GS Design Manual 081209

26
GEOTECHNICAL SERVICES DESIGN MANUAL VERSION 1.0 August 2009 DIVISION OF ENGINEERING SERVICES GEOTECHNICAL SERVICES

Transcript of GS Design Manual 081209

Page 1: GS Design Manual 081209

GEOTECHNICAL SERVICES

DESIGN MANUAL

VERSION 1.0

August 2009

DIVISION OF ENGINEERING SERVICES GEOTECHNICAL SERVICES

Page 2: GS Design Manual 081209

2

Section 1: Seismic Design Recommendations Section 1.1: Ground Motions Section 1.1.1: Definitions for Developing Deterministic Acceleration Response Spectrum (ARS) -Fault Parameters from the 2007 Fault Database

MMax Maximum moment magnitude of fault - The largest earthquake a fault is capable of generating and is related to the seismic moment. For details see Map Report.

FRV Faults identified as a Reverse Fault “R” in the 2007 Fault Database, enter “1” into the Deterministic Response Spectrum Spreadsheet, enter “0” otherwise.

FNM Faults identified as a Normal Fault “N” in the 2007 Fault Database, enter “1” into the Deterministic Response Spectrum Spreadsheet, enter “0” otherwise.

δ Fault dip angle (deg)

ZTOR Depth to top of rupture (km).

ZBOT Depth to top of rupture (km)

-Distance

RRUP Closest distance (km) to the fault rupture plane, as shown in Section 1.1.6.

RJB Joyner-Boore distance - The shortest horizontal distance (km) to the surface projection of the rupture area. Think of this as the nearest horizontal distance to the area directly overlying the fault. RJB is zero if the site is located within that area.

RJB (RJB = 0) RJB Site A Site B RX Horizontal distance (km) to the fault trace or surface projection of the top of

rupture plane. It is measured perpendicular to the fault (or the fictitious extension of the fault). The diagram below shows when a site is located on the footwall side (Site A) or on hanging wall side (Site B). When using the Deterministic Response Spectrum Spreadsheet, the “hanging wall factor” is applied to the resulting ARS when the appropriate check box is selected.

RX RX Site A Site B Footwall Hanging Wall

Page 3: GS Design Manual 081209

3

- Site Conditions

VS30 Average small strain shear wave velocity for the top 30m (100 feet) at the site (see Section 1.1.7 for details).

Z1.0 Depth to rock with shear wave velocity = 1.0 km/sec [Caltrans defined deep sedimentary basin locations only; refer to SDC (Appendix B, Figures B.5-B.11)].

Z2.5 Depth to rock with shear wave velocity = 2.5 km/sec [Caltrans defined deep sedimentary basin locations only; refer to SDC (Appendix B, Figures B.5-B.11)].

Section 1.1.2: Design Policy and Available Design Tools The current Caltrans Seismic Design Criteria (SDC) & Appendix B and Memo To Designer (MTD) Section 20 shall be the basis for specifying the minimum seismic design requirements for Ordinary Standard Bridges (defined in SDC Section 1.1). For important bridges (defined in MTD Section 20-1), the requirements of project specific design criteria shall be met. Several design tools and references are available for use in developing response spectra for seismic design recommendations. These tools include, but are not limited to:

• 2009 Caltrans Deterministic PGA Map and ARS Online Report • 2007 Fault Database • 2008 United States Geological Survey (USGS) National Seismic Hazard Map, 2008

National Seismic Hazard Map Gridded Data (static) and the 2009 USGS Interactive Deaggregation Tool (Beta)

• Fault & Site Data Input Sheet • Deterministic Response Spectrum Spreadsheet • Probabilistic Response Spectrum Spreadsheet (based on the USGS data) • Preliminary spectral acceleration and displacement curves (SDC, Appendix B) • Caltrans ARS Online

A complete list of these documents and tools with web links are listed in the Caltrans ARS Online website, http://dap3.dot.ca.gov/shake_stable/technical.php. Section 1.1.3: Deterministic Acceleration Response Spectrum (ARS) The 2009 Caltrans Deterministic PGA Map and ARS Online Report (hereafter referred to as the "Map Report") was updated in 2009 and consists of a report describing the methodologies used in creating the Caltrans Deterministic Peak Ground Acceleration (PGA) Map, the fault database, references and notes. It also discusses the development of the ARS Online web tool, a user-friendly web tool that can be used in confirming the design response spectrum described herein. The faults shown listed in the 2007 Fault Database and discussed in the Map Report are based primarily on the California Geologic Survey (Bryant, 2005) fault database and are considered active within the late Quaternary (within the past 700,000 years) and capable of producing a maximum moment magnitude (MMax) earthquake of 6.0 or greater. The deterministic map shows PGA contours assuming a VS30 of 760 meters per second. The contours were calculated

