Case Studies in Thermal Engineering 35 (2022) 102129

Contents lists available at ScienceDirect

Case Studies in Thermal Engineering

journal homepage: www.elsevier.com/locate/csite

Numerical analysis of injection and spray characteristics of diesel

fuel and rapeseed oil in a diesel engine
Vladimir Markov a, Bowen Sa a, *, Sergey Devyanin b, Leonid Grekhov a,
Vsevolod Neverov a, Jianhui Zhao c
a
Department of Combined Engines and Alternative Power Plants, Bauman Moscow State Technical University, Moscow, 105005, Russia
b
Department of Tractors and Automobiles, Moscow Timiryazev Agricultural Academy, Russian State Agrarian University, Moscow, 127550, Russia
c
College of Power and Energy Engineering, Harbin Engineering University, Harbin, 150001, China

A R T I C L E I N F O A B S T R A C T

Keywords: Vegetable oil-based biofuels are promising carbon-neutral fuels to replace fossil fuels. The study
Diesel engine focuses on a comparative analysis of injection and spray characteristics of DF and RO and the
Alternative fuel relationship between spray and emission formation. Steady nozzle flows were simulated in Ansys
Vegetable oil Fluent. Free sprays were modeled with 0D spray models, and the relationships between atomi­
Nozzle flow zation parameters and TKE at the orifice exit were analyzed. Sprays of DF/RO blends in real
Spray
combustion chamber conditions were simulated with a multi-zone spray model in DIESEL-RK.
Atomization
The influence of RO content on fuel atomization and fuel-air mixture formation were analyzed.
Compared with DF, RO has lower injection velocity and TKE at the orifice exit. The empirical
correlations of mean droplet diameters and spray angle with TKE were developed by using a
complex TKE/ν0.5 with correlation coefficients above 0.988. The increase of the RO content from
0 to 100% leads to a reduction in spray angle by 24% and an increase in spray tip penetration by
15.5%, SMD by 62.1%, and the mass fraction of fuel injected onto the combustion chamber walls
from 17.1 to 42%. A correlation between NOx emissions and SMD was fitted with a correlation
coefficient of 0.9968.

Abbreviations and nomenclature

DF diesel fuel
RO rapeseed oil
SVO straight vegetable oils
VOB vegetable oil biodiesels
TKE turbulent kinetic energy
CVC constant volume chamber
CFD computational fluid dynamic
VF2 volume fraction of cavitation vapors
ρl density of fuel liquid

* Corresponding author. Department of Combined Engines and Alternative Power Plants, Bauman Moscow State Technical University, 5, 2-ya Baumanskaya Str.,
Moscow, 105005, Russia.
E-mail address: [email protected] (B. Sa).

https://doi.org/10.1016/j.csite.2022.102129
Received 22 January 2022; Received in revised form 13 May 2022; Accepted 14 May 2022
Available online 18 May 2022
2214-157X/© 2022 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
V. Markov et al. Case Studies in Thermal Engineering 35 (2022) 102129

ρv density of fuel vapor

ρg density of ambient gas
β spray angle
Ltip spray tip penetration
Vinj injection velocity
AMD arithmetic mean diameter
VMD volume mean diameter
SMD Sauter mean diameter
MMD median mean diameter
BTE thermal efficiency
CI compression ignition
IC internal combustion
pin injection pressure
pamb back pressure
ps saturation pressure of fuel vapor
μl dynamic viscosity of fuel liquid
νl kinematic viscosity of fuel liquid
σl surface tension coefficient of fuel liquid

1. Introduction
Increasing energy consumption and climate change, including global warming, force the promotion of alternatives to fossil fuels,
including zero-carbon and carbon-neutral fuels. Among different alternative fuels, straight vegetable oils (SVO) [1] and their de­
rivatives such as methyl, ethyl, and butyl ethers – vegetable oil biodiesels (VOB) [2] and hydrotreated vegetable oils (HVO) [3] are
attracting more and more attention because of their biodegradability, renewability, and carbon neutrality. Attributed to the similarity
of physical and chemical properties between vegetable oil-based biofuels and petroleum diesel fuel (DF), it is more appropriate to use
vegetable oil-based biofuels as a fuel substitution in compression ignition (CI) engines and their power plants, which will continue to
play a significant (central) role in the energy and transport sectors for a long time [4].
Despite the advantages of VOB and HVO over SVO, such as reduced viscosity, increased volatility, increased cetane number, and
increased calorific value during transesterification, as well as the problems arising during fueling diesel engines with SVO, research on
the application of SVO in CI engines continues due to the high economic and environmental costs of vegetable oil transesterification [5,
6]. In addition, it is more profitable to use SVO in the agricultural sectors, where it is possible to realize the direct use of vegetable oils
without the cost of their transportation [7,8].
However, in comparison with petroleum DF, SVO have higher viscosity, higher density, higher surface tension, lower evaporation
property, and worse flammability [9–11], which make it challenging to use SVO as a separate fuel. The possibilities of using various
vegetable oils, including rapeseed oil (RO), sunflower oil (SuO), soybean oil (SO), palm oil (PO), Jatropha oil (JO), Castor oil (CaO),
and Karanja oil (KO), in diesel engines were reviewed in Ref. [11]. It is noted that these vegetable oils have high viscosity from 30 to 40
mm2/s at 38 ◦ C, nearly 15–20 times higher than that of DF. SVO are usually used by blending with other low-viscosity fuels, mainly
with petroleum DF [7,11]. Measurements of the kinematic viscosity of rapeseed oil (RO) blended DF [8] have shown that the fuel blend
has a viscosity close to that of DF when the fraction of RO is less than 20%. Fuel properties of lemongrass oil (LGO) blended DF and the
possibility of their use in a diesel engine have also been studied [12]. Blending LGO with DF brings density, viscosity, and cetane
number close to these of traditional diesel fuel. The negative influence of LGO on engine performance can be compensated by opti­
mizing fuel injection parameters.
In comparison with petroleum DF, SVO and SVO blended DF can reduce the emission of toxic substances in exhaust gases, such as
soot [13], carcinogenic aromatic hydrocarbon compounds (AHC) [11], and sulfides [14], owing to the relatively high oxygen content
and absence of sulfur and AHC in SVO. There is no consistent conclusion on carbon monoxide and nitrogen oxide emissions, which
depend on oil feedstock sources, engine specifics, operating conditions [11]. In CI engines, fuel injection and atomization processes
that predetermine the quality of the subsequent fuel-air mixture formation and combustion processes have a critical role in engine
thermal performance and emission characteristics. The high viscosity, density, and surface tension of SVO deteriorate the quality of
injection and atomization processes, which leads to a decrease in combustion efficiency and brake thermal efficiency (BTE) of engines
fueled with SVO and SVO blended biofuels [14]. Coke and deposition formation in fuel injector systems is intensified, and carbon
deposition in the combustion chamber is observed due to the high viscosity and density of vegetable oil-based biofuels when they are
used in diesel engines [15].
Injection and spray of vegetable oil-based fuels have been studied experimentally and numerically. In Ref. [16], the characteristics
of the fuel injection and fuel-air mixture formation processes for vegetable oil-based fuels in diesel engines have been analyzed, and the
improvement directions of these processes have been proposed. A flow visualization system has been employed to investigate the
nozzle internal flow and cavitation flow patterns of VOB and DF in an enlarged nozzle with a cylindrical orifice under different in­
jection pressures and atmospheric spray conditions [17]. Compared with DF, VOB had lower volume flow rate and injection velocity
under the same injection pressure, and supercavitation for VOB occurred under higher injection pressure. With increasing injection

