Towards a better reliability of risk assessment: Development of a qualitative & quantitative risk evaluation model (QREM) for different trades of construction works in Hong Kong

Since the safety professionals are the key decision makers dealing with project safety and risk assessment in the construction industry, their perceptions of safety risk would directly affect the reliability of risk assessment. The safety professionals generally tend to heavily rely on their own past experiences to make subjective decisions on risk assessment without systematic decision making. Indeed, understanding of the underlying principles of risk assessment is significant. In this study, the qualitative analysis on the safety professionals’ beliefs of risk assessment and their perceptions towards risk assessment, including their recognitions of possible accident causes, the degree of differentiations on their perceptions of risk levels of different trades of works, recognitions of the occurrence of different types of accidents, and their inter-relationships with safety performance in terms of accident rates will be explored in the Stage 1.     

Effective Gene Selection Method Using Bayesian Discriminant Based Criterion and Genetic Algorithms

Microarray gene expression data usually consist of a large amount of genes. Among these genes, only a small fraction is informative for performing cancer diagnostic tests. This paper focuses on effective identification of informative genes. A newly developed gene selection criterion using the concept of Bayesian discriminant is used. The criterion measures the classification ability of a feature set. Excellent gene selection results are then made possible. Apart from the cost function, this paper addresses the drawback of conventional sequential forward search (SFS) method. New genetic algorithms based Bayesian discriminant criterion is designed. The proposed strategies have been thoroughly evaluated on three kinds of cancer diagnoses based on the classification results of three typical classifiers which are a multilayer perception model (MLP), a support vector machine model (SVM), and a 3-nearest neighbor rule classifier (3-NN). The obtained results show that the proposed strategies can improve the performance of gene selection substantially. The experimental results also indicate that the proposed methods are very robust under all the investigated cases.

Novel catalyst-support interaction for direct formic acid fuel cell anodes: Pd electrodeposition on surface-modified graphite felt

The electrodeposition of Pd on graphite felt (GF, thickness ~3 mm in uncompressed state) was studied and the resulting catalyst was compared with Pt-Ru/GF for the electro-oxidation of formic acid. A micellar solution composed of the non-ionic surfactant Triton X-102 and an aqueous phase containing PdCl₂ were utilized for the galvanostatic electrodeposition of Pd nanoparticles. The presence of the surfactant during electrodeposition coupled with pretreatment of the GF surface by a Shipley-type method (PdCl₂ + SnCl₂ solution) creating nucleation sites had a major impact on the Pd catalyst morphology and penetration throughout the electrode thickness, affecting, therefore, the electrocatalytic activity toward formic acid oxidation. It was found that large (~1,000 nm) Pd particles with smooth surface favored the indirect CO_ad pathway, while Pd nanoparticles (diameter <40 nm) with rough surface, formed with surfactant and pretreatment, were much more active leading to the direct non-CO_ad pathway. Due to pretreatment the GF surface has been modified and the effective catalytic system could be described as Pd/SnO₂–Pd(PdO)/GF with possible electronic interaction between support and catalyst. In direct formic acid fuel cell (DFAFC) experiments at 333 K and 1 M HCOOH, the peak power density using the Pd/GF anode reached 852 W m^?2 (57 g m^?2 Pd) compared to 392 W m^?2 (40 g m^?2 Pd) with a commercial Pd catalyst-coated membrane (CCM). The long-term stability of Pd-based anodes was poor and inferior to Pt–Ru (4:1 at. ratio) prepared and tested under identical conditions.     

Selection of Highly Proliferative and Multipotent Meniscus Progenitors through Differential Adhesion to Fibronectin: A Novel Approach in Meniscus Tissue Engineering

Meniscus injuries can be highly debilitating and lead to knee osteoarthritis. Progenitor cells from the meniscus could be a superior cell type for meniscus repair and tissue-engineering. The purpose of this study is to characterize meniscus progenitor cells isolated by differential adhesion to fibronectin (FN-prog). Human osteoarthritic menisci were digested, and FN-prog were selected by differential adhesion to fibronectin. Multilineage differentiation, population doubling time, colony formation, and MSC surface markers were assessed in the FN-prog and the total meniscus population (Men). Colony formation was compared between outer and inner zone meniscus digest. Chondrogenic pellet cultures were performed for redifferentiation. FN-prog demonstrated multipotency. The outer zone FN-prog formed more colonies than the inner zone FN-prog. FN-prog displayed more colony formation and a higher proliferation rate than Men. FN-prog redifferentiated in pellet culture and mostly adhered to the MSC surface marker profile, except for HLA-DR receptor expression. This is the first study that demonstrates differential adhesion to fibronectin for the isolation of a progenitor-like population from the meniscus. The high proliferation rates and ability to form meniscus extracellular matrix upon redifferentiation, together with the broad availability of osteoarthritis meniscus tissue, make FN-prog a promising cell type for clinical translation in meniscus tissue-engineering.     

