Scalability analysis in gracefully-degradable large systems

The scalability of large degradable homogeneous multiprocessors is analyzed. The objective is to assess the limitations, imposed by reliability considerations, on the number of processors. The analysis of the mean-time-to-failure and the mission-time shows that, for a given value of the coverage factor, there exists a value of the number of processors at which these measures are maximal. As the system size is increased beyond this value, the reliability of the system becomes a rapidly decreasing function of the number of processors. For computations with linear speed-up, the amount of reliable computational work is constant for large system-sizes. When the speed-up is not linear, this amount is a decreasing function of the number of processors. Therefore, for large system-sizes and same technology, increasing the number of processors results in a decrease of the average amount of reliable computational work the system can deliver. Graceful degradation in large fault-tolerant systems is not scalable     

An evaluation of medium-grain dataflow code

In this paper, we study several issues related to the medium grain dataflow model of execution. We present bottom-up compilation of medium grainclusters from a fine grain dataflow graph. We compare thebasic block and thedependence sets algorithms that partition dataflow graphs into clusters. For an extensive set of benchmarks we assess the average number of instructions in a cluster and the reduction in matching operations compared with fine grain dataflow execution. We study the performance of medium grain dataflow when several architectural parameters, such as the number of processors, matching cost, and network latency, are varied. The results indicate that medium grain execution offers a good speedup over the fine grain model, that it is scalable, and tolerates network latency and high matching costs well. Medium grain execution can benefit from a higher output bandwidth of a processor and fainally, a simple superscalar processor with an issue rate of two is sufficient to exploit the internal parallelism of a cluster. This work is supported in part by NSF Grants CCR-9010240 and MIP-9113268.     

Comparison of total energy systems using gas turbines and diesel engines for combined cooling

A gas turbine engine was used to drive the compressor of a vapour compression cycle so that the usually wasted energy in the exhaust gases was partially recovered and used in the generator of an absorption cycle. The cooling effect was therefore boosted. The degree of energy utilization was further enhanced when the energy released from the absorber and condenser of both cycles was recovered in the form of hot water, which could be used for different applications. The performance parameters for this combined system, such as the cooling effect, total heat recovered and performance effectiveness ratio, were calculated for various evaporator and condenser temperatures. It was found that a system driven by a gas turbine gives a better performance than a diesel engine system under similar operating conditions.     

Nonlinear water diffusion in nonsaturated concrete

The equations governing drying and wetting of concrete are formulated, assuming the diffusivity and other material parameters to be dependent on pore humidity, temperature and degree of hydration. By fitting of computer solutions for slabs, cylinders and spheres against numerous test data available in the literature it is found that the diffusion coefficient decreases about 10 to 20 times when passing from 0.9 to 0.6 pore humidity. The problem is thus strongly nonlinear. Dependence on temperature is found to agree satisfactorily with the rate process theory. Effect of temperature on equilibrium pore humidity is also studied. The aging effect is defined by means of an equivalent hydration period. To enable easy prediction of drying of simple bodies, charts for the solution in terms of nondimensional variables are presented. Finally, correlation to the diffusion in a saturated concrete is discussed.     

The generalized finite point method

In this paper we propose a new mesh-less method based on a sub-domain collocation approach. By reducing the size of the sub-domains the method becomes similar to the well-known finite point method (FPM) and thus it can be regarded as the generalized form of finite point method (GFPM). However, unlike the FPM, the equilibrium equations are weakly satisfied on the sub-domains. It is shown that the accuracy of the results is dependent on the sizes of the sub-domains. To find an optimal size for a sub-domain we propose a patch test procedure in which a set of polynomials of higher order than those chosen for the approximations/interpolations are used as the exact solution and a suitable error norm is minimized through a size tuning procedure. In this paper we have employed the GFPM in elasto-static problems. We give the results of the size optimization in a series of tables for further use. Also the results of the integrations on a generic sub-domain are given as a series of library functions for those who want to use GFPM as a cheap and fast integral-based mesh-less method. The performance of GFPM has been demonstrated by solving several sample problems.     