2
V. Markov et al. Case Studies in Thermal Engineering 35 (2022) 102129

pressure, the velocity coefficient of DF changed dramatically, but the velocity coefficient of VOB was almost a constant. Som S et al.
[18] have simulated the cavitation flow of soybean oil biodiesel and DF in a full-production mini-sac multi-orifice nozzle and the spray
development of these fuels in a constant volume chamber (CVC) under injection pressures of 110 and 130 MPa by statistically coupling
the nozzle flow modeling with spray modeling. Significant differences in the nozzle flow and spray characteristics between biodiesel
and DF were observed. Unlike DF cavitation, which developed up to the orifice exit, biodiesel cavitation was formed only in a limited
volume far before the orifice exit due to the low saturation pressure of biodiesel. In comparison with DF, biodiesel had lower injection
velocity, mass flow rate, and turbulent kinetic energy (TKE) at the orifice exit, which resulted in longer spray liquid penetration,
narrower spray angle, and bigger Sauter mean diameter (SMD) of droplets for biodiesel under non-evaporating conditions. The same
results have also been obtained by Yu S et al. [19] through experimentally and numerically comparing nozzle flow and spray char­
acteristics between DF and biodiesel in a single-orifice nozzle under injection pressures of 50 and 90 MPa and atmospheric conditions.
Di Blasio G et al. [3] experimentally investigated the multiple injection strategies on combustion and emission characteristics of HVO.
It was found that at high engine loads the difference in ignition delay and combustion duration between HVO and DF was minimal and
one of the main reasons is that the spray characteristics of HVO and DF were very close. However, at low engine loads the difference in
spray characteristics was enlarged due to higher evaporation property of HVO. Sathiyamoorthi R et al. [20] have experimentally
investigated the spray characteristics of Palmarosa oil (PO) biodiesel and DF in a CVC under different injection pressure. Increasing the
content of PO biodiesel in biodiesel/diesel blends led to an increase in spray tip penetration and SMD and a reduction in the spray area.
Higher injection pressure produced smaller fuel droplets. Kegl B and Lešnik L [21] have employed phenomenological models for spray
penetration and spray angle to analyze the spray behavior differences between DF and rapeseed oil biodiesel. It has also been found
that in comparison with fuel types, the chamber pressure has a more significant effect on spray penetration and spray angle. The impact
of biodiesel fraction in DF and Karanja oil biodiesel blends on spray evolutions was studied in a CVC under different chamber ambient
pressures in Ref. [22]. Droplet size increased with increasing the biodiesel fraction, and the atomization was enhanced with increasing
ambient pressure. The spray characteristics of the gasoline/biodiesel blends with different biodiesel content were experimentally
investigated in a CVC under non-evaporating conditions [23]. The fuel blend with lower biodiesel content had a larger spray angle. The
sprays of soybean oil biodiesel and DF in a CVC under diesel-like operating (evaporating) conditions were captured with Schlieren
high-speed imaging and Mie-scatter imaging [24]. The two fuels had almost the same vapor-phase penetration, but the maximum
liquid-phase penetration of biodiesel was higher than that of DF. The same result has also been reported for Jatropha oil biodiesel and
DF in Ref. [25]. Moreover, an analytical prediction model was used to understand the effect of fuel properties on spray vapor pene­
tration. Using the computational fluid dynamic (CFD) approach, Ishak MHH et al. [26] numerically investigated the characteristics of
cavitation flow of a biofuel blend from vegetable oils and diesel fuel in injector nozzles with circle and elliptical shapes and the spray
development in a constant volume chamber. The biofuel blend was prepared by blending melaleuca cajuput oil with refined palm oil
that is used to reduce the viscosity. Compared to fuel properties, nozzle geometry shape had a more significant effect on cavitation
morphology inside the nozzle. The biofuel blend had a lower injection velocity and narrower spray angle under the same nozzle shape.
Using a spray visualization system, Qi DH et al. [27] experimentally investigated spray characteristics of hybrid fuels consisting of
diesel fuel, tung oil, and ethanol under an injection pressure of 25 MPa and atmospheric conditions. Increasing the content of tung oil
led to an increase in the spray penetration and Sauter mean diameter (SMD) due to increased density and viscosity, but adverse effects
were found with increasing the content of ethanol as a result of decreased density and viscosity. Bhikuning A et al. [28] experimentally
investigated the spray characteristics of waste cooking oil (WCO), bio-hydro fined diesel oil (BHD), and n-tridecane in a CVC under
non-evaporating/evaporating conditions at different injection pressures of 50–150 MPa. WCO had longer tip penetration, narrower
spray angle, and larger SMD than BHD and tridecane due to the higher viscosity, density, boiling point temperature of WCO. An
increase in injection pressures caused an increase in spray tip penetration and spray angle and a decrease in SMD and can narrow the
differences in spray tip penetration and spray angle between fuels. Atomization quality was improved as the ambient temperature
increased. Moreover, an increase in ambient temperatures reduced the differences in spray tip penetration and spray angle between
WCO and n-tridecane. In general, spray development is sentive to fuel properties. Hoang AT and Le AT [15] have investigated the spray
and combustion characteristics of Jatropha oil (JO), preheated JO, and DF with using clean injectors and injectors after 300 h of
endurance test. Deposits were easily formed in the nozzle orifices with JO due to its high density, viscosity, and surface tension.
Deposits formed in the nozzle orifices deteriorated atomization quality (increased spray penetration and reduced spray angle) and
combustion quality, which in turn accelerated the formation of deposits in the nozzle orifices and the walls of combustion chamber.
Preheating of JO can enhance the atomization of JO and decrease the formation of deposits.
The literature analysis presented above shows that a lot of research is devoted to the properties of vegetable oil-based biofuels and
their application in CI engines, as well as the nozzle internal and atomization characteristics of VOB and their blends. However, the
characteristics of internal flow and atomization of SVO and their blends, as well as the interaction between the atomization of these
fuels and emission formations, have not been studied sufficiently. The performance indicators of CI engines, to a large extent, are
related to the parameters of fuel flow discharging from the nozzle orifice, such as mass flow, injection velocity, and turbulence
[29–31]. The turbulence of fuel flow inside the injector nozzle and the turbulence of the fuel jet in the combustion chamber signifi­
cantly impact the quality of the atomization process. The higher the turbulence intensity of the fuel flow at the nozzle exit, the faster
the disintegration of the fuel jet and the finer the fuel droplets generated [31,32]. The fuel supply and atomization processes pre­
determine the characteristics of the subsequent fuel-air formation and combustion processes and have a noticeable effect on indicators
of fuel economy and emissions of engines.
In order to fill a gap in the relevant literature, this paper aims to perform a numerical investigation and comparative analysis of the
nozzle flow and atomization characteristics of DF and SVO. RO is selected as SVO investigated because, in the present, RO and its
derives are regarded as one of the prospective biofuels [8,33]. In 2021, the global production of major vegetable oils exceeded 200

3
V. Markov et al. Case Studies in Thermal Engineering 35 (2022) 102129

million tons, including 28.3 million tons of rapeseed oil [34]. In the first stage of the present study, flows of DF and RO in a diesel
injector nozzle were simulated in the software package Ansys Fluent and the flow parameters of these fuels in the orifice were obtained.
In the second stage, the atomization of DF and RO was calculated by using 0D spray models, and correlations between atomization
parameters and TKE at the orifice exit were proposed. In the third stage, the atomization of DF/RO blends in the combustion chamber
conditions of a diesel engine was modeled by using the software DIESEL-RK [35] developed by professor Kuleshov AS in Bauman
Moscow State Technical University. Finally, the relationship between NOx emissions and the mean diameter of fuel droplets was
analyzed.

2. Materials and research methods

2.1. Nozzle flow simulation models
The flow characteristics inside the injector nozzle are sensitive to fuel properties and predominate the subsequent fuel atomization.
To analyze the effect of SVO on the fuel supply process, steady flows of SVO and DF inside an injector nozzle at a full needle lift have
been simulated in the software package Ansys Fluent 17.2. Steady flows in the nozzle at a full needle lift were modeled because the
main portion of the fuel is injected into the combustion chamber under high needle lifts, where the needle-seat flow area is higher than
the total area of nozzle orifices, and the throttling effect is mainly contributed by the nozzle orifices. The homogeneous multiphase
approach has been employed to model the two-phase flow inside the nozzle with considering the generation of cavitation vapor. In this
approach, all phases have the same local pressure. The primary phase is fuel liquid and the secondary phase is fuel vapor. The mass
transfer between the primary and secondary phases was simulated using a modified Schnerr-Sauer cavitation model, which uses a
modified Rayleigh-Plesset bubble dynamic equation, to describe the formation, growth, and collapse of cavitation bubbles [36]:
( )1/3 √̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅
⃒ ⃒̅
ρ ρ 4π 2 ⃒⃒ps + 0.195ρl k − p⃒⃒
Rv = v l Nb α v (1 − αv )4/3 ⃒ for pv < p; (1)
ρm 3 3⃒ ρl

( )1/3 √̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅
⃒ ⃒̅
ρv ρl 4π 2 ⃒⃒ps + 0.195ρl k − p⃒⃒
Rv = Ccond Nb α v (1 − αv )4/3 ⃒ for pv ≥ p, (2)
ρm 3 3⃒ ρ
l

