Prediction Model of Elastic Modulus for Granular Road Bases

The estimation of elastic modulus for road bases is the primary objective of this research which is implemented a significant role in transmitting the vertical loading to the pavement foundation layers. In this study, the effect of weathering conditions on the stiffness of base course is investigated and implied the durability test by subjecting the prepared samples to a different numbers of wet-dry cycles (0,2, 4, 6, 8 and 10). A conventional base materials of local natural gravel aggregate and treated base materials with recycled concrete aggregate RCA at different percentages (0%, 25%, 50% 75% and 100%) is adopted in this research. The elastic characteristics are estimated in terms of elastic modulus. Elastic modulus are estimated by passing the ultrasonic pulse velocity through the untreated and treated base materials laboratory specimens. This test can be used to study the elastic modulus properties of base materials. A multiple linear regression analysis is used for prediction the elastic modulus using the SPSS (software ver.21). Elastic Modulus (kPa) is the dependent variable whereas the independent variable are; No. of wetdry cycle and Percent (%) of RCA stabilizer. The obtained results for elastic modulus (Es) of granular base material layer showed increasing in elastic modulus with percentage of RCA%., results revealed that the (Es) values reached a maximum value of (6927kPa) for 100%. For the OMC’s values increases due to the percentage increment of RCA in granular base material mixture, this increment in water contents is refer to high absorption capacity of the paste clinging to the RCA. On other side the dry density decrease gradually with adding percentage of (RCA) in granular base material mixture.


Introduction
The behaviour and performance of pavements against distresses rutting has significantly influence by the design standard (including the estimation of resilient elastic modulus) and construction of the granular base layer. Achieving the long-life of pavements is depending on the control of granular base layer construction quality (Titi, et.al 2018). Also the growth in axle loads and number of vehicles in industrial and developing countries has make road agencies to modify their design standards and to adopt designs that provide roads with higher load bearing capacity. Also Titi, and Matar (2018) predicted a model of the resilient modulus of base granular layer using repeated triaxial test using gravel /crushed grave and crushed stone (Titi and Matar, 2018).
The elastic modulus of granular pavement layers is considered as the most important parameters in pavement design procedures. Optimum pavement design requires to determine the appropriate stiffness value for each pavement layer.
Stabilize and improve the engineering behaviour of granular pavement layers is considered as important in the desired engineering properties of compacted granular layers construction, not only its moisture content and dry density. Various studies have been achieved over the years with respect to engineering properties of granular base materials and the effect of (wet-dry cycles) on such properties. These preceding studies concentrated on mechanical strength properties and how such these properties are affected by weathering impact (Biswal, et.al. 2018).

Experimental Work
The experimental work of this study includes the laboratory tests for (30) samples of granular base material prepared from locally Al-Nibaee aggregate material treated with crushed concrete waste material; recycled concrete aggregate (RCA). Also the durability test for several wet-dry cycles; 0, 2, 4, 6,8, 10 and 12 were carried out.

Materials
The materials used are locally available in Iraq taken from the hot mix plant of Al-Nibaee quarry at Al-Taji used in road paving works. The aggregate used is crushed aggregate and Recycled concrete aggregate (RCA) from stocks of construction waste materials.  (1) and (2) respectively.  The fine aggregate refers to a combination of natural sand (river sand). The physical composition and chemical properties of the fine aggregate are shown in Tables (1) and (2)    RCA tends to be very angular and rough due to the crushing of the virgin aggregate particles and the presence of cement paste that continues to cling to the surfaces of the aggregate. It has higher absorption capacity of water than natural aggregate due to the porous nature of the cement paste portion of the recycled aggregates so at equal water to cement ratio RCA concrete gave lower compressive strength than control concrete made from natural aggregate as well as it has lower elastic modulus than control and similar flexural strength (Alkaissi, et.al, 2016). The Specific Gravity of RCA is lower than natural aggregate due to the crushed mortar present in and on the aggregate particles which make it less dense (low dry density) than NA because of its porosity and entrained air voids. The loss for RCA in L.A. Abrasion Mass Loss is usually higher than NA (Alkaissi, et.al, 2016). In terms of durability properties, RCA concrete has high chloride contents (Sharma and Singla, 2014) that may effect on the durability of the new concrete and the corrosion of steel in new concrete.

Sample Preparation
1) Determination of relative proportions of cement and the aggregate as well as percent of each aggregate size fractions involving filler.
3) After the bulk samples are sieved, a particle-size distribution is established for each material that facilitated reconstruction of replicate samples with identical gradations.

Preparation of Aggregate
The aggregates and filler for conventional materials and non-conventional materials (recycled concrete aggregate (RCA)) are prepared using the same procedure. The weight of each aggregate size and filler is determined by multiplying its percent by the desired weight of final mix. The use of RCA is about (25, 50, 75 and 100) percentage replaced from coarse aggregate.

 The Process of Mixing
The total weight of the batch was approximately 5500 gm to produce a testing sample. (9) samples with different moisture contents are prepared for moisture-density testing of untreated materials aggregate treated with (RCA) materials. Just before compaction, the dry fine fraction, passing the No. 4 sieve and mineral filler, was added to the coarse fraction to get more appropriate mix. The combined material is then mixed until it becomes uniform in color and texture. In case of treated material with cement, the cement is added to the fine fraction and then mixed until it reaching uniform mix; finally it mixed with the coarse fraction. The final mixture has been compacted into the mold using modified Proctor compaction effort using a mechanical compactor in accordance with (ASTM D698 -12). The test machines are shown in plate (2). The modified Proctor procedure requires compaction of the samples in five layers; each lift consists of 56 blows. The specimen is subsequently dried to constant weight in an oven at (105-110) °C for 24 hours to facilitate calculation of gravimetric moisture content and dry density. The moisture content determinations are performed in accordance with (ASTM D 2216 -05). These values were plotted to determine the optimum moisture content (OMC) and maximum dry density (MDD), see Plate (3).

