Investigation of the Dielectric and Thermal Properties of Non-Edible Cottonseed Oil by Infusing h-BN Nanoparticles

Vegetable oils have emerged as insulating fluids in transformer applications and as a prominent and effective alternative for traditional dielectric fluids. However, most of vegetable oils are edible causing their application on a large scale to be limited. In the present work, a novel non-edible vegetable oil is developed as an insulating fluid. The developed oil is oxidation-inhibited cottonseed oil (CSO) based nanofluids. Tertiary butylhydroquinone was used as antioxidant. The concept of nanofluids was used to overcome the limited dielectric and thermal properties of cottonseed oil. Hexagonal Boron Nitride (h-BN) nanoparticles at low weight fractions (0.01 - 0.1 wt%) were proposed as nanofillers to achieve adequate dielectric strength and improved thermal conductivity. Stability of prepared CSO based nanofluids was analyzed using Ultraviolet–visible (UV-Vis) spectroscopy. Then, the prepared nanofluids were tested for dielectric and thermal properties under a temperature range between 45 °C and 90 °C. The dielectric properties include breakdown strengths under AC and lightning impulse voltages, dielectric constant, dissipation factor, and resistivity, while thermal properties include thermal conductivity and thermogram analysis. The dielectric and thermal properties were significantly improved in CSO based nanofluids. The creation of electric double layer at nanoparticle/oil interface and the lattice vibration of nanoparticles were used to clarify the obtained results. The proposed CSO based h-BN nanofluids open up a great opportunity in both natural ester insulating fluid applications and thermal energy management systems.

It is important to maintain the key properties of insulating fluids used in transformers. These fluids have dual roles, where they perform as an electrical insulator and heat transfer agent. The dielectric strength of these fluids should be adequate to withstand the potential range of electrical stress imposed during service. On the other hand, combinations of their viscosity, thermal conductivity, and specific heat should be sufficient to transfer heat from transformer to surroundings. In addition, insulating fluids should have a sufficiently high flash point and fire point to meet safety standards. Finally, they should be resistant to oxidation and deterioration.
Owing to sustainability, biodegradability and pollutionfree nature of natural ester insulation oils, many researchers and industries investigated such oils for transformer applications [1]- [3]. As a result, several international standards were devoted to these oils such as ASTM D6871, IEC 62770, IEEE C57.155-2014, ASTM D6871, IEEE C57.147-2018, and IEC 63012. These standards endorse modified and blended NEIOs in transformers, assist equipment manufacturers and service companies to determine the suitability of unused NEIOs received from suppliers, and assist transformer operators in assessing and maintaining NEIOs in a serviceable condition.
Turning vegetable oil feedstock into a competitive commercial transformer grade isolating oil with better inherent properties imposes numerous practical consequences and challenges. The first challenge is the oxidation tendency of vegetable oils with corresponding aging and impact on various oil properties [4]. The second and important challenge is the dependence of vegetable oils on edible products. Accordingly, this paper aims to develop a novel non-edible vegetable oil as an insulating fluid.
In India, agro-technological innovation and biotechnology offers tremendous potential for sustainable and cost-effective production of cotton [5], whereas cottonseed oil is not as common as edible oil. On the other side, CSO oil is biodegradable and non-toxic making it an environmentally friendly fluid. In addition, it has high flash point and fire point [6]. Despite CSO has these favorable characteristics, it has a high level of un-saturated fatty acid, leading to a greater risk of oxidation. However, to resolve this limitation, the base feedstock can be upgraded to higher oleic content by means of genetically modified CSO [7] or by using antioxidant additives.
Blending and dispersing stable ultra-fine nanoparticles, such as metals, oxides, nitrides, carbide ceramics, and carbon nanotubes, into fluids forms what is called nanofluids. These nanofluids had several distinct characteristics, including dielectric and thermal characteristics [8]- [12]. Nanoparticles can be classified into conductors, semiconductors, and insulators. The high surface area of two-dimensional materials leads to an efficient heat transfer phenomenon and is a better alternative for the application with nanofluids. The h-BN or graphene nanoparticles have a multifaceted feature among various 2D materials. However, h-BN overtakes other nanofillers and is an enticing material with high thermal conductivity and electrical insulation properties [13], [14]. Furthermore, theoretical studies demonstrate that the planes of h-BN can achieve high thermal conductivity [15]. Table 1 shows, few previous experimental studies on nanoinfused natural esters. These studies show that there is a considerable percentage enhancement in the electro-thermal characteristics of nanofluids compare to corresponding base fluids.
In the present investigation, it is aimed to develop non-edible CSO as natural ester insulating fluids through inclusion of antioxidant and filling with h-BN nanofillers. Inclusion of antioxidant will enable to resist the oxidation process of the oil, while filling with h-BN nanofillers will enable to enhance the dielectric and thermal properties of the oil. Tertiary butylhydroquinone (TBHQ) has been used as antioxidant. Four dosage levels (0.01, 0.02, 0.05, and 0.1 wt%) of h-BN nanoadditives have been introduced into base TBHQ cottonseed oil. Comprehensive overview on synthesis, characterization and stability analysis of CSO based h-BN nanofluids is presented. Breakdown strengths under AC and lightning impulse voltages are evaluated. Also, other dielectric properties such as dielectric constant, dissipation factor, and resistivity are measured. In addition, thermal properties including thermal conductivity and thermogram analysis are obtained. All these results are compared to that of the base CSO. Finally, physical mechanisms behind the obtained results are discussed considering the effects of 2D material (h-BN) and the role of electric double layer at nanoparticle/oil interface.