Page 4: GS Design Manual 081209

4

using the arithmetic average of two Next Generation Attenuation (NGA) ground motion prediction equations developed by Chiou & Youngs and Campbell & Bozorgnia (CY-CB) in 2008. The combined models are referred to as the CY-CB ground motion prediction equation (CY-CB GMPE). The contour map shows the California deterministic seismic ground motion hazard as defined in Appendix B of the SDC, except that the map does not include near-field or additional amplification due to soil and/or basin effects. The fault locations and estimated PGA values shown on the map are for preliminary design only. Utilization of the Caltrans Deterministic PGA Map for preliminary design requires standardized spectral curves so that an entire spectrum can be estimated from a single PGA value shown on the Map Report. The standardized spectral curves will be discussed later in this section. The newly adopted CY-CB GMPE along with the additional complexity of using various site parameters led to the development of the ARS Online web tool that generates design ARS curves to be compared to the results generated with the Deterministic Response Spectrum Spreadsheet. The preferred method of developing an acceleration response spectrum (ARS) curve is the direct use of the CY-CB GMPE using the Deterministic Response Spectrum Spreadsheet available at the web link provided in Section 1.1.2. The Deterministic Response Spectrum Spreadsheet allows the geotechnical designer to develop an ARS curve based on several input parameters, including fault characteristics, site-to-fault distances and site parameters (e.g. VS30, Z1.0 & Z2.5). The results generated from the Deterministic Response Spectrum Spreadsheet should then be checked using ARS Online. The web tool provides a simple, user-friendly interface for developing ARS curves based on the methods described above and in the SDC (Appendix B). In anticipation that new fault data may be identified or existing fault data may be modified after the implementation of the ARS Online web tool, it is the responsibility of the geotechnical professional to verify that the fault information shown on the ARS Online map and the results generated by the ARS Online are valid. One limitation of ARS Online is that it does not incorporate the errata changes in its calculations. However, it does alert users with flags that changes have been made to the fault database. In these cases, ARS Online results will not match the results correctly generated using the Deterministic Response Spectrum Spreadsheet with fault errata. Updates and errata for the Map Report including the fault database are posted at http://dap3.dot.ca.gov/shake_stable/Errata_Report_81109.pdf. Preliminary ARS curves are available in the SDC (Appendix B) and may be used for preliminary design only when inadequate subsurface data precludes the determination of more specific design curves using the Deterministic Response Spectrum Spreadsheet. For details, refer to the SDC (Appendix B). Section 1.1.4: - Quality Control/Quality Assurance The Department has adopted new seismic design procedures that require training and independent reviews to ensure that the new design procedures are followed correctly. Geotechnical Services has implemented a quality control and quality assurance policy, detailed in the “Quality Control/Quality Assurance for the 2009 Seismic Design Procedures,” dated August 12, 2009 (http://dap3.dot.ca.gov/shake_stable/references/QC_QA_Implementation_Memo_81209.pdf). Effective September 30, 2009, all Geotechnical Services staff and consultants providing seismic recommendations for bridge structures shall use the procedures presented in this Geotechnical

Page 5: GS Design Manual 081209

5

Services Design Manual and shall ensure that quality control and quality assurance procedures are followed. At the option of the geotechnical designer, the new seismic design procedures may be adopted as early as August 6, 2009 as long as all QC/QA activities are followed. Section 1.1.5: Procedures for Developing a Deterministic ARS The step-by-step procedures and examples for developing an acceleration response spectrum (ARS) are based on the current criteria in the SDC, Appendix B and require the user to be familiar with the Caltrans Deterministic PGA Map and ARS Online Report (a.k.a. "Map Report"), 2007 Fault Database, the ARS Online web tool, and Deterministic Response Spectrum Spreadsheet tool. The Fault & Site Data Input Sheet is also available to help users organize the necessary data discussed below. This document can also serve as a checklist to ensure that all relevant site factors and fault parameters have been considered. It is available at the web link provided in Section 1.1.2. The steps provided below are used in developing a deterministic ARS.

a) Locate the site on the Caltrans Deterministic PGA Map and determine the controlling fault(s) in the area. • The Caltrans Deterministic PGA Map shows faults that Caltrans considers “active”

and the corresponding peak ground accelerations (PGA) for sites with a VS30 of 760 m/s. The fault accuracy shown on the Caltrans Deterministic PGA Map (and ARS Online) is assumed equivalent to that of a map with an approximate scale of 1:750,000 (1 inch = 12 miles). The PGA map does not include site factors (e.g. soil amplification, near-fault, basin, etc) as defined in the SDC. Those factors must be applied to accurately determine the design ground motion. In order to facilitate this process, ARS Online and spreadsheet tools are available to determine the design ground motion.

b) Locate the site on the ARS Online map and verify that the site is properly located using the satellite map view. • The ARS Online tool displays the active faults from the Caltrans Deterministic PGA

Map. The simplified fault traces shown on the ARS Online tool are for estimating ground motions only and must NOT to be used for fault surface rupture determination. Uncertainties in the locations of the digital fault traces shown on the map are approximately ± 1 km. Uncertainty of the Google satellite map view is less than 140 meters. (Caution: when using the ARS Online “search” option (e.g. search by name, bridge number, dist-co-rte, etc) to locate your site, the user shall verify that the site identified is actually the project site. There have been a few instances where ARS Online “search” option has incorrectly located bridge sites).

c) Obtain the required fault parameters from the fault database contained in the 2007 Fault Database. The required fault parameters for input into the Deterministic Response Spectrum Spreadsheet are the Maximum Moment Magnitude, fault type, fault dip angle, dip direction, and the top of rupture plane. • Since the fault database was locked down in 2007, a number of changes to fault

parameters were identified and evaluated to determine if the change was significant. If the fault parameter change was determined to be significant, then it is listed in the fault errata file (http://dap3.dot.ca.gov/shake_stable/references/Errata_072909.pdf). Users must review the errata to ensure the latest fault information is used. One limitation of ARS Online is that it does not incorporate the errata changes in its

Page 6: GS Design Manual 081209

6

calculations. However, it does alert users with flags that changes have been made to the fault database. In these cases, ARS Online results will not match the results correctly generated using the Deterministic Response Spectrum Spreadsheet with fault errata.

d) Calculate or measure the distances RX, RRUP, RJB. (Refer to Section 1.1.6 for equations and examples for calculating the R distances) • It is acceptable to use ARS Online “ruler” tool for determining the R distances to be

used in estimating ground motions. However, for sites in close proximity to a simplified fault trace (e.g. less than 1.6 km or 1.0 miles), additional geologic investigations and research (e.g. site visits, using detailed fault maps, etc) by an experienced geo-professional is required to evaluate and report the potential for surface fault rupture.

e) Draw a sketch in both plan and profile view that shows the site and fault-related geometry (e.g. footwall and hanging wall side of fault, perpendicular and/or off-set distance to fault).

f) Calculate or use the measured VS30 value for the site (Refer to Section 1.1.7 for equations and methods for determining VS30).

