Design and Construction Guidelines for Geosynthetic-Reinforced Soil Bridge Abutments with a Flexible Facing (2006)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

A rational design method and construction guidelines for geosynthetic-reinforced soil (GRS) bridge abutments with a flexible facing have been developed in this study. The design method and construction guidelines were established based on the findings of (1) a literature study, (2) full-scale experiments, and (3) an analytical study. Each study and the major findings are summarized below. Literature Study The literature study included reviewing and synthesizing the measured behavior and lessons learned from case histories of GRS bridge-supporting structures with flexible facings, including six in-service GRS bridge abutments and six full-scale field experi- ments of GRS bridge piers and abutments. The following were observed about the per- formance, design, and construction of the GRS bridge-supporting structures: • GRS bridge abutments with a flexible facing are a viable alternative to conventional bridge abutments. All six in-service GRS abutments performed satisfactorily under service loads. The maximum settlements and maximum lateral displacements for all the abutments met the tolerable movement criteria based on experience with real bridges—100 mm (4 in.) for settlement and 50 mm (2 in.) for lateral displacement. • With a well-graded and well-compacted granular backfill and closely spaced rein- forcement (e.g., 0.2 m of vertical spacing), the load-carrying capacity of a GRS bridge-supporting structure can be as high as 900 kPa. The load-carrying capacity will be significantly smaller (down to 120 to 140 kPa) when the backfill is of a low density or the reinforcement is not of sufficient length or strength. • With a well-graded and well-compacted granular backfill, the maximum settle- ment of the loading slab (sill) and the maximum lateral movement of the wall face can be very small under service loads. With a lower quality backfill, the move- ments will be significantly larger. • Fill placement density plays a major role in the performance of the GRS structures. • Preloading can significantly reduce post-construction settlement of a GRS abut- ment by a factor of 2 to 6, depending on the initial placement density. If there is a SUMMARY DESIGN AND CONSTRUCTION GUIDELINES FOR GEOSYNTHETIC-REINFORCED SOIL BRIDGE ABUTMENTS WITH A FLEXIBLE FACING

significant difference in the height of the two abutments supporting a bridge, preloading is an effective way to reduce differential settlement. • With a well-graded and well-compacted granular backfill, long-term creep under service loads can be negligibly small. • For all the bridge-supporting structures, the maximum tensile strains in the rein- forcement were in the range of 0.1 percent to 1.6 percent under service loads, with larger maximum strains being associated with lower strength backfill. • Reinforcement length and reinforcement type appear to have only a secondary effect on the performance characteristics. • The “sill clear distance” (i.e., the distance between front edge of sill and back face of wall facing) used in the cases varies widely, from 0.2 m to 2.2 m. A small sill clear distance can only be used with well-compacted backfill, especially near the wall face. A larger sill clear distance may result in a longer bridge deck (thus higher costs) and may compromise stability if the reinforcement is not of sufficient length. Full-Scale Experiments The full-scale experiments, referred to as the NCHRP test abutments, consisted of two test sections: the Amoco test section and the Mirafi test section, in a back-to-back configuration. The two test sections differed only in the type of reinforcement. The main features of the test sections were as follows: • Abutment height: 4.65 m (15.25 ft) • Sill: 0.9 m  4.5 m (3 ft x 15 ft) • Clear distance: 0.15 m (6 in.) • Reinforcement type: Amoco test section: Amoco 2044 (Tult  70 kN/m) Mirafi test section: Mirafi 500x (Tult  21 kN/m) • Reinforcement length: 3.5 m (10 ft) • Reinforcement spacing: 0.2 m (8 in.) • Facing: Cinder blocks, without mechanical connection be- tween blocks • Backfill: A non-plastic silty sand (SP-SM), with internal friction angle  34.8° from standard direct shear tests; field placement density  100 percent of T-99. • Loading: Vertical loads applied to sill in 50 kPa increments. The loading was terminated at 814 kPa for the Amoco test section (as the loading rams reached their maximum extension) and 414 kPa for the Mirafi test section (because of “excessive” deformation). The measured performance and observed behavior of the NCHRP test abutments are presented in Chapter 2. Highlights of the measured performance and observed behavior follow. Load-Carrying Capacity and Ductility • As the loading was being terminated, 814 kPa applied pressure for the Amoco test section and 414 kPa for the Mirafi test section, the Mirafi test section had approached a bearing failure condition while the Amoco test section appeared to 2

