UNCLASSIFIED

AD NUMBER

AD874828

NEW LIMITATION CHANGE

TOApproved for public release, distributionunlimited

FROMDistribution authorized to U.S. Gov't.agencies only; Administrative/OperationalUse; MAY 1970. Other requests shall bereferred to Naval Civil Engineering Lab.,Port Hueneme, CA.

AUTHORITY

USNBC ltr, 19 Jun 1978

THIS PAGE IS UNCLASSIFIED

THIS REPORT HAS BEEN DELIMITED

AND CLEARED FOR PUBLIC RELEASE

UNDER DOD DIRECTIVE 5200,20 AND

NO RESTRICTIONS ARE IMPOSED UPON

ITS USE AND DISCLOSURE,

DISTRIBUTION STATEMENT A

APPROVED FOR PUBLIC RELEASEI

DISTRIBUTION UNLIMITED,

'7N4

CR69.024

NAVAL CIVIL ENGINEERING LARORATORYPortHueneme, California

VI Sponsored by? -NAVAL ELECTRONIC SYSTEMS COMMAND

LUL

DESIGN OF PRESSURIZED FUEL

STORAGE TANKS TO RESIST

AIR OVERPRESSURES

May 1970

An Investigation Conducted by

T. Y. LIN AND ASSOCIATESVan Nuys, California 91401 DN62399-69-C-0017

V K SEP 28 1970

Each transmittal of this document 6utside the =liE iagencies of the U. S. Government must haveprior approval of NAVELEX CODE PME-117.

akt&*U2Qd

DESIGN OF PRESSURIZED FUEL STORAGE

TANKS TO RESIST AIR OVERPRESSURE

by

WILLIAM E. GATES

for

U.S. NAVAL CIVIL ENGINEERING LABORATORY

under

CONTRACT NBy 62399-69-C-0017

withT.Y. LIN AND ASSOCIATES

Consulting Engineers

14656 Oxnard StreetVan Nuys, California 91401

May 1970

Each transmittal of this document outside the agenciesof the U.S. Government must have prior approval ofNAVELEX CODE PME-117.

FOREWORD

This report constitutes a portion of the ProjectClarinet-Sanguine facilities research program which wasaccomplished through NCEL by contractural and in-houseeffort. The facilities research program was sponsoredby the Earth Sciences Division of the Office of NavalResearch (ONR) under their field projects program.ONR was charged by the Sanguine Division of the NavalElectronic Systems Command with the general managementof the overall research program for Project Sanguine.

This facilities research program consists of workunits as follows:

Facility System StudyAir Entrainment SystemWater WellsAir System ComponentsFuel Storage ContainersComputer Output DisplayHardness of Buried CableVertical Grounds

These work units encompass areas in which improvedtechnology was needed or in which significant cost re-duction might be achieved through improvement in extantknowledge. The research program at NCEL was initiatedin early November, 1968.

General direction of the program was provided byDr. T.P. Quinn and his staff of the Earth Sciences Divi-sion, ONR, with the counsel of Mr. W.J. Bobisch of theNaval Facilities Engineering Command. Mr. J.R. Allgood ofNCEL served as project coordinator and Mr. S.K. Takahashiserved as contract monitor for the work described in thisreport.

For T. Y. Lin and Associates the Project Super-visor was Mr. W. E. Gates who authored this report.Contributors to the report were Dr. S. Yaghamai andMr. L. P. Prunotto.

i

ABSTRACT

This report presents a design procedure and cost op-timization curves for cylindrical steel tanks with spher-ical end caps that are buried with their axis parallel tothe ground surface. The fuel storage tanks are designedto resist high overpressures associated with nuclear airblast. Internal pressurization has been utilized to min-imize material and fabrication costs and increase thebuckling capacity of the tanks.