Explicit analysis technique for nonlinear inverse mechanics problems

Nonlinear material models are conventionally used in forward analysis to predict the global mechanical response of boundary value problems. Such models are not expected to exactly reproduce global experimental response in all cases. Accordingly, the measured global response at specific domain or surface points can guide an inverse nonlinear structural analysis to successively recover a representative material model. By assuming an initial set of stress–strain data points, the load–displacement response at the control points is computed in a forward incremental analysis without iterations. This analysis retains the stresses at the integration points. The corresponding strains are not expected to be accurate since the computed displacements are not anticipated to match the measured displacements at the control points. Therefore, a conjugate incremental displacement analysis is performed at the same load steps to correct for displacements and strains everywhere by matching the measured displacements at the control points. It is found that the predicted stress–strain data set at the most highly stressed integration point provides the most accurate representation of an improved material model. This data set is used in the next two-phase incremental analysis pass as the material model. The process is repeated until the forward analysis phase reproduces the measured displacements at the control points requiring no corrections. Accordingly, the selection of a single stress–strain data set yields an explicit recovery of the nonlinear elasticity material model without any interpolation or averaging schemes, which significantly reduces data storage and computational effort. The applicability of the present explicit approach is demonstrated on simple mechanical models, a skeletal structural system and a 2D finite element mesh.     

Evaluation of the role of transverse reinforcement in confining tension lap splices in high strength concrete

Reports on the very brittle and splitting mode, of failure of tension lap splices anchored in high strength concrete (HSC) specimens and the lower normalized bond stress\(\left( {{u \mathord{\left/ {\vphantom {u {\sqrt {f'_c } }}} \right. \kern-\nulldelimiterspace} {\sqrt {f'_c } }}} \right)\) for high strength than for normal strength concrete, instigated several research projects aiming at recommending mechanisms to provide confinement and ductility for bars or splices anchored in HSC. The state-of-the-art report of ACI Committee 363, ACI 363R-84, defines high strength concrete as concrete with compressive strength above 6,000 psi (41.4 MPa).The main objective of the research reported in this paper was to assess the effect of transverse reinforcement on bond strength and ductility of the mode of failure of tension lap splices anchored in HSC beam specimens with nominal concrete compressive strength 55 MPa (8,000 psi), and to recommend a transverse reinforcement design requirement related to the ACI transverse reinforcement index K_tr.To meet the objective fifteen full-scale HSC beam specimens were tested. No pozzolanic material such as silica fume or fly ash was added to achieve high strength to avoid any effect such material might impose, on structural behavior. Each beam was designed with bars spliced in a constant moment region at midspan. The variables used in the investigation were bar size, and the amount of transverse reinforcement placed in the splice region.     

Very high gain AlGaAs/GaAs pnp heterojunction bipolar transistor

The effect of the emitter Al mole fraction and emitter-base junction type on the performance of AlGaAs/GaAs pnp heterojunction transistors was investigated experimentally. A current gain as high as 1360 was obtained for a structure with an Al/sub 0.4/Ga/sub 0.6/As emitter and compositionally abrupt emitter-base junction; this was attributed to reduced surface recombination.<>     

Gibbs function as a correlating performance parameter for fuels used in gas turbines

Correlating performance parameters for fuels, such as ring carbon, aromatic percentage, carbon number, C/H mass ratio and hydrogen percentage, have proved to have relative correlating deficiencies. Although hydrogen percentage is the most successful, so far, it is postulated that the Gibbs function is more powerful owing to its direct relation with the chemical potential, which causes reactions to take place. This parameter gave favourable trends when checked against different previous works on gas turbine combustors especially with flame radiation, smoke and liner temperature. However, improvement over hydrogen content is marginal.     

Monolithic integration of complementary HBTs by selective MOVPE

The monolithic integration of n-p-n and p-n-p heterojunction bipolar transistors (HBTs) through the use of selective metal organic vapor phase epitaxial regrowth is discussed. This was accomplished by masking, patterning, and etching a p-n-p HBT wafer and then selectively regrowing an n-p-n structure in the etched areas. The selective epitaxial regrowth did not degrade the current gain of the p-n-p structure. Several complementary amplifier circuits were fabricated and tested successfully, demonstrating the feasibility of a monolithic complementary HBT technology