3be still sufficiently stable. For a typical design pressure of 200 kPa, the safety mar- gin in terms of load-carrying capacity is considered “acceptable” for the Amoco test section and “marginally acceptable” for the Mirafi test section. • With a sufficiently strong reinforcement, the Amoco test section was still exhibit- ing a near-linear load-settlement relationship at 814 kPa, about four times the typ- ical design pressure of 200 kPa, although the deformation had become fairly large—a clear indication of a high ductility of the abutment system. With a weak reinforcement, however, the ductility was significantly compromised. As the applied pressure increased beyond 200 kPa, the rate of settlement continued to increase with increasing applied pressure. Sill Settlement and Angular Distortion • The settlement of the sill in the Mirafi test section was about 80 percent greater than in the Amoco test section under an applied pressure of 200 kPa. As the load- ing was terminated, 814 kPa for the Amoco test section and 414 kPa for the Mirafi test section, the average sill settlement of the two test sections was comparable. • The sill settlement of the Amoco test section under 200 kPa met the typical settle- ment criterion of 100 mm (4 in.); while in the Mirafi test section, the sill settlement was likely to compromise ride quality, but would still be considered “tolerable.” • For an 18-m (60-ft)-long single-span bridge, the “maximum possible” angular dis- tortions for both test sections were below the typical acceptable criterion of 1:200. Lateral Movement of Abutment Wall • For both test sections, the abutment wall moved outward with the maximum movement occurring near the top of the wall. The maximum lateral displacements at 200 kPa were somewhat below the typical maximum lateral movement crite- rion of 50 mm (2 in.). Contact Pressure on Foundation Level • The measured contact pressure was the highest beneath the wall face and decreased nearly linearly with the distance from the wall face. • The 2V:1H pyramidal distribution method suggested by the NHI manual yielded approximately the “average” measured contact pressure, although the calculated pressures were somewhat higher than the measured average values at higher applied pressures. Observed Behavior • A tension crack was observed on the wall crest in both load test sections. The tension crack was first observed near where the reinforcement ended at an applied pressure of 150 to 200 kPa. The location of the tension crack suggested that the assumption of a rigid reinforced soil mass in the existing design methods for evaluating external stability is a sound assumption. If an upper wall had been constructed over the test abutment, as in the case of typical bridge abutments, the tension crack would not have been visible and would perhaps be less likely to occur.

• Under higher applied loads, the facing blocks in the top three courses were pushed outward as the sill tilted counter-clockwise toward the wall face. This suggests that (1) a sill clear distance of 0.15 m, a minimum value stipulated by the NHI manual, may be too small, and (2) it may be beneficial to increase the connection strength in the top three to four courses of the facing during construction. Analytical Study The Analytical Tool A finite element software program, DYNA3D, written by Hallquist and Whirley in 1989 (along with its PC version, LS-DYNA), was selected as the analytical tool for the study. The capability of DYNA3D/LS-DYNA for analyzing the performance of segmental fac- ing GRS bridge abutments was evaluated critically. The evaluation included comparing the analytical results with measured data of five well-instrumented full-scale experi- ments. The comparisons are presented in NCHRP Web-Only Document 81. Very good agreements between the analytical results and all the measured quantities (including fail- ure loads, when applicable) were obtained. Parametric Analysis An extensive parametric analysis was conducted by using the analytical tool. For the parametric analysis, a bridge abutment configuration and a set of material properties were selected as the base case. The performance characteristics of the base case as affected by soil density (i.e., soil stiffness and friction angle), reinforcement stiffness, reinforcement spacing, reinforcement truncation, and the sill clear distance were inves- tigated. The results of the performance analysis (presented fully in Chapter 2) are sum- marized below. • Effect of reinforcement spacing: For a fill with φ  34°, the effect of reinforcement spacing on the abutment performance was very small when the applied pressure was less than 100 kPa. At 200 kPa, there was a 25 percent increase in sill settlement as the spacing increased from 0.2 m to 0.4 m, and another 25 percent increase as the spacing increased from 0.4 m to 0.6 m. The increase in settlement because of an increase in reinforcement spacing was higher at 400 kPa than at 200 kPa. • Effect of soil stiffness and strength: At reinforcement spacing of 0.2 m, the sill settlement of an abutment with 34° soil friction angle was 23 percent and 35 per- cent higher than those with 37° and 40° friction angles, respectively. The effect was more pronounced for reinforcement spacing of 0.4 m. The performance of an abutment with φ  34° and reinforcement spacing  0.2 m was very similar to an abutment with φ  37° and reinforcement spacing  0.4 m, suggesting that more closely spaced reinforcement had an effect similar to denser fill compaction. • Effect of reinforcement stiffness: For a soil with φ  34° and reinforcement spacing  0.2 m, the sill settlement decreased 43 percent at 200 kPa as the rein- forcement stiffness increased from 530 kN/m to 5,300 kN/m. The sill settlement increased about 250 percent as the reinforcement stiffness decreased from 530 kN/m to 53 kN/m. • Effect of sill clear distance: The sill settlement increased about 20 percent when the sill clear distance increased from 0 to 0.15 m and increased another 10 percent when the sill clear distance increased from 0.15 to 0.3 m. 4

