Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 63
63 CHAPTER 4 Conclusions and Suggested Research the same project. The information available that accompanied 4.1 Conclusions the soil nail load-test data was in general insufﬁcient to study This study was conducted in the following main steps: other aspects (e.g., construction methods) that may affect the (i) review of guidance procedures and speciﬁcations for the variability of soil nail pullout resistance. In addition, a database design and construction of SNWs; (ii) compilation of soil nail of soil nail loads based on instrumented SNWs was created. load-testing data for developing pullout resistance informa- Resistance factors for elements that are common to other tion, and load data from instrumented walls for developing retaining systems (e.g., factor for the nominal tensile resist- load statistics; (iii) development of databases for pullout ance of steel bars) were adopted from the AASHTO LRFD resistance and loads in SNWs; (iv) development of resistance Bridge Design Specifications (AASHTO, 2007) for consis- factors based on reliability methods and on the aforementioned tency. Current values were found to be acceptable for the databases; and (v) comparison of designs using the LRFD and design of SNWS. These resistance factors are presented in ASD methods. Table 4-1. The review of existing procedures for the design and con- The calibration of the resistance factor for soil nail pullout struction of SNWs was focused on U.S. practice, although the was conducted using reliability methods and the resistance review also included international references. LRFD factors and load databases mentioned above. The calibration was con- developed for comparable types of retaining structures were ducted using the procedures suggested for developing load and also reviewed, including interim editions and the latest edition resistance factors in general geotechnical and structural design of the AASHTO LRFD Bridge Design Speciﬁcations (AASHTO, (Allen et al., 2005). In this approach, several steps were followed, 2007). Because the design of SNWs as conducted in the United from selecting a target reliability index that is consistent with the States is based on limit-equilibrium methods (i.e., related to level of structural redundancy of SNWs, to a Monte Carlo limit states of overall stability), the load combination selected simulation to estimate pullout resistance factors. to design SNWs was the service limit state, consistent with the For each soil/rock material considered in the pullout resist- approach currently adopted in AASHTO (2007) for the limit ance database, statistical parameters were obtained for the bias of pullout resistance and loads in SNWs. In addition, the states of overall stability. database of soil nail loads allowed an estimation of the statis- A signiﬁcant amount of soil nail load-test data was collected tical parameters for the bias of loads. Both load and resistance from several sources. After several results were eliminated due were considered to be random variables having lognormal to lack of information or inconsistencies, a database of nail distributions. pullout resistance was compiled to support the calibration of The target reliability index was selected based on a compar- pullout resistance factors. The volume of pullout resistance ison of SNWs with other substructures that have a compa- data was sufﬁcient to create data subsets for three subsurface rable level of structural redundancy and for which target conditions, namely predominantly sandy soils, clayey soils, and reliability indices have been proposed. The reliability selected weathered rock. More data points were available from projects for SNWs was 2.33, which is consistent with the value used of SNWs constructed in sandy soils than in clayey soils and for the calibration of resistance parameters for pullout in MSE weathered rock. To reduce potential scatter in the database due walls (Allen et al., 2005). SNWs and MSE walls have compa- to variable levels of workmanship and equipment among dif- rable reinforcement densities (i.e., number of reinforcement ferent contractors, data points were selected, as much as pos- elements per unit of wall area), comparable reinforcement sible, from the same contractor using the same equipment at
OCR for page 63