II. MATERIALS AND METHODOLOGY
The h-BN nanoparticles were procured from Intelligent Materials Pvt. Ltd. (NANOSHEL Stock No: NS6130-02-260 CAS: 10043-11-5). The characteristics of the obtained h-BN nanoparticles are as follows. The size of the h-BN nanoparticles is in the range of 50-70 nm, Young's modulus of 14-60 GPa, and a density of 2.29 g/cm 3 . The h-BN nanofluids were prepared using the below mentioned process with various weight percentages, 0.01, 0.02, 0.05, and 0.1. Fig. 1a and Fig. 1b illustrate the visualization of the powdered sample and produced nanofluids at various h-BN weight percentages, respectively. VOLUME 8, 2020 Two-step method was implemented for preparing nanofluids in the context of its advantages compared to one-step method, where it has lower cost, more productivity, and reduced particle waste. CSO was used for the present implementation as the base fluid. This vegetable oil was provided by local supplier in India from M/s S S Odunavar Industries. The TBHQ antioxidant was used as a synthetic antioxidant due to its effectiveness at high temperatures in comparison to other synthetic antioxidants [2]. It was supplied from Sigma Aldrich. A 0.02% of TBHQ was added to CSO and the solution was stirred at 80 • C for 1.5 hour with a specific interval after each 30 minutes. The addition of antioxidants suppresses the decomposing of the oil at high temperatures. The measured quantity of h-BN is then added to the solution and stirred again for 30 minutes at the same temperature. The stable dispersion is then achieved by sonication for a period of 1 hour by a probe sonicator.
The boron nitride nanosheets consists of covalently bonded hexagonal B and N (sp2-hybridization). These h-BN layers are stacked on the top of each other, where the adjacent layers are held together by a weak force of van der Waals, as shown in Fig. 1c. The h-BN nanosheets consist of mono-layered h-BN planes. The 2D h-BN nanosheets are exfoliated from bulk h-BN by long term sonication similar to the process adopted in [20]. Fig. 1d shows pictorially exfoliation of h-BN nanosheets.  was measured using a UV-Vis spectrometer (Chemito Spectrascan UV-2700).
According to IEC156 and ASTM D3300 standards, the AC breakdown and impulse breakdown voltages of the prepared CSO/nanofluids were measured at room temperature. Based on these standards; electrode configurations, voltage increasing method/voltage ramp rate, time intervals, and other useful information are detailed in Table 2. ADTR instrument from ELTEL was used to measure loss tangent, dielectric constant, and resistivity of CSO/nanofluids at a standard operating frequency of 50 Hz and at four specific temperatures (45 • C, 60 • C, 75 • C, and 90 • C). For dielectric constant and loss tangent measurements, the testing voltage was set at 500 V, 50 Hz AC, while for resistivity measurements, it was set at 500 V, DC. The precision of the measurement device is ±0.1%, ±1%, and 2-5% for dielectric constant, loss tangent, and resistivity, respectively. Thermal conductivity of CSO/nanofluids was quantified by the KD2 PRO equipment from Decagon using the transient hot-wire method. Following factory calibration, the accuracy is about ±5% for the used sensor as given in the instrument specifications. To maintain the accuracy in the test results, the used sensor is crosschecked with a standard material before testing thermal conductivity of samples. Measurements were conducted at four different temperatures, 35 • C, 45 • C, 55 • C, and 65 • C.
For further thermal conductivity verification of CSO/nanofluids, thermal response tests were conducted using the experimental setup demonstrated in Fig. 2a. In brief, 100 ml of CSO/nanofluids was put into a beaker. Then, the electric current preheats CSO/nanofluids at a certain fixed power level for 26 minutes and the specimen was allowed to cool down naturally for the next 26 minutes. The heating and cooling response were recorded. Moreover, the thermal VOLUME 8, 2020 conductivity of CSO/nanofluids was validated through analyzing surface temperature images captured by the thermal imaging camera Fluke Ti125. Fig. 2b shows this experimental setup in detail. The experimental process includes heating 100 ml of various samples in a beaker under a certain fixed power level for 45 minutes and capturing surface temperature images at a specified time interval. The heating method is similar to that performed in thermal response tests.