where Rv is the mass transfer rate from the liquid phase to the vapor phase per volume; αv is the volume fraction of the vapor phase; ρm,
ρv, and ρl are the density of the mixture phase, vapor phase, and liquid phase, respectively; Nb is the number of bubbles; ps is the
saturation pressure of fuel vapors; k is TKE; p is the pressure in the external liquid phase; Ccond is the condensation coefficient. Nb and
Ccond are functions of cavitation number and are determined based on experimental data [37]. The slip between phases was not taken
into account because of the relatively high flow velocity inside the nozzle [38]. The momentum conservation equations were described
for the mixture phase and closed by a turbulence model – a realizable k-ε model. Enhanced wall treatment (EWT) was used to obtain an
accurate prediction of the effect of fuel viscosity on the solution of near-wall regions. The temperature inside the nozzle was assumed to
be a constant [38]. Detailed descriptions of these models have been given in our previous works [39–41]. Validation of these models
based on the experiment of Winklhofer E for an enlarged nozzle with a rectangular cross-section orifice [37] and the experiment of
Sforzo BA for a real-size nozzle with a circular cross-section divergent orifice [42] has been performed in our previous study [39]. The
well-validated nozzle models, high mesh resolution, and accurate representation of fuel properties can provide the necessary calcu­
lation accuracy to identify the difference in flow characteristics between the investigated fuels.
In the present study, the injector used is a diesel injector named FDM-22 produced by Noginsk Fuel Equipment Plant (Russia,
Moscow). The FDM-22 injector is assembled with a valve-covered-orifice (VCO) nozzle No. 171.07.00 manufactured by Altai Precision
Components Plant (Russia, Moscow) and used in the D-245.12S diesel engine manufactured by Minsk Motor Plant (Belarus, Minsk).
The main specifics of the injector nozzle have been given in Table 1. Given the geometric symmetry of the nozzle used, a 72◦ sector of
the nozzle was employed to save computational costs. The mesh model of the nozzle sector has been constructed in ICEM CFD with grid
sizes selected according to mesh dependency analysis carried out for the same nozzle in our previous work [39]. Eventually, the nozzle
sector at the maximum needle lift was comprised of 135 million constructed hexahedral grid cells. The geometry model and con­
structed mesh of the nozzle sector at the maximum needle lift have been presented in Fig. 1. The pressure at the inlet to the calculation
domain (pin ) was selected as 51,5 40, and 20 MPa, which corresponded to the pressure before the inlet to injector when the fuel in­
jection system (FIS) of the D-245.12S diesel engine was operated in different operating conditions. The pressure at the outlet from the
calculation domain (pamb ) was set to 8.878 MPa. This pressure was a gas pressure in the cylinder during fuel injection. The fuel
temperature was set to 40 ◦ C, which corresponded to the fuel temperature in the high-pressure pipeline of FIS.

2.2. Spray models

Fuel spray models for CI engine applications can be divided into zero-dimensional (0D) models, quasi-dimensional multi-zone

Table 1
Specifics of the injector nozzle used.

Orifice number Orifice diameter, mm Orifice length, mm Maximum needle lift, mm

5 0.35 1.1 0.32

4
V. Markov et al. Case Studies in Thermal Engineering 35 (2022) 102129

Fig. 1. Schematics of the nozzle geometry model (a) and the nozzle mesh model (b).

models, and multi-dimensional CFD models. 0D models predict free spray characteristics (such as spray cone angle, atomization
fineness and spray tip penetration) by using empirical correlations versus several dimensionless criteria or parameters derived from
injection velocity, nozzle geometry, fuel properties, and spray ambient conditions. In quasi-dimensional multi-zone models, fuel sprays
are divided into different zones. Thermodynamic parameters and evaporation rate are calculated for every elementary zone, and the
mass transfer between zones is usually ignored. Spray geometry is determined with the help of empirical correlations. Interactions
between spray and walls and the effect of swirl vortex can be included by using similarity criterion or empirical correlations. These
spray models are also known as phenomenological spray models and are widely used in scientific research and engineering practice
attributed to their high computational efficiency and the convenience of introducing these models into engine simulation software.
Moreover, it is appropriate to use 0D models to conduct a comparative analysis of free non-evaporating spray characteristics in a
stagnant gas ambient (i.e., without taking into account the interaction between fuel spray and walls) and to use quasi-dimensional
multi-zone models to predict evaporating spray characteristics under combustion chamber condition with taking into account
spray/wall interaction and variations of ambient conditions. Studies on spray characteristics of biofuels by using 0D models and quasi-
dimensional multi-zone models have been reported in works [43–47].

2.2.1. 0D spray models of stabilized free spray

Variations of nozzle flow characteristics induced by fuel property differences inevitably change the spray characteristics. In this
study, to analyze the atomization fineness (diameters of fuel droplets) and spray angle of SVO and DF in a stagnant gas ambient during
the stabilized stage of spray development and analyze the relationship between these parameters and the turbulence of fuel flow at the
orifice exit, 0D spray models developed by Lyshevsky AS [48] and Kutovoy VA [49] have been used. Since these 0D spray models were
created based on the data from experiments carried out on an injector close to the investigated injector in terms of geometric di­
mensions, therefore, additional verification of these models was not performed. In the Lyshevsky model [48], parameters of atomi­
zation fineness – arithmetic mean diameter (AMD), volume mean diameter (VMD), and Sauter mean diameter (SMD) are calculated by
following correlations
( / )− 0.266 0.0733
AMD = 1.8dorif Weρg ρl M ; (3)

( / )−
(4)
0.266
VMD = 2.21dorif Weρg ρl M 0.0733 ;

( / )−
(5)
0.266
SMD = 2.68dorif Weρg ρl M 0.0733 ,

where dorif is the diameter of nozzle orifice; ρl and ρg are the density of fuel liquid and density of ambient gas, respectively; We and M
are the Weber number and the reciprocal of the Laplace number (La), respectively, based on properties of fuel liquid (σl – surface
tension coefficient of fuel liquid, μl – dynamic viscosity of fuel liquid), injection velocity Vinj , and orifice diameter:

5
V. Markov et al. Case Studies in Thermal Engineering 35 (2022) 102129

/
We = Vinj 2 ρl dorif σl ; (6)
/( )
M = μl 2
ρl dorif σ l . (7)

It is needed to note that the criterion M is equivalent to the Weber number divided by the square of the Reynolds number. The
injection velocity Vinj is determined by the well-known Bernoulli’s equation for inner tube flow:
√̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅
Vinj = 2(pin − pamb )/ρl , (8)

where pin and pamb are injection pressure and ambient pressure, respectively. Kutovoy VA has developed a correlation for calculating
median mean diameter (MMD) of fuel droplets, which has the following form [49]:
[ ]
( / ) p1 ( / )0.125
MMD = d0 1 + dorif d2 + 2(1 − pave / (2p1 )) ( / ) lorif dorif 0.15
(νl /ν0 ) , (9)
pave + (p1 /π) d2 dorif

where lorif is the orifice length; pave is the average injection pressure (in the calculation, this value is equal to pin ); νl is the kinematic
viscosity of fuel liquid; the constants have the following values: d0 = 16.5 μm, d2 = 0.3 mm, p1 = 15Mpa, and ν0 = 5.23 mm2/s.
Aside from atomization fineness parameters described above, another important parameter for spray geometry is spray angle β,
which characterizes the coverage of the combustion chamber by fuel jets and the uniformity of the distribution of fuel droplets over the
combustion chamber. In the present study, the spray angle is calculated by using the correlation developed by Lyshevsky AS [48],
which correlates the tangent value of half spray angle β/2 based on the main section of the fuel spray to the Weber number We,
criterion M, and the ratio of gas density ρg to liquid density ρl :
/ ( / )0.5
tan(β / 2) = 0.0112We0.32 M 0.07 ρg ρl . (10)

2.2.2. Quasi-dimensional multi-zone spray model in DIESEL-RK

In order to analyze the spray and atomization characteristics of different fuels in real combustion chamber conditions of the D-
245.12S diesel engine during the transient injection process, an engine model of D-245.12S created and validated in the software
DIESEL-RK developed by professor Kuleshov AS (Bauman Moscow State Technical University) has been used. The main specifics of the
D-245.12S diesel engine have been given in Table 2. It is worth noting that in the combustion chamber of this diesel engine, the fuel-air
mixture is formed by combining spatial atomization with the fuel film evaporation on the hot surface of the piston bowl, ensuring the
stable ignition of RO and DF-RO blends. DIESEL-RK is an engine simulation tool based on a thermodynamic engine model combined
with a multi-zone spray combustion model and designed for computational studies of the working processes in internal combustion
(IC) engines of different classes and purposes. DIESEL-RK has gained high recognition in Russia and other countries because of its
excellent computational efficiency, high computational accuracy, great capacity of physical processes embedded, and optimization
tools. Since the main purpose of this work is to analyze the spray characteristics of different fuels, only the spray model (the RK spray
model) is described here. A detailed description of DIESEL-RK can be found on the software website [35].
The RK spray model was firstly developed by professor Razleytsev NF and then improved and perfected by professor Kuleshov AS
[50–52]. The RK spray model takes into account fuel injection characteristics (including multiphase injection strategies), fuel at­
omization fineness, fuel spray direction, dynamics of fuel spray development, interactions of the fuel spray with the air vortex and wall,
formation of near-wall flow, and interactions between fuel sprays. This spray model is a quasi-dimensional multi-zone spray model, in
which fuel spray is divided into several characteristic zones with different evaporation and combustion conditions. Fig. 2 shows the
schematic of the fuel spray structure in the combustion chamber of the D-245.12S diesel engine. Before the spray impinges on the
combustion chamber wall, the fuel spray is described with three characteristic zones: dense axial core, dense forward front, and dilute
spray shell. The spray tip penetration is calculated with the empirical relations developed by Lyshevsky AS and modified by Razleytsev
NF [48,50]:
( / )− 0.6
Ltip = 10.37dorif We0.25 M 0.4 ρl ρg (τs /τb )0.7 exp( − 0.2τs / τb ), if τs ≤ τb ; (11)
/( )
Ltip = dorif 0.5 Vinj ave
0.5
ρg 0.5 We0.105 M 0.08 τs 0.5 2.63ρl 0.5 , if τs > τb (12)

Table 2
Specifics of the D-245.12S diesel engine.

