Abrasion resistance of IV-RCC used to construct spillway concrete steps of South African dams

This dissertation outlines the investigation on abrasion resistance of roller compacted concrete designed for the spillway steps of dams owned by the Department of Water and Sanitation. Literature on abrasion resistance of concrete has been reviewed and factors affecting abrasion resistance are discussed. Among these include hardness of concrete, aggregate/paste bond, type of aggregates, concrete compressive strength, curing, surface treatment and the quality of the concrete surface which is influenced by the finishing method. Available test methods as outlined by ASTM International Standards are also discussed in the literature review but only two methods (sandblasting method and wire brush method) were used in this research. These have been chosen because of their availability (at the UCT Civil Engineering Laboratory) and relevancy to this research. The sandblasting method is the most relevant in this research because it stimulates similar abrasion action as waterborne action on the concrete surface. The other method has been chosen for its availability in the UCT lab, to compare its results with the sandblasting results, and to evaluate whether if it can be used to assess abrasion resistance of hydraulic structures. A two-way experimental approach was adopted to achieve results for this research. The first approach was a laboratory investigation, where concrete cubes were produced under controlled conditions and tested for abrasion resistance, compressive strength, oxygen permeability index (OPI) and Sorptivity test. The durability index tests were conducted to get a general evaluation of the concrete’s microstructure. If the OPI values are acceptable (more than 9) then the concrete’s general microstructure would be accepted to be durable. Materials (aggregates) from De Hoop Dam and those readily available at the UCT laboratory were used to cast different types of concrete to be investigated and compared. The types of concrete included conventional concrete, fly ash concrete and slag (GGBS) concrete. DWS uses fly ash in the majority of their mass concrete projects due to its pozzolanic properties which help slowing down the heat of hydration. The slag was included in these mixes in order to have available results when it is used in future projects as it is highly likely to be used for areas in the KwaZulu-Natal province which have a less reactive type of slag in terms of heat of hydration. The raising of the Hazelmere Dam, close to Durban, is currently using GGBS as an extender. The second approach of this research was based on abrasion testing of cores drilled from the stepped wall of De Hoop Dam. These cores were drilled and taken to UCT laboratory for investigations with the same test methods used for laboratory cubes. A comparison was drawn between the test results of the laboratory made cubes and those drilled from the dam spillway, and a correlation was investigated between these two. Furthermore, the sandblast test method was compared with the wire brush method to assess which one produces more reliable results. The abrasion results showed small differences in abrasion resistance performance between the conventional concrete and the roller compacted concrete mixes. In fact when the percentage of the extender was adjusted from 70% to 65% these two concrete mixes performed equally. It was confirmed that beyond 45% replacement of Portland cement with fly ash, there is no improvement in abrasion resistance of concrete. However, there were no significant differences between abrasion resistant values of concrete with 45% or 65% fly ash replacements. It was observed that abrasion resistance of concrete was compromised beyond 65% fly ash replacement. Aggregates with sub-angular shape and rough texture have proven to yield high abrasion resistant concrete. The concrete made with blends of GGBS and Portland cement showed slightly improvement in abrasion resistance of concrete when compared to fly ash concrete. The cores drilled from the structure had abrasion resistance values 3 times more than the laboratory produced cubes and this was consistent with both the sandblasting and wire brush method, indicating that the actual structure has adequate abrasion resistance. The sandblasting method has proven to be reliable and more sensitive and it is recommended for assessment of abrasion resistance of hydraulic structures, while the wire brush method can be used for general quality control. Recommendations based on the results and literature study will be made available to the Department of Water and Sanitation to update their specifications, which was one of the key goals of this research