g) Determine Z1.0, if applicable (Refer to Section 1.1.8 for details on estimating Z1.0). h) Determine Z2.5, if applicable (Refer to Section 1.1.8 for details on estimating Z2.5). i) Input fault and site parameters into the Deterministic Response Spectrum Spreadsheet

tool for each possible controlling fault and create an ARS curve for each case. • For sites where the Cascadia Subduction Zone Fault is one of the controlling faults or

where the site is located in the Eastern California Shear Zone [SDC (Appendix B, Figure B.1)], use the modified process described in the SDC, Appendix B to create an ARS curve (Refer to Section 1.1.11 &1.1.12 for details).

j) If applicable, apply “near-fault factors” (as defined in SDC, Appendix B) and “hanging wall factors” to the ARS curve developed with the Deterministic Response Spectrum Spreadsheet.

k) Create the minimum deterministic response spectrum for California (as defined in SDC) using the Deterministic Response Spectrum Spreadsheet tool.

l) Compare the site ARS curve(s) to the minimum deterministic response spectrum. Use the upper envelope of the spectral values for the deterministic design response spectrum.

m) Check the calculated design spectrum using ARS Online. • If a significant discrepancy exists between the two methods then the user should

check inputs for both methods to identify possible input errors. If the discrepancy cannot be resolved, then refer to Section 1.1.14 for details.

Page 7: GS Design Manual 081209

7

Section 1.1.6: Calculating Distances (RX, RRUP & RJB) For Use in ARS Spreadsheet Tool Three different site-to-fault distances must be determined.

• RX is the horizontal distance to the fault trace or surface projection of the top of rupture plane. It is measured perpendicular to the fault (or the fictitious extension of the fault),

• RRUP is the shortest distance to the fault rupture plane, • RJB is the distance to the vertical projection of the fault rupture plane.

This section provides six general cases for determining the various distances from the site to the fault rupture plane for use in the verification spreadsheet tools. Cases 1 through 4 assume that a site is situated similar to Site A (footwall side of fault). Cases 5 and 6 assume that a site is located as in Site B or C (hanging wall side of fault). For Sites D, E and F (shown below), which are offset from the fault plane, additional correction factors or measurements will be needed to determine the final RX, RRUP & RJB distances (see Scenario 3 at the end of Section 1.1.6). Plan View Site F Site D Site E

Site A Site C

Site B Oblique View

Exposed trace or vertical projection of the fault at surface

Vertical Projection of Dipping Fault

Depth to Bottom of Fault Rupture

Projected Fault Width

) Fault Dip Angle

Site A

Site C

Site ESite F

Vertical Projection of Dipping Fault

Exposed trace or vertical projection of the fault at surface Site B

Site D

Page 8: GS Design Manual 081209

8

Scenario 1: Site is located on either side of vertical fault (Case 1 - 2) or site is located on footwall side of dipping fault (Case 3 – 4). -Case 1: Fault dip = vertical (fault exposed at surface: ZTOR = 0) For vertical, strike-slip faults (or other faults with vertical dips) that intersect the surface, RJB, RRUP, and RX are the same as the map distance. Site RJB, RRUP, RX Fault -Case 2: Fault dip = vertical (buried fault: ZTOR > 0) For vertical, strike-slip faults that do not intersect the surface (ZTOR > 0), RJB and RX are the same as the map distance, but RRUP must be calculated (see below) RJB, RX Site RRUP ZTOR Fault

( ) ( )2X

2TORRUP RZ R +=

Page 9: GS Design Manual 081209

9

Fault dip not vertical and site on footwall side For dipping faults, the calculations are different depending on whether the site is down-dip (hanging wall side) or up-dip (footwall side), and if the fault intersects the surface (ZTOR = 0). For Cases 3 and 4, the sites are on the footwall side and are distinguished by whether or not the fault intersects the surface. -Case 3. Site is on footwall side of the fault, ZTOR = 0 RJB, RRUP, RX Site δ Footwall Hanging wall -Case 4. Site is on the footwall side of the fault, ZTOR > 0 RJB and RX are the same as the map distance and RRUP must be calculated. RJB, RX

Site ZTOR

. RRUP ) δ

( ) ( )2X

2TORRUP RZ R +=

Page 10: GS Design Manual 081209

10

Scenario 2: Sites on hanging wall side of fault (Cases 5 & 6) In Cases 5 and 6, the site is on the hanging wall side of the fault and more calculations are required to determine the RX, RRUP & RJB distances than the previous cases. To assist in determining which of the following cases apply, some other pertinent distance parameters (FWP, X1, & X2) may be calculated to assist in determining the required RX, RRUP & RJB distances. Plan View Fault Width (FWP) Bottom of Dipping Fault (Vert. Projection) Exposed Dipping Fault at surface Site C Site B -Case 5. Site is on the hanging wall side of the fault, ZTOR =0. In Case 5, the depth to top of rupture (ZTOR) is zero. Here it is necessary to know the depth to the bottom of fault rupture (ZBOT) in order to calculate the projected fault width (FWP), and to determine the X1 distance (shown below). The X1 distance is a geometric factor used only to determine which of the following cases applies. Case 5a. Rx is less than or equal to the sum of FWP and X1: Rx, where Rx ≤ FWP + X1 Fault Width (FWP) X1

RJB

RRUP

δtan

ZFW BOTP =

δtanZX BOT1 ∗=

0RJB = (Site B, when situated directly over fault plane)

PJB FW-RxR = where RJB ≥ 0, similar to Site C.