Parameter Value

Engine type 4-cylinder, 4-stroke, direct injection, turbocharged, compression ignition

Total cylinder capacity, L 4.32
Compression ratio 15.1
Rated speed, rpm 2400
Rated power, kW 80
Fuel injection system In-line plunger pump
Initial injection pressure, MPa 21.0

6
V. Markov et al. Case Studies in Thermal Engineering 35 (2022) 102129

Fig. 2. Calculation schematic of fuel spray structure in RK spray model: 1 – dense axial core; 2 – dense forward front; 3 – dilute spray shell; 4 – axial conical core of
near-wall flows; 5 – dense core of near-wall flows on piston bowl surface; 6 – forward front of near-wall flows; 7 – dilute shell of near-wall flows.

where Vinj_ave is the average injection velocity; τs is the current time after the start of injection; τb is the jet breakup time and calculated
as
/( )
τb = 500dorif ρg 0.2 We0.29 M 0.64 Vinj ave ρl 0.2 . (13)

The spray angle β is modeled as follows [48,50]:

/ ( / )0.5
tan(β / 2) = 0.0075We0.32 M 0.07 ρg ρl (τs /τb )0.24 exp(0.07τs / τb ), if τs ≤ τb ; (14)

/ ( / )0.5 ( /( ))0.12
tan(β / 2) = 0.008We0.32 M 0.07 ρg ρl ρl dorif 3 σl τs 2 , if τs > τb . (15)

In every calculation time step, the fuel distribution among the characteristic zones is calculated based on the injection profile and
the elementary fuel portion injected at every time step [52]. After spray impingement on walls, the following characteristic zones are
generated: axial conical core of near-wall flows, dense core of near-wall flows on piston bowl surface, forward front of near-wall flows,
and dilute shell of near-wall flows. The evolution of near-wall flows in radial and tangential directions is calculated by taking into
account the impingement angle between the fuel spray and the wall in both two planes and the effect of the swirl vortex. The fuel
distribution among the near-wall flow zones is calculated based on the volume of these zones [52].
The fuel evaporation is modeled for the zones of intensive heat transfer, and the fuel evaporation in the core zones is neglected due
to insignificant heat transfer. The evaporation rate in each evaporating zone is a sum of the evaporation rate of a single fuel droplet,
which is calculated based on the equation of Sreznevsky BI:

dk 2 = d0 2 − K τu , (16)

Table 3
Physicochemical properties of DF, RO, and DF-RO blends.

Property Fuel

DF 80% DF + 20% 60% DF + 40% 40% DF + 60% 20% DF + 80% RO

RO RO RO RO

Molar mass, kg/mol 223.5 355.4 487.3 619.2 751.1 883.0

Density (40 оС), kg/m3 822.7 837.4 852.1 866.9 881.6 896.3
Kinematic viscosity (40 оС), mm2/s 2.4 4.6 8.6 15.1 25.5 41.5
Surface tension (40 оС), 10− 3N/m 26.4 27.5 28.7 29.8 31.0 32.1
Lower calorific value, MJ/kg 42.5 41.5 40.5 39.5 38.5 37.5
Cetane number 46 44 42 40 38 36
Conditional activation energy of preflame reactions, 22.0 22.4 22.8 23.2 23.6 24.0
kJ/mol
Vaporization enthalpy, kJ/kg 250 260 270 280 290 300
Specific heat at injector temperature, J/(kg⋅K) 1850 1830 1810 1790 1770 1750
Saturation pressure at 480 К/710 К, bar 0.047/ 0.0380/ 0.0285/ 0.0190/ 0.0095/ 0.000/
1.616 1.297 0.978 0.659 0.340 0.021
Element weight content,
С 0.870 0.850 0.830 0.810 0.790 0.770
Н 0.126 0.125 0.124 0.123 0.122 0.121
О 0.004 0.025 0.046 0.067 0.088 0.109

7
V. Markov et al. Case Studies in Thermal Engineering 35 (2022) 102129

where dk is the current droplet diameter; d0 is the initial droplet diameter determined by formula (3); τu is the current time from the
droplet enters the zone; K is the evaporation constant. K is calculated from the equation:
/
K = 4 × 106 NuD Dfv ps ρf , (17)

where NuD is the Nusselt number for diffusion process; Dfv is the diffusion coefficient for fuel vapor; ps is the saturation pressure of fuel
vapor. The overall evaporation rate is achieved by summing evaporation rates in all evaporating characteristics zones.

2.3. Fuel properties

In the study, nozzle flow characteristics of DF and RO and spray characteristics of DF, RO, and DF-RO blends have been simulated
and analyzed. The data on the composition and lower calorific value of these fuels were taken from Ref. [41]. The density and viscosity
of these fuels were determined from the correlations proposed in Ref. [53]. The data on the surface tension of these fuels were taken
from Refs. [54,55]. The saturation pressure of DF vapor was taken from Ref. [54], and the saturation pressure of RO vapor was
calculated based on the Clausius Clapeyron equation and Fragment based-approach with taking into account the composition of RO
[56]. The vaporization enthalpy and specific heat of DF were determined based on the correlations proposed in Ref. [54]. The
vaporization enthalpy and specific heat of RO were determined from data in Refs. [57,58]. Some fuel physicochemical properties used
in simulations have been summarized in Table 3.

Fig. 3. Distribution of the volume fraction of cavitation vapors (VF2), the velocity (U), and TKE of DF and RO in the longitudinal section of the nozzle orifice under an
injection pressure of 51.5 MPa and backpressure of 8.878 MPa.

8
V. Markov et al. Case Studies in Thermal Engineering 35 (2022) 102129

3. Results and discussion

3.1. Nozzle internal flow
Steady flows of petroleum DF and RO in the injector nozzle have been simulated at the full needle lift and under injection pressures
of 51.5, 40, and 20 MPa and constant backpressure of 8.878 MPa by using the numerical models presented in section 2.1. The dis­
tributions of the following basic flow parameters inside the nozzle have been achieved: flow velocity, turbulent kinetic energy (TKE),
and volume fraction of cavitation vapors (VF2). Different distribution structures of flow parameters have been observed between DF
and RO. Fig. 3 shows the distributions of these flow parameters in the longitudinal section of the nozzle orifice under an injection
pressure of 51.5 MPa and backpressure of 8.878 MPa. Different cavitation patterns inside the orifice are generated for DF and RO
attributed to the significant difference in saturation pressure between DF and RO (DF is prone to cavitate). The cavitation of DF reaches
up to the end region of the orifice, but the cavitation vapors of RO have completely collapsed upstream of the middle of the orifice. The
intensive cavitation of DF inside the upper region of the orifice, narrowing the flow area, coupled with the smaller flow resistance for
DF, leads to a significantly higher velocity in the lower region of the orifice for DF in comparison with RO. The generation, devel­
opment, and dissipation of turbulence (TKE) inside the orifice are closely related to the evolution of cavitation. Downstream of the
cavitation zone, the flow area is suddenly largened, which contributes to producing vortex and increasing TKE. The generated TKE of
the fuel flow dissipates during its further flow inside the orifice. The cavitation collapse occurs in the entrance region of the orifice for
RO and in the end region of the orifice for DF, so RO has higher TKE in the entrance region, while DF has higher TKE in the end region,
as can be seen from Fig. 3.
In order to quantitatively analyze the flow characteristics of DF and RO inside the nozzle orifice, the original calculated data have
been processed by using a uniaxial coordinate system. The center point of the inlet cross-section of the orifice is selected as the original
point of this system. The coordinate axis (lp) coincides with the orifice axis. Thus, each cross-section of the orifice has its coordinate
value, and the orifice exit has a coordinate lp = 1.1 mm (the orifice length is 1.1 mm). Flow parameters have been averaged over each
cross-section of the orifice. Fig. 4 shows the averaged axial velocity, TKE, and volume fraction of cavitation vapors for DF and RO in
different cross-sections of the orifice. It can be found that in comparison with RO, DF has higher velocity in all cross-sections at all
injection pressures. The averaged velocity at the orifice exit for RO at an injection pressure of 51.5 MPa is close to that for DF at an
injection pressure of only 40 MPa. Under high injection pressures (51.5 and 40 MPa), the averaged TKE of RO is obviously higher than
that of DF in the middle region of the orifice but is lower than that of DF at the orifice exit. This can be explained by the effect of
cavitation on turbulence development inside the orifice, as discussed above. The cavitation intensity for DF is significantly higher than
that of RO at injection pressures of 51.5 and 40 MPa. The collapse of DF cavitation in the end region of the orifice induces turbulence

Fig. 4. Change of axial velocity, TKE, and volume fraction of cavitation vapors (VF2) averaged over the cross-section of the nozzle orifice along the orifice axis.