The design procedure is believed to be valid foroverpressures ranging from 10 psi to 3000 psi. The over-all reasonableness of the design procedure has been con-firmed by experimental model tests on buried steel cylin-ders and pressurized tanks in granular materials.

ii

TABLE OF CONTENTS

Page

Foreword iAbstract iiTable of Contents iiiList of Figures viiList of Symbols xi

Section 1

INTRODUCTION

1-1 Background 1-1

1-2 Objective of Report 1-2

* 1-3 Scope of Report 1-4

1-4 Organization of Report 1-5

1-5 Conclusions, Recommendations, andDesign Criteria 1-7

Section 2

RECOMMENDED DESIGN PROCEDURE

2-1 Introduction 2-1

2-2 Nuclear Weapons Effects 2-12-2.1 Air Blast Pressures 2-12-2.2 Crater Induced Ground Shock 2-52-2.3 Air Blast Induced Ground Motion 2-72-2.4 Attenuation of Dynamic Stress

with Depth 2-72-2.5 Pressure Amplification in

Saturated Soils 2 -102-2.6 Soil Arching 2-122-,2.7 Dynamic Amplification 2-18

iii

Page

2-3 Yield Load Capacity of Tank 2-22

2-4 Buckling 2-24

2-5 Internal Pressurization 2-25

2-6 Dead Load of Soil 2-25

2-7 Design Procedure 2-252-7.1 Proportioning of Tank Thickness 2-252-7.2 Discussion of Design"Procedure 2-312-7.3 Backfill Requirements 2-322-7.4 Design for Special Soil Conditions 2-32

2-8 Design Examples 2-33

2-9 Corrosion Protection 2-45

2-10 Mechanical Consideration for InternalPressurization 2-462-10.1 Compressor 2-472-10.2 Piping 2-472-10.3 Valves 2-472-10.4 Access 2-48

2-11 Site Investigation 2-48

Section 3

OPTIMIZATION STUDY

3-1 introduction 3-1

3-2 Cost Factors 3-13-2.1 Tank Costs 3-13-2.2 Excavation Costs 3-23-2.3 Backfill Costs 3-53-2.4 Pressurization Costs 3-5

3-3 Parameters Considered in Study 3-7

iv

Page

3-4 Results from Case Studies 3-103-4.1 Case 1 - Optimum Costs as a Func-

tion of Overpressure & Tank Volume 3-10- 3-4.2 Case 2 - Influence of Water Table

Depth on Costs 3-153-4. 3 Case 3 - Influence of Internal

Pressure on Tank Costs 3-213-1J.4 Case 4 - Influence of Tank Yield

Strength 3-213-4.5 Case 5 - Influence of Soil Condi-

tions on Excavation Costs 3-25

3-5 List of Major Findings and Conclusions 3-30

References R-1

R-2

Appendix A

COMPUTER PROGRAM FOR DESIGN AND COST OPTIMIZATION

A-i Introduction A-i

A-2 Computer Program A-2

A-3 Sample Problems A-3

Appendix B

COMPARISON OF EXPERIMENTAL RESULTS

WITH EMPIRICAL EQUATIONS

B.-l. Test Description B-1

B-2 Tank Stresses B-i

B-3 Arching B-6

B-4 Buckling B-6

v

Page

Appendix C

RECOMMENDED SOILS INVESTIGATION

C-1 Introduction C-1

C-2 Design Requirements C-1

C-3 Recommended Soils Engineering andEngineering Geology Investigation C-1C-3.1 General C-1C-3.2 Site Selection C-2C-3.3 Geological Survey C-3C-3.4 Drilling and Sampling Program C-4C-3.5 Geophysical Survey C-10C-3.6 Laboratory Testing C-11

C-4 Engineering Analysis and ReportPreparation c-18

C-5 Services During Construction C-20

Distribution List D-1

vi

LIST OF FIGURES

Figure No. Title Page

1-1 Longitudinal Section of BuriedFuel Tank 1-3

1-2 Tank Thickness as a Function ofPeak Overpressure Level and TankDiameter 1-6

1-3 Finite Element Model for HighWater Table 1-10

1-4 Finite Element Model for RockSite 1-12

2-1 Buried Tank Notation 2-2

2-2 Normalized Overpressure WaveForms - IMT Surface Burst 2-3

2-3 Triangular Representations ofOverpressure - Time Curves 2-4

2-4 Duration of Data for Represen-tation of Overpressure - TimeCurves - IMT Surface Burst 2-6

