Sciknow Publications Ltd.

Research Journal of Computation and Mathematics

Attribution 3.0 Unported (CC BY 3.0)

RJCM 2013, 1(2): 28-32 DOI: 10.12966/rjcm.07.02.2013

Stepwise Optimization Modeling of Asphalt Concrete Mix Design

Saad Issa Sarsam*
Professor of Transportation Engineering,Department of Civil Engineering, College of Engineering,University of Baghdad, Baghdad IRAQ *Corresponding author (E-mail: saadisasarsam@coeng.uobaghdad.edu.iq )

Abstract - This paper presents computerized methods for asphalt concrete job mix formula design using stepwise linear fit. The method cares for equation smoothing and also the location of the equation in the specifications area so that it is neither near the upper limits nor near the lower limits. The program find all possible equations and print the best six (e.g. 1 to e.g.6) ascended according to the sum of errors squares. Decision can be given to the computer to choose one of the six equations for economical purposes. If all the six equations are near the specification limits, the stepwise linear fit method is called for optimization and help in final suggestion of the mix design. After several months with every day execution of the program for the method; it proved that it is efficient and quick enough to be used in such mix design. Keywords - Asphalt Concrete, Job Mix Formula, Optimization, Stepwise Linear Fit

1. Introduction
The asphalt concrete mix for base, binder and wearing courses, is composed basically of aggregate, coarse, medium, and fine in sizes, mineral filler and asphalt cement. The several mineral constituents are to be sized, uniformly graded and combined in such proportions that the resulting blend (job mix formula) meets the grading requirements for the specific layer type under contract. Such combinations are considered at 100% by weight of total aggregate in the mix. The job mix formula should not vary from the lower limit of the specification requirements on the sieve to the higher limit on the adjacent sieve but should be uniformly graded. Asphalt concrete mix design equation is usually found manually using trial and error method. Although good results may be obtained, it is certainly not the best and also consumes time. Difficulties appear using computer are the high process time for finding the optimum solution and the decision factors affecting the optimality sum of which are logical. The asphalt concrete is designed to produce a stable material of maximum durability and minimum cost. Such requirements are specified in terms of a target grading and permitted range of asphalt content, stiffness and voids. Various sizes of aggregates (coarse, medium, fine, sand, and filler) are combined theoretically using trial and error method to get the final gradation (job mix formula) which must satisfy the specification requirements. It was felt that using a computer program in such analysis may give more accurate and fast results.

2. Background
The setting up of a mathematical model to combine the aggregates in order to meet a desired grading was studied by [1]. The process used is a linear programming technique. He concluded that optimum cost mixes can be achieved by using the obtained model to select and combine aggregates, filler and bitumen. The program developed can mix fine material sizes. A study presented by [2] describes the application of general curve fitting procedure for AASHO Road test which can be used for linear or non-linear models (least square residuals and minimum absolute residuals). He presented results that are obtained when the test data are summarized with models that are simpler in form and more comprehensive using a computer program than those employed in previous test report calculated by traditional methods. The possibility of using the mathematical modeling technique in combining the aggregates of different sizes to meet the specification requirements has been reported by [3]; he concluded that modeling the aggregate combination is efficient and quick for application in the mix design. The specification requirements for various asphalt concrete layers were presented by [4], it was stated that the job mix formula should not vary from the lower limit of the specification requirements on the sieve to the higher limit on the adjacent sieve but should be uniformly graded. Usually a dense gradation formula is employed in executing pavement of asphalt concrete type as illustrated in table 1. A computerized method for asphalt concrete mix design

Research Journal of Computation and Mathematics (2013) 28-32

29

using least square parabola fit technique was presented by [5]; he concluded that it is efficient and quick enough for such purpose.

