1 Introduction

The present world is focused on the use of waste material in a sustainable way to make a greener earth for the future generations. Different studies were incorporated to use industrial waste and agricultural by-product in modifying concrete. Rice husk ash (RHA), an agricultural byproduct, is yielded during the burning process of rice hull. RHA with more than 80% silica content in shape of non-crystalline silica is produced when it is burnt at less than 700 °C (Rashid et al. 2010). This high silica percentage can make RHA a sustainable replacement of ordinary Portland cement.

Rice hull is the outer shell of the rice grain. During the husking process of paddy, about 30% rice hulls is produced (Givi et al. 2010). It is used as biomass for energy production in paddy milling process as well as in various power plants. During the burning process, about 20% of it converts into RHA (Givi et al. 2010; Rashid et al. 2010). About 738.2 million tons of paddy is produced in the world each year of which about 40 million tons are RHA (Rice Market Monitor 2016). This huge amount of RHA is usually stored at temporary locations that may cause environmental havocs. Moreover, RHA cannot be naturally degraded due to its siliceous compositions (Zerbino et al. 2011). On the other hand, cement industries are responsible for about 7% of the total equivalent CO2 emissions (Zerbino et al. 2011), and one ton of cement production is responsible for about one ton of equivalent CO2 emission (Khan et al. 2012). The use of RHA in substitution of cement can reduce the CO2 footprint as well as the ecological hazards (Ramesh and Kavitha 2014; Chakraborty and Goswami 2015; Swaminathen and Ravi 2016).

Riceland Foods, Inc., a farmer-family owned business, is the largest rice miller in the U.S. that produces 125 million bushels of paddy annually. A significant amount of that paddy produces a huge amount of RHA. Generally, Riceland Foods considers RHA as a waste product and disposes it to nearby custom-designed landfill, which is an economical burden to the company and may pose an environmental threat to local communities. This RHA contains about 70% amorphous silica that makes it a potential source of pozzolanic materials, and it can be used as a replacement of cementitious material. But, its application as a pozzolanic material has not been investigated yet other than the current study. The present study primarily focuses on the effects of silica content of RHA in RHA-modified concrete.

The strength development of concrete is primarily governed by the strength of cement gel. The cement gel is nothing but the crystallized compound of calcium silicate hydrate (C-S-H) and calcium aluminate hydrate (C-A-H). The Ordinary Portland Cement (OPC) generally consists of tricalcium silicate (C3S) and dicalcium silicate (C2S). During the hydration process, C3S and C2S react with water and produce C-S-H gel and calcium hydroxide (CH), as shown in Eqs. (1) and (2). The produced CH reacts with alumina (Al2O3) and generates C-A-H as shown in Eq. (3). Both C-S-H and C-A-H are responsible for strength development of cement paste. When a pozzolanic material having silica (SiO2) comes in contact with the CH produced from C3S and C2S contributes to the amount of cement paste by forming C-S-H gel as shown in Eq. (4) (Givi et al. 2010). It is reported that, free CH remains after strength development of cement paste (Dunstan 2011). If pozzolanic material (e.g., RHA) is partially used with cement, it will react with the excess free lime and produce more cement gel. It will fill up the void spaces and densify the concrete. Thus, partial replacement of cement by RHA can produce stronger concrete.

(1)
(2)
(3)
(4)

Previous laboratory studies reported that RHA imparted the durability of concrete (Givi et al. 2010; Rashid et al. 2010). Some researchers (e.g., de Sensale 2006; Givi et al. 2010) have reported that the use of RHA as a partial replacement of cement significantly improves the compressive, tensile and flexural strengths of concrete. The maximum long-term compressive strength of concrete can be achieved when up to a 30% of cement is replaced by RHA (Givi et al. 2010). RHA as partial replacement of cement also improves flexural and tensile strengths of RHA modified concrete (Zhang et al. 1996; Habeeb and Fayyadh 2009). Moreover, depending on the type of aggregate, 12–15% replacement of cement by RHA may prevent the expansion of concrete due to alkali-silica reaction (Givi et al. 2010). It can be noted that alkali-silica reaction can cause serious cracking in concrete, resulting in critical structural problems and is a concern for premature concrete pavement failures in northwest part of the state of Arkansas. Thus, RHA could be an inexpensive source of pozzolan and can be used in construction activities as an alternative of OPC or other commonly used industry by-products such as fly ash.

The main objective of the present study was to evaluate strength properties of RHA-modified concrete with partial replacements (0%, 10% and 20% by weight) of OPC by RHA. Selected chemical and physical properties of RHA were also investigated in this study. In addition, the suitability of the RHA in concrete pavement was also evaluated by following the alkali-silica reactivity (ASR) test method on RHA-modified mortar bars.

