You are on page 1of 6

OPTIMAL COST ANALYSIS OF REINFORCED EARTH WALLS

P. K. Basudhar
Professor, Department of Civil Engineering, Indian Institute of Technology Kanpur, Kanpur-208016, India

Amol Vashistha
Former M. Tech Student, Department of Civil Engineering, Indian Institute of Technology Kanpur, Kanpur-208016

Kousik Deb and Arindam Dey


Research Scholar, Department of Civil Engineering, Indian Institute of Technology Kanpur, Kanpur-208016, India

Sourav De
M. Tech Student, Department of Civil Engineering, Indian Institute of Technology Kanpur, Kanpur-208016, India

Abstract
This paper deals with optimal cost design of geosynthetic reinforced earth retaining walls. Choosing the length and strength of the reinforcement as a feasible design vector the Sequential Unconstrained Minimization Technique (SUMT) has been used to arrive at an optimal cost of the reinforced earth wall. Optimal cost tables are presented for different combinations of the loading and the developed procedure is validated by taking up an example problem. It has been found from the typical example problem that saving of the order of 7-8% can be made over the standard design of mechanically stabilized earth (MSE) walls with the aid of design charts. Key words: Mechanically stabilized earth (MSE), geosynthetic reinforced earth retaining wall, optimal cost design, Sequential Unconstrained Minimization technique.

INTRODUCTION
Retaining walls as earth structures are frequently constructed for a variety of applications, most common being bridge abutments and road construction. When selecting a retaining wall type, mechanically stabilized earth (MSE) walls should always be considered. MSE walls are basically composed of some type of reinforcing elements, e.g. geosynthetics in the soil fill to resist lateral earth pressures. The use of geogrids or geotextiles rather than metallic strips (ties) is a further development of the Reinforced earth concept. Geosynthetics offer a variable and often economical alternative to metallic reinforcements for both permanent and temporary walls, especially under certain environmental conditions. The reinforced earth retaining walls are very cost effective than conventional concrete retaining walls too. Furthermore, these systems are more flexible than the conventional earth retaining walls such as reinforced concrete-cantilever or gravity walls. Therefore, they are suitable for sites with poor foundations and seismically active areas. There are lot of studies have been undertaken to develop methodologies for the analysis reinforced earth walls (Broms, 1978; Bonoparte et al., 1987; Leshchinsky and Perry, 1987; Schmertmann et al., 1990). However, no work has been undertaken to develop methods for their optimum design. Such an attempt has been made in this paper.

STATEMENT OF THE PROBLEM


Figure 1 shows the cross-section of a reinforced earth retaining wall. The main item is the soil backfill which is placed as the wall is constructed. For the problem considered herein, the backfill is limited to cohesion less, free

draining material, i.e. sand; having density and angle of internal friction . The backfill is strengthened by reinforcing members, which in the present case are geotextiles and geogrids. The facings, additionally, enhance the aesthetic appeal of the completed wall. For geogrid wall, Modular Concrete Unit (MCU) is the most common facing used. They are popular because of their aesthetic appeal, easy availability and relatively low cost. Geotextiles are more commonly used in warp-around-faced MES walls. Such facing are used for temporary structures, for walls that are subjected to significant post-construction settlements and where aesthetic requirements are low. Given the wall geometry, the soil properties and the reinforcement properties, the problem is to define the length and strength of the reinforcement that gives the least cost estimation for construction.

Figure 1: Cross-section of MSE wall

ANALYSIS
Optimal cost design retaining walls when reinforced with either geogrids or geotextiles is carried out. Standard design procedures developed by Federal Highway Authority (FHWA) have been adopted. This particular analysis procedure is based on the Rankine Approach. The same is cast in the frame of optimal design procedure in the following sections. Design Guidelines Input Parameters The input parameters for the optional design procedures are presented in Table 1 due to the lack of available manufacturers data, the values of the allowable tensile strength of the geosynthetic are interpolated. Design Variable The design vectors chosen for the formulation are related to the reinforcement. They are: Reinforcement;

D (1) = l = Length of the

D ( 2) = Ta = Long-term allowable strength of the reinforcement.

Objective Function The objective function is the cost estimate for the proposed design of the geosynthetic reinforced retaining wall. The costs involved per meter length of the wall are as follows:

Cost of leveling Pad = Cost of the wall fill =

c1
c2 f / g hd l

Cost of the geosynthetic used = MCU face units cost =

c3 nl l

c 4 hd (only for Geogrid Wall)

Engineering and testing cost = Installation cost =

c 5 hd

c 6 hd

The objective function is the sum of all the above costs and, thus, is a function of the length and strength of the reinforcement which are the design variables.