5• Effect of truncated reinforcement at base: For a soil with φ  34° and reinforce- ment spacing  0.4 m, the effect was insignificant. Load-Carrying Capacity Analysis To determine the allowable sill pressures for the recommended design method, 72 additional analyses were performed using the LS-DYNA code. The effects of the fol- lowing parameters were investigated: sill type (integrated sill and isolated sill), sill width (0.8 m, 1.5 m, and 2.5 m), reinforcement spacing (0.2 m and 0.4 m), soil friction angle (34°, 37°, and 40°), and foundation (6-m thick medium sand foundation and rigid foundation). The results of the load-carrying analysis (presented more fully in Chapter 2 of this report) are summarized below. • For reinforcement spacing of 0.2 m, none of the abutments suffered from any sta- bility problems up to an applied pressure of 1,000 kPa. • For reinforcement spacing of 0.4 m, most of the abutments encountered facing fail- ure (i.e., the top two to three courses of facing blocks “fell off” the wall face) when the applied pressure exceeded 500 kPa to 970 kPa (depending on the geometric condition and material properties of the abutment). Only those abutments with sill width  0.8 m and soil friction angle  37° and 40° did not encounter facing fail- ure up to an applied pressure of 1,000 kPa. Nonetheless, there was no catastrophic failure in any of the abutments up to 1,000 kPa applied pressure. • The differences in the magnitude of the performance characteristics for φ between 34° and 37° were generally greater than those between 37° and 40°. This suggests that increasing the soil friction angle (by selecting a better fill type and/or with bet- ter compaction efforts) to improve the performance characteristics will be more efficient for soils with a lower friction angle than for soils with a higher friction angle. • The effect of reinforcement spacing on sill settlement and maximum lateral displacement of wall face was significant, especially at higher applied pressure (i.e., greater than 200 kPa). • A glaring difference between an integrated sill and an isolated sill was in the rota- tion of the sill. The integrated sills experienced counter-clockwise tilting; while the isolated sills generally experienced clockwise rotation, except for a larger sill width (2.5 m) where the rotations were clockwise. • Over a rigid foundation, the abutments tend to have significantly smaller sill settlements, smaller maximum lateral wall displacements, smaller sill lateral movements, and smaller sill rotations (except for isolated sills) than the abutments situated over a medium sand foundation. The allowable bearing pressures of GRS abutments were evaluated using the results of the 36 analyses with a medium sand foundation, because GRS abutments offered more conservative allowable bearing pressures than those abutments with a rigid foun- dation. Two performance criteria were examined. One criterion involves a limiting sill settlement, where the allowable bearing pressure is corresponding to a sill settlement of 1 percent of the lower wall height (i.e., 1%H). The other criterion involves the dis- tribution of the critical shear strain in the reinforced soil mass, where the allowable bearing pressure corresponds to a limiting condition in which a triangular critical shear strain distribution reaches the back edge of the sill (i.e., the heel of the sill).