A non-solvent approach for high-stiffness all-cellulose biocomposites based on pure wood cellulose

All-cellulose composites are commonly prepared using cellulose solvents. In this study, moldable all-cellulose I wood fiber materials of high cellulose purity (97%) were successfully compression molded. Water is the only processing aid. The material is interesting as a “green” biocomposite for industrial applications. Dissolving wood fiber pulps (Eucalyptus hardwood and conifer softwood) are used and the influence of pulp origin, beating and pressing temperature (20–180 °C) on supramolecular cellulose nanostructure is studied by solid state CP/MAS ¹³C NMR. Average molar mass is determined by SEC to assess process degradation effects. Mechanical properties are determined in tensile tests. High-density composites were obtained with a Young’s modulus of up to 13 GPa. In addition, nanoscale cellulose fibril aggregation was confirmed due to processing, and resulted in a less moisture sensitive material.     

Allelopathic effects ofPolygonum aviculare L.

We previously reported thatPolygonum aviculare has a strong allelopathic action againstCynodon dactylon (L.) Pers. and other test species. Moreover, we found that several phenolic compounds appeared to be important allelochemics in this activity. We have now isolated other potential inhibitors fromP. aviculare residues and soil underPolygonum stands, and none of these occurred in soil underC. dactylon stands. GC-MS analysis revealed that these additional inhibitors are long-chain fatty acids with 14–22 carbons. Nine were identified inP. aviculare residues and seven in soil underP. aviculare. Sodium salts of all the identified fatty acids inhibited seedling growth ofC. dactylon and at least some test strains of the nitrogen-fixing bacteria,Azotobacter andRhizobium.     

Content-based image retrieval using growing hierarchical self-organizing quadtree map

In this paper, a growing hierarchical self-organizing quadtree map (GHSOQM) is proposed and used for a content-based image retrieval (CBIR) system. The incorporation of GHSOQM in a CBIR system organizes images in a hierarchical structure. The retrieval time by GHSOQM is less than that by using direct image comparison using a flat structure. Furthermore, the ability of incremental learning enables GHSOQM to be a prospective neural-network-based approach for CBIR systems. We also propose feature matrices, image distance and relevance feedback for region-based images in the GHSOQM-based CBIR system. Experimental results strongly demonstrate the effectiveness of the proposed system.     

Determination of Bridge Design Parameters through Field Evaluation of the Route 601 Bridge Utilizing Fiber-Reinforced Polymer Girders

The Route 601 Bridge in Sugar Grove, Virginia, spans 11.89?m (39?ft) over Dickey Creek. The bridge is the first to use the Strongwell 91.4?cm (36?in.) deep fiber-reinforced polymer double web beam in a vehicular bridge superstructure. Construction of the new bridge was completed in October 2001 and field testing was undertaken shortly thereafter, as well as in June of 2002, to assess any potential changes in structural performance. This paper details the field evaluation of the Route 601 Bridge. Using midspan deflection and strain data from the October 2001 and June 2002 field tests, AASHTO bridge design parameters were determined—namely, wheel load distribution factor g, dynamic load allowance IM, and maximum deflection. The wheel load distribution factor was determined to be S/4, a dynamic load allowance was determined to be 0.36, and the maximum deflection of the bridge was L/1,110. Deflection results were lower than the AASHTO L/800 limit. This discrepancy is attributed to partial composite action of the deck-to-girder connections, bearing restraint at the supports, and contribution of guardrail stiffness. It was found that diaphragm removal had a small effect on the wheel load distribution factor.     

Assessment of Time of Concentration Formulas for Overland Flow

Despite the importance of overland time of concentration on the design discharge, engineers are often bewildered by the array of formulas available in the literature without knowing the accuracy of each formula. This assessment covers nine formulas published between 1946 and 1993, which are intended for overland flow only that is subjected to uniform rain. The assessment compares the estimates from the formulas with experimental values that are derived under the same conditions for two surfaces: concrete and grass. The assessment shows that formulas that do not account for the rainfall intensity are only valid for a limited range of rainfall intensities. The formulas that account for the rainfall intensity generally show better agreement with the experimental data. Finally, the assessment gives two rankings of the formulas for the two surfaces in accordance to their accuracy as compared to the experimental data. The formula that has the best accuracy for both surfaces is the Chen and Wong formula.