2-5 Change in Pressure - Time Curveswith Depth 2-8

2-6 Tank Below Ground Water Table 2-11

2-7 Shallow Excavation with MoundedBackfill 2-13

2-8 Soil Arching 2-14

2-9 Graph of Arching Function 2-19

2-10 Natural Modes of Vibration forTank Cross Section 2-21

2-11 Pressure Distribution for DesignCapacity of Tank 2-23

vii

Figure No. Title Page

3-1 Total Cost Per Pound of T-lSteel Plate - IncludingMaterial, Fabrication, Plac-ing and Corrosion Protection 3-3

3-2 Excavation Costs Vs. Depth

of Total Excavation 3-4

3-3 Excavation and Backfill 3-6

3-4 Parameters of Tank Opti-

mization Study 3-9

3-5 Total Tank Cost as a Functionof Peak Overpressure Level andTank Diameter - Case 1 3-12

3-6 Total Cost Per Tank Installedas a Function of Peak Over-Pressure Level and Tank Diam-eter - Case 1 3-13

3-7 Normalized Total Tank Costs asa Function of Peak OverpressureLevel and Tank Diameter - Case 1 3-14

3-8 Plate Thickness as a Function ofWater Table Depth and Tank Diam-eter - Case 2 (p = 3000 psi) 3-16

3-9 Plate Thickness as a Function ofOverpressure Level and Tank Diam-eter - Case 2 (Full Water Table) 3-17

3-10 Total Cost Per Tank as a Functionof Water Table Depth and TankDiameter - Case 2 (po = 3000 p:si) 3-18

3-11 Total Cost Per Tank as a Functionof Water Table Depth and TankDiameter - Case 2 (Po = 1000 psi) 3-19

3-12 Total Cost Per Tank as a Functionof Overpressure Level and TankDiameter - Case 2 (zw = 0) 3-20

viii

Figure No. Title Page

3-13 Normalized Total Tank Costs asa Function of Water Table Depthand Tank Diameter - Case 2(po = 1000 psi) 3-22

3-14 Plate Thickness as a Function ofInternal Pressure, Overpressureand Tank Diameter - Case 3 3-23

3-15 Total Cost Per Tank as a Functionof Internal Pressure, Overpressureand Tank Diameter - Case 3 3-24

3-16 Plate Thickness as a Function ofYield Stress, Overpressure Leveland Tank Diameter - Case 4 3-26

3-17 Total Cost Per Tank as a Functionof Yield Stress, OverpressureLevel and Tank Diameter - Case It 3-27

3-18 Normalized Total Tank Costs as aFunction of Peak OverpressureLevel and Tank Diameter - Case 4 3-28

3-19 Excavation and Backfill Costs forRock Site Versus Sand or Clay SoilConditions - Case 5 3-29

ix

APPENDIX A

Figure No. Title Page

A-1 General Logic Diagram for Designand Cost Optimization Program(TKOPT) A-2

APPENDIX B

B-i Model Test of InternallyPressurized Buried Fuel Tank -

Static and Dynamic Tests B-2

B-2 Comparison of Tank Stresses FromStatic Test Results and EmpiricalFormula 3

B-3 Comparison of Tank Stresses FromDynamic Test Results and EmpiricalFormula B-5