δsin R R XRUP ∗= for both Sites B & C

δ Site C

ZBOT

Site B

Page 11: GS Design Manual 081209

11

Case 5b. Rx is greater than the sum of FWP and X1: Rx, where Rx > FWP + X1 Fault Width (FWP) X1 Site RRUP

tanδZFW BOT

P =

RJB δtanZX BOT1 ∗=

⎟⎟⎟

⎜⎜⎜

⎛−=−=

tanδZRFWRR BOT

x PxJB ( ) ( )2JB

2BOTRUP RZ R +=

Case 6. Site is on the hanging wall side of the fault, ZTOR > 0. If the fault does not intersect the surface and the site is on the hanging wall side (down-dip side) of the fault, there are three possible cases. To assist in determining which case applies, a geometric factor X2 is calculated. Case 6a. Site is located at or less than X2 from the vertical projection of fault (top of rupture):

0RJB =

ZTOR δtanZX TOR2 ∗=

( ) ( )2

TOR2

XRUP ZR R +=

δ

ZBOT

Rx

RRUP

X2

) δ

Site

Page 12: GS Design Manual 081209

12

Case 6b. Site is between X2 and the outside edge X1: RX, where Rx < FWp + X1 and Rx> X2 X2 ZTOR

RRUP ZBOT FWp __

)(tanδ

Z -ZFW TOR BOTP =

δtanZX TOR2 ∗=

δtanZX BOT1 ∗=

PJB FW-RxR = where RJB ≥ 0, similar to Site C; (For Site B, RJB = 0).

⎟⎠⎞⎜

⎝⎛

⎥⎦⎤

⎢⎣⎡+= δδ

TOR

x RUP tanZ R sin R

Case 6c. Site to fault distance is greater than the sum of projected fault width (FWP) and the X1

distance as shown below. RX, where Rx > FWp + X1

FWP RJB X1 Site ZTOR ) δ ZBOT RRUP

)(tanδ

Z -ZFW TOR BOTP =

PJB FW-RxR =

( ) ( )2JB

2BOTRUP RZ R +=

RJB X1

Site C

) δ

Site B

Page 13: GS Design Manual 081209

13

Scenario 3: Offset Corrections for final RX, RJB and RRUP distances (if applicable) Cases 1 thru 6 in the preceding section assume that a site is oriented in a manner similar to Site A, B or C, where the shortest distance from the site to the surface expression of the fault are perpendicular to each other. When the site is offset from the fault similar to Sites D, E or F, then some additional measurements will be needed to determine the final RJB and RRUP distances. For Sites D, E and F, create a fictitious extension of the fault plane in the direction of the site so that the site could be classified similarly to Sites A, B or C. Using the methodologies given in the previous sections, calculate preliminary RJB and RRUP distances. Using the fictitious extension of the fault plane, determine the offset distance from the end of the fault plane to the site (see figures shown below). After the distances are determined, apply simple trigonometry to determine the final RJB and RRUP distances (see formulas below). For sites D, E and F, RX is the perpendicular distance to the fictitious extension of the fault trace (or surface projection of the fault trace) as shown below: RX (final) (Plan View) Site D . Site E Site F

Site A

Site C Site B Oblique View Fictitious Extension of Fault Trace at surface

Exposed trace or vertical projection of the fault at surface Vertical Projection

of Dipping Fault

Depth to Bottom of Fault Rupture

Projected Fault Width

) Fault Dip Angle

Site A

Site B

Site E

Site C

Vertical Projection of Dipping Fault

Exposed trace or vertical projection of the fault at surface

RRUP (final) RJB (final)

RX (final)

RRUP (final) RJB (final)

Offset (Site F)

Site FSite D

Page 14: GS Design Manual 081209

14

Final Distance Measurements For Sites D, E & F

( ) ( )22RUP(final)RUP OffsetR R += For Site D, E & F

( ) ( )22JB(final)JB OffsetR R += For Site D & F

Offset R (final)JB = For Site E Section 1.1.7: Determination of Shear Wave Velocity and VS30 The average small strain shear wave velocity (Vs30) for the upper 30 meters (100 feet) of the soil/rock profile is required to determine the design ground motion at the ground surface of a project site using the attenuation relationships included in the SDC. within the upper 30m (100 ft) is:

n

nS

VD

VD

VD

VD

V++++

=...

ft 100

3

3

2

2

1

130 ;

The shear wave velocity (Vs) of each soil or rock layer may be measured in-situ or, where applicable, estimated based on empirical correlations with other parameters (e.g. field or lab data). In-situ measurements of Vs using geophysical methods, where feasible, are relatively simple and preferred for the purpose of estimating ground motions. The table below provides brief descriptions of some common geophysical methods for measuring Vs.

Geophysical Test Brief Description

PS Suspension logging

Shear wave measurements are made in an open or cased borehole. Shear wave velocity measurements are usually made in one-meter increments. VS30 is calculated as the reciprocal of the summed travel times over 30 meters.

Down-hole seismic Shear wave measurements may be made in a similar fashion to suspension logging. The difference is that the seismic source is always at the surface, and the receiver may be lowered into the borehole at discrete intervals. It is possible using down-hole seismic to measure VS30 directly by placing the seismic source at the top of the borehole and the source at 30 meters. The depth to receiver divided by the measured travel time is VS30.

Seismic CPT cone CPT seismic cone method is essentially the same as the down-hole seismic method. The source is at the surface, but the receiver is in the CPT cone. Thus no predrilled borehole is necessary. Measurements may be made at discrete intervals, or a single measurement may be made with the cone at a depth of 30 meters or maximum penetration if less than 30 meters.

Rayleigh Wave Inversion

There are a variety of geophysical methods available to determine VS30. Examples include Spectral Analysis of Surface Waves (SASW) and variants, seismic refraction and seismic reflection. If geophysical methods are used, the results of the test will usually be provided to the geotechnical designer by the geophysics branch or geophysicist.

where D is the layer thickness (ft) and V is the shear wave velocity (ft/s) for that layer. Note: VS30 input for attenuation models must be converted into meters/sec.