9
V. Markov et al. Case Studies in Thermal Engineering 35 (2022) 102129

growth, resulting in higher averaged TKE at the orifice exit. At an injection pressure of 20 MPa, cavitation is not formed for both fuels
(not plotted). At this pressure, DF has a higher average TKE in all cross-sections of the office due to a highviscosity of RO. In any case,
DF always has greater velocity and TKE at the orifice exit.
Although the number of studies on the flow characteristics of SVO in injector nozzles is limited, similar results have been reported
in research works on the flow characteristics of VOB and DF in nozzles of diesel injectors [17–19]. Based on the comparison of the
results of the present study with the results in work [19], it should be noted the relatively more significant difference in cavitation
intension between SVO and DF compared to the difference between VOB and DF under the same pressure conditions due to the
property differences between SVO and VOB. In other words, it is not suitable to quantitatively express the nozzle flow characteristics of
SVO with those of VOB, which is especially important during injection system design.

3.2. Effect of TKE at the orifice exit on parameters of free spray

The simulation results on the nozzle flow of DF and RO show that the fuel properties have a significant impact on the flow pa­
rameters at the nozzle orifice exit. The difference in properties between DF and RO results in a noticeable variation in TKE at the orifice
exit during substituting DF with RO. The aforementioned flow parameters in the fuel injection system predetermine the dispersity
(fineness) of fuel atomization.
Sprays of DF and RO in a stagnant gas ambient (free spray) have been modeled by using the spray models described in section 2.2.1.
The injection pressures and ambient back pressure are the same as those in nozzle flow simulations. The obtained results about the
atomization fineness parameters (mean diameters of fuel droplets) and spray angle have been given in Table 4. It can be found that
compared with DF, RO always produces larger droplet diameters and smaller spray angles under all injection pressures studied. This is
consistent with the experimental results on the effect of fuel properties on atomization reported in Refs. [19,22,27]. TKE of the fuel jet
at the orifice exit has also been included in Table 4 in order to analyze the relationship between the fuel jet turbulence at the orifice exit
and the atomization parameters. AMD and MMD have been plotted with respect to TKE at the orifice exit in Fig. 5. It is noted that as
TKE increases from 500 to 2500 m2/s2, AMD and MMD decrease by about half. An increase in the fuel jet turbulence contributes to
accelerating the primary breakup process, resulting in decreased diameters of fuel droplets. Decreased SMD induced by increased
turbulence of fuel jet has been reported for identical fuel in Ref. [59] and for VOB and DF in Ref. [18].
In order to quantitatively evaluate the interrelation between fuel jet turbulence at the orifice exit and atomization fineness pa­
rameters, the authors have proposed a complex that represents the quotient of the averaged TKE at the orifice exit and the kinetic
viscosity of fuel to the power of n (TKE/νn). The use of the proposed complex makes it possible to obtain correlation dependencies for
determining the mean diameters of fuel spray droplets during injection and atomization. Eventually, the following empirical corre­
lations for atomization fineness parameters have been fitted based on the data presented in Table 4 and have been presented in Table 5.
The correlation coefficient is close to unit one, which means the correlations for atomization fineness parameters have been suc­
cessfully created by using the complex TKE/ν0.5. The complex has the form of TKE/νn because the increase of fuel jet turbulence
contributes to accelerating the process of spray breakup, while the high fuel viscosity delays the breakup of fuel droplets. Therefore,
atomization fineness should be directly proportional to TKE and inversely proportional to fuel viscosity. The exponent n in the de­
nominator of the complex TKE/νn is determined as 0.5 because the kinematic viscosity of the studied fuels varies widely. At a tem­
perature of 40 ◦ C, the kinematic viscosity of DF is 2.4 mm2/s, while the kinematic viscosity of RO is 41.5 mm2/s. The exponent of 0.5 is
optimal to represent the dependence of the atomization fineness on the kinematic viscosity of fuel. Fig. 6 shows the modeled points of
atomization fineness parameters and lines plotted according to the developed correlations for mean diameters of spray droplets. The
good coincidence between the lines and points and the high correlation coefficient (0.989–0.994) indicate that the developed cor­
relations can be used to determine the mean diameters of the spray droplets during fuel injection.
The effect of TKE at the orifice exit on the spray angle has been analyzed by using the proposed complex TKE/ν0.5. The modeled
values of the tangent of half spray angle (tan(β/2)) with respect to TKE/ν0.5 have been presented in Fig. 6. The increase of TKE and the
decrease of fuel viscosity result in a higher spray angle. By fitting these data, a correlation of spray angle with TKE/ν0.5 has been
obtained:
( / )0.2749
tan(β / 2) = 4.41 × 10− 3 TKE υ0.5 . (22)

The correlation coefficient is 0.9968. This means that the developed correlation perfectly reproduces the dependence of spray angle
on fuel jet turbulence and fuel viscosity and can be used to determine the spray angle during fuel injection.

Table 4
The modeled atomization fineness parameters and spray angle for the free spray of DF and RO and TKE at the orifice exit under different injection pressure pin .

Parameters pin = 51.5 MPa pin = 40 MPa pin = 20 MPa

DF RO DF RO DF RO

AMD, μm 25.1 40.8 27.3 44.4 35.8 58.3

VMD, μm 30.8 50.1 33.5 54.5 44.0 71.6
SMD, μm 37.4 60.8 40.6 66.1 53.4 86.8
MMD, μm 31.4 48.1 35.4 54.3 50.8 77.8
Spray angle β, ◦ 26.3 16.2 23.9 14.6 17.3 10.6
TKE at the orifice exit, m2/s2 2574.4 2208.1 2057.6 1483.2 623.4 371.4

10
V. Markov et al. Case Studies in Thermal Engineering 35 (2022) 102129

Fig. 5. Relationship between TKE at the orifice exit and atomization fineness parameters. (Points are the modeled diameters, and lines are the fitted trend lines.)

Table 5
The developed correlations between atomization fineness parameters and the complex TKE/ν0.5 and their correlation coefficient R.

Correlation R

AMD = 165.7 − 9.85 ln(TKE /ν0.5 ) [μm] (18) 0.9940

VMD = 203.4 − 12.09 ln(TKE /ν0.5 ) [μm] (19) 0.9941
SMD = 246.7 − 14.66 ln(TKE /ν0.5 ) [μm] (20) 0.9943
MMD = 220.5 − 13.25 ln(TKE /ν0.5 ) [μm] (21) 0.9886

Fig. 6. Relationship of atomization fineness parameters (AMD, VMD, SMD, MMD) and tan(β/2) with TKE/ν0.5. (Points are the modeled results, and lines are plotted
according to the developed correlations for mean diameters and spray angle.)

Table 6
Fuel injection profile for the operating condition with an engine speed of 2400 rpm and a cycle fuel delivery of 80 mm3.

Crank angle, ◦ Injection rate, m3/s Crank angle, ◦ Injection rate, m3/s Crank angle, ◦
Injection rate, m3/s
5 5
0 0 6.0 8.6 × 10− 12.0 8.8 × 10−
5 5 5
1.0 2.2 × 10− 7.0 8.5 × 10− 13.0 8.2 × 10−
5 5 5
2.0 5.6 × 10− 8.0 8.4 × 10− 14.0 6.4 × 10−
5 5 5
3.0 7.6 × 10− 9.0 9.5 × 10− 15.0 4.4 × 10−
5 5
4.0 8.0 × 10− 10.0 9.4 × 10− 16.0 0
5 5
5.0 8.4 × 10− 11.0 9.2 × 10−

11
V. Markov et al. Case Studies in Thermal Engineering 35 (2022) 102129

3.3. Spray in real combustion chamber conditions during the transient injection process
The nozzle flow and spray characteristics of petroleum DF and RO, considered above, were obtained by simulating the steady flow
of these fuels inside the injector nozzle of the D-245.12S diesel engine at the full needle lift. Under the real operating conditions of a
diesel engine, the fuel injection system operates in an impulse mode – the fuel is supplied at the end of the compression stroke. In this
regard, it is advisable to study the spray characteristics of DF and RO in real combustion chamber conditions of the D-245.12S diesel
engine and during the transient injection process. Spray simulations of DF, RO, and DF/RO blends have been performed by using the
diesel engine model of D-245.12S created in DIESEL-RK. The diesel engine was modeled at the rated operating condition with an
engine speed of 2400 rpm. The fuel injection profile used was given in Table 6, and the cycle fuel delivery was 80 mm3. The volume
fraction of RO in DF/RO blends was changed from 0% (pure DF) to 100% (pure RO).
The characteristics of the atomization and mixture formation processes for the investigated fuels in the rated operating condition
have been obtained. Fig. 7 shows the spray tip penetration, SMD, and spray angle for different DF/RO blends at the end of injection. It
is noted that the characteristics of spray and atomization processes are deteriorated with increasing the content of RO in fuel blends. As
the RO content increases from 0% to 100%, the SMD increases from 24.3 to 39.4 μm (by 62.1%), the spray tip increases from 41.9 to
48.4 (by 15.5%), and the spray angle reduces from 20.8 to 15.8◦ (by 24.0%). The fuel mass fraction in the different characteristics
zones of the fuel spray at the end of injection has also been included in Fig. 7. With increasing the RO content from 0% to 100%, the fuel
mass fraction in the dense core zones of the spray increases from 5.8 to 23.1%, the fuel mass fraction in the dilute shell zones of the
spray decreases from 77.1 to 34.9%, and the fuel mass fraction in the near-wall flow zones increases from 17.1 to 42.0%. These results
mean that the increase of the RO content in fuel blends delays the process of fuel-air mixture formation and increases the amount of
fuel sprayed onto the combustion chamber walls.
The presented data indicate that from the point of view of the quality of the fuel atomization and fuel-air mixture formation, the
operation of the investigated diesel engine is possible at any RO content in DF/RO blends. At the same time, a number of studies on
diesel engines fueled with pure SVO or blends of DF and SVO have shown that the long-term operation of an engine on these fuels can
be accompanied by coking of nozzle orifices and carbon deposits on combustion chamber walls [11,60]. In this regard, diesel engines
are often fueled with not pure SVO but on petroleum DF with small additives of SVO [7,8,11,14].
It is known that the quality of fuel atomization and fuel-air mixture formation has a significant effect on the indicators of fuel
efficiency and exhaust emissions of diesel engines [29–31]. In particular, the emissions of nitrogen oxides (NOx), as the main gaseous
toxic component in the exhaust gas, are directly related to the uniformity of the distribution of fuel droplets over the combustion
chamber [29]. The most active oxidation of air nitrogen occurs in the combustion chamber zones with the minimum excess air co­
efficients and the maximum combustion temperatures.

