The experimental work was followed by the characterization of engineering properties of locally collected raw soil, bentonite, and flyash. Different compositions were prepared from the raw soil with varying percentage of bentonite and flyash then compaction test was carried out to obtain the optimum moisture content (OMC) and maximum dry density (MDD). Two confirmatory tests were conducted after the trial work utilizing the Taguchi Factorial Design (TFD) method and the Sunflower Optimization Technique (SOT). Analysis of variance then emerged as the key factor determining the optimized value of OMC and MDD under ideal conditions (ANOVA). In the end, the impervious clay core properties of the existing structures were utilized to validate the optimized data for OMC and MDD obtained from TFD and SOT.
2.1 Problem Statement
The core is basically an impervious barrier constructed with natural unprocessed material like puddle clay that provides sufficient finesse as well as reduces the permeability of the structure. Blending or other processing of composite may sometimes be desirable to improve the clay core properties.The central puddle-clay core is impermeable but its structural strength is low. So, the present work tries to build an impermeable core with soil-bentonite-FA composite such that clay core provides strength as well as reduce seepage of the embankment. Figure 1 shows a schematic diagram of embankment with central core.
In the present work, a low-cost clay core composite was attempted to be made using bentonite and flyash combined with different percentages of raw soil. OMC and MDD of the different combinations were optimized to get the optimum desirable values that can be fit to construct a low permeable and high strength clay core.
2.2 Taguchi Methodology
Taguchi methods are statistical methods developed by Genichi Taguchi to improve the quality of manufactured goods and more recently also applied to engineering (Karna and Sahai 2012). The primary objective of the Taguchi method is to design a reliable system that is robust in the face of uncontrolled circumstances (Taguchi 1978; Taguchi 1987; Taguchi and Phadke 1989). The technique seeks to optimize the design parameters so that the system response is robust, that is, insensitive to noise factors, which are challenging or impossible to control (Taguchi and Phadke 1989). MINITAB 18 is a tool to maximize and minimize the optimal values under Taguchi method using following equation;
$$S/N=-10\text{*}\text{log}\left(\frac{1}{n}\sum \left(\frac{1}{{x}^{2}}\right)\right)$$
1
$$S/N=-10\text{*}\text{log}\left(\frac{1}{n}\sum \left({x}^{2}\right)\right)$$
2
where, \(x\) represents the results from the control factors. Eq. (1) gives the result for higher the best whereas Eq. (2) gives results lower the best (Akbari et al. 2014).
2.3 Sunflower optimization (SFO) technique
Sunflower Optimization (SFO) Algorithm was formulated by Gomez et al. (2018). They considered the unusual behavior of sunflowers in their hunt for the optimal orientation towards the sun. A sunflower's life cycle is consistent: like the needles of a clock, they arise and accompany the sun every day
According to the inverse square law, the amount of radiation is inversely related to the square of the distance.
The quantity of heat Q that each plant receives is then determined by:
$${Q}_{i}=\frac{P}{4\pi {r}_{i}^{2}}$$
3
where ri is the distance between the current source and plant i and P is the source's power.
The sunflowers face the sun in the following directions:
\(\overrightarrow{{S}_{i}}=\frac{{x}^{*}-{x}_{i}}{‖{x}^{*}-{x}_{i}‖}\) , i = 1,2,….,np (4)
Calculations are made to determine the sunflowers' stride in direction s using:
$${d}_{i}=\alpha \times {p}_{i}\left(‖{x}_{i}+{x}_{x-1}‖\right)\times ‖{x}_{i}+{x}_{i-1}‖$$
5
Where \(\alpha\) is the constant value defining an "inertial" movement of the plants, \({p}_{i}\left(‖{x}_{i}+{x}_{i-1}‖\right)\) is the probability of pollination, i.e., the sunflower i pollinates with its nearest neighbouri − 1, creating a new individual at a random position that changes with the distance between the blooms. In other words, those who are closest to the sun will move more slowly in pursuit of a local refinement, while those who are farther away would move normally. Additionally, in order to avoid excluding places that are likely to be candidates for the global minimum, the greatest step that any person may take must be limited. Here, the maximum step is defined as
$${ d}_{max}=\frac{‖{x}_{max}-{x}_{min}‖}{2\times {M}_{pop}}$$
6
Where Mpop is the total number of plants in the population and xmax and xmin are the values' upper and lower limits.
It will be a new plantation called:
$$\overrightarrow{{x}_{i+1}}=\overrightarrow{{x}_{i}}+{d}_{i}\times \overrightarrow{{S}_{i}}$$
7
The algorithm is initiated by creating an even or random population. Irrespective of the individual’s rating, the current algorithm helps us to find out the changing sun. After several runs, similar to the sunflowers, all the additional entities will face the sun and move in a controlled random manner. Paramount plants will pollinate around the sun (Malik 2021).