Page 15: GS Design Manual 081209

15

In-situ measurements of Vs are generally required for project sites underlain by profile types that fall within the Type F category (as listed in SDC, Figure B.12), and in many cases Type E, or when site-specific dynamic ground response analyses are performed. For these cases, Vs measurements may need to extend to depths greater than 30 m depending on the site conditions and the appropriate depth of the input design base motion. Shear Wave Velocity Correlations In the absence of in-situ measurements, Vs for most soil layers and very soft to soft rock layers can be estimated using established correlations. For cohesive soils, empirical correlations with laboratory measured undrained shear strength are preferred. For cohesionless soils, correlations with either the SPT blow count value, N60 (blow count corrected for hammer efficiency but not for overburden), or the CPT tip resistance, qt, may be used. Recommended shear wave velocity correlations for soil layers for use in the development of ground motions for State projects are presented below. A correlation with SPT blow counts is also provided for estimating Vs for young sedimentary rocks. For stronger rock, the mass shear wave velocity for distinct zones may be evaluated based on correlations with other engineering and physical properties of rock mass and rock cores measured in the field or laboratory, or estimated by experienced geo-professionals using geologic correlations. Such properties include, but are not limited to, uniaxial compressive strength of rock, RQD of rock mass, elastic modulus, poisson’s ratio and ultrasonic shear velocity (see Mayne et al, 2001). Well documented and established empirical VS30 correlations specific to the earth material under consideration, or to the project site or the general area with similar earth material may be used by experienced geo-professionals provided adequate justifications of their use and the pertinent references are included in the structure foundation report. In 2007, UC Davis compiled published correlations between shear wave velocity and common in-situ geotechnical test parameters and presented recommended correlations for various soil types. The recommended correlations were reviewed and compared by Caltrans for consistency with the various parameters used to define the profile types and their ranges (after ATC-32, 1996) included in the SDC. The recommended correlations were also compared to a limited amount of P-S log data to ensure acceptable performance. The correlations provided below are the recommended correlations for State project sites, where applicable. The geo-professionals should be aware of the limitations of each correlation used. For example, penetration of the SPT sampler in earth material may be limited or affected by the presence of large particles (e.g. gravel, cobbles, boulders or rock fragments). Correlations, in particular using SPT data, should only be used with test data that are reliable and representative of the actual site conditions. If correlations are not applicable (e.g. SPT correlation used in a thick, coarse gravel deposit) or not available in a region, then in-situ measurements are recommended. The correlation equations summarized below determine shear wave velocity in meters/sec. The correlations recommended below for cohesionless and cohesive soils should not exceed 380 m/sec and 310 m/sec respectively, unless verified with in-situ measurements.

Page 16: GS Design Manual 081209

16

- Cohesionless Soils The shear wave velocity of cohesionless soil layers may be estimated by using correlations with SPT blow counts or CPT tip resistance. When using SPT blow counts, the correlation adapted by Sykora (1987) after Sykora and Stokoe (1983) is recommended to estimate shear wave velocity for cohesionless soil layers:

( ) 29.060 5.100 NVS =

Where, N60 is the SPT blow counts corrected only for the hammer energy. When using CPT penetration data, the correlation adapted by Mayne (2007) after Baldi et al, (1989) is recommended to estimate shear wave velocity for cohesionless soil layers:

( ) ( ) 27.013.0 ' 277 votS qV σ=

Where, qt and σ’vo are the CPT tip resistance and effective overburden pressure, respectively, both in MPa. - Cohesive Soils The correlation developed by Dickenson (1994) that uses undrained shear strength to estimate shear wave velocity is recommended for cohesive soils layers:

( ) 475.0/ 203 auS pSV = Where, Su is the undrained shear strength of cohesive soils and pa is the atmospheric pressure in the same unit as Su. In the absence of measured undrained shear strength data, the shear wave velocity of cohesive soil layers may be estimated by using the CPT correlation developed by Mayne and Rix (1995):

( ) 627.0 75.1 tS qV =

Where, qt is the measured CPT tip resistance (kPa). When analyzing previous investigations where measured undrained shear strength or CPT tip resistance data is not available, the SPT based correlation developed by Ohta and Goto (1978) may be used to estimate the shear wave velocity of cohesive soil layers:

( ) 333.060 9.86 NVS =

Page 17: GS Design Manual 081209

17

- Young Sedimentary Rocks For young sedimentary rock (Tertiary deposits), Imai and Tonouchi reviewed over a hundred tests using SPT data. For these types of rock, the correlation developed by Imai and Tonouchi (1982) for “Tertiary Sand/Clay” may be used to estimate shear wave velocity.

( ) 319.060 109 NVS =

If not measured directly, the Vs value estimated using the SPT correlation for young sedimentary rock layers should be limited to 560 m/sec. - Other Rocks While there are numerous studies that correlate shear wave velocity to in-situ geotechnical testing of soil, there are relatively few studies that correlate shear wave velocity to physical properties of rock. Two notable studies that may be useful to geo-professionals in developing approximations of shear wave velocities based on physical properties of rock are provided below. A study by Fumal (1978) correlated shear wave velocity to weathering, hardness, fracture spacing, and lithology based on data from 27 sites in the upland areas of the San Francisco (Bay Area) region. Another study by Fumal and Tinsley (1985) extended the 1978 study to include 84 sites in the Los Angeles region. Some physical properties of the rock were more important than others depending on lithology, soil texture, rock hardness, but fracture spacing was found to be the most important factor affecting shear wave velocity. A thorough review of these studies will significantly aid geo-professionals in their estimations of shear velocity. In the absence of in-situ measurements of Vs, the VS30 for competent rocks in California should be limited to 760 m/sec. - Estimating VS30 for sites with subsurface information less than 100 ft (30 m) When analyzing subsurface investigations where soil boring depths of 100 feet (30 meters) where not available, an extrapolation equation exist for estimating VS30 assuming that Vs could be estimated at shallower depths and no significant changes in the subsurface will occur to the extrapolated depth of 100 feet. The following equation was developed based on work done by David Boore (2004). [ ] (d) s 30 V d) (0.015 - 1.45 ∗∗=SV ; Section 1.1.8: Determine Depth to Z1.0 and Z2.5 Several significant deep sedimentary basins exist in California. In these regions, a depth to rock (Z) parameter allows for a better prediction of the ground motion. The sedimentary basins are

where d is the depth in meters to the bottom of the known soil column and Vs(d) is the time averaged velocity (m/sec) for the known soil column.

