The rheological and stochastic characteristics of the fracture of elastomers are hereunder discussed to begin with. For the failure process of elastomers, two hypotheses are proposed, viz. (a) The flow of the polymer chains in polymeric solids may be effective to disperse the stress concentration around a flaw in the solid. (b) The probability to grow a catastrophic flaw in the solid may increase by the increase in time interval during the flow. These two hypotheses will bring reverse effects on the strength of the polymer. A theory on fracture mechanism based on the hypothesis (b) is introduced, and, on this theory, the creep failure process of some elastomers are analyzed according to the theory. In this analysis, two material constants on fracture, by which the failure process is characterized, are obtained by the experiment. As another application of the theory, the distribution of the breaking stretch ratios in uniaxial extension of constant rate is estimated by the constants obtained by the creep fracture. There is good agreement between the theoretical and the experimental results, but there are cases where the fracture mechanism assumed by hypothesis (a) is also important. It is considered practically that both the mechanisms occur simultaneously. Secondly, the importance of the research on the strength in multiaxial stressed state is emphasized for the sake of considering the fracture phenomena of elastomers in relation to the micro structures. Some data from a biaxial extension testing of valcanized SBR is introduced and, as a fracture criterion, the following formula is presented λi≥λB (i=1 or 2) where λi is stretch ratio along the principal axes (axis 1 and 2 have orthogonal direction to each other), λB is a critical stretch ratio and independent value on any combinations of λi (i=1, 2). This simple result is discussed in relation to its mechanical property obtained by the biaxial tensile testing.
Hereunder is discussed application of Weissenberg rheogoniometer to the study of viscoelastic properties of linear polymer solutions. From the normal stress difference and shear stress measured at varying shear rates, it is possible to obtain the steady-flow viscosity and the steady-state compliance as functions of shear rate, which will give the energy dissipation and the energy storage of the polymer solution. Using monodisperse poly (α-methyl styrenes), the effect of molecular weight and molecular weight distribution on both steady-flow viscosity and steady-state compliance are discussed in comparison with the current molecular theories on viscoelasticity of linear polymer solutions. Moreover it is pointed out that the current continuum theories on the three dimensional viscoelastic properties of linear polymer solutions are not satisfactory even in a shear strain region where the solution behaves linearly.
When vulcanized rubber is strained by external force at lower temperature, e. g. lower than its glass transition temperature, it is considered that there occurs cleavage in the network of polymer chains. When natural rubber has been strained by external force to 200% of its original bulk at -70°C, and is left in nitrogen at room temperature for about four hours, its cross-linking density increases about 17% as high. The relation between the cross-linking density and the rate of deformation due to the strain shows the Maxwellian type. Experiments have revealed the fact, on the other hand, that when vulcanized rubber is impregnated with radical acceptor its cross-linking density gets remarkably reduced in the rate of increment. It is conceivable, all things considered, that some mechanical reaction has taken place on the degraded rubber radicals to recombine.
The dielectric behavior of crosslinked polymers containing 11-membered ring (β-relaxation) has been investigated in connection with the dynamic mechanical properties. Control of the mobility of the ring structure and its homologous structures involved in the crosslinked polymers was attempted by modification of the substituents at the 1-or 1, 2-position of diallyl succinate, and the diallyl succinate monomers were derived from the succinic acid and its derivatives: succinic acid (designated as DASu), methyl succinic (DAMeSu) acid; cis-1, 2-dicarboxylic acids of cyclopropane (30C), cyclobutane (40C) and cyclohexane (60C); dihydrohymic acid and phthalic acid (HHAC and DAP). The results of the investigation are as follows: All the crosslinked polymers exhibit β-relaxation at 20°C when measured at 30KHz, which is presumably associated with the local motion of side chains of the polymers. The energy of activation is ΔH≠=11-14Kcal/mole, ΔS≠=10-20cal/mole. 60C shows an abomalous behavior presumably owing to the existence of cyclohexyl ring which flips between the chair-chair forms.
