific energy ratio” expressed in Equation 2.5.

B = ^C^p,f(T_b,f− T_f) +h_vap,f

C_p,g(T_g− T_b,f) ^(2.5) Where C_p,f and C_p,g are the specific heats at constant pressure of fuel and of gas, T_f is the initial fuel temperature and T_g is the gas temperature. A higher latent heat of vaporization of the fuel (h_vap,f) requires more energy from the entrained gas to vaporize fuel droplets. On the other hand, a higher fuel volatility – i.e. a lower boiling point (T_b,f) – facilitates fuel droplet evaporation by acting on the sensible heat contributions in the numerator and the denominator of Equation 2.5.

(25)

The aforementioned considerations explain the observations reported in previous experimental studies [44, 53, 54] regarding the characteristics of alcohol fuel sprays.

Higgins et al. [53] and Saccullo et al. [44] performed optical studies at conditions typical of heavy-duty engine operation, observing slightly shorter liquid penetrations of ethanol and methanol sprays compared to diesel sprays. Wu et al. [54] reported that the difference in liquid length between acetone-butanol-ethanol, n-butanol and diesel sprays diminishes at increasing gas temperatures, becoming comparable at 1200 K. Higgins’ correlation in Equation 2.3 [53] explains these observations:

alcohol fuels have a significantly higher latent heat of vaporization compared to diesel, balancing the impact of volatility during the vaporization process. Lower in-cylinder gas temperatures reduce the sensible heat contribution in the denominator of Equation 2.5, thereby increasing the impact of the fuel’s latent heat of vaporization and volatility on the spray liquid penetration. Wu et al. [54] also observed that higher gas temperatures drive a decrease in the fuel viscosity and surface tension as well as an increase in the vapor pressure, promoting fuel spray breakup and vaporization.

2.3 Combustion of short-chain alcohol fuels in CI engines

The study of Siebers and Edwards [24] provides a paramount insight into the autoignition characteristics of ethanol and methanol sprays under conditions typical of heavy-duty diesel engine operation: ethanol ignition is found to be weakly dependent on the in-cylinder pressure, while being sensitive to in-cylinder temperature variations. Two temperature thresholds are identified: (i) 950 K as the minimum temperature necessary to trigger compression-ignition of methanol and ethanol, and (ii) 1100 K as the temperature threshold enabling ignition delays comparable to diesel combustion. At intermediate temperatures between 950 K and 1100 K, a high degree of premixing at ignition occurs due to the long ignition delays of methanol and ethanol sprays, causing excessive peak pressure rise rates.

In order to enable the use of short-chain alcohol fuels – like methanol and ethanol – in compression ignition engines, different technical solutions have been investigated in the literature, including:

• Injecting a pilot fuel prone to autoignite at the end of the compression stroke, triggering the ignition of premixed alcohol fuels [55,56] – i.e. the “conventional”

dual-fuel combustion concept;

• Adopting high compression ratios [57] and/or using an ignition improver [58];

• Increasing the engine intake temperature [59].

The “conventional” dual-fuel combustion concept [55, 56], with fully premixed ethanol as main fuel and direct injected diesel as pilot fuel, retains the benefits of low temperature combustion with respect to NOx and soot emission reduction.

On the other hand, higher CO and HC emissions affect engine operation, owing to

(26)

2.3. Combustion of short-chain alcohol fuels in CI engines 13

the resistance of ethanol to autoignition and unburned fuel pockets in the crevice volumes [56]. Moreover, excessive pressure rise rates at high load pose a limit to the ethanol ratio [55,56]. For the same reason, engine operation is limited when adopting high compression ratios – e.g. 27:1 [57]. In this case, NO_x emissions are likely to increase due to the high in-cylinder temperature levels during combustion [57].

The use of ignition improvers may mitigate pressure rise rates and unburned fuel emissions by reducing the autoignition temperature of additized ethanol. However, this solution does not solve the issue with high NO_x emissions [58]. Moreover, the higher fuel production cost due to the additive price shall be taken into account when evaluating this option.

Setting a high intake temperature is a solution potentially enabling an increase of the ethanol ratio throughout the entire engine load range even with standard compression ratios. However, this solution may add performance costs due to the energy needed to warm-up the intake charge. Blumreiter [59] implemented high- temperature stoichiometric combustion of ethanol in a compression-ignition engine, reporting stable combustion and emissions of CO, HC and NOx within acceptable levels to enable the use of a three-way catalyst. The use of exhaust heat recovery can boost the engine thermal efficiency. On the other hand, additional technical solutions [59] are needed for cold-start and engine operation at low load.

