Fatigue tests on the three strand roller chains having various fitting conditions between middle link plates and pins were made. Results are as follows:
1. The fatigue strength of treble strand M-type roller chains (standard) having clearance between middle link plates and pins is twice as that of single strand chains.
2. The fatigue strength of treble strand P-type roller chains having interference between middle link plates and pins is 2.4 times as large as that of single strand chains.
3. The endurance ratio of multiple strand roller chains is smaller than that of single strand roller chains.
4. The fatigue strength reduction factor due to increase of strands is nearly independent to the number of cycles to estimate the fatigue strength.
5. The fatigue strength reduction factor takes the smallest value at P-type. M-type is the next. The value is the largest at L-type.
6. The load distribution for the each strand seems to be nearly uniform at P-type, considering from the form of destruction. In the other types, it seems to be not uniform.

View full abstract

As a part of the research program probing into the cause of failures in the nose suspension type traction gears, measurements of fillet stresses occurring in both sides of gear tooth surfaces of a running electric locomotive were carried out using wire resistance strain gauges with the following results;
1. Nose suspension type traction gears are not free from occurrence of uneven contact at the sides; actually the whole width of tooth surface does not share the load uniformly.
2. Maximum fillet stress of gear develops at the side nearer to the motor when the motor is preceding the wheel in the running direction.
3. The proportion of the fillet stress at the side nearer to the motor to the calculated fillet stress on the assumption of uniform tooth contact is 3.9 maximum for EF 58 locomotive and 11 maximum for DF 50. Maximum measured stress caused by uneven contact, even when the actual output torque was small, turned out often larger than the calculated value assuming uniform tooth contact under maximum load.
4. When the wheel precedes the motor in the running direction, it is clearly observed in the DF 50 locomotive having less weight of motor that the movement of tooth contact from the wheel side to the motor side takes place.
5. In the quill drive transmission mechanism, uneven contact was observed by measuring the fillet stresses of gear at both sides of the tooth surface when the shafts of pinion and gear were misaligned.

View full abstract

Creep characteristics of 18-8 stainless steel at room temperature were studied. The specimens were fabricated from 20mm diameter bars, which were cold drawn and solution treated at 1050°C. Constant load creep tests were made under the stresses from 12.5kg/mm² to 50.0kg/mm². Durations of tests were about 2000 hours and the test under 20kg/mm², which is still continued, were conducted over 17000 hours.
Under the stresses even lower than the yield strength, relatively large creep was observed and this is advancing even over 10000 hours. In the case of 20kg/mm², instantial elongation was 0.24 per cent, and total elongation reached to 1.3 per cent at 15500th hour. The creep rate which showed 0.091 per cent per 1000 hours in the interval between 1000th and 2000th hour, was reduced to 0.098 per cent per 1000 hours between 10000th and 17000th hour.
For the test which were conducted under the higher stresses than 20kg/mm², instantial elongation and initial creep rate was greatly increased with stress, but the creep rate at 1000th hour or after decreased to the similar order of the creep under 20kg/mm².
Plottings of logarithm of permanent elongation vs. logarithm of time showed good liner relationship for all these tests and any tendency to deviate from the line to upper side at longer time range, was not detected. It may be concluded that the room temperature creep of this material consist mainly with transient creep, expressed by the formula of dε/dt=αt^-m, and no steady creep is contained.
From practical point of view, these creep should be taken into account for the design of parts of installations, which should be confined to the small strain. For this purpose, the stresses which produce a definite strain at a certain time were determined from this results. The stress for 0.2 per cent at 1000th hour is about 75 per cent of the yield strength determined by routine test.

View full abstract

For the several years, the theory of photoplasticity has been studied for a method by which the elastic-plastic problem of metal can be solved experimentally.
It is the important task in this research to select the applied test piece material. We make the experiments of simple tension, simple compression and uniform bending about the four kinds of celluloid which differ reciprocally the camphor content in regard to the dry base of nitrocellulose and about the two directions of the four celluloid materials which are rolling direction and perpendicular to it, then obtain the stress-strain diagrams, fringe order-strain diagrams and so the properties of the celluloid.
Then the celluloid is found a suitable material that is applicable in the solution of the elastic-plastic problem of some metallic materials.

View full abstract

Oil-quenched and tempered 13-Cr steel SUS-2 was tested at a temperature of 500°C under constant stresses, stepwise changing stresses and periodic stresses of rectangular wave. Behaviours of the material under changing stresses were compared with those under constant stresses.
It was found that, strictly speaking, the law of“the mechanical equation of state in solid” was not valid for the material tested. The strain rate after the abrupt increase in stress was higher than that of the reference test under the constant stress equal to that after the stress change, and the strain rate was lower after the abrupt decrease in stress. This tendency was accentuated when the first stress change was delayed from primary creep to secondary creep, and further to tertiary creep. The behaviour of this material may be estimated with a small error by the use of the law of“the mechanical equation of state in solid, ”when the stress change is given during earlier stage of creep. The earlier the time of stress change, the smaller the error induced in the estimation.
This is quite the same in the tests under periodic stress. The deformation in primary and secondary creep under periodic stress could be estimated from the data of constant stress creep tests with sufficient accuracy. This is applicable for the prediction of creep limit under varying stresses from the data of constant stress creep tests.

View full abstract

In this experimental research, 0.097-0.82% carbon steels were studied under the repeated tension impact test. We made the experiment using the universal repeated impact testing machine, and further with the charpy impact testing machine beyond the limits of the capacity of the former.
From the result of the tests, it may be said that the repeated tension impact test discontinuously changes from impact test to fatigue test. In the diagram of the relation between repeated tension impact energy and number of impacts to failure, two curves came out as shown in Fig. 4 and Fig. 5, and they were discontinuous. The fracture of specimens was cap and cone on one curve, and on the other, fatigue fracture. By measuring the deformation of specimens, we knew the change of α, β and ψ with the impact energy as shown Figs. 7. 8 and. 9.

View full abstract