2 Research Significance

RHA produced by Riceland Foods Inc. has the potential to be used in construction work, but its properties as construction material have not been investigated yet. This RHA could be an alternative cementitious material of cement or fly ash. The use of local RHA as a supplementary cementitious material (SCM) would aid local farmers to be economically sustainable by reducing its disposal cost and profiting from its sales to construction and paving contractors.

3 Materials Used

An OPC Type-I with a specific gravity of 3.15 was used in this research. Locally available stone sand with a fineness modulus (FM) of 2.2 and crushed stone having a nominal maximum size (NMS) of 25 mm (1 in.) were used as fine aggregate and coarse aggregate, respectively. The specific gravity values of fine aggregate and coarse aggregate were 2.58 and 2.61, respectively, and their absorption values were 0.3% and 0.9%, respectively. The volumetric mix design recommended in American Concrete Institute (ACI) 211.1 was followed. For a desired compressive strength of 31 MPa (4,500 psi), it yielded a mix proportion of 1.0, 1.3, and 3.0 of cement, fine aggregate and coarse aggregate, respectively. RHA was collected from Riceland Foods Inc. plant located in Stuttgart, Arkansas, and its selected chemical and physical properties are presented in Table 1. It is seen that this RHA contains more than 90% reactive oxides, which makes it a possible source of pozzolan. However, the loss on ignition (LOI) value is slightly higher than the American Society for Testing and Materials (ASTM) specification limit.

Table 1. Chemical and physical properties of RHA
Full size table

4 Experimental Methods

In this study, two different percentages (10% and 20%) RHA were used to substitute the OPC. The control sample (0% RHA) was tested for comparing test data with the other two samples (10% and 20% RHA). Values of slump, air content and unit weight of fresh concrete were estimated in accordance with ASTM C143, ASTM C231 and ASTM C138, respectively. Apart from these, mechanical properties such as compressive strength, tensile strength, flexural strength, modulus of elasticity, and Poisson’s ratio of hardened concrete cured for different days were also estimated in the laboratory. Compressive and tensile strength of concrete were determined using 150 mm × 300 mm (6 in × 12 in.) cylinders after 7, 14, 21 and 28 days of curing time in accordance with ASTM C39-04a. Two replicate specimens for each testing condition of the RHA-modified samples were casted in plastic molds and cured for 24 h in a moist condition at 24°C. For the flexural strength testing, a 600 mm (24 in.) long beam with a cross sectional area of 150 mm × 150 mm (6 in × 6 in) was casted for each RHA-modified concrete samples. Modulus of elasticity and Poisson’s ratio were measured after 28 days of curing period using the same specimens casted for the compressive strength. The test setup for measuring the Poisson’s ratio is shown in Fig. 1. To determine the Poisson’s ratio, an electrical strain gage and a compressometer were used to measure the lateral strain and the vertical strain, respectively. A portable strain indicator connected to the strain gages through wires displayed the lateral strains. The compressometer readings were noted from a dial gage. All cylinders and beams were demolded and kept in a water bath at room temperature of 24°C for curing until the age of testing (Fig. 2).

Fig. 1.
figure 1

Setup for measuring Poisson’s ratio with Compression machine.

Full size image
Fig. 2.
figure 2

Water bath curing at room temperature

Full size image

The specific surface area, an indicator of fineness, of RHA was estimated by following the Brunauer, Emmett and Teller (BET) method, which used experimental data collected by a NOVA 2200e analyzer (Fig. 3). The BET surface area was calculated using Eq. (5).

$$ \frac{1}{{{\text{W}}\left( {\frac{\text{P}}{{{\text{P}}_{0} }}} \right)^{ - 1} }} = \frac{1}{{{\text{W}}_{\text{m}} {\text{C}}}} + \frac{{{\text{C}} - 1}}{{{\text{W}}_{\text{m}} {\text{C}}}}\left( {\frac{\text{P}}{{{\text{P}}_{0} }}} \right) $$
(5)
Fig. 3.
figure 3

NOVA 2200e multi-station high speed gas sorption analyzer version 10.05

Full size image

Where,

W:

= weight of gas adsorbed,

P/P0 :

= relative pressure,

P:

= equilibrium adsorption pressure,

P0 :

= saturation vapor pressure,

Wm :

= weight of adsorbate as monolayer, and

C:

= the BET constant.

To get more accurate results, the multi-point BET mode was used with nitrogen as the adsorbate gas. Liquid nitrogen was used to provide the required test temperature (77°K) for nitrogen adsorption isotherms (Sing 2011; Mohamed et al. 2015).

