Investigation on Electrical, Thermal, and Mechanical Properties of Silicone Rubber ATH Nanocomposites

In the present study, investigations have been carried out by choosing the nano-sized ATH (30–50 nm) particles as filler content in silicone rubber to acquire a deep insight to understand the use of nano ATH fillers in silicone rubber on its performance, to use it as insulation structure as an outdoor insulator. Variation in the surface condition of silicone rubber nanocomposites due to corona aging and the recovery characteristics analysis carried out through contact angle measurement. Nano ATH added in silicone rubber, up to 3 wt% of filler, shows increased dielectric constant with less loss factor. Water droplet-initiated discharges on silicone rubber nano ATH composites under DC voltage measured using UHF technique and the damage caused zone is analyzed through AFM studies and by surface potential measurement, which indicates higher surface roughness and increased charge trap depth. It is observed that negative DC voltage has higher CIV followed by positive DC and AC voltage. Laser Induced breakdown spectroscopy (LIBS) studies show a direct correlation between plasma temperature and material surface hardness. Also, on laser shining on the surface of the specimen, its temperature was measured on the backside of the specimen through thermal imaging confirming increased diffusivity of temperature. Laser flash analysis shows a direct correlation between filler content in base polymer and thermal conductivity and diffusivity. Higher storage modulus and activation energies measured from the dynamic mechanical analysis indicate the significance of prepared composites as an excellent mechanical support structure for the power system network.

DC voltages (iv) variation in surface potential decay and  pare the polymer composites. The ATH nanoparticles were 96 kept in the oven at 150 • C for 24 hours to remove moisture. 97 Then the dried nanoparticles with different weight percent-98 ages were measured and mixed with the ethanol solvent for 99 30 minutes. For efficient dispersion of particles in solvent, 100 the sonication process was carried out for 30 minutes by 101 using an ultra-sonicator with the rating of 750 W, 20 kHz, 102 and 230 V using 9s ON and 9s OFF mode. The properly 103 dispersed nanoparticle solvent was then added to the part-104 A base RTV silicone rubber (RTV8112, Momentive, USA) 105 and stirred for another hour. The mixture was kept in the 106 oven at 100 • C to remove excess solvent. A curing agent 107 (Hardener RTV9858) was added at a 1:10 ratio to the solvent-108 free solution. The degassing was performed by keeping the 109 mixture in a vacuum to remove trapped air bubbles. The 110 final mixture was kept in the compression mold for 24 hours 111 at 0.1MPa, at room temperature. Using the same procedure, 112 silicone rubber nanocomposite samples containing 0, 1, 3, 5, 113 and 10 wt% of nano ATH were synthesized and are labeled 114 as S0, S1, S3, S5, and S10 respectively. The static contact angle of pure silicone rubber and its com-118 posites were measured with a goniometer by placing a 10 µL 119 deionized water droplet over the specimen surface. The aver-120 age of five independent readings was obtained for each spec-121 imen and used for comparison. To understand the influence 122 of added fillers on hydrophobicity recovery characteristics, 123 specimens were exposed to AC corona for 30 minutes using 124 a multi-needle electrode system, as recommended by the 125 CIGRE working group D1.14 [19].  Fig. 1 shows the setup for measuring corona inception volt-128 age (CIV), which includes a signal generator (Tektronix 129 3051C), a high voltage amplifier (Trek type 20/20 C), and 130 stainless-steel electrodes with a 45 • cut at the tip, as per 131 IEC 60112. These electrodes were mounted on a 2 mm thick 132 silicone rubber test specimen. Throughout the experiment, a 133 10 mm distance was maintained between the electrodes. The 134 high voltage source was connected to one electrode through 135 a 10 M resistor, while the other was grounded. A 20 µL 136 deionized water droplet was placed between the electrodes. 137   For measuring the dielectric response, samples with 177 a diameter of 20 mm and a thickness of 2 mm were used. 178 The dielectric properties were analyzed using a Novo control 179 technology broadband dielectric/ impedance spectrometer 180 (Alpha-A High-performance frequency analyzer). The exper-181 iments were carried out at room temperature within a span of 182 frequency range from 0.1 Hz to 1 MHz.
The elemental analysis of a material can be determined ana-185 lytically using LIBS, regardless of its condition. This detec-186 tion method has several advantages including low sample 187 usage, high sensitivity, non-destructive nature and the ability 188 to be deployed remotely. Fig. 3 depicts the LIBS experimen-189 tal setup. The laser pulse was generated using a Q-switch 190 driven Nd 3+ : YAG laser (LAB-150-10-S2K, Quanta-Ray 191 LAB series, Spectra-Physics) with a repetition rate of 10 Hz 192 and a pulse duration of 10 ns. The principle of operation of 193 LIBS and the method of acquiring spectra were detailed in 194 the earlier study [18].  The LFA test was conducted with Netzsch LFA 467 Hyper-206 flash (Netzsch GmbH, Selb, Germany) equipment. A xenon 207 flash lamp was used as a heating source. For the analysis, 208 samples with a diameter of 12 mm and a thickness of 2 mm 209 were used, and the temperature range was selected as 30 to 210 150 • C. The thermal conductivity and thermal diffusivity 211 parameters were derived from the LFA data. rubber composites subjected to corona discharges. The addi-226 tion of filler shows a marginal reduction in the contact 227 angle. The SR have high contact angle due to the low sur-228 face energy of the polar methyl groups in their side chain 229 of SR having surface resistant to water and shows high 230 hydrophobicity [21].
231 Fig 4b and c shows the initial and complete hydropho-232 bicity recovery characteristics of SR and composites after 233 the corona exposure. The contact angle of silicone rubber 234 composites was reduced by an average of 20 to 30 degrees 235 immediately after the corona exposure. All the specimens 236 recovered their hydrophobicity almost in 20 hours but not 237 near equal to the virgin specimen and recovered to their 238 original contact angle after 25 days. The recovery rate is high 239 with the virgin specimen S0. The hydrophobicity of silicone 240 rubber recovery is by the diffusion of low molecular weight 241 (LMW) components to the surface. The slow recovery of 242 silicone rubber composites is due to the immobilization of 243 LMW components by the fillers enhancing the recovery time 244 high [21]. It was noticed that the measured CIV is high under DC 250 voltages compared to AC voltage. The reason could be due 251 to the water droplet oscillation caused by AC voltage [22]. 252 Also, AC voltage polarizes the water droplet on the specimen, 253 causing more stress at the triple-junction initiating corona 254 discharge [23]. Under DC voltages, the negative DC CIV is 255 higher than positive DC CIV. Charge accumulating on water 256 droplets causes corona discharge under positive DC voltage, 257 lowering CIV voltage. On the other hand, under negative 258 DC the charge accumulates along with electrostatic force, 259 changing the morphology of the water droplet causing higher 260 voltage to initiate discharges under negative DC voltage [24]. 261   in an unfilled sample is high, and adding fillers prevent the 300 samples from the damage occurring due to the discharges.

