Mohammed E. Haque

ANN back-propagation prediction model for fracture toughness in microalloy steel

Fracture toughness of microalloy steel was evaluated as per ASTM E399 standard. Artificial neural network back-propagation model was developed to predict the behavior of fracture toughness and tensile strength as a function of microstructure. Both fracture toughness and tensile strength were found to increase with the increase in martensite content in a dual phase microstructure of microalloy steel. The primary objective of the ANN Back-propagation (BP) prediction model was to validate and extend the application of microalloy steels for various engineering applications. ANN training model was found to be in good agreement with the experimental results. The ANN training model has been used to predict the best/optimum toughness properties in terms of intercritical annealing temperature and martensite content. This can be used as a practical tool for predicting the fracture toughness in other series of steels comprising dual-phase microstructures and also to optimize strength and ductility properties.

Prediction of corrosion-fatigue behavior of DP steel through artificial neural network

Abstract Corrosion–fatigue crack growth (da/dN) of dual phase (DP) steel was analyzed using an artificial neural network (ANN) based model. The training data consisted of corrosion–fatigue crack growth rates at varying stress intensity ranges (ΔK) for martensite contents between 32 and 76%. The ANN model exhibited excellent comparison with the experimental results. Since a large number of variables are used during training the model, it will provide a reliable and useful predictor for corrosion–fatigue crack growth (FCG) in DP steels.

Time-Space-Activity Conflict Detection Using 4D Visualization in Multi-storied Construction Project

This study investigated whether a 4D model can help project participants of construction sites to detect possible errors in construction process. Identifying problems in work sequence and elements are possible when necessary measures are implemented to rectify the problems in advance using a 4D model. The common problem associated with a construction project is a time-space conflict which may lead to project delays and cost overrun. Construction space management and activity sequencing are important aspects for timely project completion within an estimated budget. This research explored the effective use of 4D visualization that could help overcome timespace problems. In order to fulfill this research goal, a 4D model was developed that incorporated spatial requirements along a chronological schedule of events. The model produced logical evidence that a 4D model could effectively be usedin a construction site to detect time-space conflicts.

Service learning: engineering, construction science, and the experiential curriculum

Service Learning is a form of experiential learning that utilizes the context of a community service project to practice academic skills. In recent years Service learning has been included in many academic disciplines throughout the United States. Incorporating a hands-on component into the curriculums of construction science is challenging for the instructor. Most often these service-learning activities are accompanied by lessons in the classroom creating a pedagogical synergy not found in singular teaching/learning activities in the traditional classroom setting. A difficulty arises when professors want to develop service-learning projects for courses in structures. This paper examines the current status of service learning in the Construction Science curriculum and proposes a service-learning project for Construction Science students at Texas A&M University. Interest in Service Learning has occurred at the university.

Mechanical behavior of powder metallurgy steel—Experimental investigation and artificial neural network-based prediction model

Mechanical properties of high-density powder metallurgy (PM) steels have been evaluated using standard tests, and a theoretical model using the artificial neural network (ANN) has been developed. Various heat treatments were carried out to study their influence on mechanical properties, viz. endurance limit (EL), yield strength (YS), and hardness, and also on the carbon content in PM steel. The material containing 0.47% C that was quenched and tempered at 503 K (QT 503 K) showed the optimum combination of yield strength/ultimate tensile strength (YS/UTS) and EL. The ANN-based model showed excellent agreement with experimental results. Prediction models based on the ANN are demonstrated for YS as well as for the EL as a function of heat treatment (ranging from QT 400 K to QT 900 K) and percent carbon (%C) (between 0.1 and 0.5). This would help the materials engineer suitably design the heat-treatment schedule to obtain the desired/best combination of fatigue and strength properties.

