40 Job Mix
40 Job Mix
40 Job Mix
ABSTRACT: This paper presents a computerized method for asphalt concrete job-mix formula design; The Excel
Solver using Linear Programming by GRG Nonlinear method was adopted in this paper. The solver can handle
problems that involve many variable cells and can help to find combinations of variables that maximize or minimize a
target cell. İt also specifies one or more constraints-conditions that must be met for the solution to be valid. It proved
that it is efficient and quick enough to be used in such mix design.
I. INTRODUCTION
Aggregates are classified as coarse, fine, and filler. The function of the coarse aggregates in contributing to the stability
of a bituminous paving mixture is largely due to interlocking and frictional resistance of adjacent particles. Similarly,
fines or sand contributes to stability failure function in filling the voids between coarse aggregates. Mineral filler is
largely visualized as a void filling agent. Crushed aggregates and sharp sands produce higher stability of the mix when
compared with gravel and rounded sands.
Gradation has a profound elect on mix performance. It might be reasonable to believe that the best gradation is one that
produces maximum density. This would involve a particle arrangement where smaller particles are packed between
larger particles, thus reducing the void space between particles. This creates more particle-to-particle contact, which in
bituminous pavements would increase stability and reduce water interaction. However, some minimum amount of void
space is necessary to:
provide adequate volume for the binder to occupy,
promote rapid drainage, and
Provide resistance to frost action for base and sub base courses.
The proper selection of different sizes of aggregates to insure better performing mix is known as aggregate blending.
There are several methods available for aggregate blending which can broadly classify into three categories: a)
Graphical Methods; b) Trial and Errors and c) Methods which involve Optimization Techniques.
The graphical methods are applied for early stage of asphalt construction and still popular among engineers due to its
simplicity and rapidity. Even these methods can be applied in the field to quickly assess the proper aggregate
proportioning. The several popular graphical methods are Triangular Chart Method, Asphalt Institute Method and
Routhfutch Method. Graphical methods are limited by number of aggregate sizes. Asphalt Institute graphical method
and the triangular chart method cannot accommodate more than two and three sizes of aggregates respectively. The
results obtained from the graphical methods are roughly accurate and cannot be directly used. Although graphical
methods can be used as an initial tool for aggregate proportioning and the solution obtained can be further optimized by
trial and error method. The use of trial and error method also become complex with increases the number of different
sizes of aggregates. Also, the trial and error methods and graphical methods cannot be used to optimize cost. The
aggregate blending problem which affects a large number of aggregates and more than one constraint cannot be solved
by trial & error or graphical methods. These complex aggregate blending problems require more accurate and
mathematical approach for acceptable results. To address these issues several optimization tools are developed by
continuous research in the field. This Paper Presents the Proportioning of Aggregates by Analytical Method i.e., Linear
Programming by GRG Nonlinear method using EXCEL SOLVER.
The concept of parabola fit using least square method was adopted by Saad Issa Sarsam. [1]. The method uses parabola
fit method to find all possible equations that combines the various material gradations as per the specification’s
requirements, the optimization process will print six formulas ascended according to the sum of errors squares. In [2]
Dr. Talal Hussien Fadhil has taken ten samples from ten different text books and papers. Each one contains three types
of aggregates; coarse, fine, and filler. The samples were solved individually by seven different methods, five by
graphical method, the sixth method was solved by using MATLAB, and the last method was solved by using Excel
sheet. In this research, it has been found that the equal distance method would be considered as an accurate, fast and
even easy method and can be used for any no of aggregates. In [3], Fouad A. Ahmed developed an approach to
determine the area bounded by irregular curves. Two simple formulas are dived to calculate the area in orthogonal and
polar coordinate system. Practical applications shoe that this technique gives higher accuracy than the conventional
method where offsets are taken at equal intercepts. In [4], Khaled A. Kandil and Al-Sobky investigated the use of fuzzy
triangular membership functions to develop a linear program model that can be used to get the optimum aggregate
blend. A model was developed to provide the optimum blend taking into consideration: design range, tolerances of job
mix formula and variability associated with the percent of each sieve. An experimental investigation was developed to
evaluate the variability associated with the percent of each sieve to be taken into formulated and the main factors
affecting this process were discussed.