Polypyromellitimides show excellent thermal stability due to their rigid aromatic chain molecules. In the present paper, the viscoelastic properties of such polymers are correlated to their rigid chain structure. The viscoelastic properties of polypyromellitimides prepared with various degrees of cyclization from polypyromellitamic acid film are measured with a tensile oscillating apparatus at 5Hz in the range of -60∼90°C. And the viscoelastic properties of polypyromellitimide, Kapton H film, are also measured at a higher temperature range of 20∼500°C by means of a torsional pendulum. Three mechanical ralaxations were found at -60°C and 70°C by tensile measurement, and at 60°C and 330°C by torsional measurement. The lowest one is ascribed to the absorbed water as previously pointed out. The middle one is found also to be related to the absorbed water. The highest one, which is revealed in this work for the first time, is assigned to rearrangement of the oriented molecules at higher temperatures.
Birefringence was measured in the creep recovery of preoriented PVC cast films both of the unplasticized sort and of the plasticized sort. The PVC films were stretched under dead load at various temperatures ranging below and above glass transition temperature. The measurements of temperature dependence of orientaion birefringence were conducted between room temperature and 160°C at a constant rate of temperature increase. The experimental findings are as follows: (1) Birefringence increases almost linearly between room temperature and Tg (2) Tg is independent of the temperature at which the films are stretched. (3) Another transition above Tg exists at 105°C, which does not depend on the stretching temperature, and is affected by plasticizer. (4) Transition nearby the stretching temperature can be visualized easily at Tori using birefringence measurement. The initial increase in birefringence is explained as having been caused by thermal expansion. The 105°C-transition is proposed to be a lower limit of melting point, Tm(min), of a microcrystalline structure formed by stretching. Tori corresponds to one of the state of the polymer association being observable under stretching between Tm(min) and Tm(max) which is the conventional primary transition.
The dynamic mechanical tests of samples of ethylene-vinyl acetate copolymers with various contents of vinyl acetate monomer (VAc) ranging from 0 to 25.3wt% made at varying temperatures from 5°C to 80°C at frequencies ranging from 1 to 100 cycles/sec., and their dynamic tensile modulus E', E" were obtained. These were made the data for examining each sample with respect to its principle of superposition of time and temperature. It was found that the larger the melt index of the sample was the wider was the scatter of these data in it. The shift distance log aT was larger than those given from WLF equation. The values of log aT2 computed from the following equation gave good straight lines on a plot of log aT2 vs. 1/T; logaT=(1-φ)logaT1+φlogaT2 where log aT and log aT1 represent respective shift distance obtained from experiment and from WLF equation, and in present trial, φ was assumed to represent volume fraction of crystalline part of each sample. Time-concentration principle was also examined with a convenient assumption that one monomer composition in the copolymer act as a diluent or a low molecular plasticizer. It seemed that the principle of superposing time and concentration was formally applicable to the data of copolymer by regarding the content of one monomer composition as concentration.
