UDK 629.331:62-714:536.2:620.

3 ISSN 1580-2949
Original scientific article/Izvirni znanstveni ~lanek MTAEC9, 52(4)507(2018)

T. GANESAN, P. SEENI KANNAN: HEAT-TRANSFER ENHANCEMENT OF NANOFLUIDS IN A CAR RADIATOR

507–512

HEAT-TRANSFER ENHANCEMENT OF NANOFLUIDS IN A CAR

RADIATOR
IZBOLJ[ANJE PRENOSA TOPLOTE NANOTEKO^IN V
AVTOMOBILSKEM RADIATORJU

Thangasamy Ganesan1, Pauldurai Seeni Kannan2

1Department of Mechanical Engineering, AR College of Engineering & Technology, Kadayam, Tamilnadu 627 423, India
2Department of Mechanical Engineering, AAA College of Engineering and Technology, Sivakasi, Tamilnadu 626 123, India
ganesanjoshi2012@gmail.com

Prejem rokopisa – received: 2017-10-07; sprejem za objavo – accepted for publication: 2018-02-22

doi:10.17222/mit.2017.171

In recent times, enhancing the efficiency of automobiles has become very important. As part of this effort, increasing the
efficiency of the heat transfer in the radiator of an automobile is also crucial. The present approach involves an addition of
nanoparticles to the cooling liquid. Most of the current results were obtained using a few types of nanoparticles. Only a limited
amount of work involved hybrid nanofluids including several types of nanoparticles. Multi-walled carbon nanotubes (MWCNT)
were used along with aluminium-oxide nanoparticles to form a nanofluid with the base fluid (ethylene glycol (EG) + distilled
water (DW) – 1:1 ratio) having volume concentrations of (0.03, 0.06, 0.09 and 0.12) %. The characterization of the
nanoparticles was carried out. The experiments were carried out at inlet temperatures of 40–75 °C, with increments of 5 °C. The
volume flow rate was varied from 0.6 m3/h to 0.96 m3/h, with increments of 0.12 m3/h. An increase in the heat transfer of 35 %
was achieved.
Keywords: computational fluid dynamics, nanofluids, alumina, multi-walled carbon nanotubes

V zadnjem ~asu je postalo zelo pomembno pove~anje zmogljivosti avtomobilov, hkrati s tem pa tudi pove~anje u~inkovitosti
prenosa toplote v avtomobilskih radiatorjih. V prispevku so avtorji dodali nanodelce v hladilno sredstvo (kapljevino). Ve~ina
podanih rezultatov se nana{a na uporabo nekaj razli~nih vrst nanodelcev. Delno pa so avtorji raziskovali tudi hibridna hladilna
sredstva oziroma vpliv isto~asnega dodatka ve~ razli~nih vrst nanodelcev. Uporabili so dodatek ve~stenskih ogljikovih nanocevk
(MWCNT; angl.: Multi Walled Carbon Nano Tubes) v kombinaciji z nanodelci Al oksida k osnovnemu hladilnemu sredstvu
(Etilen Glikol (EG) + destilirana voda (DW) – v razmerju 1:1) v volumskih koncentracijah (0,03, 0,06, 0,09 and 0,12) %. Izvedli
so karakterizacijo nanodelcev. Preizkuse so izvajali pri vstopnih temperaturah hladilnega sredstva od 40 °C do 75 °C v korakih
po 5 °C. Hitrost volumskega pretoka je varirala od 0,6 m3/h do 0,96 m3/h v korakih po 0,12 m3/h. Dosegli so pove~anje prenosa
toplote do 35 %.
Klju~ne besede: ra~unalni{ka dinamika teko~in (kapljevin in plinov), nanoteko~ine, aluminijev oksid, ve~stenske ogljikove
nanocevke