APPENDIX C

C-i Soil Sampler C-5

C-2 Piston Sampler C-7

C-3 Unconfined Compression andTriaxial Compression Tests C-12

C-4 Direct Shear and Friction Tests C-13

C-5 Percolation Test C-14

C-6 Consolidation Tests C-16

C-7 Dynamic Triaxial Test C-17

x

LIST OF SYMBOLS

A = arching factor

A = geometry factor,o g

A = maximum active arching

A = plan area of tank

B = buoyant uplift force

C = soil cover thickness

C = propagation velocity of peak stress wave5

Ct = compression wave velocity of the tank material

C = seismic velocity of the soil layer above the water1 table

C = seismic velocity below the water table2

D = mean diameter of tank

DF = dynamic amplification factor

D = total positive phase duration of the pressureP pulse

do = deDth of burial, from ground surface to crown

E = soil cover thickness

E = dynamic secant modulus of the soil in ones dimensional compression

Et = elastic modulus of the tank

e = Naperian constant

f = yield stress

H = depth of excavation

I = moment of inertia of the tank section

L = length of tank

xi

(L = depth factorw

M = confined compression (secant) modulus of soil =c E s/0.577

n = experimentally determined constant

P = overpressure

Pcr = critical pressure

PDL = dead load pressure

Pe = effective vertical interface pressure at crown oftank

Pi = internal pressure

PO = peak surface overpressure

Pv = uniform vertical pressure in free-field at eleva-tion of crown

Pz = vertical free-field soil pressure at a depth, z

Pz = vertical soil pressure at water table

r = radius of tank cylinder

S = plan perimeter of tank

T = natural period of tank vibration

t = thickness of tank, time

td = positive time duration of equivalent triangularair blast impulse

ti = positive time duration of triangular pulse whichhas the same total impulse as the actualpressure-time curve

t = positive time duration of triangular impulse

50 which has the same time ordinate at 50 percent of

peak overpressure as the actual pressure-timecurve

t = positive time duration of triangular impulsewhich has the same slope as the initial decayportion of the pressure-time curve

xii

I:

<1

W = weapon yield

z = depth below ground surface

z = depth of water table below ground surface

1.5 Ma = Ln C

EtI/D3

a = depth attenuation factors unit weight of unsaturated soil

Ysw unit weight of saturated soil

Y unit weight of tank material

Y unit weight of water

A = horizontal expansion of tank diameterX

A = vertical expansion of tank diameter

6 = relative deflection between tank and free-fieldsoil

=1 ductility factor

V = Poisson's ratio

s = mass density of the soil

Pt =mass density of the tank

= unit mass of soil above the water table

= unit mass of saturated soil below the water table2

a h =hoop stress

C = meridional stressm

a = yield stress of tank materialy

M= A c

ge

xiii

* = geometry -stiffness factorI

= angle of internal friction of the soil

xiv

~Section 1

~INTRODUCTION

liqu-i Background

The safe underground storage of large quantities of! liquid fuel is a key factor in the design of certain de-fense facilities which must maintain full operation in

the event of a nuclear attack. In high-overpressure re-gions of a nuclear blast, underground fuel tanks design-ed according to normal practice will require-thick steelshells heavily reinforced with stiffeners to prevent in-ward buckling of the vessels. The material and fabrica-tion costs of such structures are very high.

Preliminary studies indicated that major savings inmaterial costs could be achieved by internally pressur-izing the fuel tanks. In essence, the vessels are pre-stressed in tension to partially compensate for the highcompressive forces induced in them under nuclear blast.Internal pressurization also stiffens the vessel, thusmaking it less susceptible to buckling.

A program of investigation was set up to achievethe following objectives:

1. To determine the feasibility of using pressur-ized flexible tanks buried near the earth'ssurface as an economical and blast-resistantmethod for storing fuel.

2. To develop design criteria for undergroundfuel storage facilities in the high over-pressure regions of a nuclear blast.

3. To develop a finite element computer programwhich was applicable to the analysis of under-ground liquid storage containers. This pro-gram would serve as an analytical check onthe design.