In the previous paper, the viscoelastic and rheo-optical properties of ethylene-propylene block copolymers were reported and compared with those of polyethylene-polypropylene blends, suggesting the effect of the introduction of a primary bond between the different crystalline polymers on their physical properties. In the present work, ethylene-isoprene (A-B type) block copolymers having different compositions have been synthesized and their viscoelastic and rheo-optical properties have been studied to clarify the nature of block copolymers composed of crystalline and amorphous polymers. The samples of block copolymers were prepared with the catalyst system of Al (C2H5)3 and TiCl4 in n-hexane. The compositions of the copolymers were determined from the ratio of the infrared absorbances of the 830cm-1 and 720cm-1 bands. The temperature dispersion curves of storage Young's modulus E' measured with“Vibron” Model DDV II at 110Hz greatly depend upon the composition. The decrease of E' with increasing temperature is much restrained by the presence of the polyethylene component, even at small contents. The loss modulus E" versus temperature curves for copolymers show clear absorption peaks below 0°C, near-95°C (designated to γ-absorption peak due to polyethylene) and-45°C (designated to α-absorption due to glass transition temperature of polyisoprene). The temperature Tmax at which the absorption maximum is observed and the intensity of the former absorption are independent of the composition, supporting the idea that the γ-dispersion of polyethylene originates from the motion of the chain end. In the case of latter, the temperature Tmax is independent of the composition, but the intensity is dependent on the amount of polyisoprene. The birefringence of the block copolymers was measured simultaneously with the stress relaxation measurement at 25°C. The higher the content of polyethylene component, the longer is the time at which the strain-optical coefficient reaches its equilibrium value and the higher is the equilibrium value. This result may support the interpretation that the time dependence of strain-optical coefficient for polyethylene mainly depends on that of the orientation of the crystalline phase, which was concluded previously from the result of dynamic birefringence.
In order to clarify the deformation mechanism of crystalline polymers, their structural changes during their uniaxial drawing were studied by electron microscopy, by the X-ray method at small and wide angles and by thermal analysis. Polyethylene (PE) was chosen for study as highly crystalline polymer, and polyethylene terephthalate (PET) as polymer of low crystallinity. In the case of the drawing of a commercial PE film having the so-called a-axis orientation along the machine direction, the long X-ray spacing increased at first with increasing draw ratio, showing maximum increase of ca. 14%, which corresponded with the macroscopic elongation at the yield point, and further elongation gave rise to rapid drop in the elongation range 100-120%, and then to levelling off. This long constant spacing was dependent only on the drawing temperature, irrespective of the original spacing, suggesting the melt-recrystallization mechanism occurring in the drawing process. The necking behavior was analysed theoretically by assuming that melting occurred by the mechanical force applied to the lamellar crystals. The results suggest that the melting occurs through interlamellar tie chains from which local unfolding takes place. Another type of (complete) unfolding may occur on drawing PET film of a low crystallinity, which results in formation of fringe micelle structure.
The process of solid state extrusion of high density polyethylene was analyzed by assuming pure plastic behavior of the polymer. The agreement between the values calculated on the basis of the tensile data and that observed in extrusion experiments confirmed the assumed plastic nature of the polymer. The equation examined will be efficient as a starting point for detailed studies of solid state extrusion. Solid state extrusion of 6-nylon was also tried. It was successful only at the relatively low degrees of processing even at the temperatures close to the melting point. This fact makes contrast to that of polyethylene, in which extrudates of high degrees of processing was obtainable at the temperatures above 80°C.
The dielectric dispersions in copolymers of the methyl acrylate with styrene and of the ethyl methacrylate with styrene were measured at the temperatures from -75°C-140°C in the frequency range of 30c/s∼100kc/s. In the copolymers of methyl acrylate with styrene, two distinct tanδ maxima based on α-and β-relaxation processes were observed. The tanδ maximum corresponding to α-relaxation shifted to higher temperatures with decrease of methyl acrylate content, while the one corresponding to β-relaxation shifted to lower temperatures with decreasing methyl acrylate content. The latter effect is attributed to the dipole-dipole interactions of ester side chains. On the other hand, in the ethyl methacrylate-styrene copolymers, the change in composition did not affect the position of tanδ maximum corresponding to β-relaxation. The values of apparent activation energy of dielectric relaxation were obtained for methyl acrylate-styrene copolymers.
The glass transition phenomena of liquid are hereunder discussed in terms of the non-equilibrium statistical mechanics. The hitherto accepted theories such as free volume theory and entropy theory are criticized from the standpoint of molecular dynamics of shear viscosity. The molecular expression of the shear viscosity is given in terms of time-correlation function of the pair configuration distribution function. It is clarified that the elementary molecular process of viscosity is fluctuation of the pair distribution function whose dynamical aspects are discussed by using Langevin-type stochastic equation. It is suggested that the generalized random force of the stochastic equation is related to mobility of a molecule which was discussed in the free volume theory, and that the concept of the cooperative molecular rearrangement used in the entropy theory is substantially equivalent to the one of the fluctuation of the pair configuration distribution function of the present theory.