Page 18: GS Design Manual 081209

18

shown in SDC, Appendix B (Figures B.5-B.11) and are also shown on ARS Online. These figures have contours that show the vertical depth to rock with a particular shear wave velocity. When a site is located within one of these basins, then the following must be determined:

• Z1.0: the vertical depth to rock (meters) with a shear wave velocity of 1000 m/s • Z2.5: the vertical depth to rock (km) with a shear wave velocity of 2500 m/s

Section 1.1.9: Input parameters into Deterministic Response Spectrum Spreadsheet Using the information gathered in the preceding sections, input the required parameters into the Deterministic Response Spectrum Spreadsheet tool and develop the deterministic ARS. Also, use the spreadsheet to develop the minimum deterministic response spectrum for California or the Eastern California Shear Zone (as defined in the SDC).

Section 1.1.10: Near-Fault Factors An adjustment factor for near-fault effects shall be applied to the calculated spectral acceleration values when a site is located within 25 km (15.5 miles) of a causative fault. The distance shall be the shortest distance to the rupture plane (RRUP). The increase is 20% for sites within 15 km of a causative fault and tapers linearly to zero at 25 km. The percent increase or amplification in

Page 19: GS Design Manual 081209

19

spectral acceleration values corresponds to periods longer than 1-second and linearly tapers to zero at a period of 0.5-second (see the figures below).

Near-Fault Factor with Respect to Distance

1

1.1

1.2

1.3

0 5 10 15 20 25 30

Distance (km)

Am

plifi

catio

n

Near-Fault Factor with Respect to Period

1

1.1

1.2

1.3

0 1 2 3 4 5

Period (s)

Am

plifi

catio

n(fo

r 15k

m o

r les

s)

The Deterministic Response Spectrum Spreadsheet and ARS Online can both apply this factor when using either tool. Users need to verify that the factor is applied correctly with both tools. There are two cases where near fault factor is not applied. Refer to Section 1.1.11 for details. Section 1.1.11: Minimum Spectrum for California and Eastern California Shear Zone A minimum deterministic spectrum is imposed statewide, in recognition of the potential for earthquakes to occur on faults previously unknown or thought inactive. This minimum spectrum is determined by using the Deterministic Response Spectrum Spreadsheet for an assumed maximum moment magnitude 6.5, vertical strike-slip event occurring at a distance of 12 km (7.5 miles), except for the Eastern California Shear Zone (ECSZ). The ECSZ (as defined in the SDC, Figure B.2) minimum spectrum is determined by using the Deterministic Response Spectrum Spreadsheet and is based on a vertical, strike-slip fault with a moment magnitude of 7.6 and a distance of 10 km. A near-fault adjustment factor shall NOT be applied to the minimum spectra defined above. Section 1.1.12: Cascadia Subduction Zone Fault (Special Case for Northern California) The Cascadia Subduction Zone, which is a large shallow dipping thrust fault zone, is a unique seismic source that can generate large ground motions in northwestern California. For this fault zone, a separate attenuation model and modified analysis is required. The deterministic spectrum for the Cascadia Subduction zone is defined as the median spectrum from the Youngs et al. (1997) ground motion prediction equation, with the added criterion that where the Youngs et al. spectrum is less than the average of the CY-CB GMPE (both without the hanging wall term applied), an arithmetic average of the Youngs et al. and CY-CB GMPE is used. The Deterministic Response Spectrum Spreadsheet contains an alternate spreadsheet tab to facilitate use of the procedure mentioned above. First, the user will input fault and site data using the spreadsheet “input-compare” tab and the procedures listed in previous sections. Once the input data have been entered, the user must select the alternate spreadsheet tab “Youngs 97” to analyze and compare spectra using the procedure mentioned above in order to select the appropriate design response spectrum.

Page 20: GS Design Manual 081209

20

Section 1.1.13: Compare Deterministic Spectra Compare the calculated deterministic response spectrum developed with the Deterministic Response Spectrum Spreadsheet for each of the causative faults near the site and compare the results to the minimum response spectrum for California (as defined in the SDC). In cases where more than one fault contributes to the maximum spectral values across the period spectrum, use the upper envelope of the spectral values shall be used for the design response spectrum. Section 1.1.14: Compare Deterministic Design Spectrum with ARS On-Line tool Compare the calculated deterministic response spectrum developed with the Deterministic Response Spectrum Spreadsheet to the results from the ARS On-line web tool. If a discrepancy of more than 10% exists between the two methods then check the calculations and the web tool to determine if an input error caused the discrepancy between the two methods. If the large discrepancy cannot be resolved then alert the “checker” who will be verifying the calculations. If the checker also identifies a discrepancy between the two methods, then the checker shall make a note in the “Quality Control/Quality Assurance Documentation for Seismic Design Recommendations”, dated July 30, 2009. A copy of the QC/QA document and comments describing the discrepancy should also be submitted to [email protected] so that the ARS Online development team can review the potential bug and help resolve the issue. Section 1.1.15: Probabilistic ARS and USGS Data The probabilistic response spectrum to be used for design of bridge structures is based on data from the 2008 USGS National Seismic Hazard Map for the 5% in 50 years probability of exceedance (or 975 year return period). Documentation, maps, data and tools are available at the USGS website, http://earthquake.usgs.gov/research/hazmaps/. During the development of ARS Online and these seismic procedures, three different data sources from USGS became available:

1. 2007 USGS spectral acceleration data (static). Spectral acceleration data (for VS30 = 760 m/s) at 10 spectral periods were provided to Caltrans from USGS at VS30 = 760 m/s. Not available publicly.