3. Job Mix Formula Design

All of the gradation specification requirements for base, binder or wearing courses were stored in a file. The program first asks about the kind of specification to follow, then it gets it out from the file and tries to find an optimum job mix formula according to specific objective function and certain constrains. The objective function is to find the smoothest

curve inside the specification, while constrains are: a.) The formula curve should be inside the specification area. b.) The formula curve should be away from both extremes of the specifications (i.e. in the middle of the specification area if possible) so that the largest range of tolerance could be obtained. c.) Logical constrains from economical point of view, satisfying all about objective function and constrains, may give a job mix formula that is not economical, for example it require high percentages of filler material. The user may then ask for another alternative.

Table 1. SCRB (2004) Specification Requirements for Asphalt Concrete

Sieve size (mm) 37.5 25 19 12.5 9.5 4.75 2.0 1.0 0.6 0.425 0.250 0.180 0.125 0.075 Percent finer by weight Base course A 100-90 95-77 90-68 83-55 75-47 65-33 50-20 ----------------30-10 -------22-5 --------10-3 Base course B 100 100-87 95-80 90-70 85-65 75-50 65-33 ----------------40-17 ---------25-10 ---------10-3 Base course C 100 100 100-92 95-82 92-75 82-60 70-42 ----------------45-20 --------28-10 --------10-3 Binder course 100 100 100-88 87-65 80-55 64-37 45-23 34-17 27-13 ------20-8 ------15-6 10-5 Wearing course 100 100 100 95-75 88-65 75-50 55-32 42-24 35-18 -----25-10 ------20-8 12-6

4. Mathematical Derivation of Job Mix Formula

To get the optimum objective function, Stepwise optimization method was implemented. The method tries to keep the track Y = r. U(x) + (1 - r). L(x) There for discrete points: Yi = r i. U (xi) + (1 - r i). L (xi) (2) (1)

in the middle of the specification area. The assumed smooth function is derived from the specification limits curves. Suppose the upper specification curve is U(x) and the lower specification curve is L(x), then the job mix formula is assumed to be: n ERRK = (Y Yi) 2 i= 1 n ERRK = [ r i .U (xi) + (1-r i) . L (x i) Y i ] 2 i =1

(3)

(4)

Of course U (xi) and L (xi) values are from specification file. At the first point (x1, y1), r1 is chosen as (0.5) so that (Y) curve becomes in the middle of the specifications area. All possible combinations of points were calculated. For each set of calculated points, sum of squares errors is calculated as:

(r1) is assumed as (0.5), values of (r2) remain (0.5) if Yi is exactly at the mid distance between U (x1) and L (xi), otherwise, next (r) (i.e. r2) is updated as:

30

Saad Issa Sarsam: Stepwise Optimization Modeling of Asphalt Concrete Mix Design

r i+1 U (xi) - Yi ______ = _________ 1 r i+1 Yi - L (xi)

(5)

U (xi) - Yi r i+1 = ___________ U (xi) L (xi)

(6)

So the program proceeds updating value of (r) till the last point, and calculates ERRK for each set of data. After that, it follows the determination of best six curves, and then the best one. Although stepwise optimization method does not give absolutely smoothest curve, it tries to maintain job mix formula in the middle of the specifications area and so it verifies all constrains of the problem. At the end of this paper, one example is solved for job mix formula using stepwise

optimization method. It can be seen that it is sited in the middle of the specifications area. Figure 1 shows the six selected equation depending on the given grain size analysis, while Table 2 demonstrates the percentages of each type of mineral aggregates required for each equation as supplied by the software. On the other hand, Table 3 shows the grain size distribution for each of the six equations selected by the stepwise optimization procedure and exhibited as an output.