It is the object of the present study to obtain clear knowledge of the relations in the polypropylene melt between its linear viscoelasticity and its non-linear steady capillary flow, paying particular attention to the elastic properties in its capillary flow By representing the linear viscoelasticity numerically with zero-shear viscosity, ηo, and steady-state compliance, Je0, evaluation has been made of the properties concerning the elasticity of polymer in the capillary flow, such as non-Newtonianity, the entrance pressure loss, the end correction, the Barus effect and the melt fracture. The steady flow viscosity, η, the entrance pressure loss, Po, the critical shear stress, τc, and the critical shear rate, γc, at which melt fracture begins to occur, are subject to ηo as follows: logη∝logηo, logPo∝logηo, τc∝-logηo, logγc∝-logηo. From the well-known relationship between ηo and the average molecular weight, Mw, these quantities are governed by Mw. Meanwhile, for such quantities as structural viscosity index, N, end correction coefficient, ν, and elastic pressure loss ratio, Po/P, following correlations hold: N∝log(ηo, Je0), logν∝log (ηo2·Je0), Po/P∝log(ηo2·Je0). As ηo and Je0 are respectively determined mainly by Mw and the molecular weight distribution, M.W.D., these quantities are governed by both Mw and M.W.D. Physical meanings of ηo·Je0 and ηo2·Je0 are respectively mean relaxation time and a measure of stored energy in steady flow. Barus effect has positive correlation to Je0, ν and Po/P The symbol employed here, ∝, means the positive correlation.
Hereunder is presented a report of studies made of the melt viscosities of the fractions of linear and ethyl-branched polyethylenes having very narrow molecular weight distributions. The Newtonian melt viscosities, ηo, of the linear fractions were related to their weight-average molecular weight Mw, by the following equation, at 190°C. ηo=2.39×10-15Mw3.42 It was possible to reduce the relation between the viscosities and the shear rates of these fractions to a single master curve irrespective of their molecular weights, using the reduced variables, η/ηo and ηoMwγ for the ordinate and the abscissa, respectively. Meanwhile, this master curve was different from that of the polyethylenes having broader molecular weight distributions: the reduced viscosity of the former fell more sharply than the latter as the reduced hear rate increased. It was found that in the branched fractions having 10-25 ethyl-branches per 1000 chain atoms, the Newtonian viscosities were about half a decade greater than that of the linear fractions at the same molecular weight. It was found also that the critical shear rates, the shear rates of the onset of the non-Newtonian flow, were lower than the linear homologues.
It seems that plasticity and viscosity in the flow caused by deformation will change from one to the other continuously according to the temperature or concentration of diluent added to the material. A preliminary investigation was reported to verify the possibility of continuous change from plasticity to viscosity. The samples used were Poly (vinyl chloride) plasticized with diethyl phthalate (DEP). The difference between plasticity and viscosity was decided by creep and its recovery data, and the flow curves obtained from the stress-strain data. The measurements of the creep and its recovery were made at temperatures ranging from 40°C to 70°C. The stressstrain data were obtained at the same temperatures and at strain rates between 0.014min-1. and 2.23min-1. The temperature and concentration region where the flow behavior would change from viscosity to plasticity were decisively located, but it was necessary to examine closely whether the usual definition of plasticity and viscosity was properly applicable or not to the amorphous polymers.