2. 2008 National Seismic Hazard Map Gridded Data (static). Spectral acceleration data (for VS30 = 760 m/s) at four spectral periods (0 s, 0.2 s, 0.3 s and 1 s) are available publicly on USGS website at http://earthquake.usgs.gov/research/hazmaps/products_data/2008/data/.

3. 2009 Probabilistic Seismic Hazard Analysis (PHSA) Interactive Deaggregation website (beta). Ten spectral acceleration values are available publicly on the USGS web tool at http://eqint.cr.usgs.gov/deaggint/2008/index.php.

In 2007, source 1 data was made available to the ARS Development Team for use with ARS Online. In 2008, source 2 data was made publicly available on the USGS website and is used to verify ARS Online design spectrum. Since source 1 & 2 data are for sites with a VS30 = 760 m/s, the ARS Online development team developed soil amplification factors (per the SDC, Appendix B) and built those factors into ARS Online and the Probabilistic Response Spectrum Spreadsheet. In 2009, source 3 data (USGS Interactive Deaggregation Tool) was made publicly available in the form of interactive web tool. The map-based web tool allows input of

Page 21: GS Design Manual 081209

21

latitude/longitude coordinates, return period, VS30 and spectral period. However, the development of a response spectrum requires 10 separate deaggregation calculations, so the process is relatively tedious when compared to ARS Online. During testing of the new seismic procedures and ARS Online, variations of 30% or more between source 1 & 2 data and source 3 data were observed in soft soil due to possible differences in how soil amplification factors are applied. For rock sites (VS30 = 760 m/s), all three source data produced nearly identical results. As a result, a criterion was developed for determining if ARS Online or the 2009 Interactive Deaggregation is appropriate for development of probabilistic spectrum (see Section 1.1.16). Regardless of the source data used, basin factors (as described in Section 1.1.8) and near-fault factor (as described in Section 1.1.10) shall be used in the probabilistic spectrum. To determine the appropriate distance to use for near-fault factor, a site hazard deaggregation using the 2009 USGS Interactive Deaggregation Tool (Beta) shall be generated at the period closest to the fundamental period of the proposed structure. When evaluating the results from the site hazard deaggregation, use the minimum of the R(mean) or R(mode) distance (from peak R, M bin). An example of a USGS deaggregation and the R(mean) and R(mode) distance is provided below.

Page 22: GS Design Manual 081209

22

Section 1.1.16: Procedure for developing probabilistic response spectrum:

a) Locate site using the ARS Online web tool. b) Mark the site using the appropriate button or input the Latitude and Longitude

coordinates. c) Input VS30. d) Generate probabilistic ARS curve by pushing the “calculate” button. e) Go to the USGS website with Interactive Deaggregation Tool (beta): Input the data from

steps b and c with a 5% in 50yr probability of exceedance near the fundamental period of the structure and determine the minimum R (mean) or R (mode) distance.

f) Go back to ARS Online window: Enter the minimum R (mean) or R (mode) distance into the appropriate input box to adjust the ARS for near-fault effect.

g) Generate spectral acceleration data by selecting the “tabular data” button h) For sites with VS30 > 300 m/s, it is acceptable to use ARS Online for generating

probabilistic spectrum. i) For sites with VS30 < 300 m/s, compare the ARS Online results to the 2008 USGS

Interactive Deaggregation Tool (Beta). If a discrepancy of more than 10% (for periods >0.5 sec) is identified between ARS Online results and the 2008 USGS Deaggregation Tool (Beta), then the 2008 USGS interactive Deaggregation Tool (Beta) results must be used, otherwise use the ARS Online results.

j) If ARS Online is appropriate for use, then the probabilistic spectrum generated from ARS Online needs to be verified with the Probabilistic Response Spectrum Spreadsheet tool. Refer to Section 1.1.17a for details.

k) If the USGS Deaggregation Tool results are used, place the spectral acceleration data into the Probabilistic Response Spectrum Spreadsheet so that the near-fault factor and basin factor can be easily applied to the results. Refer to Section 1.1.17b for details.

Section 1.1.17: Procedure for verifying ARS Online results or to adjust the USGS Deaggregation results:

a) To verify the ARS Online results: • Open the Probabilistic Response Spectrum Spreadsheet tool. • Enter the latitude and longitude coordinates of site (obtained earlier from ARS

Online). • Input VS30. • Enter the appropriate R distance determined from the USGS Deaggregation Tool. • Input Depth to Z1.0 (meter) & Z2.5 (kilometers), if applicable. • Copy ARS Online “tabular data” and paste data into spreadsheet. • Compare ARS Online tabular data with four spectral values from USGS. • If a discrepancy of 10% or less between the USGS spectral acceleration values and

the ARS Online results, then the probabilistic ARS curve developed with ARS Online is acceptable for design.

Page 23: GS Design Manual 081209

23

b) To adjust and compare the USGS Interactive Deaggregation Tool (Beta) results to ARS Online: • Open the Probabilistic Response Spectrum Spreadsheet tool. • Enter the latitude and longitude coordinates of site (obtained earlier from ARS

Online). • Input VS30 • Enter the appropriate R distance determined from the USGS Deaggregation Tool. • Input Depth to Z1.0 (meter) & Z2.5 (kilometers), if applicable. • Copy ARS Online “tabular data” and paste data into spreadsheet. • Enter the deaggregation results into the “INPUT USGS Deagg. Spectral Accel.”

Column (see figure below). Near-fault and basin effects are automatically applied in the adjacent columns, but verify that they are applied correctly.

• If a discrepancy of more than 10% (for periods >0.5 sec) is identified between ARS Online results and the 2008 USGS Deaggregation Tool (Beta), then the 2008 USGS interactive Deaggregation Tool (Beta) results must be used, otherwise use the ARS Online results.

