International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056

Volume: 05 Issue: 11 | Nov 2018 www.irjet.net p-ISSN: 2395-0072

AUTOMOBILE RADIATOR DESIGN AND VALIDATION

Ruchit Doshi1, Shakshi Himmatramka2, Janam Sanghavi3, Jahnavi Patel4, Vinit Katira5
1,2,3Graduate Student, Mechanical Engineering, DJ Sanghvi College of Engineering, Mumbai, India
4U.G. Student, Electronics and Telecommunications Engineering, DJ Sanghvi College of Engineering, Mumbai, India
5Assistant Professor, Mechanical Engineering, DJ Sanghvi College of Engineering, Mumbai, India

---------------------------------------------------------------------------***----------------------------------------------------------------------
Abstract:- The purpose of this research is to primarily design a cooling system for a formula styled race-car which is entirely
designed, manufactured and tested by graduate students from colleges across the world. The primary requirements of the
vehicle are to have high acceleration, low weights and an ability to endure. Keeping in mind the following requirements and
some governing rules made by the sanctioning bodies, the students begin their design procedure. The engine used by us for the
vehicle is a 4-stroke single cylinder 390 cc one, opted for its excellent power to weight ratio. The approach towards the design
of the cooling system has not only been a theoretical one, but also a pragmatic one. This research begins with experimentally
determining some important data of the engine. This data is used to determine the quantified requirements of the system
during the various modes of vehicle operation. Based on the targeted values, the sizing of the radiator is done using basic heat
transfer concepts. Further optimization is done using CFD analysis. The reason for not opting for a conventional wind tunnel is
that it does not holistically validate the system's capability. This research involves the development of a procedure to validate
the system in its operational state on the vehicle. Various sensors are deployed around the vehicle to acquire real time data
which is further processed to determine the actual efficiency of the system. Based on the above approach, a fruitful way of
validating the cooling system has been established and the co-relation between the expected and acquired results showed that
both the values commensurate. The theoretical and CFD results were able to accurately predict the actual heat transfer at
higher RPM and at lower RPM the predicted values were more than the actual heat rejection.

KeyWords: Heat transfer, CFD simulation, wind tunnel, on-track validation, data acquisition.

1. Introduction simulations in Star CCM+. The radiator was then

manufactured and tested practically with various
Formula Student is an SAE affiliated student level design sensors installed and data was logged in dynamic
competition, where students across the world design condition. Comparison is made between the above
and fabricate a formula styled race car. The car is results to determine variation in actual performance and
required to be light weight, quick and enduring. The designed performance.
official team of our college has been participating in
these competitions since 2014. A combustion power- 3. Flowchart:
train with a single cylinder 390 cc engine has been used
to power the vehicle. The students are required to abide
to the rule book made by the respective sanctioning
bodies of the competitions. As per the rules, the coolant
to be used is distilled water and hence this paper is
based on the research carried out by us to design a
custom cooling system for the vehicle. Basic heat
transfer concepts along with C.F.D analysis will be used
for the optimization of the same. The inputs for the
calculations and further analysis have been Fig. 1: Methodology Flowchart
experimentally obtained and the complete validation of
the design has been physically verified using a 4. Experiments:
standalone D.A.Q system.
4.1. Experiment to determine engine load to cooling
2. Methodology: system:
Various experiments were performed to determine the The engine used for powering the vehicle had values of
engine characteristics. The data obtained from these stock power and torque as 32 kW and 35 Nm
experiments was then used for theoretical calculations, respectively. Before starting the fundamental design and
CFD and validation. Initially, the radiator was sized calculations, certain engine parameters are required.
considering the engine load to cooling system by Hence, an experiment was performed for calculating the
theoretical calculations. These calculations were also change in the temperature of the coolant across the
used to determine radiator performance at various air engine at different crankshaft rotation speeds. A system
velocities passing through radiator and mass flow rate of comprising of the engine, two thermistors and a
coolant. These values were cross verified using CFD pressure cap in series was used as shown in the figure

© 2018, IRJET | Impact Factor value: 7.211 | ISO 9001:2008 Certified Journal | Page 1358
 International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 05 Issue: 11 | Nov 2018 www.irjet.net p-ISSN: 2395-0072

below for calculating the change the temperature of the with the pump which is geared to the engine crankshaft.
coolant. The thermistor readings were logged with the This test was hence performed to find mass flow rate of
help of a data logger. The ΔT has been used for the water at different crankshaft rotation speed. The
calculating the cooling load of the engine. engine speeds at which the experiment was performed
were from 2000 rpm to 7500 rpm. A digital flow-meter
was used in series with the engine, radiator, two
thermistors and a pressure cap. Following is the set-up
of the experiment performed.