Observations have been performed of liquid drops suspended in several chain polymer solutions with respect to their small deformation individually. Two flow systems have been set up in which the drops or bubbles are subject to streaming with pure shear and to uniform streaming at low constant speed. In the first case, the theoretical formula derived by Taylor is compared with the experimental results, which show that though the deformation is always linear with the applied shear, both larger and smaller deformations than those expected theoretically take place. It has become evident that there are other important factors than the shear rate, interfacial tension and viscosity, i. e. concentration of the solution, C, and the degree of polymerization of the solution, n for the same kind of liquid system. The diagrams of the dimensionless deformation vs. shear rate indicate that they contribute to the drop deformation in the form Cn. For the second case of uniform streaming, gas bubbles deform into various shapes. Qualitative interpretation of the shape is also possible in terms of Cn.
In connection with the studies that have been made on the viscoelastic behaviour of polymeric liquids under high pressure, the concentration dependence of the change of viscosity, ηp/ηo, and the critical concentration, Cc, under static pressures up to 900kg/cm2 for polydimethyl siloxane-toluene system are discussed in this paper. The measurement was made over a wide range of polymer concentration from 0.10 to 1.00 at 40 and 80°C. In this system, Cc is found at about 0.30 and seems to be independent of pressure and temperature within the experimental error. As observed in various polymeric liquids previously studied, log(ηp/ηo) is proportional to Pn at each concentration. Below Cc, log(ηp/ηo) takes an approximately constant value independent of concentration, and in the range of concentration from Cc to about 0.60, double logarithmic plot ηp/ηo against C is represented by a straight line having a positive slope under each pressure. This result is consistent with that obtained of polystyrene solution in toluene where the concentration was lower than 0.50. Over the range of higher concentration near 1.00, this plot is concave, that is, the effect of pressure on viscosity becomes larger with the increase of concentration.
The dynamic properties of concentrated polymer solutions have been investigated by taking into consideration the effect of deforming entanglement networks. According to the Hayashi theory, there exists the box type region of relaxation spectrum due to the entanglement of polymer chains in the relaxation time region which is longer than the wedge type region corresponding to the mechanism described in the Rouse-Zimm theory. The viscoelastic properties of polymer solutions are affected by the molecular weight distributions of solute (polymer), so that the molecular weight distribution should be taken into consideration in calculation of the relaxation spectrum. (In this paper), it is assumed that the relaxation spectrum is based on the Hayashi theory that the effects of the molecular weight distribution are to be calculated by the Ninomiya-Fujita method, and so we have computed the complex modulus G' and G" for the values of entanglement density as function of angular frequency. Comparing these curves, theoretical and experimental, the following conclusions have been obtained. (1) As for the polyisoprene-toluene system which follows the M3.4 law of viscosity, the experimental curves of its complex modulus approximately agree with the theoretical curves. (2) As for the Alg-aqueous solution, in spite of the fact that its viscosity is far more dependent on its molecular weight than it is the case with the polyisoprene-toluene system, and so the agreement of curves in its case between the experimental and the theoretical is short of being complete, the theoretical curves have been corrected to such extent, by taking into consideration the extreme degree of the dependence of its viscosity on its molecular weight, as to make the theoretical curves practically agree with the experimental curves.
Taking into account the effects of the flow history on the rate of formation and breaking of junction points, the memory function in Lodge's equation is modified as follows: μ(t;t')=∫∞-∞H(λ)/λR(t';λ)/R0(λ)e-1/λ[(t-t')+β∫tt'h(t", t')dt"]d ln λ h(t;t')=√Tr(C-1(t')-I) here H(λ), λ and β are relaxation spectrum in the linear region, relaxation time and a dimension less parameter respectively. R(t';λ) and R0(λ) are the rate of formation of junction points at time t' in the nonlinear and linear region, respectively. C-1(t') is the Finger strain tensor and I is the unit tensor. Using this memory function, the integral constitutive equation is represented by τ=pI+∫t-∞μ(t;t')[(l+ε)(C-1(t')-I)+ε(C(t')-I)]dt' where τ, p and ε are stress tensor, hydrostatic pressure and dimensionless parameter respectively. C(t') is the Cauchy-Green strain tensor. In the case of R(t';λ)=R0(λ), this constitutive equation makes the following predictions. (1) The non-Newtonian viscosity curve for the broad box type relaxation spectrum is nearly the same as the Bueche-Harding stand curve7). (2) The non-Newtonian viscosity curve for the narrow box type relaxation spectrum (including the case of single relaxation time) properly represents non-Newtonian viscosities of monodisperse polystyrene melts8)9). (3) The first normal stress coefficient is proportional to γ-1.5 for large shear rate γ.
