A Research On

ANALYSIS OF DIESEL ENGINE COOLING SYSTEM

Aman Pokharana , Anil Jaiswal , Lokesh Lohar ,

Kostubh Vyas , Ketan darji , Chirag Gour

Students Of
DEPARTMENT OF MECHANICAL ENGINEERING , GITS , UDAIPUR

Abstract
The cooling system of the designed diesel engine was investigated using specially prepared
models. For selected initializations of a cooling fluid in the available cooling jacket, quantitative
estimates of the levels of the heat flow to the coolant and of the temperature of the cooling
surfaces of various cylinders were made. Changing the terms of the initiation of the cooling fluid
flow (for example, the places where the coolant penetrates into the jacket) is accompanied by a
significant redistribution of the convective component of the heat sink and this influences the
temperature of the cooling surfaces. It is shown that the uniformity of the cylinder cooling and
the intensity of cooling of the cylinder heads can be controlled by choice of the place of coolant
penetration into the jacket for the considered design of a crankcase and cylinder heads.
Keywords: diesel engine; cylinder head; crankcase; cylinder liner; cooling system; redistribution
of coolant; FEA; thermal stress.

1. Introduction
The degree of speeding up of the IC engine is sometimes limited by the failure of existing
cooling systems to prevent the local temperature rise of individual parts [1]. Most susceptible to
overheating are the bottom of the cylinder head, the upper belt of the cylinder liner, the piston
crown, the upper compression ring and the exhaust valve cap. The stresses from moving thermal
field in these elements, due to the periodic nature of the work of the internal combustion engine,
may give rise to fatigue crack propagation with specific mechanisms, more dangerous than crack
propagation mechanisms in static thermal fields [2]. Non uniform heating of the structure gives
rise to thermal stresses and thermal displacements. Thermal displacements of the main bearings
and the distortion of the cylinder liners cause increased engine wear [3]. Improving the thermal
state of these parts can be achieved by levelling the unevenness of the temperature fields and the
restriction of their absolute value. During engine design a uniform distribution is achieved of the
coolant flow through the cooling cavities each cylinder [4, 5]. Modern possibilities of the
settlement analysis of movement of heat flows allow the thermal load on the engine to be
reduced by controlling the convective component of heat flows using the changing flow pattern
of the coolant in the cooling system [6].

2. Review
Both experimental and computational methods are used for the analysis of the thermal state of
the IC engine components. The thermometry method and the method of depositions are
experimental methods. Thermometry provides data about the temperature in the points of interest
for heat-loaded details. Thermocouples, mounted directly onto the metal parts, give evidence of a
change in the temperature over time at the point of the thermocouple installation, but do not
allow an estimation of the temperature gradients in the space. The installation of thermocouples
causes some difficulties and sometimes the thermocouple itself can affect the nature of the heat
distribution. The method of depositions is the experimental method that allows the determination
of the nature of fluid flow within a cylinder [7]. The essence of the method is as follows: in the
engine before full-scale tests pour water with a high content of calcium and iron. During engine
operation limescale is formed on the walls of the sleeve. The limescale is almost absent from the
sleeve in the areas with heavy liquid traffic, while in areas with less traffic there is a thicker layer
of limescale. In that way it is possible to make a judgement about the presence of dead zones in
the cooling system. If such zones are present, they may provoke the steam stopper and impair
motor cooling. Definition of the vortex nature of the flow of coolant by experimental methods is
difficult. Determining the local hydrodynamic parameters of the coolant flow, taking into
account the turbulence and vortex flow pattern in the channels, is possible using numerical
methods based on CFD (Computational Fluid Dynamics) models, such as the finite element
method, implemented with the help of CAE (Computer Aided Engineering) technologies [8, 9].
3D model of engine cooling cavity is created in a specialized CAD (Computeraided Design)
program. The power of the irreversible loss of heat energy, the consumption of the coolant and
the temperature of the engine compartment are taken as initial data for the calculation. A library
of the materials,integrated into the application package can be used to detail the thermal
properties of materials of the modeled design [10, 11].
The results of the thermal calculation by the finite element method can be represented as a
velocity field of fluid flow in the inner space of the cooling system in the form of heat or thermal
field flow. Modern computer technologies allow the creation of a model of the heat-loaded parts
in the CAD program, using appropriate CAE software to simulate the thermal load on the
workpiece. The CFD modelling results of a separate head cylinder coolant jacket are presented in
articles [12, 13]. In works [14-16] the cooling flow models of engines coolant jackets are
considered. Reduction of input data requirements in work [17] have achieved by underhood
modeling. Adequacy of numerical models is estimated by comparison with some experimental
results [13, 18]. CAE software to simulate the thermal load on the workpiece. The CFD
modelling results of a separate head cylinder coolant jacket are presented in articles [12, 13].
In works [14-16] the cooling flow models of engines coolant jackets are considered. Reduction
of input data requirements in work [17] have achieved by underhood modeling. Adequacy of
numerical models is estimated by comparison with some experimental results [13, 18].