Fig 1: Experimental Setup

Table 1: Temperature difference across engine at

different engine RPM

Engine Coolant
(T1) (T2)
Speed Temperature Sensor
Fig 2: Experimental Setup
2500 70 Engine70 68.5
Table 2: Mass flow rate of coolant at various engine RPM
3500 76 76.3 72.6
RPM Mass flow rate (LPM) Mass flow rate (kg/s)
4500 78 77.8 73.7
2500 10.78 0.18326
5500 80 79.85 75.1
2800 13.76 0.23392
6500 82 82 76.6
3500 14.2 0.24157
7500 82 82.4 77.5
4500 19.79 0.33643

5000 21.65 0.36805

Temperature across the engine
5500 22.575 0.38378
85
6000 23.5 0.3995
Coolant Temperature

80
7500 27.57 0.46869
75
T1
70
T2
65 5. Theoretical Calculations to determine radiator
2500 3500 4500 5500 6500 7500 sizing and performance at different load:
Engine Speed in RPM 5.1. Formulae:

Chart 1: Temperature difference across engine at

different engine RPM

4.2. Experiment to determine mass flow rate of

coolant at various engine RPM:

This experiment was conducted to determine the mass

flow rate of the water as a function of engine crankshaft
rotation speed. The mass flow rate of water is varied

© 2018, IRJET | Impact Factor value: 7.211 | ISO 9001:2008 Certified Journal | Page 1359
 International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 05 Issue: 11 | Nov 2018 www.irjet.net p-ISSN: 2395-0072

5.2. Water-side calculations:

5.4. NTU Method:

5.3. Air-side calculations:

Where,

Height-H

Length-L

Width-W

Thickness-t

Cross-sectional area-Acs

Surface Area-As

Perimeter-P

Hydraulic diameter-Dh

Mass flow rate-m

© 2018, IRJET | Impact Factor value: 7.211 | ISO 9001:2008 Certified Journal | Page 1360
 International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 05 Issue: 11 | Nov 2018 www.irjet.net p-ISSN: 2395-0072

Discharge-Qd c-cold fluid(air)

Velocity-V R-radiator

Reynolds Number-Re C-core

Factor for Nusselt number-α Known data:

Nusselt number-Nu Water parameters:

Convective Heat Transfer Coefficient-h ρh=960.84kg/m3

Frontal area-A µh=2.9×10-4kg/ms

Fin constant-mf kh=0.67862W/mK

Corrected length-Lc Prh=1.7438

Fin density-ρf Cph=4.08×103J/kgK

Base surface area-Ab Air Parameters:

Total base surface area-Abase ρc=1.1459kg/m3

Conductive heat transfer co-efficient-k kc=0.0267W/mK

Efficiency-η Prc=0.71317

Overall Efficiency-ηo νc=1.65×10-5

Overall heat transfer co-efficient-U µc=1.89×10-5

Number of transfer units-NTU Cpc=1.01×103J/KgK

Capacity ratio-C Tube parameters:

Effectiveness-ε Wt=0.0166m

Temperature-T Lt=0.002m

Density-ρ tt=0.0003

Dynamic viscosity-µ Fin Parameters:

Kinematic viscosity-ν Hf=0.0001m

Prantdl number-Pr Lf=0.008m

Specific heat at constant pressure-Cp ρf=189

Number-n Variable Parameters:

Velocity Inside the radiator-Vr nt, nC, Hr, mh, Vc, nr

Dynamic pressure drop-Pd Using the above calculations, heat rejected by radiator is
been calculated at different air velocities and engine
Subscripts: RPM. The results of these calculations and engine load
experimentation values were analyzed and accordingly
t-tube radiator specifications were finalized.
f-fin

h-hot fluid(water)

© 2018, IRJET | Impact Factor value: 7.211 | ISO 9001:2008 Certified Journal | Page 1361
 International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 05 Issue: 11 | Nov 2018 www.irjet.net p-ISSN: 2395-0072