It was determined that verification of the aboveobjectives could most readily be achieved through aseries of static and dynamic tests on model fuel stor-age containers. The tests were designed to simulate

1-1

the soil-structure-liquid interaction under nuclear blastloading. Results from the tests were used to:

1. Evaluate the blast-resistant propertiesprovided by internal pressurization.

2. Establish design criteria for prototypestructures based on the failure modesof the model and on the use of modelsimilitude laws.

3. Provide experimental data which couldbe used to verify the analytical modeland computational techniques used inthe computer program.

The results of the model tests and the analyticalprocedure used in the computer program are presented inseparate reports. (6,17)

1-2 Objective of Report

The first objective of this report is to present aprocedure for the design of shallow buried, internallypressurized, fuel storage tanks constructed from steelcylinders and hemispherical end caps. The tanks areoriented horizontally and located at a depth of one di-ameter as shown in Figure 1-1. The environmental con-ditions to which the design procedure must be applicableare those associated with surface or near-surface nucleardetonations at ranges which produce overpressures as highas 3000 psi. In addition, dead weight of soil backfilland buoyant forces of high water table are included whenappropriate in the environmental loading.

The soil conditions considered are typical of those

known to exist at potential sites for Naval defense fa-cilities.

The second objective is to provide cost optimiza-tion data on pressurized fuel storage tanks for variousoverpressures as a function of tank volume.

It was required that the design procedure and opti-mization data be presented in such a form that it couldbe immediately useful in a.design office.

1-2

!."

I

NUCLEAR AIR BLAST PRESSURE

po(

do= DWATER TABLE (VARIABLE DEPTH)

INTERNALFUEL TANKPRESSURE

L= 2D

Figure I-I LONGITUDINAL SECTION OF

BURIED FUEL TANK

1-3

1-3 Scope of Report

To pursue the first objective, that of providing adesign procedure for buried fuel storage containers, theresults of' model tests on horizontally buried, internallypressurized steel tanks and horizontally buried steel cyl-inders were reviewed. Most of the available test data isfor horizontal steel cylinders in sand (10) unpressurized,and without end cap constraint. To supplement these re-sults, theoretical studies on buried fuel tanks were car-ried out by dynamic finite element methods on a linearlyelastic tank, liquid, and soil system. The model was sub-jected to a static internal pressure and dynamic externalpressure applied at the surface of the soil. Second orderstiffness terms normally neglected in small deflection the-ory were included in the analysis in order to approximatethe stiffening effect of internal pressurization upon thesteel shell.

Nuclear blastloading used in the design of buriedstructures is adequately covered in numerous referenceson nuclear weapons effects (011,12,13 ,14) Thus only a sum-mary of the pertinent weapon effects is presented in thismanual for the convenience of the designer. Attention isgiven to propagation of a surface air blast pressure pulsedownward through the soil and around the tank. Both at-tenuation with depth and arching effects are considered inthe design procedure.

Particular attention has been given to the problemsassociated with high water table since this is known tobe a common occurrence at many of the proposed sites.

The design criteria for determining the wall thick-ness of the tank is based upon the requirement that thetank must be capable of surviving repeated nuclear attackwithout failure. This limits the maximum combined mem-brane stresses to the values just under the yield stressof the tank steel. Experimental data (6) have shown thatthere is a significant reserve capacity in the tank to re-sist loads well above the yield capacity, and failure inthe form of leakage does not occur even after post elas-tic buckling. Under static loads produced by internalpressurization the tank is designed with a safety factorof 2.0.

The design procedure for determining the wall thick-ness of the tank is expressed as a function of the over-pressure, weapon yield, dynamic amplification factor,

1-4

internal pressure, soil dead loads, stress attenuation,with depth, arching factor, hydrostatic loads, depths ofburial, tank diameter, elastic soil properties and elas-tic tank properties. The procedure is presented as adetailed step-by-step method that can be used by a de-signer having only limited familiarity with protectiveconstruction. However, it is assumed that the designeris familiar with steel pressure vessel design.

To further simplify the designer's task a computerprogram was developed which will solve for the tank wallthickness. The program was used to generate a designchart giving tank thickness as a function of diameterand overpressure. (See Figure 1-2).

