Research on Balanced Content of Magnesium Oxide in Roller Compacted Concrete

In order to accelerate the application of magnesium oxide (MgO) expanding admixture in roller compacted concrete (RCC), aiming at common three-grading RCC used in water conservancy and hydropower engineering, the autoclave expansion of mortar specimens and the compressive strength of the autoclaved mortar specimens were studied by means of autoclave test and compressive strength test. Results showed that the balanced MgO content in RCC, which cannot be determined by the MgO content corresponding to 0.5% autoclave expansion rate of mortar specimens or the inflection point of the curve of autoclave expansion changing with MgO content like normal concrete, should be determined by the MgO content corresponding to the inflection point of the curve that compressive strength of autoclaved mortar specimens changes withMgO content to ensure longterm safe operation of RCC dams.


Introduction
A large number of experimental studies and engineering practices show that the magnesium oxide (MgO) concrete, which is formed by mixing with MgO expansion agent in concrete, has excellent delayed microexpansion characteristics. is delayed microexpansion can compensate for the volume shrinkage of dam concrete during the long temperature drop.
When MgO concrete damming technology is applied to actual engineering in China, the added MgO content in normal concrete is generally determined by autoclave test referring to the Chinese National Standard Autoclave Method for Soundness of Portland Cement (GB/T750-1992) (hereinafter referred to as GB/T750-1992) to take mortar whose material composition is closer to the concrete than neat cement paste as the autoclave specimens. Accordingly, this method is named mortar autoclave method [26,27]. RCC which is commonly used in dams of modern water conservancy and hydropower engineering contains fly ash up to 55-70%, which is significantly higher than 25-45% in normal concrete.
en, compared with normal concrete, how to determine the balanced MgO content in RCC? At present, there are few reports about this problem, and it will be discussed in this paper so that it is helpful to apply MgO expansion agent to RCC.

Materials.
Cement used in the test was P·O 42.5 ordinary Portland cement, whose MgO content is 1.81%, water requirement of normal consistency is 25.4%, specific surface area is 325.7 m 2 /kg, and soundness is qualified. e quality of cement conforms to the Chinese National Standard Common Portland Cement (GB175-2007). Fly ash, whose fineness is 13.6%, SO 3 content is 0.63%, ignition loss is 1.24%, and water demand ratio is 102.8%, meets grade II ash standard stipulated in the Chinese Electric Power Industry Standard Technical Specification of Fly Ash for Use in Hydraulic Concrete (DL/T5055-2007). Sand was made of artificial limestone sand from a reservoir construction site in Guizhou province, and its fineness modulus is 3.04 and particle gradation is good. e sand meets the requirements of Chinese Electric Power Industry Standard Construction Specification for Hydraulic Roller Compacted Concrete (DL/ T5112-2009). e admixture was water reducing agent, whose water reduction rate is 22%, and it conforms to the Chinese Electric Power Industry Standard Technical Code for Chemical Admixtures for Hydraulic Concrete (DL/T5100-2014). e chemical composition and the physical properties of the MgO expanding admixture are shown in Tables 1 and 2, which meet the Chinese Electric Power Industry Standard Technical specification of magnesium oxide expansive for use in hydraulic concrete (DL/T5296-2013).

Test Methods.
Tests included the autoclave test and compressive strength test in this paper.
Before the autoclave test is carried out, the autoclave specimens need to be prepared according to the regulations in GB/T750-1992. Water requirement for preparing mortar specimens is converted by the water-binder ratio of 0.5 : 1, cement-sand ratio of 1 : 3.74, and cementitious material (cement + fly ash) of 196 kg/m 3 , referring to the mix proportion of the three-grading RCC used in an actual dam project. e proportion of fly ash in the total amount of cementitious material is 60%, and the added MgO amount is calculated as a percentage of the total amount of cementitious materials. e dimension of all autoclave specimens is the standard size of 25 mm × 25 mm × 280 mm specified in GB/T750-1992. e autoclave test was carried out in turn according to the following steps [28]: (1) e prepared autoclave specimens are kept in a standard curing room with a temperature of 20 ± 2°C and a humidity of not less than 95% RH for 48 h and then removed from molds. (2) e demoulded specimens are placed into a constant temperature room of 20 ± 2°C for 1-2 h, and then initial length of specimens is measured by the length comparator and marked as L s ( Figure 1). (3) e measured L s specimens are put into a boiling box and boiled for 3 h. When the water temperature drops to indoor temperature, the specimens are taken out and then put into the constant temperature room of 20 ± 2°C for 12 ± 3 h. e length of specimens after boiling is measured by the length comparator and marked as L m . (4) e measured L m specimens are placed into an autoclave kettle under a pressure of 2.0 ± 0.05 MPa for 3 h, as specified in GB/T750-1992 ( Figure 2). ereafter, the autoclave kettle is closed, and the steam in the kettle is released after the pressure falls below 0.1 MPa. When the kettle temperature reduces to 20 ± 2°C, the specimens are taken out and then placed into the constant temperature room of 20 ± 2°C for 12 ± 3 h again ( Figure 3). e length of autoclaved specimens is measured by the same length comparator and marked as L f . (5) e autoclave expansion is calculated as follows: where E is the autoclave expansion rate and L is the effective length of the specimen.
According to the provisions of GB/T750-1992, the soundness of the specimen is judged to be qualified when the autoclave expansion rate of a specimen does not exceed 0.

Autoclave Expansion and Compressive Strength.
When different amounts of MgO were added, the autoclave expansion rate of mortar specimens and the compressive strength of autoclaved specimens are listed in Tables 3 and 4, respectively. Meanwhile, the curve of autoclave expansion rate varying with added MgO content and the curve of compressive strength varying with added MgO content are drew as shown in Figure 5.