6. CFD Analysis of the Radiator: 6.3. Mesh:

The 3D model of the radiator was created in Solid Works In order to reduce the CPU time and the load on the
and was exported as a single surface in the form of an computer, the core section of the geometric model was
IGS file. The initial designs were modeled as one section given a finer mesh then the other regions of the
and features that didn’t take part in the internal flow geometry.
were suppressed.
The base size for the core section was set to 0.64m
Meshing and CFD analysis were done using Star CCM+.
Star CCM+ is a comprehensive engineering simulation The base size for the air region was set to 2.56m
package for solving problems involving flow (of fluids or
solids), heat transfer and stress. 6.4. Physical models:

6.1. Geometry: For the radiator, the physical model was selected
deciding the type of flow.
The core of the radiator was represented by a cuboid
which was then given a porous media. The core is of the As there are two fluids in the analysis, different models
dimension from the theoretical calculations that is were created.
255×210×30mm. The header above the radiator has an
Three dimensional
inlet pipe and the header below had the outlet pipe
which directed the flow of water. Two cuboids were Gradient
created in front and behind the radiator to represent the
air region. Steady flow
Before the meshing and selecting physical model, the Constant density
inlet and outlet for the two fluids and Interfaces between
different bodies were created. In order to set up different Segregated fluid flow
boundaries a new region was created.
Turbulent
6.2. Meshing:
K-epsilon turbulence
The meshing models chosen for the flow analysis were
Two Layer All y + Wall Treatment
Trimmer and Surface Remesher
Realizable K- Epsilon two Layer

Reynolds- Average Navier- Stokes

Segregated Fluid Temperature

Air physical model:

Gas

Water physical model:

Fig 3: Geometry for CFD simulations. Liquid

6.5. Boundary Conditions:

Coolant inlet Mass flow rate

Coolant outlet Pressure outlet

Air inlet Velocity inlet

Air outlet Pressure outlet

Fig 4: Final mesh

© 2018, IRJET | Impact Factor value: 7.211 | ISO 9001:2008 Certified Journal | Page 1362
 International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 05 Issue: 11 | Nov 2018 www.irjet.net p-ISSN: 2395-0072

6.6. Inertial and viscous resistance: 6.7. Results:

Water side resistance:

The pressure drop across the tube was calculated by

Darcy-Wiesbach equation

Pressure drop =

Where,

fD= coefficient of friction = 0.0014 (Moody’s chart)

ρ = density = 960.84kg/m3

Dh = = 1.45×10-3m Fig 5: Temperature distribution on the water side

V=

Ht = 0.3m

Mass flow Mass flow Velocity of Pressure

rate of water rate of water water in drop
from engine in a tube tube

0.3995kg/s 0.01kg/s 0.39m/s 505.19Pa

0.24157kg/s 0.00603kg/s 0.2363m/s 885.6Pa

To compute porous media coefficients

X= heat exchanger thickness through which pressure

drop takes place Fig 6: Temperature distribution on the air side

D= viscous coefficient

µ= viscosity

C= inertial coefficient

Ρ= Density

By substituting the values,

And solving simultaneously,

D= 86415557

C=110.7 Fig 7: Residual Graph

Air side resistance: The above results are used to cross verify the data
acquired from theoretical calculations.
This data was provided by the manufacturer
7. On-track validation:
D=456
The experiment was performed to find the temperature
C=92 and pressure drop across the radiator at different
vehicle and engine speeds. The temperature drop helped

© 2018, IRJET | Impact Factor value: 7.211 | ISO 9001:2008 Certified Journal | Page 1363
 International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 05 Issue: 11 | Nov 2018 www.irjet.net p-ISSN: 2395-0072

us determine the actual heat rejected by the radiator and

the dynamic pressure difference was used to calculate
the air velocity passing through the radiator.

The radiator was equipped with two thermistors, one at

coolant inlet pipe and the other at coolant outlet pipe,
both pipe having same diameter. Two differential
pressure sensors one for static pressure drop and one
for dynamic pressure drop were also installed in the
radiator. To measure the dynamic pressure drop, the
probes were positioned perpendicular to the airflow at
the upstream and the downstream face of the radiator
and to measure static pressure drop, the probes were
positioned parallel to the airflow.
Fig 9: Thermistors’ position on radiator

Fig 10: Pressure sensors installed on the radiator.

