Thermomechanical properties of elastomeric-filled composites

The aim of this research project was to modify the matrix polymer’s relaxation processes by the inclusion of filler and to characterise the interfacial interaction in between matrix polymer and filler. Three types of elastomers were selected for this research: a purely amorphous elastomer (poly(ethylene-co-propylene) diene monomer) (EPDM), a semi-crystalline elastomer (poly(ethylene-co-methyl acrylate)) (EMA) and a thermoplastic elastomer (thermoplastic polyurethane) (TPU). These elastomers possessed reinforcement via chemical crosslinking, crystalline molecular segments and hydrogen bonding. Fillers in the form of aluminas (with EPDM) and fumed silicas (with EMAs and TPUs) were added to the elastomers to form composites. Materials were prepared via a solvent casting method. Elastomers were dissolved in an appropriate solvent and fillers were added. A crosslinking agent was also added to the dissolved EPDM. Fillers were dispersed with ultrasonication after which dissolved elastomerfiller compositions were dried then pressed into films.<br><br>EPDM composites underwent thermally-induced crosslinking. Characterizing techniques including environmental scanning electron microscopy (ESEM), thermogravimetry and analytical techniques including dynamic force thermomechanometry (df-TM), static force thermomechanometry (sf-TM) and modulated force thermomechanometry (mf-TM) were used to examine the interaction between filler and elastomer. Analysis was performed on data from sf-TM (4-element model for creep analysis data, Kohlrausch-Williams-Watts model for recovery analysis data) and mf-TM (Debye, Cole-Cole, Cole-Davidson and Havriliak-Negami relaxation equations). Mastercurves were created from multi-frequency mf-TM data. Sinusoidal frequency mode was determined to be the best method for obtaining multi-frequency data and the HN-based mastercurve method was the best method for constructing mastercurves. The accuracy of three different shift factor models (Williams-Landel-Ferry, Vogel-Fulcher-Tammann-Hesse and arctan based V’ant model) was compared. Super-mastercurves were constructed from mastercurves. ESEM micrographs of EPDM materials revealed that greater compatibility between elastomer and filler resulted in better dispersion of filler. All elastomers were observed to have thermal reinforcement provided by the presence of fillers. Greater compatibility between elastomer and filler resulted in increased thermal stability, as seen in EPDM composites.<br><br>Thermal reinforcement was observed to increase with increasing silica volume fraction with the Abstract xxiii temperature of 50 %.w/w loss increasing by approximately 20 °C for the EMA-based materials and by 70 °C for the TPU-based materials at the highest filler volume fraction. Stress-strain analysis data indicated that composites were tougher with greater compatibility between filler and elastomer matrix. The Young’s moduli and strength of all elastomer based systems were observed to increase, but toughness decreased with increasing filler volume fraction. Creep and recovery analysis showed that greater compatibility between filler and elastomer, and increased filler volume fraction resulted in more resistance to elastic and viscoelastic deformation, and viscous flow. Modelling of creep analysis data revealed that the presence of any filler resulted in increased resistance to instantaneous and time dependant uncoiling and extension of molecules. Increasing filler volume fraction resulted in greater resistance to instantaneous and time dependant uncoiling of molecules, and irreversible flow. Materials analysed with a reduced force had greater resistance to all modes of deformation as indicated by their parameter values being greater than those obtained at a higher force. Modelling of recovery analysis data revealed greater compatibility between filler and elastomer resulted in less resistance to recovery and less constraints to the operation of relaxation modes. Resistance to recovery decreased with increasing filler volume fraction. Relaxation times decreased with increasing filler volume fraction. Parameter values obtained from analyses conducted with reduced analysis force indicated that a lower analysis force resulted in less resistance to viscoelastic recovery, longer relaxation times and fewer constraints to the operation of relaxation modes. Modulated force thermomechanometry characterised the elasticity and damping properties of materials.<br><br>The storage modulus was observed to increase with increasing filler volume fraction; however increased compatibility between filler and elastomer resulted in reduced E′ values. The temperature of failure increased with compatibility and increasing filler volume fraction. The loss modulus was observed to increase with increasing filler volume fraction; however increased compatibility between filler and elastomer resulted in lower loss of energy. The tan d values decreased with the addition of any filler and increasing filler volume fraction; however increased compatibility between filler and elastomer reduced the tan d by a lesser amount. Mastercurves were constructed from the isothermal multi-frequency mf-TM data for the EMA and TPU series of materials. They possessed frequency ranges of approximately 10-9 – 1017 Hz. The storage modulus mastercurve data values were observed to increase with increasing filler volume fraction for all materials. The storage modulus values plateaued at Abstract xxiv 10 %.v/v filler. The loss modulus mastercurves of both EMA and TPU-based materials possessed a peak in the loss modulus data corresponding to Tg. <br><br>The materials plateaued at approximately the same modulus values at the highest transposed frequencies. The modulus data obtained at the highest analysis frequencies exhibited constructive interference with the effects most pronounced in the highest filled composites. The WLF model most accurately modelled shift factor values with respect to temperature. The Wicket error function was used to determine the accuracy of modelling with the relaxation equations. The HN equation modelled the mastercurve data most accurately. The zero and infinite modulus HN parameter values increased with increasing silica volume fraction. The infinite modulus values plateaued at 10 %.v/v filler at the highest transposed frequencies. The relaxation time parameter values increased with increasing filler volume fraction while the a and b shape parameters were observed to overall decrease. Mastercurves could not be constructed for EPDM-based materials as they could not be analysed at sufficiently high frequencies, so single frequency mf-TM data was used with the HN relaxation equation. The addition of compatibilized filler resulted in an increase in E∞ modulus values while the uncompatabilzed filler caused a reduction. The a shape parameter was observed to decrease with the addition of all fillers with the untreated filler resulting in the greatest reduction. The b shape parameter was observed to slightly decrease with the addition of the compatibilized fillers but increase with the addition of the uncompatabilzed filler. The calculated relaxation times were observed to decrease with decreasing compatibility between filler and elastomer.