with a low concentration of nanoparticles. The papers

1 INTRODUCTION published so far indicate improvement in the heat
transfer due to single nanoparticles. Further research can
Only a part of the heat energy produced with com- be done using hybrid nanofluids, which are mixtures of
bustion is converted to the mechanical energy to power more than one type of nanoparticles. It is expected that
the automobiles while the remaining energy is rejected hybrid nanofluids will improve the heat transfer.
and unutilized. A part of the energy is removed by the
cooling water surrounding the engine and then taken to
the radiator while the heat is released into the atmo- 2 LITERATURE SURVEY
sphere. If effective cooling is not done, the efficiency of
the engine can be reduced and, in extreme cases, engine A brief survey of the work done so far is presented
may be damaged by overheating. The cooling system below.
normally consists of a cooling medium surrounding the K. S Suganthi et al.1 prepared a nanofluid using a
engine cylinder, a pump allowing the circulation of the spherical ZnO by mixing it with water and propylene
cooling medium, a radiator with a cooling fan and a glycol (PG). The heat-absorption capacity of their nano-
thermostat. In the present work, an attempt to improve fluid was compared with that of the solution of just water
the effectiveness of the heat transfer in the radiator was and PG. The nanofluid showed a higher heat absorption
made. With the advances made in the field of nanotech- capacity than the solution of water and PG. The
nology, we now have nanoparticles, which can be mixed nanofluid showed an increase of 16 % in the case of heat
with the fluid to form a homogeneous mixture. These transfer under indicator conditions with 2 % volume of
mixtures are called nanofluids and they have better ZnO of 70 nm. B. Ilhan et al.2 used hexagonal boron ni-
thermal properties. They enhance the heat transfer even tride (HBN) to make a nanofluid using a volume con-

Materiali in tehnologije / Materials and technology 52 (2018) 4, 507–512 507

T. GANESAN, P. SEENI KANNAN: HEAT-TRANSFER ENHANCEMENT OF NANOFLUIDS IN A CAR RADIATOR

centration of 0.03–0.3 % to make nano sodium dodecyl

sulphate (SDS) and polyvinyl pyrrolidon was used as the
surfactant. The HBN nanofluild had a higher thermal
conductivity compared to the base fluid. M. A. A. Beha-
badi et al.3 conducted experiments using multi-walled
carbon nanotubes (MWCNT). They were mixed with
water with weight concentrations of 0.05, 0.1 and 0.2.
The MWCNT nanofluid showed an enhanced heat-trans-
fer rate. The pressure drop in this case was higher than
that of the base fluid. W. H. Azmi et al.4 compared the
performance of a water-based nanofluid and ethylene-
glycol water-based nanofluid. The water-based nanofluid
showed a higher heat-transfer rate. M. B. Bigdeli et al.5
carried out a study of thermophysical properties of a
nanofluid. The study shows the nanofluid having better
Figure 1: Circuit diagram of the experimental set-up
thermal properties than the base fluids. However, the
details of the concentration and size of the nanoparticles
Table 1: Experimental set-up specifications
were not reported. M. Tamizi et al.6 studied the thermo-
physical properties of a nanofluid with different volume Width: 370 mm; Height: 400 mm; Tube size: 20
Radiator
fractions. It was shown that the heat-transfer rate was × 2 mm; Fin thickness: 1 mm; Fin pitch: 1 mm
higher for a higher volume fraction. S. Suresh et al.7 used Capacity: 14 l; Material: stainless steel;
Tank
a hybrid nanofluid in their study. The nanofluid was Diameter: 260 mm; Height: 300 mm
made with the nanoparticles of aluminium oxide and Type: Epson; Number of heaters: 3; Power:
Heater 2 kW; Thermostat: range of 0–75 °C (Girsheg);
copper. The experiment showed an improvement of 25 % Accuracy: ±0.5 °C
for the heat transfer compared to the base fluid. D. Mad- Made by CRI; Capacity: 0.5 hp; Speed:
hesh et al.8 used copper and titanium nanoparticles with Pump
2800 min–1; Water capacity: 1080 l/hr
volume concentrations varying from 0.1 % to 1 % in a Thermo- Type: digital; Make: Nanotech; Range:
radiator and heat-transfer rates increased by up to 48.5 % meter 0–400 °C; Accuracy: ±0.5 °C
were attained. N. Bozorgan et al.9 conducted a numerical Power: 80 W (motor); Number of speed rates:
investigation of Al2O3 nanoparticles with water as the Fan 2; Diameter: 400 mm; Number of blades: 6;
Type of blade: curved
base fluid using empirical relations to determine the
Rotameter Make:
Acrylic; Type: micro; Capacity: 0.6–6
properties such as density, viscosity, specific heat and m3/h
thermal conductivity. G. Huminic et al.10 also studied the Model: AVM 06; Display: LCD; Make: BEE
Anemo-
thermal properties of a nanofluid and found the heat- meter Tech-Max; Show value: 9999; Accuracy: ±
transfer rate increasing with an addition of nanoparticles. (2.0 % reading + 50 characters)
Type: rubber; Length: 1552.5 mm (radiator to
Hose tank: 900 mm, tank to pump: 450 mm,
rotameter to radiator: 202.5 mm)
3. EXPERIMENTAL PART
with increments of 0.12 m3/h. The temperature of the
Figure 1 shows a diagram of the experimental set-up
cooling fluid at the inlet was varied from 40 °C to 75 °C.
and Figure 2 includes a photograph of the set-up.
The thermocouple used to measure the temperature had
The set-up included the radiator from a Maruthi Alto an accuracy of ±0.5 °C. A pump of 0.5 HP was used to
car. The engine had the power of 85 BHP. There were circulate the cooling fluid from the storage tank. An
also the cooling fan, the water tank, the heater, and the
control for the heater used to maintain the temperature of
the inlet of the water, which was constant for the com-
ponents in the set-up. Instruments for measuring the
temperature of the water at the inlet and outlet were
provided. Two additional thermocouples were provided
to measure the temperature of the air, flowing across the
radiator. They were fixed to the air inlet and outlet of the
radiator.
A gate valve was provided to regulate the flow of the
water through the radiator. Three heater controllers were
provided to heat the water to be sent to the radiator to the
steady state temperature of the inlet. The radiator was
placed inside a duct and a fan was used to provide the
flow at two different speeds. Experiments were carried
out for different flow rates of 0.6 m3/h and 0.96 m3/h Figure 2: Experimental set-up photograph