The bearing pressures obtained from the two performance criteria formed the basis for the recommended allowable bearing pressures in the recommended design method presented in Chapter 3. Major Refinements and Revisions to the NHI Design Method The recommended design method adopted the format and basic methodology of the NHI design method for mechanically stabilized earth (MSE) abutments. Fourteen spe- cific refinements and revisions to the NHI design method are stipulated in the recom- mended design method. The refinements and revisions and the basis for each refinement and revision are presented in Chapter 3. The major refinements and revisions to the NHI design method are as follows: • The allowable bearing pressure of the bridge sill on a load-bearing wall (the lower wall) of a GRS abutment is determined as a function of the friction angle of the fill, reinforcement vertical spacing, sill width, and sill type (isolated sill or inte- grated sill). A simple three-step procedure is provided for determining the allow- able bearing pressures under various design conditions. • The default value for reinforcement vertical spacing is set as 0.2 m. To ensure sat- isfactory performance and an adequate margin of stability, reinforcement spacing greater than 0.4 m is not recommended for GRS abutments under any conditions. • To provide improved appearance and greater flexibility in construction, a front batter of 1/35 to 1/40 from the vertical is recommended for a segmental abutment wall facing. A typical setback of 5 to 6 mm between successive courses of facing blocks is recommended for blocks 200 mm (8 in.) high. • The reinforcement length may be “truncated” in the bottom part of the wall if the foundation is “competent.” The recommended configuration of the truncation is reinforcement length  0.35 H at the foundation level (H  total height of the abutment wall) and increases upward at a 45° angle. The allowable bearing pres- sure of the sill, as determined by the three-step procedure, should be reduced by 10 percent for truncated-base walls. Permitting truncated reinforcement typically will produce significant savings when excavation is involved in the construction of the load-bearing wall of a bridge abutment. • A recommended sill clear distance between the back face of the facing and the front edge of the sill is 0.3 m (12 in.). The recommended clear distance is a result of finite element analysis with the consideration that the soil immediately behind the facing is usually of a lower compacted density because a heavy compactor is not permitted close to the wall face. • For most bridge abutments, a relatively high-intensity load is applied close to the wall face. To ensure that the foundation soil beneath the abutment will have a suf- ficient safety margin against bearing failure, a revision is made to check the con- tact pressure over a more critical region—within the “influence length” D1 (as defined in Chapter 3) behind the wall face or the reinforcement length in the lower wall, whichever is smaller. In the current NHI manual, the contact pressure is the average pressure over the entire reinforced zone (with eccentricity correction). • If the bearing capacity of the foundation soil supporting the bridge abutment is found only marginally acceptable or slightly unacceptable, it is recommended that a reinforced soil foundation (RSF) be employed to increase its bearing capacity and reduce potential settlement. A typical RSF is formed by excavating a pit that is 0.5 * L deep (L  reinforcement length in the load-bearing wall) and replacing it with compacted road base material reinforced by the same reinforcement to be 6

7used in the load-bearing wall at 0.3 m vertical spacing. The lateral extent of the RSF should at least cover the vertical projection of the reinforced soil area and should extend no less than 0.25 * L in front of the wall face. • Both a minimum ultimate tensile strength and a minimum tensile stiffness of the reinforcement should be specified to ensure sufficient tensile resistance at the ser- vice loads, to provide adequate ductility, and to ensure a sufficient safety margin against rupture failure. A recommended procedure for determining the required minimum tensile stiffness (at 1.0% strain) and the minimum ultimate tensile strength has been stipulated. • It is recommended to extend the reinforcement lengths in both the upper and lower walls—at least the top three layers in each wall—to approximately 1.5 m beyond the end of the approach slab to enhance the integration effect of the abutment walls with the approach embankment, so as to eliminate the bridge “bumps”—a chronic problem in many bridges. • Connection strength is not a design concern as long as (1) the reinforcement spacing is kept no greater than 0.2 m, (2) the selected fill is compacted to meet the specification stipulated in the recommended construction guidelines, and (3) the applied pressure does not exceed the recommended design pressures in the recommended design method. Recommended Construction Guidelines The recommended construction guidelines were established based on the guidelines for segmental GRS walls as provided by various agencies (including AASHTO, NCMA, FHWA, CTI, SAGP, and JR) as well as the authors’ observations and experi- ences with the construction of GRS walls and abutments. The construction guidelines focus on GRS abutments with a segmental concrete block facing. Only the basic con- struction guidelines for three types of flexible facing (i.e., geotextile-wrapped, timber, and natural rock facing) are presented.