Environmental engineers and scientists are concerned with environmentally sustainable products due to the exponential increase in industrial activities over the recent decades, according to K. Roy et al. [1]. Therefore, the need to research materials that can be used to replace constituents of concrete cannot be overemphasized. This is due to the depletion of these natural constituents and also demand for improved concrete properties. Production of cement, which is the binder in concrete, emits a lot of CO₂ in the atmosphere, with 1 ton of cement emitting 0.9 ton of CO₂ as shown by C. Chen et al. [2]. In addition, cement production is extremely energy and resource intensive, with 1.5 tons of raw materials required for a ton of production of cement, according to K. Roy et al. [1]. A potential solution in reduction of CO₂ emission from the cement industry and reduction of cement required in construction is to add supplementary materials in replacing the cement content. These materials include industrial (steel slag and fly ash) and agricultural by-products, which help in the reduction of greenhouse gases, according to L.Rajamony Laila et al. [3]. Supplementary cementitious materials (SCMs) are materials that are used to partially replace cement in the production of concrete. These materials possess pozzolanic activity, and thus they can react with the calcium hydroxide produced during cement hydration to form an additional cementing gel. Studies have proved that the use of these supplementary cementitious materials to replace cement reduces the amounts of carbon dioxide released to the atmosphere by 17%. This is because none of the material is derived from raw materials and relatively little carbon dioxide is associated with their processing according to M. Tyrer et al [4]. Feasibility studies into use of these wastes in concrete have also been found to improve the strength properties of concrete and reduce the alkali-silica reaction on top of reducing greenhouse effect according to L. Rajamony Laila et al. [5]. Rice husk ash (RHA) is major agricultural by-product from food crop paddy. For every four tons of rice produced, a ton of rice husk ash is produced. The husk is disposed by either dumping in open area near the milling site or on the roadside to be burnt. Burning rice husk has been used for production of steam, generating about 15-20% of the ash by weight. The ash being very light is carried by wind and water in its dry state. It is difficult to coagulate and thus contributes to air and water pollution according to Barrack M [6]. Cumulative generation of ashes require a large space for disposal. This has led to environmental concerns hence the need for potential re-use. Therefore, rice husk ash (RHA) is increasingly being used as a supplementary cementitious material to replace cement in concrete production as it is a pozzolan. RHA concrete has been found comparable to normal concrete in terms of structural and mechanical behavior up to 25% level of replacement.

The strength of concrete with RHA replacements of 10 and 15% revealed that the addition of RHA in concrete increased cohesiveness and stiffness of the mixture due to the fineness of RHA. Increased RHA lowered the compressive strength of concrete; however, with 10% replacement, the strength variation was 1-2.5% hence a recommendation of 10% by A.Siddika et al. [7]. Another study by Josephim Alex et al. on the variation of RHA in concrete from 0,10,15, and 20% revealed that 10% replacement had a remarkable percentage of strength gain with 7.8% as compared to plain concrete [8]. Also, incorporating RHA in concrete was found to contribute to low ratios of chloride ion penetration up to 928 coulombs with 25% replacement, as shown by S.A Zareei et al. [9]. The current use of RHA in concrete has an emphasis on strength characteristics which is tested at 28 days.

This is in conformity with the current civil engineering practice of designing ordinary RC structures whereby a major focus is given to ULS (Ultimate Limit State), i.e., load-bearing capacity. Simplistic methods are often employed for the second limit state, i.e., SLS (Serviceability Limit State). This latter state covers criteria arenecessary for the functional and intended purpose of the structure and user’s comfort. The most commonly used criterion for SLS is ultimate deflection, hence keeping other conditions such as criteria on crack width, structural vibrations, among others, unresolved. Failure to take into account all SLS criteria in design and construction phases may only save a certain amount of money at the initial phase; however the same or even more will be dispensed after several years for inevitable maintenance, according to Havlasek et al. [10].

Creep in concrete is defined as deformation under a sustained load that takes place in the direction of the load. Aggregates usually record very little creep, but they do influence the creep through a restraining effect on the magnitude of creep. The paste is responsible for creep. Quality and amount of paste are important factors in creep. Other additional factors influencing creep are water content curing conditions, relative humidity, specimen sizes, etc. In general, the creep rate of concrete increases with applied stress, as shown by A.A. Olayinka et al. [11].

The use of SCMs in concrete affects the strain behavior of such concretes. This is because the SCMs modify the pore size distribution and characteristics of interferes. As reported, the addition of metakaolin reduces creep, whereas Blast furnace slag increases the creep, as shown by M. Shariq et al. [12]. Creep of RHA concrete studied on cylinders showed a decrease in creep with an increasing % of RHA. This was attributed to the pore microstructure of RHA as smaller capillary pores were observed on RHA concrete as compared to plain concrete. RHA addition in concrete made it denser and impervious hence lower creep values as shown by Z. Hai He et al. [13].