508 Materiali in tehnologije / Materials and technology 52 (2018) 4, 507–512

 T. GANESAN, P. SEENI KANNAN: HEAT-TRANSFER ENHANCEMENT OF NANOFLUIDS IN A CAR RADIATOR

anemometer was used to measure the velocity of the air

through the duct. Readings were taken at four points and
the operating average velocity was calculated. Table 1
shows the specifications of the experimental set-up for
this work.

3.1 Experimental procedure

The system was first cleaned thoroughly. Then the
tank was filled with distilled water. A test run was done
to ensure that the water circuit had no leakages. Then the
heaters were switched on for the water to reach the
required temperature. After that, the heating elements
were controlled by the temperature controller to maintain
a steady temperature. The water pump was switched on. Figure 4: SEM test report for Al2O3
After that, the radiator fan was switched on and we
waited until steady-state conditions were reached and all
the readings were taken. The temperatures at the outlet The precipitate formed was washed with distilled
and inlet of the water and air were measured. The water and dried. SEM images of the prepared sample are
velocity of the air was also measured. The experiment shown in Figures 4 and 5. A nanofluid with the required
was repeated for water-flow rates of (0.72, 0.84 and volume concentration was then prepared by dispersing a
0.96) m3/h. For the second set of experiments, a mixture specified amount of Al2O3 nanoparticles in water using
of water and ethylene glycol with a ratio of 1:1 was used. an ultrasonic vibrator (Toshiba, India) generating ultra-
The experiments were then conducted using nanofluids sonic pulses of 100 W at 36±3 kHz. To get a uniform
with volume concentrations of 0.03–0.12 %, with incre- dispersion and stable suspension, which determine the
ments of 0.03 %. final properties of nanofluids, the nanofluids were kept
under ultrasonic vibration continuously for 9 h. No
surfactant or pH changes were used as they may have
3.2 Nanofluid characterization
some influence on the effective thermal conductivity of
The characterization of nanofluids was done by eva- nanofluids. The pH values of the prepared nanofluids
luating the XRD test shown in Figures 3a and 3b. The with different concentrations were measured and found
Sheerin formula was used to calculate the nanoparticle to be around 5, which is far from the isoelectric point of
size, which was 40 nm. 9.2 for alumina nanoparticles. This value, being far from
the isoelectric point, ensures that the nanoparticles are
well dispersed and the nanofluid is stable because of
very large repulsive forces among the nanoparticles.