3. Description of physical models

Questions of heat transfer or heat exchange are the main issues in the design of internal
combustion engines. There are three types of heat transfer: a) heat conduction; b) convection, or
the transfer of heat by moving particles of matter; c) by radiation.

The calculation of heat conduction phenomena is based on Fourier's law:

dQ(Cond) = -λ (dt / dl) dS,

where T is the temperature; dQ (Cond) is the amount of energy per time unit dt passing through the
inside area dS of the body taken normal to the line l along which the energy flux flowing in
proportion to the time dt , area dS and temperature gradient; λ is the thermal conductivity
coefficient. The thermal conductivity coefficient λ is different for different materials and depends
on their properties. The thermal conductivity coefficient λ for gaseous or liquid substances
depends on the density, humidity, pressure and temperature. In technical calculations the values
of λ is chosen by the respective lookup tables. Heat transfer by convection depends on many
different factors: the nature of convection - convection free or forced, fluid flow regime - laminar
or turbulent, coolant flow rate; the direction of heat flow (heating or cooling); the physical
properties of the heat carrier - thermal conductivity, heat capacity, density, viscosity, temperature
difference, which depends on the difference in temperature between the surface and walls; the
surface area of the wall F, washed coolant; shape of the wall, its sizes and other factors.
The calculation of convective heat transfer process is produced on the basis of Newton's law,
which is expressed.

By the formula:
dQ (Conv) = α ( t – tw^l ) dS
where dQ (Conv) - heat flux, W; α - coefficient of proportionality is called the heat transfer
coefficient, W/(m2 K); t and tw^l , t - temperature fluids, respectively, and the wall, K.

The heat transfer coefficient α is determined experimentally or by an analytical method. An

analytical method is very complex and does not provide the desired accuracy. Radiation heat is
transferred from one body to another by means of radiant energy without the mediation of the
material environment. By the Stefan - Boltzmann law the amount of heat transferred by radiation
from one body to another varies in proportion to the fourth power of the temperature difference
between the bodies. The coefficient of radiation of real bodies C is less the radiation coefficient
of a blackbody Cs. The gas radiation coefficient depends on their composition, layer thickness
and temperature. Practically only C02 and H20 possess emissivity in the flue gas; the greater
their contents in the gas and the greater the thickness of the gas layer and the temperature, the
higher the emissivity. The presence of glowing incandescent soot particles (luminous
hydrocarbon flame) significantly increases the radiation of the layer. With regard to the problem
of estimating the thermal state of the diesel engine elements, all three types of heat transfer occur
simultaneously and any one of them may predominate. For each elementary area of the thermal
contact is made of an elementary heat transfer:
dQ = dQ Cond + dQ Conv + dQ Rad

forming the heat flux through the entire surface of the thermal contact S
dQ = dQ Cond + dQ Conv + dQ Rad .

From the hot gases there is heat transfer to parts of the engine walls in the form of radiation as
well as in the form of convection. The level of heat flow varies cyclically with time. From the
coolant, usually having water in its base, there is the heat transfer due to convection by water in
the liquid phase, and due to water transition to phase of the vapor state. The correct assignment
of the boundary conditions of the thermal load on the part of the hot gases and from the cooling
liquid over the entire volume of the analyzed structure determines the degree of adequacy of the
model. The creation of a numerical computer model that takes into account all of these processes
as a whole is a difficult task. This model should include at least three sub-models – a thermal
CFD model of a moving hot gas in the combustion chamber, a model of the components
surrounding the combustion chamber and a thermal CFD model of the coolant. Moreover, due to
the changing volume of the combustion chamber all these models need to be linked in time and
should take into account the transfer of heat to the sleeve through the piston rings.