Analysis of Results.
e autoclave test is based on the autoclaving principle. e specimens of cement-based materials containing MgO are placed in an environment with a high temperature and a high pressure of 2.0 ± 0.05 MPa (equivalent to a temperature of 215.7 ± 1.3°C) so as to cause expansion deformations rapidly. e soundness of cement-based materials containing MgO is determined by measuring the autoclave expansion. In this way, the expansion deformation of cement-based materials containing MgO, which would have taken years to be known in a normal temperature, can be known in a short time. It can be seen from Figure 5 that the autoclave expansion rate of mortar specimens increases with the increase of added MgO content. When added MgO content reaches 12%, there is an obvious inflection point on the curve of autoclave expansion rate varying with added MgO content. According to GB/T750-1992, the balanced MgO content is about 13% when it is determined by the MgO content corresponding to the autoclave expansion rate of 0.5% and about 12% when it is determined by the MgO content corresponding to the inflection point of the curve of autoclave expansion rate varying with added MgO content [29].
at is, the balanced MgO content which is determined by mortar specimens prepared with reference to the mix proportion of RCC is as high as 12-13%. However, according to the existing experimental research results, the balanced MgO content which is determined by mortar specimens prepared with reference to the mix proportion of normal concrete is generally 3.5-5% [26]. e former is 7-9.5% more than the latter, which is closely related to the following two reasons. Firstly, fly ash can obviously restrain the expansion deformation of the mortar mixed with MgO [21,30]. In this test, the fly ash content in mortar specimens accounted for 60% of the total amount of cementitious material, which is significantly higher than that of the fly ash content of normal concrete (generally 25-45%). Compared with the specimens with low fly ash content, the significant increase of fly ash content in the specimens leads to a significant reduction in the autoclave expansion rate. Secondly, the amount of cementitious material in RCC is generally about 20 kg/m 3 lower than that of normal concrete. When the percentage of added MgO is the same, the MgO content of mortar specimens prepared with reference to the mix proportion of RCC is naturally     Advances in Materials Science and Engineering lower than that of mortar specimens prepared by the mix proportion of normal concrete. e rapid expansion of the autoclave specimens mixed with MgO originates from the increase of the specimens volume caused by the rapid hydration of MgO to form Mg(OH) 2 crystals in a high temperature and a high pressure. erefore, the autoclave expansion of the specimen will decrease naturally as long as the content of the MgO expansion source decreases. In other words, due to the combined action of the above two factors, RCC requires more MgO content than normal concrete in order to achieve the same autoclave expansion. erefore, when the balanced MgO content in concrete is determined by the autoclave expansion rate of 0.5% or the inflection point of the curve of autoclave expansion rate varying with added MgO content, the balanced MgO content in RCC is much higher than normal concrete.
Meanwhile, the effect of MgO expansion on the compressive strength of cement-based materials, which would be seen after several years, can also be known in a short time through the autoclave test. As can be seen from Figure 5, the compressive strength of the autoclaved specimens increases gradually at first and then decreases rapidly with the increase of the added MgO content. ere is an obvious inflection point on the curve of compressive strength varying with added MgO content. e reason is also the volume expansion caused by the rapid hydration of MgO to form Mg(OH) 2 crystals. When the amount of MgO added to the cement-based materials is proper, the Mg(OH) 2 crystals formed by MgO hydration can reduce the porosity of mortar specimens, enhance their compactness, and improve their compressive strength. However, when the added MgO is excessive, the compactness and compressive strength of mortar specimens will decrease because the extrusion caused by excess Mg(OH) 2 crystals damages the internal structure of mortar specimens. In this test, the corresponding MgO content at the inflection point of the curve of compressive strength varying with added MgO content is 8%. at is, compressive strength has begun to decline although the   added MgO content has not reached 12%. is is an undesirable situation for concrete structure because long-term safe operation of buildings is the first priority for all engineers and technicians. erefore, for the mortar specimens prepared with reference to the mix proportion of RCC, the balanced MgO content in concrete should be determined by the MgO content corresponding to the inflection point of the curve of compressive strength varying with added MgO content but not the autoclave expansion rate of 0.5% or the inflection point of the curve of autoclave expansion rate varying with added MgO content.

Conclusions
In this paper, the autoclave mortar specimens were prepared to test the autoclave expansion and compressive strength by referring to the mix proportion of the threegrading RCC used in a practical water retaining dam and the regulations of GB/T750-1992. e following conclusions can be drawn by analyzing the variation rules of the autoclave expansion rate and the compressive strength of these specimens under different MgO contents.
(1) e autoclave expansion rate of mortar specimens increases with the increase of MgO content and increases rapidly when added MgO content increases to a certain value, which can be seen from the obvious inflection point on the curve of compressive strength varying with added MgO content. At the same time, the compressive strength of autoclaved mortar specimens increases gradually at first and then decreases rapidly with the increase of added MgO content, and an obvious inflection point can also be seen on the curve of compressive strength varying with added MgO content, and the MgO content corresponding to this inflection point is 4% lower than that of the curve of autoclave expansion rate varying with added MgO content. (2) When the balanced MgO content in the concrete is determined by the autoclave expansion rate of 0.5% or the inflection point of the curve of autoclave expansion rate varying with added MgO content, the balanced MgO content determined by mortar specimen prepared with reference to the mix proportion of RCC is as high as 12-13%, which is 7-9.5% more than that of mortar specimens prepared with reference to the mix proportion of normal concrete. Data Availability e data sets used to support the results of this study are available from the corresponding author upon request. Advances in Materials Science and Engineering 5