To achieve the second objective a series of comput-er solutions were carried out to design and cost evalu-ate horizontal steel tanks buried at one diameter ingranular soils. Five water table conditions were con-sidered:

1. No~water present (i.e. water table at 100

feet below ground).

2. Water at 30 feet below ground.

3. Water at 13 feet-below ground.

4. Water at 5 feet below ground.

5. Water at ground surface.

Two steel yield strengths were considered, one 90 ksithe other 36 ksi. Two cases of internal pressurizationwere evaluated, one with pressure the other without. Inall of these design optimization studies tank volume andoverpressure level were variables. The weapon yield andsoil properties were held constant.

1-4 Organization of Report

This report has been organized primarily for theconvenience of the designer. Hence, the body of the re-port contains only the design procedure, example problem,optimization curves, and design recommendations. Allsupplementary material and supporting data are presentedin the appendices.

1-5

SHOP__FELFABRICATED FARCATED6.

j,3 5* D TANK (fy 90Oksi)w

z

z14 2D

IL.0

SOIL MODULUS 337000 psi _____ __

cn

z WATER TABLE7

2 31

w

.

0 2.

-150

-90

0060

8 10 12 14 16 18 20 22 24 26 28 30DIAMETER OF TANK (FEET)

5 10 20 50 100 200 250CAPACITY OF TANK (1,000 GALS)

Figure 1- 2 TANK THICKNESS AS A FUNCTION OF PEAKOVERPRESSURE LEVEL AND TANK DIAMETER

1-6

Section 1 contains background information, state-ments of objective and scope of this report, organiza-tion of material, and a summary of the major designrecommendations.

Section 2 presents the basic formulation and step-by-step design procedure for calculation of tank thick-ness along with design examples. Attention is given tonon-granular soils such as clay, silt and rock althoughthe basic design procedure is primarily applicable togranular materials. Corrosion protection, backfillingprocedures, mechanical consideration and site soil eval-uation procedures are also discussed.

Section 3 presents a series of parametric studiesto evaluate the minimum cost for tanks under specificdesign conditions. Cost optimization curves are in-cluded for various overpressure levels and tank volumes.

Appendix A contains a general flow diagram for thecomputer program used in tank design and cost optimiza-tion. Two example computer solutions are included.

Appendix B is devoted to the correlation of exper-imental data with theory.

Appendix C presents detailed recommendation on sitesoil investigation procedures.

1-5 Conclusions, Recommendations, and Design Criteria

The significant conclusions from study of test data,design recommendations, and associated design criteriaare summarized below:

1. Internal pressurization has a distinct eco-nomical and structural benefit. It reducesthe tank thickness to such an extent thatmaterial and fabrication cost savings faroutweigh the added maintenance cost of in-ternal pressurization plus the initial costsof compressor equipment and special pressurereducer valves.

2. Internal pressurization provides an elasticrestoring force within the tank which pre-vents compressive buckling of the tank inthe elastic range. Post elastic buckling

1-7

is possible, however, soil arching developsas the tank buckles, preventing the tank~from rupturing.

3. Horizontally oriented tanks are superior tovertically oriented tanks in resisting nu-clear blast loads. The difference in per-formance is attributed to the formation ofa positive soil arch for horizontal tanks,while vertical tanks, behave as stiff in-clusions.

l. Optimum depth of burial for tanks in densefill is approximately one tank diameter.(")0

At this depth advantage can be taken ofearthcover in reducing the forces that aretransmitted to the tank by the soil aroundit. Usually there is little advantage tobe gained from a depth of burial greaterthan one diameter.

5. The design procedure presented here, as-sumes bending moments and local stressconcentration to be of second order mag-nitude due to the relatively thin crosssection required with internal pressur-ization. Any local yielding of the tankskin due to these second order effectsis permissible under nuclear blast load-ing.