Fig 8: Probes position in radiator to measure dynamic
and static pressure drop 8. Result and Conclusion:
The air velocity was calculated by dynamic pressure Mass flow rate of coolant and air velocity through
drop using the following equation, radiator, from the data acquired, were compared to that
deduced by different theoretical methods. All the data
was calculated, simulated and recorded at specific
engine RPM and vehicle velocities.
Pd - Dynamic pressure drop
It was observed that at lower vehicle speeds, the heat
Pu – Static pressure drop rejected by the radiator during practical experiment was
almost similar to the calculated value of heat rejected
V - Velocity and it was 5% lower when compared to the CFD results.
As the vehicle velocity increased, the actual heat rejected
ρ - Density by the radiator reduced and the difference between
results acquired by different methods increased up to
The sensors were connected to a data logger system and 25%. Therefore, the theoretical and CFD results were
logged the data in the dynamic condition. The engine able to accurately predict the actual heat transfer at
was initially fired at idle and then the car was made to lower vehicle speeds and at higher vehicle speeds, the
follow a circular path of constant radius while keeping predicted values were more than the actual heat
the engine speed constant. The procedure was repeated rejection. This was probably due to the separation of
for different radii while keeping the RPM same so as to flow in the sidepod.
achieve various air velocity across the radiator. Then, the
engine speed was changed and the procedure was
redone. Hence, we obtained pressure and temperature
drop values at various engine speeds and air velocities.

© 2018, IRJET | Impact Factor value: 7.211 | ISO 9001:2008 Certified Journal | Page 1364
 International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 05 Issue: 11 | Nov 2018 www.irjet.net p-ISSN: 2395-0072

Table 3: Temperature difference of coolant across radiator and cooling load values at different engine speeds and air
velocities

Theoritical CFD Practical Expirement

RPM Mass Flow Rate Air Velocity Temperature Difference Load (kW) Temperature Difference Load (kW) Temperature Difference Load (kW)
7500 0.46869 4 2.6 4959 3 5722 2.53 4826
7500 0.46869 8 3.66 6981 4.1 7820 3.42 6523
7500 0.46869 12 4.4 8392 4.8 9155 4.13 7877
7500 0.46869 16 4.97 9480 5.4 10300 5.09 9709
6500 0.401 4 2.99 4879 3.5 5712 2.31 3770
6500 0.401 8 4.19 6838 4.8 7833 3.57 5826
6500 0.401 12 5 8160 5.7 9302 3.89 6348
6500 0.401 16 5.63 9188 6.1 9955 4.39 7164
5500 0.38378 4 3.11 4857 3.4 5310 2.27 3545
5500 0.38378 8 4.35 6794 4.7 7341 3.21 5013
5500 0.38378 12 5.19 8106 5.8 9059 4.37 6825
5500 0.38378 16 5.83 9105 6.6 10308 4.75 7419

11000
Heat Rejected by Radiator (kW)

10000
9000
8000
Theoretical
7000
6000 Practical
5000 CFD
4000
3000
7500 7500 7500 7500 6500 6500 6500 6500 5500 5500 5500 5500
Engine Speed (RPM)

Chart 2: Comparison of heat rejected by the radiator calculated by different methods.

9. Applications:

This research involved the development of a procedure [6].Computational Fluid Dynamics for Engineers by
to validate the system in its operational state on the Bengt Andersson, Ronnie Andersson, Love Håkansson,
vehicle. The use of this kind of dynamic wind tunnel Mikael Mortensen, Rahman Sudiyo, and Berend van
testing allowed us to test the cooling system without the Wachem
use of a traditional wind tunnel hence, getting more
accurate results according to the relevant conditions. [7].Star CCM+ tutorials
https://thesteveportal.plm.automation.siemens.com/
This methodology can also be used for passenger cars,
trucks as well as for cooling system design of electrical [8]. Measurement and sensors fundamentals
components such as motors, high power circuits, etc. http://www.efunda.com/designstandards/sensors/sens
ors_home/sensors_home.cfm
10. References:
[9]. Computational Fluid Dynamics: Principles and
[1]. Heat and Mass Transfer by R.K. Rajput Applications by J. Blaze

[2]. A Heat Transfer Textbook by John H. Leinhard

[3]. Perry's Chemical Engineers' Handbook by Robert H.

Perry and Don W. Green

[4]. Industrial Heat-Transfer Calculations by Mohammad

G. Rasul

[5]. Heat and Mass Transfer by Rudramoorthy, R

© 2018, IRJET | Impact Factor value: 7.211 | ISO 9001:2008 Certified Journal | Page 1365