Earth quake resisting compressed stabilsed earth block buildings- an experimental and analytical study

Lakshmi Keshav*, Srisanthi V.G. and Poorna Kumar N.

The main objective is to promote compressed earth block building construction along with earthquake resisting features as a tool for sustainable, energy saving, eco-friendly and affordable housing. For this, six reduced scale 1:3 models of single room building constructed using three types solid, hollow and modified solid Compressed Stabilized Earth Blocks (CSEB) from locally available soil were subjected to Shake table tests. Three models of three different blocks were constructed with Earthquake Resisting Features [EQRF] that have sill band, lintel band and vertical bands to control the building vibration and other three models were without EQRF.

To examine the seismic capacity, the models were subjected to long-period ground motion and the test specimen were shaken repeatedly until the failure. The test results from Hi-end Data Acquisition system show that model with EQRF behaves better than without EQRF.A comparison between the results of tests and the FEM analysis by ANSYS predictions is made particularly concerning the acceleration, displacement response and the damage patterns carried out in this research. The data obtained from the experimental works were given as train set in Artificial Neural Network (ANN) and a tool was created in Matlab software for analyzing various blocks.

Full Text

Simulation study on the seismic continuous collapse impact distance of building structures

Wu Fengying, Wang Wei* and Ma Donghui

The earthquakes caused the collapse of the buildings along the street and debris accumulation causing traffic jams and seriously affecting the disaster relief work. In order to study the collapse debris accumulation of building structures under the earthquake action and the influence distance of debris accumulation, a simulation model is established based on the explicit dynamic analysis software ANSYS/LS-DYNA. Three story masonry structure as an example, simulated the continuous collapse mode of masonry structure under earthquake.

We divided impact distance of building structure collapse under earthquake into safety distance during earthquake and main impact distance of debris accumulation after earthquake and made a statistical research in two parts respectively in terms of the flying rocks phenomenon in the building collapse process. The results showed that numerical simulation method can accurately simulate the continuous seismic collapse failure model of the building structures to provide a new research idea to study the seismic collapse impact distance of building structures.

Full Text

Analysis of Machine Learning Algorithm for Identifying Key Actors in Covert Networks

Geetha R.* and Karthika Veera

Social Network Analysis has become one of the emerging approaches to analyze the social structure of real time networking patterns that are formed either formally or informally. This study presents the comparison of classical machine learning algorithms like Support Vector Machines, Decision Trees, Bayesian, k-means and Expectation Maximization for classifying and clustering the small world problem by analyzing a covert network dataset. Since the covert networks comprises of lot of unknown and incomplete data, classification becomes tedious and uncertain.

The Support Vector Machine and Decision Trees are suitable for analyzing the covert networks because of its ability to classify non-linear solutions efficiently whereas the uncertainty of covert networks can be handled with Bayesian networks. Since the clustering mechanisms uses maximum likelihood of data to find clusters, Expectation Maximization (EM) and k-means algorithms are selected for clustering the covert network attributes.

On applying these algorithms over 9/11 dataset, it is found that Bayesnet shows maximum precision and recall of about 0.963 and 0.961 respectively. Decision Tree algorithm is applied and a set of rules are generated identifying the critical attributes with 90.20% accuracy from which the critical actors can be recognized. The Bayesnet and Naïve Bayes models perform better than Support Vector Machine algorithms. All three algorithms identify a set of 5 attributes namely live together, ticket booking, flight members who board AA_11, AA_77 and UA_93 out off 65 as very crucial and highly influential in executing the organized events and social processes.

The Expectation Maximization algorithms show maximum log likelihood values and simple k-means algorithm shows minimum value for sum of squared error when the actors are grouped into 3 clusters. Thus, on examining the covert network structure with supervised and unsupervised algorithms, the key attribute of an actor is identified based on which the actor is determined to be an influential individual or the most dominant actor.

Full Text

GRG Optimization of Prestressed Rolled Steel Compact Sections

Ponnada Markandeya Raju

Structural steel is a stressed mineral resource. Its sustainability can be improved by its optimized utilization. This study is to determine the optimized parameters for prestressed the steel beams. IS Rolled Sections are taken for the analysis. Spreadsheets are developed to determine the pre-stressing force, eccentricity and optimized section parameters of the steel beams for various spans. MS EXCEL Solver is used to optimize the parameters. It is concluded that design bending compressive strength of beam inversely proportional to the span of the beam, hence load carrying capacity of the beam decreases with increase in the span but % increase in the load carrying capacity increases with the span of the beam.

Load carrying capacity increased by 13 to 122 % in rolled sections and in case of welded sections it is 8 to 50 %. Deflections decreased by 59 to 95 % in rolled sections and in case of welded sections it is 59 to 92 %. For the efficiency of pre-stressing i.e. to attain increment in the load carrying capacity and decrement in the deflection of the steel beams, the stress ratio (P/Afbd) must be in the range of 0.43 to 0.75 for Rolled sections and in case of welded sections it is in the range of 0.45 to 0.72.

Full Text

Grassland environmental hazards of the Gannan region in Gansu province of China, northeastern Qinghai-Tibetan Plateau

Zhang Xuexia, Sheng Wenfei and Qi Shanzhong*

Grassland degradation has been an environmental hazard problem and occurs frequently in China especially in the high-cold grassland region of Qinghai-Tibetan Plateau. The Gannan region, located on the northeast edge of the Qinghai-Tibetan Plateau, is a major area suffering from environmental hazard of grasslands. Natural threats and human activities, especially the latter, caused grasslands degradation of the region in the recent decades. This short communication investigates grassland environment hazards in the Gannan region such as grassland sandification and vegetation degeneracy and analyzes its causation.

Full Text