Figure 1. Best six job mix formulas using stepwise optimization method. Table 2. Type and Percentage of Materials Required for Each Equation
% of material Coarse aggregate Medium aggregate Fine aggregate Sand Mineral filler Equation 1 5 15 25 46 9 Equation 2 5 15 25 47 8 Equation 3 5 15 20 51 9 Equation 4 5 15 25 48 7 Equation 5 5 15 20 52 8 Equation 6 5 10 25 51 9

Research Journal of Computation and Mathematics (2013) 28-32

31

Table 3. Grain Size Distribution Output for Each Equation

Sieve size (mm) 25.0 19.0 12.5 9.5 4.75 2.0 0.475 0.18 0.75 Equation 1 97.5 90.62 82.93 75.14 61.34 48.70 26.47 14.61 9.56 Equation 2 97.5 90.62 82.93 75.14 61.33 48.55 25.84 13.73 8.76 Equation 3 97.5 90.62 82.93 76.14 64.93 52.83 28.23 15.22 9.77 Equation 4 97.5 90.62 82.93 75.14 61.32 48.40 25.21 12.86 7.95 Equation 5 97.5 90.62 82.93 76.14 64.92 52.68 27.68 14.34 8.97 Equation 6 97.5 92.43 86.96 80.11 66.29 52.95 28.32 15.22 9.77

Figure 2 illustrates the final setup of the selected equations and the design gradation. It shows that the tolerance of the J.M.F. is sited inside and within the specification requirements. On the other hand, Table 4 demonstrates the final output of the software, the gradation of asphalt concrete constituent materials are presented, their percentages which provides the optimized job mix formula are also shown.

Figure 2. Final Job mix formula design Asphalt concrete base course

32

Saad Issa Sarsam: Stepwise Optimization Modeling of Asphalt Concrete Mix Design

Table 4. Final Output Showing the Optimized Job Mix Formula

Percent finer by weight Sieve size (mm) Mineral Aggregates Material types Coarse 37.50 25.00 19.00 12.50 9.5 4.75 2.0 0.425 0.180 0.075 Percentages 100 50.2 20.5 0.9 0.0 0.0 0.0 0.0 0.0 0.0 5.0 Medium 100 100 64.0 19.2 0.8 0.0 0.0 0.0 0.0 0.0 15.0 Fine 100 100 100 100 80.1 27.2 2.4 1.8 0.2 0.2 35.0 Sand 100 100 100 100 100 99.0 85.0 37.0 12.2 4.2 48.0 Filler 100 100 100 100 100 100 100 100 100 84.8 7.0 Optimized Formula J. M. F. 100 90.0 82.3 76.1 70.1 56.4 44.2 23.4 12.2 7.7 J.M.F. Tolerance Upper 100 87.0 80.0 70.1 65.0 50.4 40.2 19.4 10.0 6.2 Lower 100 96.0 88.3 82.1 76.1 62.4 48.2 27.4 16.2 9.2 Specifications Max. 100 87.0 80.0 70/0 65.0 50.0 33.0 17.0 10.0 3.0 Min. 100 100 95.0 90.0 85.0 75.0 65.0 40.0 25.0 10.0

5. Conclusions
Based on the limited investigation, the following conclusions may be drawn: 1. The stepwise optimization method tries to keep track of mixing formula in the middle of the specifications area. 2. It provides the best six job mix formula after taking the specifications constrains into consideration. 3. It gives suggestion about the best job mix equation through optimization process, and allows the user to consider economic issues in deciding the final mix design equation. 4. The software output provides a summary of the tried equations and demonstrates the final optimized job mix formula with the required tolerance. 5. For several months and every day executing the program many times, it was proved that the stepwise optimization method is efficient and quick enough (less than 5

minutes) for five materials sizes mixing, and is recommended to be used always.

References
[1] J. Yeaman (1970) The design of asphaltic concrete using mathematical models and linear programming techniques Proceeding, ARRB Vol.5 Part 5 Australia. P. Irick (1963) Application of general curve fitting procedure to AASHO road test data HRB, Highway research record No. 42, USA. S. Sarsam (1997) The design of asphaltic concrete mix using a mathematical model Journal of Engineering and Technology Vol. 16 No. 3, University of technology, Baghdad, Iraq. S. Sarsam (2012) Computerized Optimization of Asphalt Concrete Mix Design Proceedings, 2012 International Conference on Management Sciences and Information Technology (MSIT2012), China. SCRB (2004) Standard specification for roads and bridges Ministry of Housing and construction, Iraq.

[2] [3]

[4]

[5]