302
Gas neutralization or bulk neutralization dominates surface 303 potential decay based on polymer type and experimental con-304 ditions [26], [27], [28]. So, in the present study under unvary-305 ing experimental conditions, the surface potential decay 306 phenomenon was attributed mainly to charge transport and 307 charge trapping performance was studied using average depth 308 and trap density [29], [30].

309
The surface potential decay measurement of each SR com-310 posite specimen was repeated three times. Fig. 9 shows the 311 surface potential decay with time for neat SR and its com-312 posites under both polarities of applied DC voltages. The sur-313 face potential decreases with time regardless of material and 314 applied voltage polarity. The reduction in surface potential as 315 a function of time (V(t)) is modeled using the exponentially 316 decaying mathematical function, i.e.
where q is the charge of the electron, L is sample thickness,  higher permittivity represents higher orientational polariza-356 tion between the filler and silicone matrix. In general, relative 357 permittivity in polymer composites is determined primarily 358 by polarisation with silicone matrix and nanofillers, as well 359 as interfacial polarisation at the matrix-nanoparticle interface. 360 The fact that nanocomposites have higher real permittivity at 361 low frequencies can be attributed to Maxwell-Wagner-Sillars 362 or interfacial polarisation at the solid interface [31].
363 Fig. 11b depicts the frequency dependence of the loss 364 tangent variation of SR and its composites. Among all the 365 specimens, specimen S0 has the highest dielectric loss. The 366 addition of fillers resulted in a significant decrease in loss 367 tangent. The loss factor increases at lower frequencies due 368 to higher values of imaginary permittivity near the relaxation 369 frequency [32].