IV. CHARACTERIZATION AND STABILITY ANALYSIS
Comprehensive characterization of h-BN nanoparticles and dispersion stability analysis of CSO nanofluids are reported in this section.

A. CHARACTERIZATION OF H-BN NANOPARTICLES
SEM samples were prepared on a carbon tape bonded on aluminum stub. A little h-BN sample was taken and placed on the carbon tape, then the loose particles were removed. After that, the stub was attached to the sample holder and put into the sample chamber for SEM analysis. Fig. 3(ac) illustrates the SEM images of h-BN nanoparticles. From the SEM analysis, the existence of h-BN sheets at higher magnifying levels is realized, the bulk structure is verified, and irregular morphology in this structure is observed.
During TEM samples preparation, h-BN nanoparticles are dispersed in ethanol. A drop of nanoparticle solution was deposited on the copper grid and was vacuum dried at room temperature before TEM imaging.  (110), respectively, representing hexagonal phase-planes of h-BN [21], [22].

B. STABILITY ANALYSIS OF H-BN DISPERSED NANOFLUIDS
Dispersion stability is a critical issue that affects the properties of nanofluids for various applications. One of the most effective ways to evaluate nanofluid stability is spectral absorption experiments. The following equation can establish the relation between absorbance and concentration: The nanofluid stabilization against the concentration of nanoparticles is strongly affected by the characteristics of the suspended particles and the base fluid. The higher levels of nanoparticles in the fluid lead to an increasing number of molecules into the fluid causing agglomeration with a subsequent increase in the weight. As a result, nanoparticles settle off the suspension and become less stable. Table 3 shows the absorbance values recorded on the first day and eighth day after preparation of nanofluids. At all concentrations of h-BN nanoparticles, there is no significant reduction in the absorbance values at the eighth day comparing to the first day, maintaining stability above 95% for all samples. This indicates that CSO nanofluids are quite stable.

