Keywords: grey relational analysis, recycled aggregate concrete, Taguchi method.

As a result of the rapid development of urban construction, many old or degraded buildings need be demolished or reconstructed as part of urban renewal plans. Thus, great amounts of construction waste is produced. Concrete waste accounts for the highest proportion. Traditionally, most concrete waste is landfilled. However, owing to the limitation of available landfill area, the saving of natural resources, and the prevention of environmental pollution, the use of recycled concrete aggregate (concrete made with recycled concrete aggregate) as much as possible has become an important issue, especially on the urgent requirements of sustainability.

Assessing the optimal mixture is important for obtaining the desired quality of recycled aggregate concrete. In the previous studies, the assessment of the optimal mixture is conducted using the algorithm of the design of experiments with an orthogonal array and described as follows. First, the key characteristic which can represent quality is determined and then the related control factors are identified which affect the performance of the characteristics. The level of the control factors is also determined to produce various mixtures (combinations of control factor levels). By selecting an appropriate orthogonal array, fewer mixtures (compared with all possible mixtures) are required to identify the optimal mixture in order to reduce the time and cost of the experiment. Several specimens are obtained for each selected mixture, and the average value is calculated as the performance of the mixture on the characteristic. Finally, both the analysis of variance and the level response chart are adopted to assess the optimal mixture and the characteristics.

However, two problems will be encountered when using the traditional algorithm: (1) the traditional algorithm takes the average value as the performance of mixture. Although each mixture produces several specimens, various uncontrolled factors will cause variance among the testing results of the specimens. This variance ought to be carefully considered to obtain a more robust evaluation. (2) The traditional algorithm can only consider a single characteristic. However, the quality ought to be represented in various aspects (multiple characteristics). To solve this problem of multiple characteristics, the traditional algorithm can only assess the individual optimal mixture on each characteristic and then determine the overall optimal mixture by engineering experience or cross analysis. But, the traditional algorithm cannot deal with too many characteristics at the same time because of the increasing complexity of calculation and the possible erroneous judgment.

Therefore, this paper aims to adopt a modified algorithm, which is integrated with a signal-to-noise ratio and a grey relational analysis, to obtain more robust evaluations and assess the optimal mixture on multiple characteristics. A set of experimental data of recycled aggregate concrete is adopted to compare the differences between the traditional algorithm and the modified algorithm. Water/cement ratio, volume ratio of recycled coarse aggregate, replacement by river sand, content of crushed brick, and cleanness of aggregate were selected as control factors on the characteristics of slump and the 28-day compressive strength.

The results show that the optimal mixture of recycled aggregate concrete is obtained as A1B1C2D2E1, indicating a water cement ratio of 0.5, a volume ratio of recycled coarse aggregate of 42.0%, 100% replacement of river sand, 0% crushed brick, and as-is aggregate.

purchase the full-text of this paper (price £20)

go to the previous paper
go to the next paper
return to the table of contents
return to the book description
purchase this book (price £130 +P&P)