Wet-Dry Test
Wet-dry cycle test, is one of durability tests that adopted in this study used to investigate the weathering conditions effects on the stiffness of the base course with selected percentage value of (0%, 25%, 50%, 75% and 100%) RCA.
Six Samples of aggregate material are prepared for 12 cycles. These samples are compacted as described above at OMCѕ for each ratio and for each compaction method, extruded from mould and cured at moist room for 7 days. After the 7days cure, all samples are submerged in potable water at room temperature for a period of (5 hours). Following the period of immersion, samples are removed from water and the mass of each sample is weighted and recorded. Then, each sample is placed in an oven at 71°C (160°F) for (42 hours) and removed; the mass is then weighed and recorded. This process of wetting and drying constitutes one cycle (48 hour). Plate (4) depicts the wetting and drying configurations for samples of treated Al-Nibaee aggregate materials for wet-dry test, as prescribed in ASTM D 559-96.

Ultrasonic Test
Ultrasonic test is a non destructive testing which has been utilized for characterization of granular base materials. It is used to measure the travel time of an ultrasonic pulse wave passing through the prepared specimens. The elastic characteristics are estimated in terms of elastic modulus. Elastic modulus are estimated by passing the ultrasonic pulse velocity through the untreated and treated base materials. This test can be used to study the stiffness of elastic modulus properties of base materials. The test is used to find the elastic modulus from the recording transit time ASTM (C 597 -02).

Optimum Moisture Content (OMC) and Maximum Dry Density (MDD)
The values of OMC and MDD for untreated and RCA treated granular base samples were depicted in Table (6) and Figure (2) illustrated the relation of water content with maximum dry density for untreated granular base material. With appropriate fitting the polynomial curve in Figure (2), the values of maximum dry density and optimum water content % were as follows:  25% RCA provided 6.20% OMC and 2.26 (gm/cm3) MDD.
 100% RCA provided 7.10% OMC and 2.19 (gm/cm3) MDD.    Vol. 2, No. 1;2020 increment in water content is refer to high percent of fines and the free water get absorbed by the residual cement exist in recycled concrete crushed aggregate. On the other hand the dry density decreased gradually with RCA percentage increment in granular base material mixture, the reason is that the RCA have a lower relative density than that of virgin aggregate as well as the lightweight of adhered cement mortar in RCA can cause a weak layer which decreased the density.

Effect of RCA on Elastic Modulus (E s )
The change in moisture and temperature occurs within road's base layers due to seasonal variation and change in water table level during the pavement design life. The wet-dry conditions may significantly affect performance of pavement. The materials used in road construction should have the ability to provide and enhanced the resistance stability and integrity over years of exposure to the damaging impact of weathering, which is one of most important factors that should be considered in the design life of pavement.
To calculate the modulus of elasticity by using velocity of ultrasonic waves, the equation used in this study illustrated below, test according to ASTM (C 597 -02) (British Standards Institution, 1988).

Statistical Analysis
Analyze the results obtained from previous experimental tests; (30) samples for wet dry cycle (durability test) with different RCA % and ultrasonic test for elastic modulus as described previously, a statistical process is adopted to infer the meaning about the underlying data results and try to describe the true behaviour and provide summaries about the experimental observation that have been made. The central of tendency and variability (spread) are estimated for the dependant variable (elastic modulus) which include the median, mode, mean, standard error of mean, standard deviations, variance, minimum and maximum variables and the Skewness and Kurtosis. Table (5) tested the resulted values of (E) for (30) sample size since the distribution is skewed -right (positive skew of (1.687 ) which indicated that the mean value of (5436.1333 kPa) was greater than the median of (3849 kPa) and this can be seen clearly in the residual histogram of elastic modulus model; Figure (5 To obtain a best fit model at level of confidence (95%), the independent variables are entered in a stepwise regression method.  Table ( 11). The predicted model summary presented in Table (9) with R 2 of 0.682. The regression coefficients obtained as presented in Table (10) indicated that RCA% in base material increased the elastic modulus and modulus decreasing with wet-dry cycle.  Also the error distribution mean value for the dependent variable elastic modulus and the P-P plot of regression standardized residual is depicted in Figure (5), (6) and (7).It can be seen that the individual data approximately fit the line which indicated the outcome variables could reveal unexplained patterns in the elastic modulus data by the fitted model. And the bivariate coefficients of correlation which determines the relationship between each of the predictor variable (Elastic Modulus and the dependent variables; percent of stabilizer (RCA) and number of Wet-Dry cycles are shown in Table (12).

Model Validation
A best fit is presented in Figure (8) for the predicted elastic modulus model with the obtained experimental results for the model of granular base materials. For building a general model with good performance a 75% data and 25% splitting strategy has provided and adopted extra experimental results from previous research (Alkaissi, et. al. 2016). The goodness of fit for predicted model and experimental estimated data for elastic modulus have been checked using chi-square test as shown below so the developed model is considered to be valid.

Conclusions
From the acquired of this research, the following conclusions can be accomplished: 1. The values of maximum dry density and optimum water content % were as follows:  25% RCA provided 6.20% OMC and 2.26 (gm/cm3) MDD.