3.3 Prepartion of Al2O3/multi-walled carbon nanofluids

The nanoparticles were purchased from Alfa Aesar,
USA. The Al2O3 nanoparticle MWCNTs were tested
with a Zetasizer machine for measuring the particle size,

Figure 3: XRD test result for: a) Al2O3, b) MWCNT Figure 5: SEM image of MWCNT

Materiali in tehnologije / Materials and technology 52 (2018) 4, 507–512 509

T. GANESAN, P. SEENI KANNAN: HEAT-TRANSFER ENHANCEMENT OF NANOFLUIDS IN A CAR RADIATOR

Figure 6: (a) Magnetic stirrer, (b) ultrasonicator

from Horibon, Japan. The machine is shown in Figure 6.

The nanofluids were prepared as per volume concen-
tration. Then a magnetic stirrer and an ultrasonicator Figure 8: Viscometer
were used to mix them without any sedimentation as
shown below. The characteristics of aluminium are given
in Table 2. switched off before attaching the spindle. The spindle
speeds available with this viscometer are in a range of
Table 2: Properties of the nanoparticles 0–100 min–1 and the shear-rate range is 0–750 s. Then
the spindle is immersed in the center of the test material
S. no. Properties Al2O3 MWCNT
so that the fluid’s level is at the immersion groove on the
1. Purity (%) 99.99 99.9
spindle’s shaft. With a disc-type spindle, it is sometimes
Approximate
2. size (nm) 50 60 necessary to tilt the spindle slightly while immersing it
3. Color White Black to avoid trapping air bubbles on its surface. To make a
4. Morphology Nearly spherical Tubes viscosity measurement, a speed of 12 min–1 is selected.
5. Density (g/cm3) 3.428 1.4 The maximum time allows the indicated reading to
stabilize. The time required for its stabilization depends
Next, an ultra-violet (UV) test was carried out to on the speed of the viscometer and the characteristics of
check whether the nanofluids were without any sedimen- the sample fluid. Now the device is switched off and the
tation. No sediment was identified. Figure 7 clearly sample is replaced by a new one. A minimum of three
indicates that the nanofluids contained only the base measurements were made for each test sample. Then the
fluid. average value was taken. Each sample constituted of a
20-ml solution. Table 3 shows the values of the viscosity
4 RESULTS AND DISCUSSION for all the samples.

Nanofluids with various volume concentrations were Table 3: Viscosity values for the samples
prepared. Then they were tested with the experimental S. No. Name of the sample Viscosity (cP)
set-up and the readings were taken. The thermal 1. 0.03 % hybrid nanofluid 23.5
properties of various samples were tested and the results 2. 0.06 % hybrid nanofluid 20.5
are shown below. 3. 0.09 % hybrid nanofluid 22.5
4. 0.12 % hybrid nanofluid 20.5
4.1 Analysis of thermal properties 5. EG+DW (1:1 ratio) 18.5
4.1.1 Measurement of viscosity values 6. DW 1.5

The viscosity of the nanofluids was measured using a For the 0.03 % sample, the thermal conductivity
Brookfield cone and plate viscometer (DV-E Visco- increases by 5 % where the heat transfer at a higher tem-
meter), supplied by Brookfield Engineering Labora- perature is considered to be the molecular-level diffe-
tories, USA, as shown in Figure 8. The device is rence of the sample. At the mean time for the 0.06 %
sample, the thermal conductivity increases by 11 %. It is
concluded that with the increasing volume concentration
of nanofluids, the heat transfer also increases. Thus, at
the 0.12 % volume concentration of nanofluids, a high
heat-transfer rate is achieved. It is also concluded that
with the hybrid nanofluids containing the base fluids, a
high heat-transfer rate is achieved. With the increasing
volume concentration of nanofluids, there is also a
Figure 7: UV spectroscopy result chance for pumping. This issue requires further research.