6. It has been specified that the tanks bedesigned to resist repeated nuclear airblast loading. To fulfill this criteriathe principal membrane stresses are heldbelow the yield stress of the tank mate-rial. The critical membrane stress, forwhich the tank is sized, is the spring-line thrust and axial compression.

7. A safety factor of 2.0 is applied to thedesign when checking for internal pres-sure loading alone. Thus for optimum de-sign the effective static external pres-sure on the crown of the tank is threetimes the internal pressure.

1-8

8. For surface overpressure levels less than200 psi it is probable that the final de-sign of the cylinder will be controlled bynon-blast-load conditions such as soil deadload forces, imposed by construction equip-ment and surface vehicular loads.

9. In areas where there is a potential for highwater table, precautions should be taken toprevent buoyant uplift. This may be accom-plished by providing a dewatering system,tie downs, or cut and cover techniques whichplace the tank well above the critical waterlevel.

10. The design procedure presented in this reportis directly applicable to granular soils ina dry or saturated state. However, for tankslocated just above the water table, a moredetailed analysis than outlined here may berequired in order to determine the possibleamplification of blast induced pressurescaused by upward reflection at the water ta-ble. Ideally, the tank should be checked bydynamic finite element procedures similar tothose used in predicting and evaluating themodel tank tests reported in reference 6. Theanalytical model would be composed of threebasic elements: soil, liquid, and tank steel.A schematic drawing of the finite elementgrid is shown in Figure 1-3.

11. There are no available test data for pressur-ized tanks buried in clay. In fact, thereare insufficient data available for horizon-tal cylinders buried in clay to formulate aspecific design procedure. (10) What data areavailable indicate that moments and thrustsinduced in the cylinders buried in clay areconsiderably higher than those which wouldbe produced under the same circumstances in acylinder buried in sand. Thus, it is recom-mended that tanks in clay or other non-granular soils be backfilled with denselycompacted granular material. The tanks canthen be treated as though they were buriedin sand, resulting in a more economical de-sign.

1-9

TANK AND SOIL FINITE ELEMENT GRIDCROSS SECTION

Po (t)

SOIL ELEMENTS

SOIL

TANK

FUEL___

WATER TABLE :~.::.::.:.. ".~

LIQUID ELEMENTS ::::::::':

Figure I-3 Ir"INITE ELEMENT MODEL

FOR HIGH WATER, TABLE

1-10

12. For rock sites it will be necessary to over-cut the excavation and backfill with densegranular material. A concrete cap should beprovided which keys into the rock to preventthe tank from rebounding out of the ground.Because of possible amplification of blastinduced pressures at the rock interface, thedesign procedure developed in this report isnot fully applicable. An ideal analysis un-der these circumstances would be by dynamicfinite element (6) utilizing a model similarto the one shown schematically in Figure 1-4.

13. Based on economic comparisons, high strengthsteels with yield stresses in the 90,000 psi

-range are more economical for pressurizedtanks at high overpressure levels than nor-mal structural steel. Material and fabrica-tion costs double in going from 36,000 psito 90,000 psi steel, while the yield strengthincreases by two and one-half times. Thus,it is possible to get more strength for lessmoney.

1-li

TANK AND SOIL FINITE ELEMENT GRIDCROSS SECTION

P0 (t)7

SOIL ELEMENTS

NORMAL ! """

BACKFILL . ....... ;: ,,.

CONCRETE

SELECTBACKFILL LQI

Fgr 4 FI EELEMENTS

ROCKOCK

TANK -...- ELEMENTS

Figure 1-4 FINITE ELEMENT MODEL

FOR ROCK SITE

1-12

Section 2

RECOMMENDED DESIGN PROCEDURE

2-1 Introduction

In this section recommendations are given for thedesign of buried fuel storage tanks to survive the envi-ronmental loadings associated with nuclear blast inducedsurface overpressures ranging up to 3000 psi. The designprocedure which is developed and demonstrated by exampleis directly applicable to internally pressurized steeltanks, horizontally buried in dry granular soil. Themethod has been extended to saturated granular soils.However, no substantiating experimental data exists forthe saturated soil case.