An alternative solution to maximize the ethanol ratio throughout the entire engine load range, while retaining the controllability of diesel combustion, is represented by the separate direct injection of pure alcohol fuels and diesel using a dual- injector system. Recent studies [38, 60] – including the present research work – highlighted benefits with respect to engine performance, combustion stability and pollutant emissions. Despite the increased complexity caused by the dual-injector system, previous industrial applications by Wärtsilä [61] and MAN [62] confirmed the potential of this concept. The additional degrees of freedom for combustion control enabled by the independent injection of two fuels with different physical and chemical properties open up to a wide research gap, that is the object of the present investigation.

(27)

(28)

3 Ethanol-diesel DICI combustion in a heavy-duty engine

3.1 Dual-injector system

The scope of the present research is to generate knowledge on a novel dual-fuel combustion system for heavy-duty engine applications powered by ethanol – without ignition or lubricity improvers – and minimal quantities of diesel fuel. As shown in Figure 3.1, the novelty of the system lays in the dual-injector configuration, with two separate direct injectors mounted in the cylinder head. Ethanol is the main fuel injected at the end of the compression stroke through a heavy-duty injector whereas diesel is the pilot fuel triggering ignition injected via a light-duty injector. The main injector for ethanol (in red) is centrally mounted, while the pilot injector for diesel (in yellow) is mounted in the periphery of the combustion chamber. The rail pressure, injected quantity of main and pilot fuel, and timing of the two injection

Figure 3.1: Dual-injector system for ethanol in red and diesel in yellow

15

(29)

16 Chapter 3. Ethanol-diesel DICI combustion in a heavy-duty engine

events are independently controlled. The purpose of the dual-injector system is to achieve mixing-controlled combustion of ethanol in a heavy-duty compression ignition engine, with a process analogous to conventional diesel combustion. For this reason, the novel dual-fuel combustion concept is addressed as “ethanol-diesel direct injection compression ignition (DICI) combustion”. In order to ensure a systematic comparison between ethanol-diesel and conventional diesel combustion, three different sets of engine tests were carried out using the system in Figure 3.1:

• Single-diesel combustion, with a single injection of diesel through the main injector without actuating the pilot injector. This is the conventional diesel combustion baseline, used to evaluate the overall combustion and performance parameters obtained with ethanol-diesel DICI combustion.

• Dual-diesel combustion, with dual injections of diesel through both the main and pilot injectors. The energy ratio between the two injections, as well as the injection timing are kept as in the dual-fuel test points. This alternative baseline is used to compare characteristics of ethanol-diesel DICI combustion to diesel combustion, focusing on the influence of physical and chemical properties of the ethanol main injection.

• Ethanol-diesel DICI combustion, as described previously in this section.

Throughout the tests, the injection timing of both main and pilot fuel was maintained at the end of the compression stroke. Alternative injection strategies typical of low temperature combustion concepts with a higher degree of premixing were not considered and are beyond the scope of the present work. It should be noted that the nozzle tip configurations of the injectors in Figure 3.1 were designed for late injections at the end of the compression stroke, in order to avoid spray impingement against the combustion chamber surfaces. The three custom-made nozzle tip configurations shown in Figure 3.2 were designed taking into consideration

Figure 3.2: Schematic of the dual-injector configurations. Source: Publica- tion III [63]

(30)

3.2. Project overview 17

the constrained mounting location of the pilot injector. The configurations have a non-symmetric spray pattern and share the same total nozzle area with one, two and four sprays at a decreasing hole diameter. By comparing the experimental results obtained with the different pilot injector configurations, the influence of the spread of the pilot fuel across the piston bowl volume is investigated with respect to ethanol combustion characteristics and engine performance.

3.2 Project overview

At the beginning of the doctoral project, experimental data on ethanol-diesel DICI combustion was not available in the literature. Studies on mixing-controlled combustion of alcohol fuels, referenced in the previous chapter, started to circulate when the project was approaching its conclusion. The unique approach followed in the present research, with a strong focus on the ethanol ignition dynamics, provides a novel contribution to the knowledge of mixing-controlled combustion of ethanol in compression ignition engines.

Investigating a combustion system with such a high degree of novelty required a variety of tools and an articulated methodology. The experimental work started with the design of the nozzle tip geometry of the pilot injector. Minor modifications to cylinder head and piston geometry were needed in order to integrate the second injector to the pre-existing conventional diesel combustion system, since the pilot injector fit in the original mounting location of the in-cylinder pressure transducer.

An alternative mounting location via a channel was arranged on the cylinder head for the pressure transducer. After finalizing the design and supply of the components, the dual-injector system prototype was assembled on the single-cylinder research engine at the KTH Internal Combustion Engine laboratory. An additional fuel system was connected to the engine, and its control system interfaced to the in-house engine control unit.