3.4. Dependence of NOx emissions on SMD

In order to analyze the dependence of the emission of nitrogen oxides from the exhaust gas of a diesel engine on the droplet size of
fuel spray, the actual experimental data [29] on the volume concentration of nitrogen oxides in the exhaust gas (CNOx) of the

Fig. 7. Spray parameters and fuel mass distribution among characteristic zones at the end of fuel injection for different DF/RO blends. Ccore – fuel mass fraction in
dense core zones; Cshell – fuel mass fraction in dilute shell zones; Cnear_wall – fuel mass fraction in neal-wall flow zones.

12
V. Markov et al. Case Studies in Thermal Engineering 35 (2022) 102129

aforementioned D-245.12S diesel engine, operating on DF, have been used (Fig. 8). The description of the experimental setup and
procedures is referred to our previous works [14,41]. These data indicate that the maximum content of nitrogen oxides in the exhaust
gas of this diesel engine was recorded in the operating modes of the external characteristic curve with the maximum fuel delivery and
the maximum mean effective pressure Pe.
It should be noted that the concentration of nitrogen oxides in the exhaust gas depends on many factors – engine speed and load,
fuel properties, parameters of fuel injection and atomization (injection pressure, number of fuel sprays, spray length, spray angle,
droplet size, etc.) [6,27,29–31]. The data on the volume concentration of nitrogen oxides in the exhaust gas, the injection pressure, and
the SMD of the fuel spray of the D-245.12S diesel engine fueled with DF in the operating modes of the external characteristic curve
have been given in Table 7. It should be noted that the dependence of NOx emissions on SMD was investigated not in one operating
mode but in a set of operating modes of the external characteristic curve. In these operating modes, NOx emissions are affected not only
by the size of fuel droplets but also by other parameters - the injection velocity, the time allotted for the combustion process, com­
bustion efficiency – combustion temperatures, etc. As a result, in the considered operating modes of the external characteristic curve, a
proportional relationship is observed between SMD and NOx emissions – with an increase in SMD, the volume concentration of ni­
trogen oxides in the exhaust gas increases. The analogous results about the variation trend of NOx emissions with respect to operating
modes of the external characteristic curve have also been reported in Refs. [61,62]. When this dependence is considered in one
operating mode, it can be the opposite – with an increase in SMD, NOx emissions decrease.
Based on the data in Table 7, a correlation analysis of the dependence of the volume concentration of nitrogen oxides СNOx in the
exhaust gas of the D-245.12C diesel engine on the droplet size of fuel spray has been carried out. The dependence of СNOx on SMD can
be well described by the following exponential formula:
CNOx = 3.0538 exp(0.2323SMD) [ppm]. (23)
In this formula, the unit of SMD is μm. The original data of СNOx and SMD and the fitted line according to this formula have been
presented in Fig. 9. The correlation coefficient of this dependence is 0.9972.
Of course, as noted above, the NOx emissions in the operating modes of the external characteristic curve depend not only on the
parameters, characterizing the quality of fuel atomization (droplet diameter, spray length, spray angle, etc.) but also on a number of
other factors. However, the analysis performed confirms that there is a close correlation between NOx emissions and fuel atomization
fineness. In this regard, in order to improve the emission performance of diesel engines, it is necessary to further improve the processes
of fuel injection and fuel atomization. This problem becomes especially critical when diesel engines are fueled with SVO, possing high
viscosity and density. The methods proposed in the present work make it possible to evaluate the parameters, characterizing the
atomization quality of these fuels and their influence on the emission indicators of diesel engines.

4. Conclusions
In the study, nozzle flow and atomization characteristics of DF and RO have been investigated. The following conclusions can be
drawn.
1. The distribution structures of flow parameters inside the nozzle orifice are different between DF and RO. The development of TKE
inside the orifice is closely related to cavitation patterns. Moreover, compared with DF, RO has lower injection velocity and TKE at
the orifice exit under all injection pressure studied.
2. The relationship between the atomization parameters (mean diameters of spray droplets and spray angle) and TKE at the orifice exit
has been investigated and analyzed for free spray of DF and RO. Correlations have been successfully developed by using the
proposed complex TKE/ν0.5 that reflects the relationship between TKE and fuel viscosity.
3. As for sprays of DF/RO blends in real combustion chamber conditions, the relationship between the basic spray parameters and fuel
types has been confirmed. When the RO content in DF/RO blends increases from 0% to 100%, at the end of injection, the SMD
increases from 24.3 to 39.4 μm, the spray length increased from 41.9 to 48.4 mm, and the spray angle decreases from 20.8◦ to 15.8◦ .
At the same time, the fuel mass fraction in the dense core zones of fuel spray increases from 5.8 to 23.1%, the fuel mass fraction in
the dilute shell zones of fuel spray decreases from 77.1 to 34.9%, and the fuel mass fraction in the near-wall flow zones increases
from 17.1 to 42.0%.
4. The dependence of the NOx emissions of the D-245.12S diesel engine on the droplet size of the fuel spray has been analyzed. An
empirical correlation with a correlation coefficient of 0.9972 has been developed.
The methods and empirical correlations proposed in the study can be used to evaluate the spray parameters of vegetable oil-based
biofuels and their influence on emission indicators. In order to expand the use of vegetable oil-based fuels in CI engines and their
installations, it is necessary to investigate the effectiveness of various methods for improving the quality of atomization and fuel-air
mixture formation in relation to specific application scenarios, such as optimization of the parameters of the fuel injection system,
optimization of nozzle geometry and implementation of a turbulator in nozzles to increase the turbulence of the fuel jet, fuel pre­
heating, the choice of an appropriate method of fuel-air mixture formation, optimization of combustion chamber geometry, the use of
fuel additives (including low viscosity fuels and nanoparticles).

Credit authorship contribution statement

Vladimir Markov: Conceptualization, Methodology, Formal analysis, Investigation, Writing – original draft, Writing – review &

13
V. Markov et al. Case Studies in Thermal Engineering 35 (2022) 102129

Fig. 8. The volume concentration of nitrogen oxides in the exhaust gas (CNOx) of the D-245.12S diesel engine at different engine speeds and different engine loads
(mean effective pressure – Pe).

Table 7
The volume concentration of nitrogen oxides in the exhaust gas (CNOx), maximum injection pressure (Pinj_max), and SMD of fuel spray droplets of the D-245.12S diesel
engine fueled with DF in the operating modes of the external characteristic curve.

Parameter Value

Engine speed, rpm 1000 1200 1400 1600 1800 2000 2200 2400
Pinj_max, MPa 34.3 38.0 41.0 43.5 46.0 48.3 50.2 51.9
SMD, μm 24.5 23.9 23.2 22.6 22.0 21.6 21.2 20.8
СNOx, ppm 930 780 660 570 510 460 420 390

Fig. 9. The dependence of volume concentration of nitrogen oxides in the exhaust gas (CNOx) on SMD for the D-245.12S diesel engine (points – original data; line –
fitted result).

editing. Bowen Sa: Project administration, Investigation, Software, Writing – original draft, Writing – review & editing. Sergey
Devyanin: Methodology, Investigation, Software, Validation. Leonid Grekhov: Funding acquisition, Formal analysis, Writing – re­
view & editing. Vsevolod Neverov: Investigation, Software. Jianhui Zhao: Formal analysis, Writing – review & editing.

14
V. Markov et al. Case Studies in Thermal Engineering 35 (2022) 102129

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to
influence the work reported in this paper.

Acknowledgments
This study was supported by Russian Science Foundation (Grant No: 21-49-00012), the Ministry of Science and Higher Education of
the Russian Federation (Grant No: 075-15-2021-1028).