4. Problem definition
One measure of the quality of the design of the cooling system of the internal combustion engine
is the achievement of uniform strength (the same level of thermal stresses) of different cylinders.
The shape of the spatial region occupied by a cooling fluid typically has a complex spatial
geometry, formed by interconnected parts of the individual cylinders. This makes an intuitive
prediction of the degree of uniformity of cooling of the engine difficult.
The problem formulation is: for a diesel engine being prepared for mass production, estimate,
using specially prepared models, the levels of temperature of the cooling surfaces and the
coefficients of variation of the heat flow of the cooling surfaces to various cylinders for some
selected initializations of a cooling fluid in the available coolant jacket. The geometry of the
crankcase, cylinder heads, piston parts and the mass flow rate of the coolant are left unchanged
in embodiments of cooling system. The locations of the coolant inlets in the cooling jacket are
subjected to variation.
5. Model
The model is based on a system of simplifying assumptions, the most significant of which are the
following: the geometric model is a fragment of a V-shaped six-cylinder diesel dimension 15/16;
examines the thermal state of the engine, with respect to which there are oscillations: the periodic
nature of the ventilation in the combustion chamber is ignored, a source of heat takes a fixed
volume of the combustion chamber; analysis of the strengths and weaknesses of the embodiment
of the object of the research is done on the basis of solutions of stationary heat and quasi-static
deformation problems.
For computational prediction the distribution of heat flow and temperatures used the finite
element method, implemented in the application package Ansys. Numerical simulations were
performed on a workstation with eight core processors and 64 GB of RAM.

At the stage of solving CFD problems the geometrical model of the object of research was
divided into four domains:
􀁸 two solid domains (one of which includes three pistons and connecting rods assembly, a
fragment of the crankcase for three cylinders, three cylinder liners, the second - three cylinder
heads, fragments of the valve control, intake and exhaust manifolds);
􀁸 a gas domain representing the spatial region occupied by fuel mixture while moving from the
inlet manifold of the cylinder head into the cylinder and in to the exhaust manifold;
􀁸 a domain of coolant which combines communicating spaces occupied by the cooling fluid in
the cylinder head, in the fragments of the crankcase, the intake and exhaust manifolds;
􀁸 domain of space inside crankcase which filled by oil mist.

The inclusion of an oil mist domain in the model has eliminated the temperature inside the
crankcase from the list of parameters which needed to be defined before calculations. An
algorithm for solving CFD problems, implemented in the Ansys package to determine heat
transfer and the flow parameters of liquid and gas flow, involves the use of the non-deformable
mesh of an Euler finite element (FE). Areas streaming domains in contact with the solid domains
(near-wall region) were broken up into five layers of prismatic elements; other areas were
divided by tetrahedrons with a typical size of 0.5 mm. The power source of the heat is
determined by assessing the total irreversible heat loss of the engine. The flow rate of the coolant
through a cylinder head performance was assessed by the power of a circulation pump. Airflow
is determined by calculation of the engine cycle. For the cylinder head domain the properties of
aluminum alloy were used, and for the crankcase - ductile iron. For solid domains the accepted
model of continuous, homogeneous, isotropic material was used. The selection of the
characteristic size of the finite element mesh for the structural calculation takes into account the
structural heterogeneity of the material crankcase [6, 19]. For the liquid-domain model one- and
two-phase medium (Homogeneous Binary Mixture) was used, corresponding to the
thermodynamic properties of water and steam [20].

6. Results
The determination of the thermal state was performed on the four various embodiments of the
coolant supply into the engine coolant jacket (Fig. 1). A fifth version of the calculation was also
carried out, which differed from the first four in respect of the air temperature in the engine
compartment: for the first four options the temperature in the engine compartment was set at
80 C, and in the fifth it was assumed equal to 0 C. Differences of performance coolant jacket
options were limited to a choice of places to supply "cold" (80 C) coolant in the cooling jacket
(coolant supply places are marked in Fig. 1 by blue arrows). The ratios of irreversible thermal
losses corresponding to the examined options are shown in fig. 1e.

Fig. 1 (a) first, (b) second, (c) third, (d) fourth embodiment of the coolant supply and (e) the
relationship between the parts of the heat loss
Estimations of the heat flows in the coolant jacket on the part of each of the three cylinders
individually were obtained. Heat flows from the coolant jacket of each cylinder are divided into
parts corresponding with the crankcase area and the area of cylinder head. Heat flows Q to the
coolant from each of the cylinders (as a fraction of the total heat loss) are shown in Table 1
calculated for the cases considered. In addition to the heat flows, Table 1 shows the average
temperature of the corresponding surfaces. For comparison of the degree of homogeneity of the
thermal state of the cylinder for the considered cooling, the estimations of the coefficients of
variation for the heat flux and averaged temperature from each of the three cylinders were
performed (Fig. 2).