In parallel with the experimental work, a three-dimensional CFD model was adopted to simulate ethanol-diesel DICI combustion. The first simulation study was performed in order to achieve an initial understanding of the influence of the dual-injection parameters on ethanol combustion. The preliminary knowledge on the operation of the dual-injector system at different engine conditions helped structure the experimental matrix for the subsequent test campaign on the single-cylinder engine.

Various engine conditions and dual-injection strategies were experimentally investigated using the three different pilot injector configurations in Figure 3.2.

Moreover, ethanol-diesel DICI combustion results were systematically compared to diesel combustion as explained in the previous section.

On completion of the test campaign, the experimental data on ethanol-diesel DICI combustion was used to validate the combustion simulation model previously implemented. The CFD model proved to have a good predictive capability with respect to ethanol ignition and combustion characteristics under the most relevant

(31)

18 Chapter 3. Ethanol-diesel DICI combustion in a heavy-duty engine

operating conditions investigated. Moreover, a detailed understanding of the interaction between diesel and ethanol sprays during ignition was achieved by coupling simulation and experimental results.

3.3 Research questions

The methodology described in the previous sections allowed to answer the following set of research questions:

1. Is it possible to achieve diesel-like diffusive combustion of pure ethanol with minimal diesel pilot fuel quantities over a broad engine load range?

2. Is it possible to operate the dual-fuel direct injection engine at full load using pure ethanol as main fuel?

3. How does ethanol-diesel DICI combustion compare to conventional diesel combustion with respect to engine performance and pollutant emissions?

4. What is the influence of the dual-injection strategy on:

a. Ethanol ignition and combustion characteristics;

b. Combustion stability;

c. Combustion completeness;

d. Engine performance;

e. Engine-out NO_x emissions?

5. What is the influence of the spread of pilot fuel charge across the piston bowl volume on:

a. Ethanol ignition and combustion characteristics;

b. Combustion stability;

c. Combustion completeness;

d. Engine performance?

6. What is the predictive capability of the adopted CFD model with respect to ethanol-diesel DICI combustion characteristics?

7. How do diesel and ethanol sprays interact during ignition under different engine conditions?

(32)

4 Experimental method

4.1 Single-cylinder engine setup

The dual-injector system described in the previous chapter was tested on the single-cylinder research engine at the KTH Internal Combustion Engine laboratory.

The adopted experimental setup is outlined in the schematic in Figure 4.1, while the main details of the engine configuration are listed in Table 4.1. It should be noted that the effective compression ratio was estimated to be between 17.7 and 18.2, that is lower than the geometric compression ratio owing to crevice volume and blowby. The engine is connected to a hydraulic brake regulating the speed and to a boosting system controlling the intake pressure and temperature. The exhaust system consists of a pressure vessel dampening the pressure pulses propagating from the engine. Two separate fuel systems were interfaced to the main and pilot injector, with two custom-built control systems allowing independent manual control of rail pressure, needle opening time and start of injection of the main and pilot fuel. The two control systems were synchronized by the encoder signal having a resolution of 0.1 crank angle degrees.

The adopted piston geometry was designed for conventional diesel combustion.

A recess was machined on the piston deck in order to accommodate the side injector, as shown in Figure 4.2. The main injector had eight fuel sprays and was kept unchanged throughout the tests. The geometrical specifications of the tested pilot injector configurations, shown in Figure 3.2, are listed in Table 4.2 in order of increasing spray number and decreasing nozzle hole diameter. The injected mass of pilot fuel at a given injection duration and rail pressure was comparable between the configurations, since they share the same total nozzle area.

Table 4.1: Engine geometry [42, 63]

Bore diameter 127 mm

Crank radius 77 mm

Connecting rod length 255 mm Geometric compression ratio 19.4

19

(33)

20 Chapter 4. Experimental method

Figure 4.1: Experimental setup. Modified from source: Publication II [42]

Figure 4.2: Dual-fuel engine piston

(34)

4.1. Single-cylinder engine setup 21

Table 4.2: Geometry specifications of the pilot injector configurations shown in Figure 3.2 [63]

Configuration Spray hole diameter Number of sprays

A 0.28 mm 1

B 0.20 mm 2

C 0.14 mm 4

4.1.1 Measurement system

The main parameters measured during engine tests are:

• In-cylinder pressure;

• Mass consumption of diesel and ethanol;

• Air-excess ratio;

• Engine-out emissions of CO, HC and NO_x.