III. METHODOLOGY
This Paper Develops a generalization of bituminous mix design for BC Grade-1 as per MORT&H.
Aggregate Blending Model: The aggregate blending can be performed through number of existing popular methods
like trial and error and graphical methods. İn these methods the selection of aggregate blending is becoming complex as
the more sizes of aggregates may vary. The main drawbacks of this process are that a number of trials are required for
the selection of different types of aggregate proportions to meet the required gradation range is having a lot of
possibilities. İt is time consuming process to achieve all the requirements of aggregate blending. The Excel Solver can
deal with this kind of problem. As the solver can handle problems that involve many variable cells and can help to find
combinations of variables that maximize or minimize a target cell. İt also specifies one or more constraints-conditions
that must be met for the solution to be valid. The solution is obtained by a set of equations considering the lower and
upper limits of the required gradation as well as the percentage of passing of each type of aggregate.
Let x1, x2, x3, x4 represent the different sizes of aggregates used for this mix.
The Equation of the form ax1+b x2+c x3+d x4≤ Pl or Pu can be written for each sieve size, where a, b, c and d passing
for that sieve size and Pl and Pu are the upper and lower gradation for that sieve size as per MORT&H specifications.
Solving the above system of equations manually is extremely difficult so, good computer programs are required to
solve this. Software like solver in excel yields the solution.
The aggregate gradation of BC Grade – I is calculated by analytic method is tabulated below in (figure3)
The results of mix design of BC Grade – I and the solution of gradation is:
x1 = 21.1694%
x2 = 17.8514%
x3 = 24.6193%
x4 = 36.3599%
V. CONCLUSION
The proportioning of aggregate is responsible for good packing and densities. By controlling the aggregate
gradation one can improve the mixture performance like resistance to rutting and fatigue.
The aggregate blending can be performed through numbers of existing method ranging from trial and errors
to more complex computation techniques.
The popular methods of aggregate blending involve graphical method, least square method, nonlinear
programming, Stimulated Annealing techniques and genetic algorithm etc.
In the present scenario every contractor is interested in the cost-effective aggregate blend. The old methods
are acceptable for rough use or to provide initial solution. But in order to obtain a cost-effective blend with
satisfactory specification requirement one needs to accommodate more sizes of aggregates. These many
constraints cannot be optimized with traditional method. Hence the more accurate and capable tools which
can resolve these problems by using Excel solver as discussed in this paper to optimize aggregate blending.
ACKNOWLEDGEMENT
I would like to express my special thanks and gratitude to my guide Prof. V. S. Korishetti for his continuous support
in our work and also thank full for my parents and my friends who have supported and encouraged.
REFERENCES
1. Khaled A. Kandil and Al-Sayed A. Al-Sobky. Aggregate Blending Model for Hot-Mix Asphalt Using Linear
Optimization (2013).
2. Y. Cengiz Toklu: Aggregate Blending Using Genetic Algorithms (2005)
3. Saad Issa Sarsam: Modeling Asphalt Concrete Mix Design Using Least Square Optimization (2015)
4. Mario T. Tabucanon, Pakorn Adulbhan and Stephen S. Y. Chen: A probabilistic programming model for blending
aggregates (1979).
5. Said M. Easa and Emre K.can: Optimization model for aggregate blending (1985).
6. Priyansh Singh and Gurpreet Singh Walia: Review of Optimization Methods for Aggregate Blending (2014).
7. Tom V. Mathew and K V Krishna Rao: Dry mix design NPTEL- IIT Bombay.
8. B. Pollington: Using Excel to solve linear programming problems.
9. Ahmed, F.A: Area computation using salient boundary points (1983).
10. Ismat M. EI Hassan: Irregular area computation (1987).
11. Tumu Swapna: Dissertion on Modelling of Bituminous mixes, JNTU Hyderabad (2013)