V. ELECTRICAL MEASUREMENTS OF CSO/h-BN DISPERSED NANOFLUIDS
Inclusive dielectric properties of CSO/nanofluids are presented in this section. These properties include breakdown VOLUME 8, 2020   Table 4 along with standard deviation and percentage enhancement at all weight percentages. Fig. 4a and Fig. 4b depicts the mean value and standard deviation of the AC and the impulse breakdown voltages against the weight percentages, respectively.

B. DIELECTRIC CONSTANT OF CSO/NANOFLUIDS
The dielectric constant corresponds to the inability of dielectric molecules to reorient themselves with an alternating gradient of the electric field. The material permittivity (ε * ) under  alternating electrical field is a complex parameter given as follows: where, ε' is the real part that gives an indication for the degree of polarization and ε'' is the imaginary part that gives an indication for the dielectric losses. 76210 VOLUME 8, 2020 The dielectric constant of CSO/nanofluids is depicted in Fig. 5. The dielectric constant reduces with the increase in temperature at each weight percentage similar to that obtained in literature [23]. At 45 • C, the results indicated a small increase in the dielectric constant with the rise in the nanoparticles weight percentage up to 0.02 wt%. This increase was only about 1.3% at 0.02 wt%. Above 0.02 wt%, the dielectric constant decreased remarkably. These trends can be explained considering the expected polarization types in nanofluids as will be discussed in section 5.5.

C. VOLUME RESISTIVITY OF CSO/NANOFLUIDS
A fluid volume resistivity is a dc measurement of its electrical resistance to leakage current. Volume resistivity ( /cm) is the ratio of direct potential gradient (V/cm) parallel to the current density (A/cm 2 ) in the sample at a given time and prescribed conditions. The existence of conductive contaminants is normally indicated by a lower resistivity. Inherently, natural esters provide a smaller volume resistivity compared to mineral oils [24]. For CSO/nanofluids, volume resistivity against weight percentages of h-BN nanoparticles at various temperatures is illustrated in Fig. 6. The volume resistivity increases with filler levels from 0.0 wt% to 0.1 wt%. This improvement is attributed to the decrease in the mobility of charge carriers either due to trapping by nanoparticles or due to losing their energy by the rigid structure created at nanoparticle/oil interface. Furthermore, the decrement in volume resistivity against temperature at all weight percentages is attributed to the availability of more free charge carriers. These concepts are explained in detail in the upcoming section 5.5.

D. DISSIPATION FACTOR OF CSO/NANOFLUIDS
Dissipation factor indicates energy loss, and thus it is responsible for dielectric overheating under specific voltage, frequency, and temperature circumstances. A high dissipation  factor demonstrates higher dielectric losses. Losses correlated with the dissipation factor should not be confused with losses correlated with transformer loading and excitation. The losses correlated with the dissipation factor remain smaller than that for loading and excitation in several orders of magnitude. Inherently, natural esters have larger dissipation factors than mineral oils [25]. The term dissipation factor is often referred as the loss tangent and is defined as: where, δ is the loss angle between the charging current I c and the effective current I l . Fig. 7 depicts the dissipation factor of CSO/nanofluids at different weight percentages of h-BN nanoparticles with varying the temperature. According to Equation (3), VOLUME 8, 2020 the dissipation factor is directly proportional to conductivity and inversely proportional to permittivity. There are three main results can be observed from Fig. 7 as follows. (1) At 45 • C, the dissipation factor decreases with the increase in the filler levels from 0.0 wt% to 0.1 wt%. (2) For base CSO fluid, there is a large increment in the dissipation factor with the increase in the temperature. (3) The dissipation factor decreases slightly against temperature at 0.01 wt% and 0.02 wt%, while it increases against temperature at 0.05 wt% and 0.1 wt%. The mechanisms behind these results will be discussed in the next section.