References
[1] D. Singh, D. Sharma, S.L. Soni, S. Sharma, P. Kumar Sharma, A. Jhalani, A review on feedstocks, production processes, and yield for different generations of
biodiesel, Fuel 262 (2020), 116553, https://doi.org/10.1016/j.fuel.2019.116553.
[2] M.S. Gad, Z. He, A.S. EL-Shafay, A.I. EL-Seesy, Combustion characteristics of a diesel engine running with Mandarin essential oil -diesel mixtures and propanol
additive under different exhaust gas recirculation: experimental investigation and numerical simulation, Case Stud. Therm. Eng. 26 (2021), 101100, https://doi.
org/10.1016/j.csite.2021.101100.
[3] G. Di Blasio, R. Ianniello, C. Beatrice, Hydrotreated vegetable oil as enabler for high-efficient and ultra-low emission vehicles in the view of 2030 targets, Fuel
(2022) 310, https://doi.org/10.1016/j.fuel.2021.122206.
[4] R.D. Reitz, H. Ogawa, R. Payri, T. Fansler, S. Kokjohn, Y. Moriyoshi, A.K. Agarwal, D. Arcoumanis, D. Assanis, C. Bae, K. Boulouchos, M. Canakci, S. Curran,
I. Denbratt, M. Gavaises, M. Guenthner, C. Hasse, Z. Huang, T. Ishiyama, B. Johansson, T.V. Johnson, G. Kalghatgi, M. Koike, S.C. Kong, A. Leipertz, P. Miles,
R. Novella, A. Onorati, M. Richter, S. Shuai, D. Siebers, W. Su, M. Trujillo, N. Uchida, B.M. Vaglieco, R.M. Wagner, H. Zhao, IJER editorial: the future of the
internal combustion engine, Int. J. Engine Res. 21 (2019) 3–10, https://doi.org/10.1177/1468087419877990.
[5] P. Dey, S. Ray, Comparative analysis of waste vegetable oil versus transesterified waste vegetable oil in diesel blend as alternative fuels for compression ignition
engine, Clean Technol. Environ. Policy 22 (2020) 1517–1530, https://doi.org/10.1007/s10098-020-01892-1.
[6] M. Costa, L. Marchitto, D. Piazzullo, M.V. Prati, Comparison between the energetic and environmental performance of a combined heat and power unit fueled
with diesel and waste vegetable oil: an experimental and numerical study, Renew. Energy 168 (2021) 791–805, https://doi.org/10.1016/j.renene.2020.12.099.
[7] M. Dabi, U.K. Saha, Application potential of vegetable oils as alternative to diesel fuels in compression ignition engines: a review, J. Energy Inst. 92 (2019)
1710–1726, https://doi.org/10.1016/j.joei.2019.01.003.
[8] D.H. Qi, C.F. Lee, C.C. Jia, P.P. Wang, S.T. Wu, Experimental investigations of combustion and emission characteristics of rapeseed oil-diesel blends in a two
cylinder agricultural diesel engine, Energy Convers. Manag. 77 (2014) 227–232, https://doi.org/10.1016/j.enconman.2013.09.023.
[9] Z. Franco, Q.D. Nguyen, Flow properties of vegetable oil–diesel fuel blends, Fuel 90 (2011) 838–843, https://doi.org/10.1016/j.fuel.2010.09.044.
[10] L. Zong, S. Ramanathan, C.-C. Chen, Fragment-based approach for estimating thermophysical properties of fats and vegetable oils for modeling biodiesel
production processes, Ind. Eng. Chem. Res. 49 (2010) 876–886, https://doi.org/10.1021/ie900513k.
[11] S. Che Mat, M.Y. Idroas, M.F. Hamid, Z.A. Zainal, Performance and emissions of straight vegetable oils and its blends as a fuel in diesel engine: a review, Renew.
Sustain. Energy Rev. 82 (2018) 808–823, https://doi.org/10.1016/j.rser.2017.09.080.
[12] S. Ramalingam, G. Sankaranarayanan, S. Senthil, R.A. Rohith, R. Santosh Kumar, Effect of Cerium oxide nanoparticles derived from biosynthesis of Azadirachta
indica on stability and performance of a research CI engine powered by Diesel-Lemongrass oil blends, Energy Environ. (2022), 0958305X221077386, https://
doi.org/10.1177/0958305X221077386.
[13] D. Qi, L. Ma, R. Chen, X. Jin, M. Xie, Effects of EGR rate on the combustion and emission characteristics of diesel-palm oil-ethanol ternary blends used in a CRDI
diesel engine with double injection strategy, Appl. Therm. Eng. 199 (2021), 117530, https://doi.org/10.1016/j.applthermaleng.2021.117530.
[14] V. Markov, V. Kamaltdinov, S. Devyanin, B. Sa, A. Zherdev, V. Furman, Investigation of the influence of different vegetable oils as a component of blended
biofuel on performance and emission characteristics of a diesel engine for agricultural machinery and commercial vehicles, Resour. 10 (2021), https://doi.org/
10.3390/resources10080074.
[15] A.T. Hoang, A.T. Le, V.V. Pham, A core correlation of spray characteristics, deposit formation, and combustion of a high-speed diesel engine fueled with
Jatropha oil and diesel fuel, Fuel 244 (2019) 159–175, https://doi.org/10.1016/j.fuel.2019.02.009.
[16] V.A. Markov, S.N. Devyanin, V.G. Semenov, V.V. Bagrov, S.A. Zykov, Motor Fuels Produced from Vegetable Oils, Lambert Academic Publishing, Riga, 2019.
[17] H.K. Suh, S.H. Park, C.S. Lee, Experimental investigation of nozzle cavitating flow characteristics for diesel and biodiesel fuels, Int. J. Automot. Technol. 9
(2008) 217–224, https://doi.org/10.1007/s12239-008-0028-3.
[18] S. Som, D.E. Longman, A.I. Ramírez, S.K. Aggarwal, A comparison of injector flow and spray characteristics of biodiesel with petrodiesel, Fuel 89 (2010)
4014–4024, https://doi.org/10.1016/j.fuel.2010.05.004.
[19] S. Yu, B. Yin, H. Jia, S. Wen, X. Li, J. Yu, Theoretical and experimental comparison of internal flow and spray characteristics between diesel and biodiesel, Fuel
208 (2017) 20–29, https://doi.org/10.1016/j.fuel.2017.06.136.
[20] R. Sathiyamoorthi, G. Sankaranarayanan, D.B. Munuswamy, Y. Devarajan, Experimental study of spray analysis for Palmarosa biodiesel-diesel blends in a
constant volume chamber, Environ. Prog. Sustain. Energy 40 (2021), https://doi.org/10.1002/ep.13696.
[21] B. Kegl, L. Lešnik, Modeling of macroscopic mineral diesel and biodiesel spray characteristics, Fuel 222 (2018) 810–820, https://doi.org/10.1016/j.
fuel.2018.02.169.
[22] J.G. Gupta, A.K. Agarwal, Macroscopic and microscopic spray characteristics of diesel and Karanja biodiesel blends, SAE Tech. Pap. 2016-01-0869 (2016),
https://doi.org/10.4271/2016-01-0869.
[23] K. Kim, O. Lim, Investigation of the spray development process of gasoline-biodiesel blended fuel sprays in a constant volume chamber, Energies 13 (2020),
https://doi.org/10.3390/en13184819.
[24] J.G. Nerva, C.L. Genzale, S. Kook, J.M. García-Oliver, L.M. Pickett, Fundamental spray and combustion measurements of soy methyl-ester biodiesel, Int. J.
Engine Res. 14 (2013) 373–390, https://doi.org/10.1177/1468087412456688.
[25] P. Boggavarapu, R.V. Ravikrishna, A comparison of evaporatingspray structure of Jatropha methyl ester and diesel, and surrogate fuels, At. Sprays. 28 (2018)
797–809, https://doi.org/10.1615/AtomizSpr.2018026885.
[26] M.H.H. Ishak, F. Ismail, S. Che Mat, M.Z. Abdullah, M.S. Abdul Aziz, M.Y. Idroas, Numerical analysis of nozzle flow and spray characteristics from different
nozzles using diesel and biofuel blends, Energies 12 (2019), https://doi.org/10.3390/en12020281.
[27] D.H. Qi, X.Q. Ding, W.B. Zhao, K. Yang, Spray characteristics and engine performance of vegetable oil–diesel–ethanol hybrid fuel, J. Energy Eng. 145 (2019),
04019011, https://doi.org/10.1061/(asce)ey.1943-7897.0000606.
[28] A. Bhikuning, R. Sugawara, E. Matsumura, J. Senda, Investigation of spray characteristics from waste cooking oil, bio-hydro fined diesel oil (BHD)and n-
tridecane in a constant volume chamber, Case Stud. Therm. Eng. 21 (2020), 100661, https://doi.org/10.1016/j.csite.2020.100661.
[29] V.A. Markov, S.P. Gladyshev, S.N. Devyanin, L.L. Mihalsky, O.V. Drobyshev, Perfection of the processes of the fuel spraying and the fuel-air mixture creating in a
high-speed diesel engine, working on the bio-fuel mixture, SAE Tech. Pap. (2009), https://doi.org/10.4271/2009-01-0845, 2009-01-0845.
[30] R. Payri, F.J. Salvador, J. Gimeno, J. de la Morena, Effects of nozzle geometry on direct injection diesel engine combustion process, Appl. Therm. Eng. 29 (2009)
2051–2060, https://doi.org/10.1016/j.applthermaleng.2008.10.009.
[31] S. Som, A.I. Ramirez, D.E. Longman, S.K. Aggarwal, Effect of nozzle orifice geometry on spray, combustion, and emission characteristics under diesel engine
conditions, Fuel 90 (2011) 1267–1276, https://doi.org/10.1016/j.fuel.2010.10.048.

