Search

Measuring system of the advancing crack length at the fatigue test has been developed using CCD sensor without stop the machine work. Two personal computers operate in its system to control machine work. First computer (PC1) controls the cyclic loading according to the setting maximum or minimum load, and interrupts to second computer (PC2) by using RS232C line at each setting number of cycles. Receive from PCl's interrupt signal, using GP-IB line, PC2 commands to read and write the video signal about crack image to digital memory and then sends the signal from it to memory of PC2. From data of memory of PC2, a graph of advanced crack length versus number of loading cycles and a log-log plots of da/dN versus delta K can be drawn.Under constant load, a-N plots become stepwise with number of loading cycles. During early crack development, one step corresponds to several cycles and finally one step corresponds to one cycle. On each stage, one step creates one striation line on the fracture surface. Also, the linear relationship between log-log plots of da/dN and delta K has verified.Under constant delta K, a-N plots become stepwise straight line and a log-log plots of da/dN versus delta K become one point for each experiment because da/dN is constant at each cycle.

The nip width and the driving torque were measured during rotation of metal and rubber-covered rollers under dead weight and with constant peripheral speed. The characteristic values of mechanical model of the cover were also decided from the measurement of the dynamic modulus and tan δ of the rubbers, and a comparison of the dynamic mechanical characteristic of model with them was made.Comparing the experimental results with the calculated ones using the method explained in the previous paper, the following conclusions were obtained:(1) In our test apparatus the center distance between the rollers changes with peripheral speed, particularly when soft rubber is used. Thus our theory is compared with the experiments of hard rubber-covered rollers.(2) The nip width during rotation is smaller than that of standing still, and it decreases with an increase in peripheral speed.(3) For soft rubber-covered rollers the rolling friction increases with an increase in peripheral speed, and for rollers covered by hard rubber it is constant or decreases in the high range of speed.(4) The experimental results do not coincide with the calculated ones. The disagreement is considered to arise from the fact that the characteristic values of mechanical model are decided on the basis of the measurement of the dynamic modulus of the cylindrical test pieces, where the free deformation of the test piece is allowed.

Our previous papers on this series described the manufacture of the compound filament which consists of cocoon fiber and synthetic one as core. The observation of the cross-section shows that the cocoon fibers are located around the core synthetic filament, being adhered by silk-gum or sericin.The compound filament must be twisted to cover the weak adhesion of the cocoon and to avoid separation of silk filaments from the core filament during de-gumming.The aspect of the compound filament resembles a multifilament yarn after twisting and degumming.In this paper, effects of twisting and de-gumming of the compound filament on the geometrical structure and the cross-sectional shape of the yarn are studied. TETORON (PET filament) is used as the core filament. The results are as follows:1) The core filament is reduced in length by twisting and de-gumming. When relatively large twist was given to the compound filament, most of the silk filaments are near the core in helical shape.2) From the change of the length of the core filament, it is possible to know how many twists should previously be given to a compound filament to keep a good contact of the silk filaments to the core filament after de-gumming.3) The cover-ratio is defined by the ratio of the total length of the parts covered by the silk filaments to the circumference of the core's cross-section. The equation of the cover-ratio is derived under the conditions that all silk filaments are located at the same distance from the center of the core filament and their directions from it are at random.where Fc′: cover-ratio S: circumferece of the cross section of the core filament s: arc length that one silk filament covers the circumference n: number of the silk filaments4) The values of cover-ratio measured are generally larger than those calculated from the above equation; especially when the compound filaments are twisted, the cover-ratio become larger, caused by the outer silk filaments thrusting themselves into the inner ones.

The energy method, used for the derivation of the tensile properties of continuous filament yarns by L. R. G. Treloar and G. Riding and by J. W. S. Hearle, is applied to the compound filament consisted of cocoon fiber and TETORON (PET filament).The tensile stress as a function of strain is expressed by the following equation:Where fy (ey): stress of the twisted compound filament at a strain of eyms: total denier of cocoonsmc: denier of core filamentM=ms+mcFs: stress of cocoon fiber that is a function of the strain es and determind experimentally from its s-s curve.Fc: stress of core filament which is a function of the strain ec and determined experimentally from the s-s curve.K=∂es/∂eye_??_=(β-γ)/(1-β)β: the twist-shrinkage of the compound filamentγ: the twist-shrinkage of TETORONThe s-s curves of cocoon fiber and TETORON and the twist-shrinkages of TETORON and the compound filament are introduced into this equation. The satisfactory agreement is found between the results calculated and observed. From this result, it may be said that the tensile behavior of the twisted compound filament is accounted satisfactorily by the above equation.

The feed tension of the core filament and the drawing tension at the reel are the most important factors on the manufacture of the compound filament. In the first place, the compound filaments were manufactured under the different feed tensions and with the different sizes. The core lengths of them, then, were measured for discussion of the feed tension.In the next place, the compound filaments and the raw silk were elongated in the drenched state and after drying the strain recovery was measured for discussion of the drawing tension. The following results were obtained:1) It was made sure that the feed tension of the core filament extends the core of the compound filament in proportion to the feed tension.2) The upper limit of the drawing tension at the reel was experimentally obtained from the relation between the drawing tension and the recovery of the elongation. In the actual manufacturing, the drawing tension was sufficiently smaller than that.3) It sometimes happens that the cohesion between the core filament and the cocoon fibers is broken by itself. This can be classified into two cases: the first is the case that the feed tension of the core filament is not suitable, the second is the case that the drawing tension at the reel is too large. Though the external views of the compound filament's in which break down of the cohesion causes are similar, they are not the same.

In this paper the mechanical behavior of a rubber-covered roller which is pressed against a metal roller under constant load and rotates with constant peripheral speed is analyzed on the basis of a viscoelastic model.The results are as follows:(1) Assuming that the behavior of rubber is represented by a model constructed of a Voigt model combined in series with a spring of modulus G0, the nip width is given by the following equation when rollers stand still.where W: load 1: rubber thickness L: roller width R: 1/R1+1/R2=1/R R1: radius of a rubber-covered roller R2: radius of a metal roller D: (G0+G)/G G: modulus of spring in Voigt model(2) When rollers rotate, the nip width decreases with its peripheral speed.(3) When rollers rotate, the pressure distribution becomes asymmetric to incline to the opposite direction of rolling.(4) The rolling friction increases with increasing peripheral speed, and the larger the instantaneous modulus, the value on the basis of the nip width at rest becomes larger.