Creep does not cause short-term failure in concrete; however, it leads to redistribution of internal forces and deformation of concrete structures in the long run. The total strain in concrete, including load-induced strain, becomes significant when determining whether concrete reaches a failure state and becomes much more critical when concrete undergoes creep in tension with shrinkage due to the low-strain carrying capacity of concrete in tension. Once the tensile strain reaches the ultimate value in concrete, cracking occurs, seriously affecting durability and serviceability. Bending creep has a crucial effect on gradual increases of deflection of concrete flexural members such as slab and beam, according to S. G. Kim et al. [14].

In practice, engineers predict creep strain using standard codes or available models developed in temperate countries. These creep models include ACI 209 (by American concrete institute), CEB-FIB (by Euro-international concrete comm-ittee), B3, and GL 2000. These models were developed by curve fitting of test results according to M. H. Kabir et al. [15].

Creep of concrete containing Iranian pozzolans showed that each pozzolan had an effect on the creep of concrete. Model results from creep prediction models were compared with experimental results; it was found that experimental results were higher than predicted values; however, the 28-day estimation model (short-term data model) gave a better prediction, according to P. Ghodhousi et al. [16]. Creep of concrete containing varying ratios of reinforcement showed that the reinforcement inhibits creep. Comparisons of the results with models showed a similar trend even though the values differed. The ACI model proved to be more accurate, and it was modified in order to incorporate the influence of reinforcement ratio, according to G. Sun et al. [17].

As has been documented, the creep of concrete changes when mineral admixtures have been used. Incorporation of concrete creep effect in the analysis of structures is deemed necessary. Once the creep behavior of concrete is understood, adjustments should be made in the design to cater for the same, according to M. Shariq et al. [12]. Furthermore, variation of results between model predictions (of most commonly used creep prediction models) and experimental results shows that the models have not taken into account the effect of mineral admixtures such as rice husk ash in concrete; hence modifications are required to the original equation.

Based on the above discussion and having found that RHA is a suitable SCM, an investigation on the creep of RHA concrete beams is presented in this paper. The creep is studied experimentally, and the results are used to develop a formula for creep prediction of 30MPa concrete beams made with 10% rice husk ash as SCM.

The trends in shrinkage and creep for both surfaces were similar, with a steep increase at early ages followed by a gentle increase at later ages, as shown in Fig. (5) below. However, the value of creep is more than twice that of shrinkage. In both the creep and shrinkage, growth continues with time but at a decreasing rate [12, 17].

The shrinkage strain is largest on top as compared with the bottom surfaces as drying takes place at the surface. The shrinkage reported is drying shrinkage, which is a continuous process when concrete is subjected to drying conditions. This shrinkage does not result from the loss of free water but the loss of the water held in gel pores. On exposure to drying, the gel water is lost progressively over a long time. The unrestrained paste loses its volume; hence it shrinks [19]. The rate of drying shrinkage decreases rapidly with time. It is observed that at 60 days, shrinkage strain contributes to almost 40% of the total deformation. Shrinkage causes unsightly cracks in concrete, which make the concrete creep more as it creates points of weakness

This study shows that the drying shrinkage is an inherent property of concrete that cannot be eliminated; however, its magnitude can be reduced to avoid damage in concrete. The magnitude can be reduced by the use of low heat Portland cement or pozzolanic cement as these have a higher capacity to withstand volume changes without cracking, as opposed to ordinary Portland cement [19].

Therefore, in experiments of bending creep strains, there is a need to properly account for shrinkage variation within the beam depth, which is mainly caused by compaction of concrete and bleeding during casting work.

The bottom creep is higher than the top creep, as shown in Fig. (6) below. This could be explained by higher shrinkage strains at the top since the magnitude of total strain is almost the same for both surfaces.

The rate of creep is much faster for the of early age (up to 21 days) than at later ages. At the age of t > 35 days, the rate of increase of creep tends to be stable. It is also observed in the 56% of the observed 60 days is achieved withing 14 days of sustained loading. The maximum creep for the 60 days was 620 microstrain (0.00062 mm/m). This creep strain being more than 400 microstrain is considered significant in concrete and cannot be ignored [20]. The creep strain in this experiment reached the secondary stage or the steady state creep. The creep strain and drying shrinkage have similar trends, however the value of creep is more than twice that of shrinkage.

Creep of concrete is largely induced by change in internal relative humidity of the concrete matrix as pore water distribution changes under loading from the region of high to low stress. This results in the transfer of load from the liquid phase to the surrounding solid. Under sustained load, the pressure of capillary water gradually disappears, followed by gel pore water. Therefore, the pressure on intercrystallite adsorbed water continues to act during the entire loading period. Hence the gradual decrease in the magnitude of creep [19].