64 Table 4-1. Summary of resistance factors for SNWs. Resistance Limit State Resistance Condition Value Factor φτ Sliding All 0.90 Soil Failure φb Basal Heave All 0.70 φs 0.75 (1) Slope does not support a structure Overall φs 0.65 (2) (3) NA Slope supports a structure Stability φs 0.9 (4) Seismic Mild steel bars—Grades 60 φT 0.56 (5) and 75 (ASTM A 615) Static High-resistance—Grade 150 φT 0.50 (5) (ASTM A 722) Nail in Tension Mild steel bars—Grades 60 φT 0.74 (5) and 75 (ASTM A 615) Seismic High-resistance—Grade 150 φT 0.67 (5) (ASTM A 722) Temporary and final facing φFF 0.67 (5) Facing Flexure reinforced shotcrete or concrete Structural Temporary and final facing φFP 0.67 (5) Facing Punching-Shear reinforced shotcrete or concrete φFH 0.50 (5) A307 Steel Bolt (ASTM A 307) Facing Headed-Stud Tensile φFH 0.59 (5) A325 Steel Bolt (ASTM A 325) φPO 0.47 (6) Sand φPO 0.51 (6) Clay Pullout Soil/Rock Type φPO 0.45 (6) Weathered Rock φPO 0.49 (6) All Notes: (1) AASHTO (2007) also considers this value when geotechnical parameters are well defined. (2) AASHTO (2007) also considers this value when geotechnical parameters are based on limited information. For temporary SNWs, use φs = 0.75. (3) Per AASHTO (2007) but subject to modification after new Standard is in place. A value φs = 1.00 (4) may be acceptable, as long as permanent deformations are calculated (see Anderson et al., 2008) and are found not to be excessive. Currently, there is no differentiation for temporary or non- critical structures under seismic loading; therefore, use φs = 1.00. (5) Calibrated from safety factors. From reliability-based calibration. Values shown correspond to a load factor γ = 1.00. (6) length/wall height ratios, and thereby comparable and rela- ance factors based on this range of load factors is presented tively high structural redundancies. in Table 4-2. The calibration proceeded using an iterative scheme in a Calibration resistance factors were subsequently used to Monte Carlo simulation. Based on the statistical parameters perform comparative designs for SNWs for a wide variety of for load and resistances selected earlier, up to 10,000 random conditions. The objective of the comparative designs was to simulations were conducted for each soil type in order to evaluate differences of the required soil nail length, as obtained generate a complete distribution of load and resistance. using computer programs with the ASD method or the LRFD Although the load factor should be selected as 1.0 for serv- method. Over 30 design cases were considered to assess the ice limit states (per current AASHTO LRFD practice, as men- effect of several key factors in the design. These factors included tioned previously), a series of pullout resistance factors was wall height, soil friction angle, bond resistance, and surcharge obtained for a range of load factors other than 1.0 to show the loads. Results of the comparative designs indicate that the effect of load factors on the pullout resistance factor for each of required soil nail length calculated using the LRFD method the soil/rock types considered. The load factors selected were and the proposed resistance factors are comparable with those λQ = 1.0, 1.35, 1.5, 1.6, and 1.75. This range represents the val- obtained with the ASD method. For all cases considered, the length difference is, on average, approximately 4% larger ues that can be commonly used for retaining structures that in the LRFD method. None of the factors appear to have a are part of bridge substructures. The calibrated pullout resist-
OCR for page 63
65 • Section 11.12.3 provides guidance and commentary that Table 4-2. Summary of pullout resistance factors for various aid in conducting evaluations of service limit states for both load factors. deformations and overall stability. • Section 11.12.4 addresses safety against soil failure and pro- Pullout Load vides guidance and commentary for conducting evaluations Resistance Factor Material Factor for the limit states of basal heave and sliding stability. λQ φ PO • Sections 11.12.5 and 11.12.6 provide guidance and commen- 1.75 0.82 tary for structural limit states—including soil nail pullout and 1.6 0.75 soil nail in tension—and all of the limit states for facings. Sand 1.5 0.70 • Finally, Sections 11.12.7 through 11.12.8 provide guidance 1.35 0.63 1.0 0.47 and commentary for conducting drainage evaluations and 1.75 0.90 providing corrosion-protection for SNWs. 1.6 0.82 Clay 1.5 0.77 1.35 0.69 4.2 Suggested Research 1.0 0.51 1.75 0.79 The results of this research project have provided a basis 1.6 0.72 for designing SNWs using the LRFD method for various Weathered Rock 1.5 0.68 soil conditions. However, some aspects related to SNW 1.35 0.61 construction and design were not addressed in this project 1.0 0.45 but can be expanded through additional research. Some of 1.75 0.85 1.6 these aspects and areas of additional research are discussed 0.78 All 1.5 0.73 below: 1.35 0.66 1.0 0.49 • Addressing limit-equilibrium problems as a service limit in Note: Reliability Index: β = 2.33 current AASHTO LRFD practice is apparently an unresolved issue and will remain unresolved until additional informa- tion or studies are available. Although this topic is of gen- greater inﬂuence than others, possibly with the exception of eral applicability for