15
V. Markov et al. Case Studies in Thermal Engineering 35 (2022) 102129

[32] F.J. Salvador, S. Ruiz, M. Crialesi-Esposito, I. Blanquer, Analysis on the effects of turbulent inflow conditions on spray primary atomization in the near-field by
direct numerical simulation, Int. J. Multiphas. Flow 102 (2018) 49–63, https://doi.org/10.1016/j.ijmultiphaseflow.2018.01.019.
[33] L.A. Raman, B. Deepanraj, S. Rajakumar, V. Sivasubramanian, Experimental investigation on performance, combustion and emission analysis of a direct
injection diesel engine fuelled with rapeseed oil biodiesel, Fuel 246 (2019) 69–74, https://doi.org/10.1016/j.fuel.2019.02.106.
[34] M. Shahbandeh, Production of Major Vegetable Oils Worldwide, Statista, 2022. https://www.statista.com/statistics/263933/production-of-vegetable-oils-
worldwide-since-2000/. (Accessed 26 February 2022).
[35] Diesel-RK, Bauman Moscow state tech. Univ, n.d. https://diesel-rk.bmstu.ru/. (Accessed 1 January 2022).
[36] G.H. Schnerr, J. Sauer, Physical and numerical modeling of unsteady cavitation dynamics, in: 4th Int. Conf. Multiph. Flow, 2001, pp. 1–12.
[37] E. Winklhofer, E. Kull, E. Kelz, A. Morozov, Comprehensive hydraulic and flow field documentation in model throttle experiments under cavitation conditions,
in: Proc. ILASS-Europe Conf., 2001, pp. 574–579. Zurich.
[38] M. Cristofaro, W. Edelbauer, P. Koukouvinis, M. Gavaises, A numerical study on the effect of cavitation erosion in a diesel injector, Appl. Math. Model. 78 (2020)
200–216, https://doi.org/10.1016/j.apm.2019.09.002.
[39] B. Sa, O. Klyus, V. Markov, V. Kamaltdinov, A numerical study of the effect of spiral counter grooves on a needle on flow turbulence in a diesel injector, Fuel 290
(2021), 120013, https://doi.org/10.1016/j.fuel.2020.120013.
[40] B. Sa, V. Markov, Y. Liu, V. Kamaltdinov, W. Qiao, Numerical investigation of the effect of multi-walled carbon nanotube additive on nozzle flow and spray
behaviors of diesel fuel, Fuel 290 (2021), 119802, https://doi.org/10.1016/j.fuel.2020.119802.
[41] V.A. Markov, B. Sa, S.N. Devyanin, A.A. Zherdev, P.R. Vallejo Maldonado, S.A. Zykov, A.D. Denisov, H.C. Ambawatte, Investigation of the performances of a
diesel engine operating on blended and emulsified biofuels from rapeseed oil, Energies 14 (2021) 6661, https://doi.org/10.3390/en14206661.
[42] B.A. Sforzo, K.E. Matusik, C.F. Powell, A.L. Kastengren, S. Daly, S. Skeen, E. Cenker, L.M. Pickett, C. Crua, J. Manin, Fuel nozzle geometry effects on cavitation
and spray behavior at diesel engine conditions, in: J. Katz (Ed.), Proc. 10th Int. Symp. Cavitation, ASME Press, 2018, https://doi.org/10.1115/1.861851_ch90,
0.
[43] V.A. Markov, S.N. Devyanin, V.I. Malchuk, Fuel Injection and Atomization in Diesel Engines, BMSTU, Moscow, 2007.
[44] P.R. Shah, A. Ganesh, Study the influence of pre-heating on atomization of straight vegetable oil through Ohnesorge number and Sauter mean diameter,
J. Energy Inst. 91 (2018) 828–834, https://doi.org/10.1016/j.joei.2017.10.006.
[45] F. Li, W. Fu, B. Yi, L. Song, T. Liu, X. Wang, C. Wang, Y. Lei, Q. Lin, Comparison of macroscopic spray characteristics between biodiesel-pentanol blends and
diesel, Exp. Therm. Fluid Sci. 98 (2018) 523–533, https://doi.org/10.1016/j.expthermflusci.2018.07.003.
[46] S. Salam, T.N. Verma, Appending empirical modelling to numerical solution for behaviour characterisation of microalgae biodiesel, Energy Convers. Manag.
180 (2019) 496–510, https://doi.org/10.1016/j.enconman.2018.11.014.
[47] Y. Yu, Experimental study on effects of ethanol-diesel fuel blended on spray characteristics under ultra-high injection pressure up to 350 MPa, Energy 186
(2019), 115768, https://doi.org/10.1016/j.energy.2019.07.098.
[48] A.S. Lyshevsky, Fuel Atomization Processes with Diesel Injectors, Mashgiz, Moscow, 1963.
[49] V.A. Kutovoy, Fuel Injection in Diesels, Mechanical Manufacture, Moscow, 1981.
[50] N.F. Razleytsev, Combustion Simulation and Optimization in Diesels, Vischa shkola, Kharkov, 1980.
[51] A.S. Kuleshov, Multi-zone di diesel spray combustion model and its application for matching the injector design with piston bowl shape, in: SAE Tech. Pap.,
2007, https://doi.org/10.4271/2007-01-1908.
[52] A.S. Kuleshov, Multi-zone di diesel spray combustion model for thermodynamic simulation of engine with PCCI and high EGR level, in: SAE Tech. Pap., 2009,
https://doi.org/10.4271/2009-01-1956.
[53] V. Markov, S. Devyanin, S. Zykov, B. Sa, Viscosity characteristics of multicomponent mixed biofuels based on vegetable oils (in Russian), Altern. Fuel Transp. 54
(2016) 33–49. https://ngvrus.ru/file/journal-history/tat-n6-54-2016.pdf.
[54] L. Grekhov, N. Ivashchenko, V. Markov, Fuel Delivery and Control System for Diesel Engine, Legion-Autodata, Moscow, 2005.
[55] B. Esteban, J.R. Riba, G. Baquero, R. Puig, A. Rius, Characterization of the surface tension of vegetable oils to be used as fuel in diesel engines, Fuel 102 (2012)
231–238, https://doi.org/10.1016/j.fuel.2012.07.042.
[56] B. Sa, V.A. Markov, V.G. Kamaltdinov, V.A. Neverov, Flow simulation of petroleum diesel fuel and rapeseed oil in the nozzle of a diesel injector, IOP Conf. Ser.
Mater. Sci. Eng. 1035 (2021) 12034, https://doi.org/10.1088/1757-899x/1035/1/012034.
[57] J. Parrilla, C. Cortés, Modelling of droplet burning for rapeseed oil as liquid fuel, in: Renew. Energy Power Qual. J., 2007, pp. 79–86, https://doi.org/10.24084/
repqj05.221.
[58] J. Carlos, J. Dantas, Specific heat of some vegetable oils by differential scanning calorimetry and microwave oven, in: II Congr. Bras. Plantas Ol. Óleos, Gorduras
e Biodiesel, 2005, pp. 610–614.
[59] V. Markov, B. Sa, V. Kamaltdinov, V. Neverov, A. Zherdev, Investigation on the effect of the flow passage geometry of diesel injector nozzle on injection process
parameters and engine performances, Energy Sci. Eng. 10 (2022) 552–577, https://doi.org/10.1002/ese3.1051.
[60] A.T. Hoang, V.V. Le, V.V. Pham, B.C. Tham, An investigation of deposit formation in the injector, spray characteristics, and performance of a diesel engine
fueled with preheated vegetable oil and diesel fuel, Energy Sources, Part A Recover, Util. Environ. Eff. 41 (2019) 2882–2894, https://doi.org/10.1080/
15567036.2019.1582731.
[61] L. Xing-cai, Y. Jian-guang, Z. Wu-gao, H. Zhen, Effect of cetane number improver on heat release rate and emissions of high speed diesel engine fueled with
ethanol–diesel blend fuel, Fuel 83 (2004) 2013–2020, https://doi.org/10.1016/j.fuel.2004.05.003.
[62] D. Singh, K.A. Subramanian, M. Juneja, K. Singh, S. Singh, R. Badola, N. Singh, Investigating the effect of fuel cetane number, oxygen content, fuel density, and
engine operating variables on NOx emissions of a heavy duty diesel engine, Environ. Prog. Sustain. Energy 36 (2017) 214–221, https://doi.org/10.1002/
ep.12439.

16