371
It was recently reported that using statistical tools in conjunc-372 tion with LIBS data can be used to identify and classify the 373 VOLUME 10, 2022 was measured using a FLIR thermal camera. The camera 408 temperature was set at a minimum of 28 • C and a maximum 409 of 40 • C.

410
By using the image processing technique, the frequency of 411 all temperatures in the image was plotted. Thermal images 412 and temperature-frequency plots of samples S0, S3, and S5 413 are shown in Fig. 13. Local heat accumulation (LHA) is one 414 of the main reasons for the erosion of SR. In Fig.13, the area 415 of local heat accumulation (visible in yellow color) is more 416 for sample S0. With higher filler loadings, a reduction in the 417 LHA area was observed. LHA allows the heat to build up 418 from one side to the other without spreading. This problem 419 can be avoided by adding nano-ATH fillers.

420
The spreading of heat all over the surface with respect 421 to length is shown in Fig. 14a. Maximum temperature was 422 observed at the spot where the laser pulses were applied. The 423 maximum temperature on samples S0, S1, S3, S5 and S10 424 were observed as 39.21 • C, 38.35 • C, 37.96 • C, 35.44 • C and 425 34.85 • C respectively. Due to more spreading of heat in filled 426 samples, a reduced maximum temperature was observed. 427 In order to measure the temperature decay rate of the SR 428 samples with respect to time, the maximum temperature was 429 measured at the midpoint of the thermal image at 0 s, 10 s, 430 20 s, and 30 s after the laser pulse had stopped. The plot-431 ted data of temperature at different time instances and their 432 curve fitted plots are shown in Fig. 14b. The decay rate was 433 calculated from the straight-line equation y = mx + c. Here 434 m indicates the slope of temperature decay.     the specimen is observed with unfilled specimen, which has 468 lower thermal conductivity.     where the applied vibrational frequency is mentioned as f, 492 R is known as the gas constant, f 0 is known as the proportion-493 ality constant and E a is the activation energy. In elastomers, 494 activation energy is the energy required for a polymer matrix 495 to change to a rubbery structure from a crystalline structure.

496
Activation energy is calculated from equation (5). Table. 3 497 shows the variation in glass transition temperature (T g ) 498 and activation energies at different frequencies. The relation 499 between glass transition temperature and frequency is shown 500 in Fig. 18 and it follows the Arrhenius relation.

501
The increase in glass transition temperature was observed 502 with the increased frequency as well as increased filler con- of fillers and this leads the composites to have higher activa-506 tion energy.

508
The important conclusions accrued based on the study are the 509 following:

510
• Increase in wt% of nano ATH with silicone rub-511 ber showed a marginal reduction in contact angle of 512 the specimen. On corona ageing of the nano ATH 513 added silicone rubber, it showed a reduced hydropho-514 bicity recovery rate compared to pristine composite 515 samples.

516
• Water droplet-initiated corona discharge on silicone rub-517 ber nanocomposites, under AC and DC voltages. corona 518 inception voltage is high under DC voltage compared 519 with AC voltage. It is also observed that under DC 520 voltages, the CIV is high under negative DC voltage and 521 the erosion of material due to water droplet discharge is 522 high near the positive electrode.

523
• Surface potential variation is very slow with the nano 524 ATH added SR samples compared to virgin samples. 525 The initial potential is almost similar for all the sam-526 ples. Right shifting in charge trap characteristics was 527 observed with the increasing wt% of fillers.

528
• DRS results show significantly higher relative permittiv-529 ity values and the loss factor is reduced drastically with 530 the addition of fillers especially with the filler loading 531 up to 5 wt% of nano ATH.

532
• Laser-induced breakdown spectroscopy (LIBS) analysis 533 clearly indicated that with the increase of filler content, 534 plasma temperature increases.

535
• on laser shining the surface of the specimen, the back-536 side of the specimen, its temperature was measured 537 through thermal imaging confirming increased diffusiv-538 ity of temperature. Also, the temperature decay rate and 539 the filler content in the composite material showed direct 540 correlation.

541
• Laser flash analysis showed higher thermal conductiv-542 ity and diffusivity with the increase in filler content in 543 silicone rubber nano ATH composites.