The ASR tests were conducted in accordance with the ASTM C1567 procedure. Three mortar bars with an effective gage length of 285 mm (11.25 in) were casted for each RHA-modified (10% and 20%) samples. All the mortar bars were kept in a moist room for 24 h, and they were then immersed in tap water for another 24 h. Afterwards, the specimens were removed from water and immersed in 1 N NaOH solution for 14 days. Throughout this period, the change of length of each mortar bar was determined by taking four (4) consecutive readings at 4, 8, 12 and 14 days. Data were collected using a linear variable differential transformer (LVDT) connected with a data storage unit as shown in Fig. 4.

Fig. 4.
figure 4

(a) Use of LVDT to measure length change (b) Data storage unit (DSU)

Full size image

5 Results and Discussion

5.1 Grain Size Distribution of RHA and BET Surface Area

The pozzolanic activity of RHA is very much dependent on the particle size and surface area (Cordeiro et al. 2011). The RHA sample was sieved in the Arkansas State University Materials laboratory with the ASTM standard sieves to find its grain size distribution (Fig. 5). From Fig. 5, it is observed that only 3% of the tested RHA sample are 45 µm or smaller than 45 µm. On the other hand, AASHTO M 321-04 requires that 90% of the sample should not be larger than 45 µm. About 90% of the RHA used in this study has particle size 600 µm indicating that it is about 13 times coarser than the specified limit. To use RHA as SCM, the particle size of RHA should be the same as cement. However, tests were continued with this coarser RHA to see the behavior of the RHA in concrete without any modification or chemical treatment since this RHA had not been previously used in concrete as SCM. To reduce the particle size of RHA, it is recommended to burn RHA further using a furnace or to grind it by the help of hammer mill or ball mill.

Fig. 5.
figure 5

Grain size distribution of RHA

Full size image

The BET surface area of the tested RHA sample was found to be 18.038 m2/g. Habeeb and Mahmud (2010) reported BET surface areas of 25.3 m2/g, 27.4 m2/g, 29.1 m2/g, and 30.4 m2/g for different sizes of RHA samples having average particle sizes of 63.8 µm, 31.3 µm, 18.3 µm, and 11.5 µm, respectively. Therefore, the higher the particle size, the lower the BET surface area. The RHA of the current study had a very low BET surface area, which indicated a coarse particle. Such a coarse RHA sample can affect the pozzolanic activity of RHA in the modified concrete (Xu et al. 2016).

5.2 Properties of Fresh Concrete

The properties of 0%, 10% and 20% RHA-modified fresh concrete samples are presented in Table 2. It is seen that slump values were about the same for 0% and 10% RHA concrete, whereas 20% RHA-modified concrete exhibited a higher slump value. A higher amount of air voids due to an increased RHA percentage could be the reason behind this phenomenon. An increment of the RHA percentage also increased the air voids, but it decreased the unit weight. It happened as the OPC was significantly finer than the RHA. Coarser particles of RHA created additional air voids in the concrete, which eventually decreased the unit weight of the fresh concrete, as presented in Table 2.

Table 2. Fresh properties of concrete
Full size table

5.3 Compressive Strength

The compressive strength values of the tested concrete cylinders are shown in Fig. 6. It is observed that the control specimen gained compressive strength up to 36 MPa. On the other hand, the maximum compressive strength of 10% RHA-modified concrete was 20.13 MPa, which was only 56% of that of the control sample. A considerable amount of compressive strength 16.20 MPa was also gained by 20% RHA-modified concrete, and it was about 45% of that of the control sample. Bulk particle of RHA might not generate sufficient amount of C-S-H gel to fill up void spaces. The presence of high air voids in RHA-modified concrete due to bulk particles caused noticeable amount of strength reduction. This type of RHA-modified concrete can be used in controlled low strength materials such as flowable fill that requires a low compressive strength of 8.3 MPa (Ayers et al. 1994; Deng and Tikalsky 2008).

Fig. 6.
figure 6

Comparison of compressive strength of RHA-modified concrete

Full size image

5.4 Tensile Strength

The tensile strength of RHA-modified concrete is shown in Fig. 7. It is seen that the rate of gain in tensile strength of RHA-modified concrete was similar to that of its compressive strength. Between the RHA-modified samples, the 10% RHA-modified concrete gained the highest tensile strength, which was 75% of that of the control specimen. The 20% RHA-modified concrete exhibited further reduction of tensile strength, which was 65% of that of the control specimen. It is seen that the effect of air voids in concrete has a significant role in reducing tensile strength of concrete.