In-cylinder pressure is the most relevant quantity measured for the characterization of ethanol combustion and the evaluation of engine performance. Crank-angle resolved in-cylinder pressure was measured via a channel-mounted pressure transducer with range between 0 to 250 bar, and a resolution of 0.1 bar. For each test case, the pressure traces of 100 subsequent engine cycles were sampled, then filtered to remove resonant oscillations in the mounting channel for the pressure transducer, and eventually ensemble-averaged.

The consumption of the main and pilot fuel was measured using two separate scales, logging the fuel mass variation over time. The resolution of the readings of pilot and main fuel scales was 0.1 g and 10 g, respectively: depending on the engine load, the mass of the pilot fuel was 1 to 2 orders of magnitude lower than the main fuel, thus necessitating a higher measurement resolution.

Intake temperature and pressure were sampled using a K-type thermocouple and a gauge pressure transducer, respectively – both mounted on the intake manifold.

The corresponding set-point values of these quantities are reported in Chapter 6 prior to the description of the experimental results.

Owing to safety concerns over the buildup of unburned ethanol in the exhaust pressure vessel, no back-pressure was applied during the tests. In this way, a positive difference between intake and exhaust pressure was established, thereby limiting the amounts of exhaust residuals in the engine combustion chamber. As a result, ethanol ignition was not facilitated by a higher residual gas temperature. Moreover, the pumping loop was excluded from the engine efficiency analysis, as the gross indicated efficiency was the parameter adopted to evaluate the performance of the ethanol-diesel combustion process.

(35)

22 Chapter 4. Experimental method

Air-excess ratio and emissions were measured downstream the exhaust pressure vessel in Figure 4.1. The lambda sensor provided an indication of the air-excess ratio based on the measured oxygen concentration in the exhaust stream. The concentration of the engine-out emissions of CO, HC – used to compute combustion efficiency – and NO_x in the exhaust stream was measured using a MEXA-7100 Horiba analyzer. The pollutant concentration in the exhaust stream was sampled on a dry basis based on an average over 120 s of data sampled at a 10 Hz frequency.

4.2 Data processing

In the previous chapter, the so-called “dual-injection strategy” was the object of multiple research questions. With this term, two parameters are addressed:

A. The main-pilot separation (MPS) shown in Equation 4.1. It is given by the difference between start of injection (SOI) of the main fuel – i.e. ethanol in the dual-fuel cases – and start of injection of the diesel pilot fuel. The MPS is positive when the pilot fuel injection starts before the main injection: both negative and positive MPS values were considered in the tests.

MPS[CAD] =SOImain− SOI_pilot (4.1) In the presentation of the experimental results outlined in Chapter 6, the SOIs in Equation 4.1 correspond to the switch-on of the voltage signals regulating the timing and duration of the injection events. Hence, the delay between voltage signal and needle opening – i.e. the actual start of fuel injection – was not considered in the calculation of the MPS. However, the delay of each of the two injection events was an offset approximately constant throughout the tests. The offset is dependent on the electronic components of the injector circuit – driving a delay between voltage and current signal – and the inertia of the mechanical components of the injector – causing a delay between current signal and needle opening.

B. The main-pilot energy ratio (MPR) – or ethanol ratio in the dual-fuel cases – shown in Equation 4.2. It is given by the ratio between the main and overall fuel injected energy. Values well above 50% have been considered in the present work.

MPR[%]= ^m^inj,main ^LHV^main

m_inj,pilot LHV_pilot+m_inj,main LHV_main· 100 (4.2) Where m_inj,pilot is the injected mass of the diesel pilot fuel, m_inj,main the injected mass of the main fuel – either ethanol or diesel –, and LHV is the corresponding lower heating value. As previously mentioned, the mass of ethanol and diesel injected per engine cycle was computed based on the mass recorded by the fuel scales. The lower heating values of ethanol and diesel are listed in Table 2.1.

Diffusive Combustion of Ethanol in a Dual-Fuel Direct Injection Compression Ignition Engine

Diff usive Combustion of Ethanol in a Dual-Fuel Direct Injection

Compression Ignition Engine

NICOLA GIRAMONDI

Diffusive Combustion of Ethanol in a Dual-Fuel Direct Injection

Compression Ignition Engine

NICOLA GIRAMONDI

Contents

Nomenclature

Abstract

Sammanfattning

List of Publications

1 Introduction

1.1 Transition towards future transport technologies

1.2 The role of bioethanol

1.3 Advanced fuels

1.4 Thesis outline

2 State of the art

2.1 Combustion concepts for heavy-duty CI engine applications

2.2 Ethanol fuel

2.3 Combustion of short-chain alcohol fuels in CI engines

3 Ethanol-diesel DICI combustion in a heavy-duty engine

3.1 Dual-injector system

3.2 Project overview

3.3 Research questions

4 Experimental method

4.1 Single-cylinder engine setup

4.2 Data processing