various bridge substructures, it will surcharge loads. The largest difference obtained in the compar- affect the design of SNWs if changes are made to the current ative analysis was approximately 8%. practice. • Discussions on the use of the computer programs GOLD- The current database of soil nail load tests can be expanded, NAIL and SNAILZ for LRFD-based design of SNWs are also relying on tests that exhibit clearly a limit state for pullout. provided in this document. This effort should help augment the current data sets not The comparative designs mentioned above have shown only for the three material types considered but also for that the design of SNWs using the LRFD method would result other soil types and conditions (e.g., gravelly soils, residual in quantities comparable to, although slightly higher (i.e., soil, loess, and typical “regional” soils). • approximately 4% increase of soil nail length on average) The current database of pullout resistance based on soil than, those obtained with the ASD method. Essentially there nail load tests can be expanded and subdivided for certain are no changes in the requirement of bar diameters, bar construction procedures that directly affect pullout capac- lengths, and facing dimensions and quantities. The use of the ity, including drilling techniques, practice for cleaning the LRFD method allows for designing SNWs with a reliability level hole, grout characteristics, etc. • that is compatible with reliability levels of other elements of a The database for loads measured in SNWs can be expanded bridge superstructure or other comparable retaining systems. for other conditions, particularly for larger surcharge loads. • Proposed speciﬁcations for the design and construction of Correlations between soil/rock properties, common field SNWs were also developed and are provided as appendices to investigation techniques [i.e., SPT as mentioned in this this report. The proposed speciﬁcations follow the format of report but also other popular ﬁeld techniques including AASHTO (2007). The proposed design speciﬁcations include cone penetration testing (CPT)], and pullout resistance can several sections: be developed as additional predictive tools. • The effect of the number and characteristics of soil nail • Sections 11.12.1 through 11.12.2 provide general descrip- load testing on the reliability of the design can be explored. tions, loading conditions, and controlling factors to be used It is reasonable to expect that conducting more veriﬁcation in the design of SNWs. tests, or increasing the test load in veriﬁcation tests beyond
OCR for page 63
66 • New soil-nailing techniques and new soil nail materials can 200% of the assumed design load, would help establish more precisely the ultimate resistance, would enhance the reliabil- be considered for possible application for transportation ity of the pullout resistance, and possibly result in more eco- projects. These innovations include self-boring nails, Glass- nomical designs. However, it is recognized that this approach Fiber Reinforced Polymer (GFRP) bars, and different head may penalize competent contractors who have considerable nail connections. • Aspects related to the seismic design of substructures that experience and have the expertise to guarantee the speciﬁed bond strength with little testing. have been recently proposed in interim editions of the • Effects of the spatial variability of subsurface conditions AASHTO LRFD Bridge Design Specifications may require on pullout resistance, which are not commonly taken into evaluation in order to adapt those changes to the design account, can be explored in more detail when enough ﬁeld of SNWs. • The current criterion for estimating lateral deformation of exploration data is available (i.e., typically much more than what is conventionally produced). While this effect may not SNWs is limited. The quantiﬁcation of the effects of soil be signiﬁcant for SNWs constructed over small areas, this nail layout on the distribution and magnitude of deforma- effect may be signiﬁcant in the use of SNWs along roadways tions is also suggested as a follow-up research topic. To this or as part of the abutments for relatively long bridges. How- end, numerical studies using the ﬁnite-element method or ever, it is recognized that a reliable quantiﬁcation of spatial comparable techniques are suggested to obtain estimates variability can only be achieved if sufficient field explo- of constructed and monitored walls. Comparisons of the ration data is available. For most project conditions, it is numerically estimated and measured wall deformations unlikely that enough geotechnical data would be available to will help calibrate the numerical methods, which can even- quantify spatial variability. tually be used to predict the deformation of future walls.