Fig. 7.
figure 7

Comparison of tensile strength of RHA-modified concrete

Full size image

5.5 Flexural Strength

Beam samples were tested after 28 days curing period to determine the modulus of rupture, which is often known as flexural strength. It is noticed that 10% and 20% RHA-modified concrete imparted 96% and 76%, respectively, of the flexural strength of the control specimen (Fig. 8). The use of RHA as a partial replacement of OPC reduced the flexural strength of RHA-modified concrete.

Fig. 8.
figure 8

Results of flexural strength of RHA-modified concrete

Full size image

5.6 Poisson’s Ratio and Modulus of Elasticity

The result of Poisson’s ratio and Modulus of elasticity of RHA-modified concrete are presented in Table 3. Poisson’s ratios increased with the increment of the RHA content. The rate of the gain of compressive or tensile strength of RHA concrete is highly correlated (negatively) with as the Poisson’s ratio. Another important parameter of concrete is modulus of elasticity. While comparing the measured modulus of elasticity with the American Concrete Institute (ACI) equation (Ec = 4700√f′c) (Table 4), it is seen that the measured Ec value is within ±22% of that obtained in the latter approach. In general, the modulus of elasticity values decreased with the increment of RHA percentage. It occurred due to less strength development in RHA concrete than the control specimen.

Table 3. Results of poisson’s ratio
Full size table
Table 4. Modulus of elasticity
Full size table

5.7 ASR Test Results

The expansion of concrete mortar bars obtained from ASR tests is shown in Fig. 9. The expansion of the control and 10% RHA-modified concrete is 0.20% and 0.25%, respectively, which exceeded the ASTM C1567 specified expansion limit (0.10%). It happened as RHA containing bulk particle size imparts higher alkali-silica expansions (Attoh-okine and Atique 2006). Visible cracks were found in the control sample as well as in 10% RHA-modified sample. However, the 20% RHA-modified samples showed reduced expansion of mortar bars after 14 days of curing. This is possibly due to the production of less expansive alkali-silica gel during hydration process of cement (de Sensale 2010) or due to lower alkali/ reactive silica ratio. Further, micro level investigation should be carried out on the RHA-modified mortar bars by scanning electron microscopy (SEM) in association with X-ray diffraction (XRD) to visualize the intermolecular bonding phenomena. This would help to describe more accurately the ASR expansion of RHA-modified mortar bar.

Fig. 9.
figure 9

Comparison of effect of ASR on RHA-modified concrete

Full size image

5.8 Summary

The maximum compressive, tensile and flexural strengths were gained by 10% RHA-modified concrete. The compressive, tensile and flexural strength gained by 10% RHA concrete were 20.13 MPa, 2.77 MPa and 4.16 MPa, respectively. Less cement gel production due to usage of bulk RHA could be a possible reason for the strength reduction in concrete. The modulus of elasticity of control, 10% and 20% RHA-modified concrete were 3.12x104 MPa, 2.56x104 MPa and 1.65x104 MPa, respectively. It is observed that the control, 10% and 20% RHA-modified concrete samples had Poisson’s ration 0.28, 0.40 and 0.55, respectively. With the increment of RHA percentage, the modulus of elasticity decreased but the Poisson’s ratio increased. The value ASR seemed to increase in the case of concrete with 10% RHA, but it did not vary significantly from the control sample when 20% RHA was used.

Based on the properties of fresh concrete and hardened concrete, it is observed that RHA-modified concrete of the current study is not suitable for typical structures such as foundation, floor, and pavement. This is mainly due to significant low strength of RHA-modified concrete even though the workability and ASR properties are favorable. The authors believe that the low strength gain in RHA-modified concrete is due to the use of significantly large RHA particles, which is observed from the gradation analysis. Concretes with such low strength can be used as controlled low strength materials (CLSM), also known as controlled density fill, flowable fill, unshrinkable fill, or lean-mix backfill (NRMCA 2000). The CLMS does not require human to compact them, and it can be used in various constructions such as utility trenches, bridge abutment, abandoned mines, and underground storage tanks.

6 Conclusions

Engineering properties of RHA-modified concrete cylinders, beams and mortar bar samples were tested in the laboratory. Significant strength reduction was observed by partial replacement of cement by RHA. A 10% replacement of cement by RHA in concrete was found to be the optimum for this study based on mechanistic properties of concrete. However, ASR test indicated that 20% cement replacement by RHA will have less ASR-related cracks. It is concluded that this RHA can be used as supplementary cementitious material in controlled low strength materials such as flowable fill. Further, it is recommended to burn RHA further using a furnace at a higher temperature or to grind it by the help of a hammer or ball mill in a controlled environment to obtain finer than 45 micros RHA.