Page 24: GS Design Manual 081209

24

When using the ARS Online results, verification of the results is an important step for ensuring quality seismic design recommendations. If a discrepancy of more than 10% exists between the ARS Online results and the four USGS spectral acceleration values, then re-check calculations and web tool to determine if an input error caused the discrepancy between the two methods. If the large discrepancy cannot be resolved then alert the “checker” who will be verifying the calculations. If the checker also identifies a discrepancy between the two methods, then the checker shall make a note in the “Quality Control/Quality Assurance Documentation for Seismic Design Recommendations”, dated July 30, 2009. A copy of the QC/QA document and comments describing the discrepancy should also be submitted to [email protected] so that the ARS Online development team can review the potential bug and help resolve the issue. Section 1.1.18: Procedures for Design Response Spectrum Compare the probabilistic and the deterministic response spectra (Section 1.1.13) across the period spectrum. The upper envelope of the spectral values shall be used for the design spectrum.

Page 25: GS Design Manual 081209

25

References Applied Technology Council, 1996, Improved Seismic Design Criteria for California Bridges: Resource Document, Publication 32-1, Redwood City, California. Boore, D., 2004, Estimating VS30 (or NEHRP Site Classes) from Shallow Velocity Models: Bulletin of the Seismological Society of America, Vol. 94, No. 2, pp. 591–597, April 2004 Boore, D., and Atkinson, G., 2008, Ground-motion prediction equations for the average horizontal component of PGA, PGV, and 5%-damped PSA at spectral periods between 0.01 s and 10.0 s: Earthquake Spectra, Vol.24, pp.99-138. Bryant, W. A. (compiler), 2005, Digital Database of Quaternary and Younger Faults from the Fault Activity Map of California, version 2.0: California Geological Survey Web Page, http://www.consrv.ca.gov/CGS/information/publications/QuaternaryFaults_ver2.htm Campbell, K., and Bozorgnia, Y., 2008, NGA ground motion model for the geometric mean horizontal component of PGA, PGV, PGD, and 5% damped linear elastic response spectra for periods ranging from 0.01 to 10 s.: Earthquake Spectra, Vol.24, pp.139-172. (http://www.eeri.org/cds_publications/catalog/product_info.php?products_id=293) Chiou, B., and Youngs, R., 2008, An NGA model for the average horizontal component of peak ground motion and response spectra: Earthquake Spectra, Vol.24, pp.173-216. (http://www.eeri.org/cds_publications/catalog/product_info.php?products_id=293) DeJong, J. T, 2007, Site Characterization – Guidelines for Estimating Vs Based on In-Situ Test, University of California at Davis, Soil Interactions Laboratory. pp 1-22. http://dap3.dot.ca.gov/shake_stable/reference/Vs30_Correlations_UC_Davis.pdf Dickenson, S.E., 1994, Dynamic Response of Soft and Deep Cohesive Soils During the Loma Prieta Earthquake of October 17, 1989: Dissertation for Doctor of Philosophy in Civil Engineering at the University of California at Berkley, pp. 104-112. Fumal, T.E., 1978, Correlations between seismic wave velocities and physical properties of near-surface geologic materials in the southern San Francisco Bay region, California: U.S. Geological Survey Open-File Report 78-1067, pp. 73-88. http://pubs.er.usgs.gov/djvu/OFR/1978/ofr_78_1067.djvu Fumal, T.E., Tinsley, J.C., 1985, Mapping Shear-Wave Velocities of Near-Surface Geologic Materials, U.S. Geological Survey Professional Paper 1360, pp. 127-149. http://pubs.er.usgs.gov/djvu/PP/pp_1360.djvu Imai T, Tonouchi K (1982) Correlation of N-value with S-wave velocity and shear modulus. In: Proceedings of the 2nd European symposium of penetration testing, Amsterdam, pp 57-72. Mayne, P.W., 2007, Cone Penetration Testing State-of-Practice. NCHRP Project 20-05 Topic 37- pp. 14. Mayne P.W., Christopher, B.R, and DeJong, Jason, 2001, Manual on Subsurface Investigations, National Highway Institute, NHI-01-031.

Page 26: GS Design Manual 081209

26

Mayne, P.W., and Rix, G.J.,1995, "Correlations between Shear Wave Velocity and Cone Tip Resistance in Natural Clays." Soils and Foundations, 35 (2), pp. 107-110. Ohta, Y., and Goto, N., 1978, "Empirical Shear Wave Velocity Equations in Terms of Characteristic Soil Indexes." Earthquake Engineering and Structural Dynamics. 6, pp. 167-187. Petersen, Mark D., Frankel, Arthur D., Harmsen, Stephen C., Mueller, Charles S., Haller, Kathleen M., Wheeler, Russell L., Wesson, Robert L., Zeng, Yuehua, Boyd, Oliver S., Perkins, David M., Luco, Nicolas, Field, Edward H., Wills, Chris J., and Rukstales, Kenneth S., 2008, Documentation for the 2008 Update of the United States National Seismic Hazard Maps: U.S. Geological Survey Open-File Report 2008–1128, pp. 61. Shantz, T., Merriam, M., 2009, Development of the Caltrans Deterministic PGA Map and Caltrans ARS Online, California Department of Transportation, Sacramento, CA. http://dap3.dot.ca.gov/shake_stable/references/Deterministic_PGA_Map_and_ARS_Online_Report_071409.pdf Sykora, D.W., 1987, "Examination of Existing Shear Wave Velocity and Shear Modulus Correlations in Soils." Department of the Army, Waterways Experiment Station, Corps of Engineers, Miscellaneous Paper GL-87-22. Youngs, R.R., S.J. Chiou, W.J. Silva, and J.R. Humphrey (1997). Strong ground motion attenuation relationships for subduction zone earthquakes, Seism. Res. Letts., v. 68, no. 1, pp. 58-73.