E. PHYSICAL MECHANISMS BEHIND DIELECTRIC PROPERTIES OF CSO/NANOFLUIDS
There are two different mechanisms that can be responsible for dielectric properties of CSO/nanofluids. The first mechanism is the ability of nanoparticles to trap charges, which consequently depends on dielectric relaxation time. Dielectric relaxation time is highly dependent on the permittivity (ε) and electrical conductivity (σ ) of the nanoparticles, together with the surrounding fluid. The relaxation time of the nanoparticles in the fluid (τ h−BN ) is taken from Laplace equation [26] and can be represented as: where, subscripts 1 and 2 denote to CSO and nanoparticles, respectively. From many literatures, the electric conductivity of CSO is apparent within the range 10 −9 − 10 −12 S.m −1 and its permittivity was measured 2.95ε o F.m −1 . For h-BN nanoparticles, the value of electric conductivity is less than 10 −13 S.m −1 [27], and their permittivity was found in the range of 4.0ε o − 4.4ε o . With the lowest range, the relaxation time (τ h−BN ) was found to be about 43.85 ms according to Equation (4). There are four modes of streamer propagation in dielectric fluids [28]. The first and second modes occur at relatively low velocities and charges can be trapped by nanoparticles. This explains the enhancement in breakdown strength under AC voltage. However, under impulse voltage application, the third and fourth modes predominate. These modes are very fast compared to the relaxation time of nanoparticles making it difficult for nanoparticles to trap charges. Therefore, the charge trapping cannot contribute to the enhancement in impulse breakdown strength.
The second mechanism responsible for dielectric properties of CSO/nanofluids is governed by nanoparticle/oil interface. According to the theory of electrical double-layer (EDL) [29], the surface of nanoparticles in contact with transformer oil will accumulate free charges. These surface charges attract the counter-ions in the transformer oil as well as repulse the co-ions. Immediately next to the surface of the charged nanoparticles, there is a sheet of counter-ions that are highly bound to the particle surface and are immobile. This layer is known as a compact layer and has a rigid structure. From the compact layer to the electrically neutral transformer oil, a second layer called the diffuse layer is formed, through which the net charge density slowly falls to zero. Ions in the diffuse layer are mobile and are influenced easily by electrostatic forces.
EDL acts towards decreasing the energy of brisk electrons produced under high electrical field and trapping them as shown in Fig. 8. Due to the uniform distribution of nanoparticles in the oil, interfacial volumes predominate creating a lot of traps in the system. Higher concentration of nanoparticles results in greater interfacial volume, leading to more traps. This improves the insulation characteristics of oil including AC breakdown strength and volume resistivity. For impulse breakdown strength, the propagation velocity is very fast with limited time for charge trapping in EDL.
Regarding dielectric constant, it can be explained considering the expected polarization types in nanofluids. According to polarization model of oil mixtures with solid particles, there are two possible types of polarization, the polarization of oil molecules and the inner polarization of solid particles. But, with decreasing the size of solid particles to nanoscale, a third polarization is originated for oil molecules which are positioned at nanoparticle/oil interface. This third polarization has limited polarizability due to the rigid and aligned structure created at this region [30], [31]. So, the relative permittivity of CSO/nanofluids can be expressed by Clausius-Mossotti equation as follows: At a low concentration of h-BN nanoparticles (0.01 wt% and 0.02 wt%), nanofluids have a slightly higher relative permittivity than that of pure CSO. According to Clausius-Mossotti equation, this higher relative permittivity is predominantly due to the increase in total polarization of h-BN nanoparticles due to the increase in the number of nanoparticles (N 2 ). On other side, at higher concentration of h-BN nanoparticles (0.05 wt% to 0.1 wt%), the amount of dispersed h-BN nanosheets per unit volume is high. This leads to an increase in the number of oil molecules at nanoparticle/oil interface (N 3 ) that have limited polarization with a subsequent decrease in the number of oil molecules at the free space (oil molecules other than that at nanoparticle/oil interface). As a result, the first polarization term in Equation (5) significantly decreases, while the second and third terms slightly increase. According to these key factors, the dielectric constant at 0.05 wt% and 0.1 wt% drops. Regarding dissipation factor, it generally decreases with the increase in the weight percentages from 0.0 wt% to 0.1 wt% as obtained at 45 • C. This may be due to charge trapping by nanoparticles and EDL with a subsequent decrease in the mobility of charge carriers and the electric conductivity. For base CSO fluid, there is an increment in the dissipation factor with the increase in the temperature as usual in dielectric fluids, due to the increase of charge carriers and subsequent increase of conductivity. The temperature-dependency of dissipation factor exhibited a slight decrease against temperature at 0.01 wt% and 0.02 wt% and an increase 76212 VOLUME 8, 2020 against temperature at 0.05 wt% and 0.1 wt%. This trend is attributed to the effect of electrophoresis, which results in a decrease in electrical conductivity against temperature for a certain temperature range and an increase in electrical conductivity against temperature for another temperature range [32]. Under AC voltage application the h-BN nanoparticles with its EDL migrate through the fluid according to Brownian motion [33] and following the AC voltage polarity. This migration reverses each half-cycle of the applied voltage. The temperature rise in this case increases Brownian motion [33]. For small weight percentages of nanoparticles, migration of nanoparticles assists in scavenging charge carriers into the fluid, thereby decreasing the conductivity and decreasing the dissipation factor. For higher weight percentages of nanoparticles, the distance between nanoparticles decreases. Increasing Brownian motion of nanoparticles with the temperature rise allows overlaps between adjacent EDLs, and thus facilitates transport of charge carriers. This increases the conductivity with a subsequent increase in the dissipation factor.

