No abstract

Photoelastic evaluation of the stress intensity factor (SIF) for a crack in a bimaterial tangential to the interface is hitherto confined to the use of only a singular stress field equation. In this paper, the multi-parameter stress field equations of Deng are simplified for use by experimentalists to evaluate the stress field parameters by digital photoelasticity using an overdeterministic least-squares technique. A bimaterial Brazillian disc with a central interface crack is selected as the model for study as different mode mixities could be easily simulated by changing the crack orientation angle. The use of SIF evaluation based on a singular stress field equation is found to be inadequate for the problems considered. On the other hand, the use of a multi-parameter stress field equation is quite successful in evaluating the stress field for various mode mixities and for two values of bimaterial constant. It is shown that the new procedure allows data collection from a larger zone, which helps to simplify data collection from experiments.

No abstract

The main sources of error in the determination of stress intensity factors (SIFs) for an interface crack in a bi-material by conventional photoelasticity are the measurement of the positional co-ordinates of the data point and the fringe order. In the present work, use of two digital photoelasticity methods for collecting these data is discussed. SIFs are evaluated using constant radius method and a least squares approach based on the singular stress field equation. The need for developing a multi-parameter stress field solution for evaluating SIF is highlighted.

Extracting performance from modern parallel architectures requires that applications be divided into many different threads of execution. Unfortunately selecting the appropriate number of threads for an application is a daunting task. Having too many threads can quickly saturate shared resources, such as cache capacity or memory bandwidth, thus degrading performance. On the other hand, having too few threads makes inefficient use of the resources available. Beyond static resource assignment, the program inputs and dynamic system state (e.g., what other applications are executing in the system) can have a significant impact on the right number of threads to use for a particular application. To address this problem we present the Thread Tailor, a dynamic system that automatically adjusts the number of threads in an application to optimize system efficiency. The Thread Tailor leverages offline analysis to estimate what type of threads will exist at runtime and the communication patterns between them. Using this information Thread Tailor dynamically combines threads to better suit the needs of the target system. Thread Tailor adjusts not only to the architecture, but also other applications in the system, and this paper demonstrates that this type of adjustment can lead to significantly better use of thread-level parallelism in real-world architectures.