D T = ( d1 , d 2 )
The objective function is minimized with respect to the design vector

F = f ( D ) = f ( d1 , d 2 )
Table 1: Input Parameters for Design of MSE Wall
Parameter Height of wall Minimum embankment depth of wall Angle of internal friction of the fill Unit weight of the fill Angle of internal friction of the backfill Unit weight of the backfill Allowable tensile strength of the geosynthetic Minimum spacing between the reinforcements Minimum embedment length of the reinforcement Design Factor of Safety: Overturning Sliding Bearing capacity Reinforcement strength Symbol H hdmin Value 3 10m 0.45m 35 3 20kN/m 30 3 18kN/m 30 60 kN/m 0.2m 1.0m 2.0 1.5 2.0 1.5 2.0 $10/m $3/1000 kg $(Ta(0.03)+2.6)/m 2 $(Ta(0.03)+2.0)/m 2 $60/m $30/m 2 $10/m 2 $50/m
2 2

f f b b Ta smin lemin

Reinforcement pullout Cost Factors: Leveling pad Wall fill Cost of Geotextile Cost of Geogrid
Cost of MCU face unit Engineering and testing cost: Geotextile wall Geogrid wall Installation cost

FSovd FSsld FSbcd FSstd FSpld c1 c2 c3gx c3gd c4 c5gx c5gd c6

Design Constraints The various design constants that are to be placed on the choice of design variables and other design parameters were as follows: The initial length of the reinforcement chosen for design is to be greater than the critical length.

g j1 = d min d1 0
The initial allowable strength of the geosynthetic is chosen to be greater than the minimum specified.

g j 2 = Ta min d 2 0
The allowable strength of the geosynthetic is not to exceed the maximum available.

g j 3 = d 2 Ta max 0
The factor of safety for overturning is to be greater than design factor of safety`1

g j 4 = FS ovd FS over 0
The factor of safety for sliding is to be greater than the corresponding design factor of safety

g j 5 = FS sld FS slid 0
The factor of safety for bearing capacity is to be greater than the corresponding design factor of safety

g j 6 = FS bcd FS bear 0
The spacing between the reinforcement is to be greater than the recommended the minimum spacing

g j 7 = s min s 0
At each level of reinforcement, the factor of safety against pullout is greater than the design factor of safety

g j 8 = FS pld FS pull 0
The problem could be stated as an optimization problem as follows: Find the decision vector Dm , such that

F = f ( Dm ) is to be minimum of F (D ) subject to g j ( Dm ) 0

j = 1,2,, n

The unconstrained minimization problem is converted to a constrained minimization problem by the formation of a composite function as

( D, rk ) = F ( D) rk
j =1

1 , where, r = interior penalty parameter k g j ( D)

The SUMT (Fox, 1971) was used in which interior penalty function methods was employed in conjunction with Powells multidimensional search and quadratic interpolation for minimizing steps. The interior penalty function

approach needs a feasible starting point for initiating the solution. Such a starting point was easy to get. Results are obtained thus for a feasible initial design vector using interior penalty function approach.

RESULTS AND DISCUSSIONS A computer program based on the formulation as described above has been developed and solutions are obtained using an iterative technique with a tolerance value of 10-3.The effectiveness of the developed
methodology for the optimum design has been demonstrated with a design example problem. Example Problem Wall description The wall is about 200m long and exposed wall height is 5m (H). Surcharge and seismic effect are ignored. The reinforced fill is imported to the site at a cost of $3 per 1000 kg. An effective angle of internal friction,f , of 34 and unit weight of 20 kN/m3 is assumed for the fill. The retained backfill is assumed to have b = 30 and b = 18kN/m3. Calculations The active earth pressure coefficient values are: Kai = 0.283 (for internal stability), Kae = 0.33 (for external stability) Design Height, hd = H + 0.45 = 5.45m Assume l / h ratio of 0.7 so, l = 3.815 4m External stability has been checked and values are found as: FSover = 5.11; FSslid = 1.91; and FSbear = 8.62. All these values are above the design factor of safety values, so the wall fulfills external stability requirements. The lateral load to be resisted by the geogrid is:

1 1 2 K ai f hd = (0.283)(20)(5.45) 2 = 84.06 kN/m 2 2


Assuming that a geogrid with along-term design strength of 20 kN/m is chosen, the safe design strength of the geogrid is given by:

Td = Ta / 1.5 = 13.33 kN/m


The approximate number of layers of the geogrid required is: 84.06/13.33 = 6.2. This number is approximated to 8 to account for practical layout consideration and 5:1 slope. Spacing of the layers is sp = 5/8 = 0.625m Cost Estimate 1. 2. 3. 4. 5. 6. Cost of the leveling pad = (200m)($10/m) = $2000 Cost of the reinforced wall fill = (200m)(5.45m)(4m)(20kN/m )/(9.8)($3)/(1000 kg) = $26670 Cost of geogrid soil reinforcement = (8 layer)(4m)(200m)($5/m2) = $32000 Cost of the MCU face units = (200m)(5.45m)($60/m ) = $65400 Engineering and testing cost = (200m)(5.45m)($10/m ) = $10900 Installation cost = (200m)(5.45m)($50/m ) = $54500
2 2 2 3

The total cost estimate = $191470 Cost estimate from optimization program = $176536 Total net saving = $14934 and the percentage saving = 7.79%

Optimum Cost Tables Typical optimum cost tables for different wall heights (3 to 10m) are presented for design parameter as specified in Table 1. The walls are designed for different loading conditions. The results are presented in Tables 2 and 3 for both Geotextile-Warp wall and Geogrid wall. The tables show that optimum costs increase for increasing wall heights and loading intensity. The optimum design vector for a particular combination of loading and wall height is almost the same for both the types of walls since they have essentially the same design procedure. The geogrid type wall shows more costs because of the additional cost of the MCU face units. The general observation is that for the same height of the wall and the loading condition, the geogrid wall shows higher costs in the range of 20 30% for low wall heights (3m to 6m), while the increase is 10 20% for height greater than 6m. Table 2: Optimum Cost Table for Geotextile-Warp Wall, qs = 20kN/m
Ht. (m) 3 4 5 6 7 8 9 10 l (m) 2.72 3.37 4.01 4.64 5.27 5.90 6.53 7.18

Ta
( kN/m) 30.00 30.00 41.26 44.31 40.88 44.87 48.87 48.50

Fill Vol. 9.39 14.99 21.84 29.94 39.28 49.85 61.67 75.40

Geotextile 2 area (m /m) 17.88 25.85 30.04 39.85 57.18 70.10 84.26 108.18

nl
4 5 5 6 8 9 10 12

Spacing (m) 0.75 0.8 1.0 1.0 0.88 0.89 0.9 0.83

Cost 2 ($/m ) 103.79 119.78 124.23 140.95 169.66 186.54 203.54 232.72

Table 3: Optimum Cost Table for Geogrid Wall, qs = 20kN/m


Ht. (m) 3 4 5 6 7 8 9 10 l (m) 2.72 3.37 4.01 4.64 5.27 5.90 6.53 7.18

Ta
( kN/m) 30.00 30.00 41.26 44.31 40.88 44.87 48.87 48.50

Fill Vol. 9.39 14.99 21.84 29.94 39.28 49.85 61.67 75.40

Geogrid 2 area (m /m) 10.88 16.85 20.04 27.85 42.18 53.10 65.26 86.18

nl
4 5 5 6 8 9 10 12

Spacing (m) 0.75 0.8 1.0 1.0 0.88 0.89 0.9 0.83

Cost 2 ($/m ) 129.57 144.89 149.74 166.23 194.22 211.23 228.49 257.52

CONCLUSIONS
Optimal cost tables are presented for different wall heights and design parameters for both Geotextile Warp wall and Geogrid wall. The geogrid walls show greater optimal costs because of the inherent design involving MCU face units. The saving is of the order of 7 8% for the optimum design procedure developed here over standard design.

REFERENCES
Bonaparte, R., Holtz, R. D., and Giroud, J. P. (1987). Soil reinforcement design using geotextiles and geogrids. Geotextile testing and the design engineer, STP 952, J. E. Fluet Jr., ed., ASTM, Philadelphia, 69116. Broms, B. B. (1978). Design of fabric reinforced retaining structures.Proc., Symp. on Earth Reinforcement, ASCE, New York, 282304. Fox, R.L. (1971). Optimization Methods for Engineering Design, Addison-Wesley, Reading, MA. Leshchinsky, D., and Perry, E. B. (1987). A design procedure for geotextile-reinforced walls. Geosynthetics 87, Industrial Fabrics Association International, St. Paul, Minn. 95107. Schmertmann, G. R., Chouery-Curtis, V. E., and Johnson, R. D. (1990).Design charts for geogrid-reinforced soil slopes. Geosynthetics 87, Industrial Fabrics Association International, St. Paul, Minn., 108120.

You might also like