7. Conclusion
For a diesel engine being prepared for mass production, using specially prepared models for
some selected initializations of a cooling fluid in the available coolant jacket, the levels of the
cooling surfaces temperature and the coefficients of variation of heat flow to cooling surfaces of
various cylinders were estimated quantitatively. It is shown that the contribution of the coolant to
the full amount of heat loss is not primary, is approximately a quarter of the heat loss and
depends on a significant number of independent parameters (engine operating conditions, coolant
flow rate, temperature of the engine compartment and many others). At the same time, changes
in the conditions of the initiation of the cooling fluid flow (places where the coolant is input into
the jacket) to the convective component of the heat loss are so significant that they affect on the
temperature of the cooling surfaces.
For the considered diesel engine it was shown that without changing the design of the crankcase
and the cylinder heads there exists the possibility of successfully managing cooling in a uniform
way considering as the object of control the averaged values of heat flow or temperature of the
cooling water surface separately for the crankcase and cylinder heads.
Fig. 2 Coefficients of variation for (a) heat flux and (b) average temperature of the water-cooled
surfaces.

It has been shown that for the considered combination of power heat loss, flow of coolant and
engine compartment temperatures the temperature coefficient of variation refrigerated surfaces
can be reduced from 4.8% to 0.3%. In the performed calculation experiments, to achieve full
cooling uniformity is not possible due to the fact that each of the cylinders has a different surface
area from which heat passes into the engine compartment. As a result of the increase in the
component of the heat flow into the engine compartment (which is achieved by lowering the
engine compartment temperature) the attainable standard of uniformity of cooling is lower,
which is reflected in the noticeable increase in the coefficient of variation of temperature (see.
Fig. 2). Note also that the list of criteria which can be improved by work on the engine's cooling
system is not confined to the cooling surface temperatures, and can be extended to thermal
stresses and strains.

References

[1] A.A. Zhilin, A.B. Zharov, Heat Transfer Numerical Modelling for IC Engine Coolant Jacket,
in: Proceeding of International scientific
symposium Avtotraktorostroenie – 2009, Moscow, MAMI. (2009) 78􀁸88. URL:
http://www.mami.ru/science/autotr2009/scientific/article
/s02/s02_13.pdf
[2] D.A. Tereshin, Fatigue Long Crack Propagation in the Steel Specimen under Moving
Temperature Field, Vestnik SUSU, Mashinostroenie. 11
(2006) 47􀁸53.
[3] V.A. Romanov, N.A. Khozeniuk, Main Bearing Operability Evaluation for Internal
Combustion Engine Taking Into Account Crankcase
Thermal State, Friction and lubrication of machine and mechanism. 3 (2013) 23􀁸26.
[4] A.M. Kriger, M.E. Diskin, M.E. Diskin, Liquid Cooling of Automotive Engine,
Machinostroenie, Moscow, 1985.
[5] R.M. Petrichenko, Convective Heat Transfer in Piston Engine, Machinostroenie, Leningrad,
1979.
[6] V.A. Romanov, The Calculation of Thermal Stress State for the Crankcase of the Diesel
Engine of the Industrial Tractor, Science of SUSU,
Preceeding. (2011) 61–65.
[7] P.A. Loshakov, Experience of Influence on Piston and Cylinder Liner Temperature of the
Transport Diesel, Yaroslavl, 2004.
[8] V.A. Romanov, E.A. Lazarev, N.A. Khozeniuk, Cooling Conditions Influence on the
Cylinder Head Stress-Strain State for Internal
Combustion Engine, International conference Prom-Engineering–2015, Chelaybinsk, SUSU.
(2015) 70􀁸77.
[9] V.A. Romanov, E.A. Lazarev, N.A. Khozeniuk, The Evaluation of the Stress-strain State for
the Cylinder Heads of High-powered Diesel
Engines using the Multiphysics ANSYS Technology, Procedia Engineering. 129 (2015) 549–
556.
[10] D. Khitrikh, Modeling of the Combustion, Heat and Mass Transfer Processes in the
Pulverized-Fuel Boiler Furnace, Ansys advantage,