A theoretical investigation has been made of the deformation of an elastic prolate spheroidal particle in Poiseuille flow by taking the deformation velocity into account. Although the elongational deformation is also considered, the leading deformation is the bending due to the buckling under the present condition in which the short-/long-axis ratio of the particle is considerablly small. It is well known that deformation due to the buckling is unstable. However, as long as the particle exists in the viscous fluid, the deflection can be determined as a function of time or of Eulerian angles. This is because resistant force is produced against buckling by the deformation velocity of the particle. In this treatment, the deflection, which is here represented in terms of curvature, is given as a special case in which the particle is much shorter than the tube radius, since otherwise the numerical method of integration must be used to solve the differential equations. The results obtained, however, are not capable of theoretcal interpretation of the radial migration, which occurrs from the wall toward the tube axis.
Transfer of heat in a steady state in non-Newtonian fluids flowing in laminar motion through circular tube with constant wall temperature has been treated theoretically and experimentally by a number of workers. However, on the distribution of temperature within the fluid, few experimental data have hitherto been published. The theoretical interpretations of the experimental data have so far been made under the assumptions that the physical properties, such as heat capacity, density, consistency index and thermal conductivity of the fluids are independent of temperature, and that the heat produced by viscous friction and the heat conduction in the axial direction are negligibly small. The temperature distribution in the radial directions were directly measured by thermistor probe. As samples of Newtonian and non-Newtonian fluids, water and aqueous solutions of sodium carboxymethylcellulose of various concentrations were respectively used. In the analysis of the latter fluid, the power law is assumed. The observed temperature distributions agree fairly well with the theoretical predictions except at the vicinity of the tube wall.
Rayleigh's problem of an idealized non-Newtonian fluid past a flat plate has been considered by using the constitutive equation of the power-law model. The velocity profiles can be reduced to a single curve by changing the scale of the axis perpendicular to a flat plate, and the exact solution of the problem can also be generated. Of the pseudoplastic fluid the velocity decreases more slowly than that of the Newtonian fluid, as its distance from the plate increasses. Of the dilatant fluid the velocity decreases faster, and becomes zero at a definite distance from the plate.
It is the object of the present study to inquire experimentally into the irregular flow behaviors of polymer melts in short capillary passages. The polymers tested are polyethylenes of high density, and three kinds of experimental methods, such as the flow property test by using KÔKA-type flow tester, the observation on flow patterns in a reservoir or on extrudates by using a test piece with flow marker of carbon black, and observation on the fringe patterns at the capillary inlet or in the capillary passage by using polarized light were used. The results obtained are as follows; (1) Four kinds of the types of extrudates, such as normal flow, spiraling flow, melt fracture flow and solid state fracture flow induced by the pressure were observed. (2) The polarizing fringe at the capillary inlet is similar to a horse's hoof in shape in the normal flow region, and the number of fringes is directly proportional to the amount of apparent shear rate at the capillary wall. (3) The polarizing fringe under spiraling or in the melt fracture flow regions has an unsteadly special shape, such as twisting or pulsation, and it seems reasonable that twisting of the polarizing fringe is induced by a small slippage or lecal high shear rate at or near the capillary inlet and causes the extrudate to flow spiraling. The pulsation of polarizing fringe is induced by the large slip or high shear rate which is generated on an inversed conical surface with a vertex angle about 30 degree in a reservoir and causes the extrudate to augment the melt fracture.
