A scheduling problem with job values given as a power function of their completion times

This paper deals with a problem of scheduling jobs on the identical parallel machines, where job values are given as a power function of the job completion times. Minimization of the total loss of job values is considered as a criterion. We establish the computational complexity of the problem – strong NP-hardness of its general version and NP-hardness of its single machine case. Moreover, we solve some special cases of the problem in polynomial time. Finally, we construct and experimentally test branch and bound algorithm (along with some elimination properties improving its efficiency) and several heuristic algorithms for the general case of the problem.     

Femtosecond laser drilling of crystalline and multicrystalline silicon for advanced solar cell fabrication

The rear contact solar cell concept has been implemented to increase the solar cell efficiency. Practically, it necessitates rapid fabrication of a large number of via holes to form low-loss current paths. It is not a trivial task to drill a number of microscopic holes through a typical Si wafer of ??200???m thickness at reasonable processing throughput and yield. In this research, a femtosecond laser is employed to drill via holes in both crystalline silicon (c-Si) and multicrystalline silicon (mc-Si) thin wafers of ??170???m thickness with various laser parameters such as number of laser shots and pulse energy. Since a significantly high pulse energy compared to ablation threshold is mainly applied, aiming to achieve a rapid drilling process, the femtosecond laser beam is subjected to complex non-linear characteristics. Therefore, the relative placement of the sample with respect to the laser focal position is also rigorously examined. While the non-linear effect at high pulse energy regime is complex, it also facilitates the drilling process in terms of achieving high-aspect ratio, for example, by extending the effective depth of focus by non-linear effect. Cross-sectional morphological analysis in conjunction with on-line emission and shadowgraph imaging are carried out in order to elucidate the drilling mechanism.     

Characterization of hydrophilic networks synthesized by group transfer polymerization

Group transfer polymerization was used to synthesize several series of hydrophilic random and model networks. Cationic random networks were prepared both in bulk and in tetrahydrofuran (THF) using a monofunctional initiator and simultaneous polymerization of monomer and branch units, while a bifanctional initiator was employed in THF for the synthesis of model networks comprising basic or acidic chains. Upon polymerization of the monomer, the latter initiator gives linear polymer chains with two “living” ends, which are subsequently interconnected to a polymer network by the addition of a branch unit. Homopolymer network star polymers were also synthesized in THF by a one‐pot procedure. The synthesis involved the use of a monofunctional initiator and the four‐step addition of the following reagents: (i) monomer, to give linear homopolymers; (ii) branch unit, to form “arm‐first” star polymers; (iii) monomer, to form secondary arms and give “in‐out” star polymers; and, finally (iv) branch unit again, to interconnect the “in‐out” stars to networks. Different networks were prepared for which the degree of polymerization (DP) of the linear chains between junction points was varied systematically. For all networks synthesized, the linear segments, the “arm‐first” and the “in‐out” stars were characterized in terms of their molecular weight (MW) and molecular weight distribution (MWD) using gel permeation chromatography (GPC). The degrees of swelling of both the random and model networks in water were measured and the effects of DP, pH, and monomer type were investigated.