People | Locations | Statistics |
---|---|---|
Naji, M. |
| |
Motta, Antonella |
| |
Aletan, Dirar |
| |
Mohamed, Tarek |
| |
Ertürk, Emre |
| |
Taccardi, Nicola |
| |
Kononenko, Denys |
| |
Petrov, R. H. | Madrid |
|
Alshaaer, Mazen | Brussels |
|
Bih, L. |
| |
Casati, R. |
| |
Muller, Hermance |
| |
Kočí, Jan | Prague |
|
Šuljagić, Marija |
| |
Kalteremidou, Kalliopi-Artemi | Brussels |
|
Azam, Siraj |
| |
Ospanova, Alyiya |
| |
Blanpain, Bart |
| |
Ali, M. A. |
| |
Popa, V. |
| |
Rančić, M. |
| |
Ollier, Nadège |
| |
Azevedo, Nuno Monteiro |
| |
Landes, Michael |
| |
Rignanese, Gian-Marco |
|
Hosny, Ossama
in Cooperation with on an Cooperation-Score of 37%
Topics
Publications (1/1 displayed)
Places of action
Organizations | Location | People |
---|
article
AI-Based Approach for Optimum Soil Stabilization
Abstract
esults from previous studies confirmed that, adding Ground Granulated Blast-Furnace Slag (GGBS)activated by hydrated lime (L) to a typical Egyptian clayey soil increases strength and decreases swelling. This paperinvestigates reaching optimum soil stabilization for clayey soil to suit safe and economic road construction.Optimum soil stabilization can be achieved mainly through two stages as proposed in this paper: stage 1: quantify theeffect of the soil stabilization parameters represented in the GGBS%, Lime%, and the curing time/condition on thestabilized soil unconfined compressive strength (UCS) and the free swelling percentage (FS%) using ArtificialNeural Networks (ANN). Stage 2: determine the optimum set of stabilization parameters by conducting backwardoptimization on the developed ANN prediction model while meeting practical design preferences, using GeneticAlgorithms (GAs). Initially a simple to use ANN add-ins (Neural Tools 5.5) for Excel was used where the UCS waspredicted with an acceptable error of 10% for both training and testing sets. A detailed error analysis was performedand showed that the maximum under and over estimate errors were less than 3% and 5.35% for training and testingrespectively. However, it is not possible to use neural tool or other ANN software packages in performing backwardanalysis to determine the optimum set of inputs that may result in a certain output. Accordingly, a more transparentANN model was developed. After training and testing the developed ANN, it can work as an optimization modelwhere the decision variables are the stabilization parameters with an objective to reach a certain UCS while keepingthe swelling percentage within a certain range. The model has been applied on a case study where it was able tocome up with the practical ranges of the lime%, GGBS%, and the curing time/condition that would satisfy therequired design criteria.[M. S. Ouf, A. Elhakeem and O. Hosny. AI-Based Approach for Optimum Soil Stabilization. J Am Sci 2012;8(12):138-145]. (ISSN: 1545-1003). http://www.jofamericanscience.org. 21 Keywords: Soil stabilization; GGBS; lime; swelling soil; Modeling; ANN; and GAs optimization. 1. Introduction The economic development of any country iscontrolled to a great extent by its highway networks.This is becoming particularly apparent in thedeveloping countries, where tremendous lengths ofroads and highways need to be constructed in order tofacilitate the development. The type and thickness ofthe pavement construction represents a largepercentage of the cost of any road/highway project.Therefore, the development and use of methods todecrease the cost of pavement construction wouldresult in great cost savings. However beforedesigning the type and the thickness of the pavement,it is essential to take into consideration the conditionsof the subgrade soil, as it carries the traffic loads aswell as the pavement loads (Bari, 1995). The traditional section for road typicallyconsists of different layers such as: surface course,binding course, base and sub-base courses. Theselayers are typically made of imported materials thatrequire transportation, environmental and other coststhat increase with the distance from the source of thematerials to the site where roads being constructed.The reliance on imported material is the mainproblem, from a sustainability and efficiency point ofview, of the traditional road and pavement design andconstruction. Due to the gradual depletion in theconventional resources, searching for a more rationalroad construction approach aimed at reducing thedependence on imported material while improvingthe quality and durability of the roads is necessary(Marjanovic et al., 2008).Many problems associated with foundations onexpansive soils, include heaving, cracking and breakup of pavements, have been reported. Thefoundations of light structures supported on theground (e.g. highways) are more affected byexpansive soil problems than heavy or deep buriedstructures (Xidakis, 1979).Roads constructed on expansive clays may beadversely affected by the behavior of the clay. Thevolume change of such clays causes up