510 Materiali in tehnologije / Materials and technology 52 (2018) 4, 507–512

 T. GANESAN, P. SEENI KANNAN: HEAT-TRANSFER ENHANCEMENT OF NANOFLUIDS IN A CAR RADIATOR

4.1.2 Thermal-conductivity measurement 4.2 Uncertainity analysis

The compact KD2 Pro controller is much more than a Experimental uncertainties were calculated using the
simple readout for time and temperature as shown in method developed by Kline and McClintock. This
Figure 9. A proprietary algorithm fits the time and method incorporates the estimated uncertainties of the
temperature data with experimental integral functions experimental measurements such as the water-inlet tem-
using a nonlinear least squares method. This fully ma- perature, water-outlet temperature and water-mass-flow
thematical solution delivers thermal conductivity to rate in the experiment. These are then compared with the
within ±10 %. The ASTM standard D5334 and IEEE final parameters of the heat-transfer rate and the calcul-
standard 442-1981 are followed. All the samples were ated Reynolds number. Using the measured uncertainties
characterized at room temperature. The KD2 Pro was of the above-mentioned parameters, the maximum
configured in the low-power mode with a 1-minute read calculated uncertainties for the heat-transfer rate were
time to prevent free convection in the fluid test samples. found to be 10 % and 9.7 % for pure water and the nano-
The accuracy of the KS-1 sensor and the associated fluids, respectively. The fluid-mass-flow rate participated
KD-2 Pro unit was verified using the glycerol verifica- significantly to the calculated uncertainty, especially in
tion standard immediately before the measurements the case of the nanofluids. The maximum uncertainty of
made on the test samples. A minimum of three measure- the calculated Reynolds number was found to be 2.8 %
ments was made for each test sample. Before the for all the cases.
beginning of each measurement, the vial with the mate-
rial was vigorously shaken until all the nanoparticles 4.3 Computational-fulid-dynamics analysis (CFD)
were in the suspension and then the shaking continued
for at least 60 additional seconds to ensure that the The heat transfer rate was also measured using the
sample was well dispersed and fully suspended. The computational-fluid-dynamics (CFD) software. Its re-
accuracy and range repeatability of this instrument are sults were also compared with the experimental results.
±1 % and ±0.2 %, respectively. The thermal conductivity It is shown that the results for the increasing heat transfer
of all six samples measured using a 0.40 mL solution for in the experimental set-up were higher than those of the
each sample is shown in Table 4. CFD analysis. In Figure 10, for the 0.09 % Al2O3 + 1 %
MWNCT nanofluid, the CFD analysis shows that a small
Table 4: Thermal-conductivity values for the samples increment in the heat-transfer rate of 19 kW is obtained
Thermal
at the given mass-flow rate (0.6 m3/hr) and the tempera-
Temperature conductivity ture of 40 °C. For the temperature of 70 °C and the
S. no Name of the sample
(°C)
(W/mK) 0.96 m3/h mass-flow rate, the maximum heat-transfer
1. 0.03 % hybrid nanofluid 28.75 0.579 rate of 56 kW is obtained with the CFD analysis. It is
2. 0.06 % hybrid nanofluid 28.72 0.551 observed that while adding 1 % MWNCT with the Al2O3
3. 0.09 % hybrid nanofluid 28.68 0.585 nanofluids, the heat-transfer rate increases by 26 %
4. 0.12 % hybrid nanofluid 29.19 0.570 according to the CFD analysis.
5. EG+DW (1:1 ratio) 28.7 0.569 Figure 11 shows the heat-transfer rate at the 0.6 m3/h
6. DW 29.04 0.529 mass-flow rate for pure distilled water, 1 % MWNCT
and 1.5 % MWNCT with distilled water, 0.09 % Al2O3
nanofluids with 1 % MWNCT and 1.5 % MWNCT.
From this graph, it is clear that the heat-transfer rate of
the 0.09 % Al2O3 nanofluid is followed by those of 1 %

Figure 9: Thermal-conductivity instrument Figure 10: CFD analysis for 0.09 % Al2O3 with 1 % MWCNT

Materiali in tehnologije / Materials and technology 52 (2018) 4, 507–512 511

T. GANESAN, P. SEENI KANNAN: HEAT-TRANSFER ENHANCEMENT OF NANOFLUIDS IN A CAR RADIATOR

MWCNT and 0.09 % Al2O3 nanofluid, the heat-transfer

rate increased by 34 %. Similarly, for 1.5 % MWCNT
and 0.09 % Al2O3 nanofluid, the heat-transfer rate in-
creased by 35 %. It was proven that MWCNT signifi-
cantly improved the heat transfer when added together
with the Al2O3 nanofluids.