VI. THERMAL ANALYSIS OF CSO/h-BN DISPERSED NANOFLUIDS
Cooling is the second role of transformer oil. Accordingly, the heat conductance of CSO/nanofluids has been examined in comparison to base CSO by measurement of thermal conductivity, thermal response analysis and thermal images. Fig. 9a shows the thermal conductivity of CSO/nanofluids against weight percentage of h-BN nanoparticles for a temperature range between 35 • C and 65 • C. With the introduction of h-BN nanoparticles, the thermal conductivity continued to increase against weight percentage until attained about 0.221 W/m.K at 0.1 wt% and 35 • C. This represents an enhancement of 33% compared to base TBHQ CSO. The significant increase in the thermal conductivity of nanofluids deviates from the classical Maxwell theory [34].

A. THERMAL CONDUCTIVITY AND THERMAL RESPONSE ANALYSIS
Also, base TBHQ CSO thermal conductivity is almost flat at temperatures ranging from 35 to 65 • C and is compatible with that obtained in literature [35]. But, the thermal conductivity of nanofluids is temperature-dependent, making nanofluids act as smart fluids capable to dissipate more heat at higher temperatures. This confirm the role of h-BN nanoparticles in improving thermal conductivity.
Regarding thermal response, it is depicted in Fig. 9b. It is clear that nanofluids are heated and cooled down faster than base fluids. This effectiveness is improved with increasing the weight percentage of nanoparticles. This trend is similar to that obtained in graphene/transformer oil nanofluids [36].
There are three different mechanisms that are responsible for the obtained enhancement in thermal conductivity of nanofluids. The first mechanism is phonon transport through solid nanoparticles, which represents the governing mechanism of heat transport in solid materials. Phonon transport is originated from lattice vibrations [37]. The phonon mean free path is in the order of few tens of nm. In nanofluids, the ballistic phonon transport predominate the heat transport process as illustrated in Fig. 9c, since the size of h-BN nanoparticles reaches near the phonon mean free path. As shown in Fig. 9c, the 2D structure of h-BN creates stacks of nanosheets that produce pathways for successful heat transfer even at low weight percentages. With increasing the temperature, the atoms within the nanostructures move faster and vibrate in their mean position increasing the number of phonons.
The second mechanism is Brownian movement of nanoparticles [33]. The enhanced thermal conductivity due to Brownian movement is expressed by the following formula [38]: where the second term accounts for the enhancement due to Brownian movement. It is clear that Brownian movement effect rises with increasing the temperature, giving increased thermal performance. In addition to the contribution of Brownian movement itself to the thermal conductance, it also helps transmit phonons between nanoparticles. VOLUME 8, 2020 The third mechanism that can contribute to enhanced thermal conductivity is the interparticle interactions due to the EDL at nanoparticle/oil interface [39]. This EDL causes a repulsion Coulomb force on nanoparticles resulting in an additional movement of nanoparticles with a subsequent enhancement in thermal conductivity. Also, EDL facilitates the phonon transport between adjacent nanoparticles, especially with the elevated interfacial area created by the 2D structure of h-BN nanoparticles [40].
It is important to point out that agglomeration of nanoparticles affects thermal conductivity of constituting nanofluids [41]. But, this effect is omitted in the present study, where average h-BN size observed by TEM and stability of nanofluids observed by UV-Vis spectroscopy indicated that samples are agglomerate free.