5D virtual constructions: Designer/constructor’s perspective

The construction activity level intricacy makes the whole construction process quite complicated and difficult for planners and builders. 4D (3D+Time) construction planning is a promising area, and it presents the ability to represent the construction process with the additional sequential dimension. It makes the virtual models easy to analyze and plan for the sequence of activities. Traditionally, architectural/construction engineering and technology education, especially construction scheduling has been dependent on bar charts and network diagrams. However, students can hardly understand the schedule-construction progress relationship using a CPM network or a bar chart. Using 4D visualization students can learn time-space relationship in construction schedule more effectively. The objective of this research was to create a user-friendly 5D (3D+time+cost) model by adding the cost of project with reference to the time line so that the planners would be able to peg the cost control measures with the schedule. The study provides a structured method and a systematic approach that will enable the students and planners to develop 5D models without worrying about the errors and software hitches. This 5D visualization model will facilitate developing a more realistic approach to the whole construction process. The entire model is integrated to produce an interactive visualization to make the process fairly easier for students as well as the construction industry professionals. In addition to this, changes to the 5D model can be done easily by triggering changes at one level. The techniques demonstrated through 5D virtual construction models can potentially be a valuable tool not only in the classroom, but also as an effective learner-centered self-directed tool to learn planning and construction processes.

Integrated Multimedia Uniform Bridge Element Identification System Database for Bridge Inspection and Maintenance

Bridge inspection information management poses a conundrum to transportation agencies across the nation. Today, there is a real need to use all media types, including text, image, sound, and video, to document bridge inspection projects, rather than only text. The development of an interactive, menu-driven, PC-based bridge inspection and maintenance database system using various information technologies to provide multimedia documentation is addressed. The system is geared to identify each bridge element uniquely using the Uniform Bridge Element Identification System (UBEIS). UBEIS is a generalized multidimensional coordinate system consisting of a string of alphanumeric characters that does not require any structural detail drawings, such as framing plan or cross sections, to identify a unique structural element. This database system enables bridge engineers to evaluate future rehabilitation needs, track the condition of structurally deficient members, and keep the rehabilitation and replacement history of the bridge. The system can be considered as a practical link between inspection and rehabilitation and acts as a warning system against bridge deterioration and potential failure.

Artificial Neural Network Model: Prediction of Mechanical Properties in Beta-Titanium Biomaterial

Theoretical prediction of mechanical properties using Artificial Neural Network (ANN) of a new beta-titanium alloy was investigated and compared with the experimental results obtained as a function of heat treatment. The cold worked biomaterial was subjected to different thermal processing cycles. The experimental values of mechanical properties were determined using MTS Landmark-servo hydraulic UTS machine. The new beta-titanium alloy demonstrated an excellent combination of strength and ductility for β-annealing and solution treatment plus aging thermal processing treatments. This data was used to train an artificial neural network (ANN) model to predict hardness. The predicted hardness values were found to demonstrate very good agreement with the experimental values.

REINFORCED CONCRETE BEAM-COLUMN DESIGN: AN ARTIFICIAL NEURAL NETWORK APPROACH

This paper investigated the suitability of an artificial neural network (ANN) for modeling a preliminary design of reinforced concrete beam-column. An ANN backpropagation model was developed to design a beam-column that predicts column cross-section and reinforcing steel requirements for a given set of inputs: concrete compressive strength, reinforcing steel strength, factored axial load, and moment. The trained ANN backpropagation model was tested with several actual sets of design data, and a comparative evaluation between the ANN model predictions and the actual design is presented.

Field Performance of Instrumented Concrete Pipe Under Deep Burial

Two 1.524-m-diameter, reinforced concrete pipes were instrumented to compare field results with design calculations. A computer program, Standard Installation Direct Design, developed by the American Concrete Pipe Association was used to design the pipe. Instrumented pipes were completely monitored until 11.9 m of cover had been placed. Measurements of soil contact pressures and vertical and horizontal deflections continued for 6 months. The computer-simulated and observed responses of the buried concrete pipe installations were compared. Pipe contact pressures measured at the invert were much smaller than those assumed from calculations of thrust and moment. Design moments were conservative compared with the experimentally measured values. Thrusts are difficult to calculate accurately and show large experimental variations.