6 REFERENCES
1
K. S. Suganthi, V. Leela Vinodhan, K. S. Rajan, ZnO-Propylene
Glycol-Water Nanofluids with Improved Properties for Potential
Applications in Renewable Energy and Thermal Management,
Colloids and Surfaces A: Physicochemical and Engineering Aspects,
506 (2016) 5, 63–73, doi:10.1016/j.colsurfa.2016.06.007
2
B. Ilhan, M. Kurt, H. Ertürk, Experimental investigation of heat
transfer enhancement and viscosity change of hBN nanofluids,
Experimental Thermal and Fluid Science, 77 (2016), 272–283,
doi:10.1016/j.expthermflusci.2016.04.024
3
Figure 11: Experimental comparison of the heat-transfer rate for M. A. A Behabadi, M. Shahidi, M. R. Aligoodarz, M. Fakoor, An
various percentages of MNCNT, Al2O3 and water (0.6 m3/h) experimental investigation on rheological properties and heat transfer
performance of MWCNT-water nanofluid flow inside vertical tubes,
Applied Thermal Engineering, 106 (2016) 5, 916–924,
MWNCT and 1.5 % MWNCT. While considering the doi:10.1016/j.applthermaleng.2016.06.076
4
comparison with the CFD analysis, we observe a high W. H. Azmi, K. Abdul Hamid, N. A. Usri, R. Mamat, K. V. Sharma,
heat-transfer rate. Heat transfer augmentation of ethylene glycol: water nanofluids and
applications – A review, International Communications in Heat and
Mass Transfer, 75 (2016), 13–23, doi:10.1016/j.icheatmasstransfer.
5 CONCLUSIONS 2016.03.018
5
M. B. Bigdeli, M. Fasano, A. Cardellini, E. Chiavazzo, P. Asinari, A
Experimental and CFD analyses were done. The review on the heat and mass transfer phenomena in nanofluid
MWCNT was added together with 0.09 % Al2O3 nano- coolants with special focus on automotive applications, 60 (2016),
1615–1633, doi:10.1016/j.rser.2016.03.027
fluids. The input temperature varied between (40, 45, 50, 6
M. Tamizi, M. Kamalv, M. Namazian, Dependency of the thermo-
55, 60, 65 and 70) °C. The mass-flow rate varied bet- physical properties of nanofluids on the excess adsorption, Inter-
ween (0.6, 0.72, 0.84 and 0.96) m3/h, respectively. It was national Journal of Heat and Mass Transfer, 99 (2016), 630–637,
concluded that when adding (0.5, 1 and 1.5) % MWCNT doi:10.1016/j.ijheatmasstransfer.2016.04.029
7
together with the nanofluids, increased heat-transfer S. Suresh, K. P. Venkitaraj, P. Selvakumar, M. Chandrasekar, Effect
rates were obtained. According to the CFD analysis, with of Al2O3–Cu/water hybrid nanofluid in heat transfer, Experimental
Thermal and Fluid Science, 38 (2012), 54–60, doi:10.1016/
the 0.5 % MWCNT + 0.09 % Al2O3 nanofluid, the heat- j.expthermflusci.2011.11.007
transfer rate increased by 3.31 %; with 1% MWCNT + 8
D. Madhesh, S. Kalaiselvam, Experimental Analysis of Hybrid
0.09 % Al2O3, the heat-transfer rate increased by 8.96 %; Nanofluid as a Coolant, Procedia Engineering, 97 (2014),
and, similarly, when 1.5 % MWCNT was added together 1667–1675, doi:10.1016/j.proeng.2014.12.317
9
with the Al2O3 nanofluid, an increase in the heat-transfer N. Bozorgan, K. Krishnakumar, N. Bozorganv, The Performance
rate of 13.814 % was obtained. In the case of the expe- Evaluation of Overall Heat Transfer and Pumping Power of g-Al2O3/
water Nanofluid as Coolant in Automotive Diesel Engine Radiator,
rimental analysis, when adding 0.5 % MWCNT together Analele Universitãlii "Eftimie Murgu" Resita, 1 (2013)
with the 0.09 % Al2O3 nanofluids, the maximum increase 10
G. Huminic, A. Huminic, Application of nanofluids in heat ex-
in the heat transfer rate of 16 % was obtained at a mass- changers: A review, Renewable and Sustainable Energy Reviews, 16
flow rate of 0.96 m3/h. With the addition of 1 % (2012) 8, 5625–5638, doi:10.1016/j.rser.2012.05.023

512 Materiali in tehnologije / Materials and technology 52 (2018) 4, 507–512