B. 6.2. THERMOGRAM ANALYSIS
The infrared optical images taken to support improved thermal conductivity in CSO/nanofluids are displayed in Fig. 10. All samples have been analyzed in the same range of temperature. The images show that the surface temperature of nanofluids at selected time period is above the base CSO and has an increasing tendency with the weight percentage. The average surface temperature recorded for base CSO and nanofluids with maximum weight percentage of 0.1 wt% were 95 • C and 104 • C, respectively, after 45 minutes of heating. This validates the better thermal conduction of nanofluids.

VII. CONCLUSION
In this article, non-edible cottonseed oil (CSO) was developed as natural ester insulating fluids through inclusion of antioxidant and filling with h-BN nanofillers. Different weight percentages of h-BN nanoparticles have been used ranging from 0.01 wt% to 0.1 wt%. The dispersion of nanoparticles and their suspension stability have been confirmed through TEM observation and UV-Vis spectroscopy. The usage of h-BN nanoparticles could simultaneously improve dielectric and thermal properties as follows: 1. An enhancement in AC breakdown voltage was achieved and attained a maximum percentage increment of 63.3% at a weight percentage of 0.1 wt%.
On the other hand, impulse breakdown voltage has been enhanced to a little extent of about 5%. These results were explained considering the charge trapping by nanoparticles themselves or by EDL created at nanoparticle/oil interface. 2. The dielectric constant of CSO/nanofluids indicated a small increase with the weight percentage of nanoparticles up to 0.02 wt%, and then a remarkable decrease above 0.02 wt%. These trends were explained considering the different polarization types in nanofluids. 3. The volume resistivity of nanofluids increased and their dissipation factor decreased against weight percentages of h-BN nanoparticles due to the charge trapping effect. 4. The thermal conductivity of nanofluids enhanced significantly compared to base CSO as validated from thermal conductivity measurement, thermal response analysis, and thermogram analysis. Also, the thermal conductivity of nanofluids was temperature-dependent capable to dissipate more heat at higher temperatures. These enhancements have been explained considering phonon transport through nanoparticles, their Brownian movement, and interparticle interactions due to the EDL at nanoparticle/oil interface.
All the obtained results and clarified mechanisms in the present study validate the superiority of CSO based h-BN nanofluids as a potential candidate either for power equipment or for management systems of thermal energy. MANZOORE ELAHI M. SOUDAGAR was born in Dharwad, India, in 1988. He received the B.Tech. and M.Tech. degrees in mechanical engineering from Visvesvaraya Technological University (VTU), Karnataka, in 2010 and 2012, respectively. He is currently pursuing the Ph.D. degree in the biodiesel production from nonedible feedstocks and effects of nano-additives on the performance and emission characteristics of diesel-biodiesel fuel, Department of Mechanical Engineering, University of Malaya, Kuala Lumpur, Malaysia. His other research interests include biodiesel production technologies, finite element analysis of porous cavities, and solar energy. VOLUME 8, 2020