Infrared dichroism of PVC films containing less than 5wt.% of plasticizers such as DMP, DOP, DOA and TCP were measured for the absorption bands at 603, 635, 690, 1426 and 1434cm-1 for PVC and for C=O stretching vibration for plasticizers. In the plot of plasticizer level vs. orientation function while the minimum value of the crystalline absorption bands was clearly observed in the orientation function at around 2wt.% of plasticizer level, no clear presence of minimum value was assured of the amorphous absorption bands. This was proposed as interpretable in terms of antiplasticization mechanism caused by small amount of plasticizers acting as pseudo-crosslinking agents in PVC. The dichroic data were utilized for the estimation of possible effects of hydrogen bonding and charge transfer force on the intermolecular force between plasticizers and PVC. It is concluded that the majority of contribution to this pseudo-crosslink is from the hydrogen bonding between PVC and carbonyl group in the plasticizer molecule, while the charge transfer force is not large enough to affect the orientation of the polymer chain.
The ultrasonic absorption in solutions of polystyrene in dibutylphtalate was measured in the frequency range from 1 to 52MHz. The measuring temperature range was from 5 to 65°C, and the concentration of the solution was up to about 8g/100ml. The molecular weight of the sample of polystyrene was 3.0×105. The ultrasonic velocity, density and absorption coefficient of the solution increased linearly with the increase of the concentration of polymer over the whole temperarture range. The plot Δα/f2 against the frequency showed a relaxation curve of single relaxation. With the temperature rise, the relaxation frequency fr increased, and the absorption maximum per wavelength μmax showed the maximum. The imaginary part of the bulk modulus K" and shear modulus, G" were calculated from the deta of ultrasonic absorption and those of Moore et al. ΔK" and ΔG" were increased were with the increase of the concentration and frequency. The ratio, ΔK"/ΔG" is far greater than 1, and the ratio of Δκ' (Bulk viscosity) to Δη' (Shear viscosity) was 2.2 at 25°C. From the temperature dependence of the partial specific compressibility and the partial molar volume, it is suggested that the polymer chains in the solution began to rotate freely above 35°C. The energy difference and the activation energy were estimated from the temperature change of the absorption on the bases of two-state model. The following values were obtained; the energy difference ΔH0=1.66kcal/mol, and the activation energy ΔH21=6.6Kcal/mol.
The Ultrasonic absorption of Polyvinylacetate in acetone solution was measured by the pulse method over 5MHz-75MHz within the temperature range from 0°C to 20°. A thermal relaxation phenomenon was observed in this frequency range. A rotationary isomer model has been presented based on the observed molecular weight dependence of compressibility at 1MHz. This model is composed of head-to-tail sequences of polymer chain bonds and involves a doubly degenerated ground state. The energy state in this model is found to be ΔH0=1.5±0.5Kcal/mol ΔH≠=4.8Kcal/mol.
The mechanical strength of unsaturated soil in the water system depends on its capillary attraction force, while its adhesive force is said to be almost equal to the force of chemical potential in its capillary attraction. According to the centrifugal method, the chemical potential of water in soil is calculated by setting the sample subject to centrifugal force. This paper concerns the relation between the strength of cohesive clay and its interstitial force. The change in volume of a sample is found to be equal to that of water drainage by centrifuging. Therefore almost no capillary attraction is observable in centrifuged cohesive sample. The yield strength of the sample qθ, which is drained by the centrifugal force, is recognized to relate to the chemical potential of water Δμ in the sample which is given in the following equation. Δμ=K·qθ where K is a fluctuation coefficient, ranging from 1 to 8. Consequently, it is suggested that the cohesive strength of clay does not depend exclusively on capillary potential, but on various other potentials including electric potential and van der Waals potential. The total potential of water in soil may be balanced with centrifugal force in process of irreversible drainage of water, and the yield strength of clay would be built by the total potential.