Figure 2-1 depicts the tank under nuclear blastloading. The ground surface pressures, Po, generated byan air blast are normally defined as a function of'weap-on yield and distance from detonation. The effectivesoil pressure at the crown of the tank, Pe, is simply theground surface pressure modified by the depth of burialand soil conditions. Due to the dynamic nature of theair blast loading, interaction between the buried tankand the surrounding soil must also be considered in ar-riving at the effective design loads.

2-2 Nuclear Weapons Effects

The nuclear weapon effects which are of major con-cern in the design of horizontally buried steel tanksare: air blast pressures, air blast induced groundshock, and directly-induced ground shock.

2-2.1 Air Blast Pressures

The general shape for the surface, air blast induc-ed, overpressure-time function is shown in Figure 2-2 asa family of normalized curves for various overpressurelevels. For design purposes a triangular pressure-timefunction, shown in Figure 2-3, may be used in place ofthe exact load function. To evaluate the normalizedoverpressure-time functions, the following data must beknown:

2-1

-P PEAK SURFACE OVERPRESSURE

GROUND SURFACE

do ZPe, EFFECTIVE PRESSURE-' Pz

Pit INTERNAL -- PiD PRESSURE

L 2D

WATER TABLE

LONGITUDINAL SECTION

- - -CROWN

SPRINGLINE

INVERT

CROSS SECTION

Figure 2-1 BURIED TANK NOTATION

2-2

1.0-

0.

0.9 P0 = Peak or shock overpressure

P =Overpressurej0.8 D+= Duration of positive overpressure

0 p0.t Time

w0.7

w 0.6a:

0

0N. . .0.4 0.5 0. 0.p0s09i.

0.-3

p

po

a.

"'p

OVERPRESSURE -TIME CURVE

0 I

0 t m t o ti Dp+

50 D50

TIME, SECONDS

Figure 2-3 TRIANGULAR REPRESENTATIONS OFOVERPRESSURE-TIME CURVES

2-4

W, weapon yield

P 0 peak surface overpressure

D + total positive phase duration of the pressurepulse

too., the positive time duration of a triangular im-pulse which has the same slope as the initialdecay portion of the pressure-time curve

t50., the positive time duration of a triangul,-r im-pulse which has the same time ordinate at 50percent of peak overpressure as the actualpressure-time curve

ti., the positive time duration of a triangularpulse which has the same total impulse as theactual pressure-time curve

Values of Dp, t., t5 0, and ti, are given in Figure2-4 for a wide range of peak overpressure levels. Thedurations shown are for a 1 MT surface burst, as indi-cated on the figure. They may be scaled for other weap-on yields by multiplying by ( ) 1/3

The triangular pulse of ti duration is normallyused when maximum response of the structure occurs afterthe pressure pulse has passed. A small triangular im-pulse of duration tw produces critical loads on verystiff structures which reach maximum response early inthe pressure history. Triangular loads of duration t50are applicable if the response of the structure is inter-mediate to those mentioned above.

If greater accuracy in analysis is desired than thatpossible with the single triangle, a multiple-trianglerepresentation of the pressure-time curve may be used.(12)

2-2.2 Crater Induced Ground Shock

The cratering action of a surface burst will producemotion of significant magnitude in buried structures lo-cated near the perimeter of the crater. These motions

2-5

5,000

5300

0 5

500 X.=.3 e. Ops

300

When 2 p 3 1'00 psi,

10 t =0.87 sec. l-s

a. When 235 p I0- O Psi,

OF OVRRSUETM CURVES- I MTSUFC BURST

t =0.872e6

will have direct bearing on the design of shock isolationequipment for personnel and equipment. However, for theoverpressures considered in this report (3000 psi), thestructural resistance of pressurized steel tanks will bemore than adequate if they are designed to resist the

* air blast induced soil pressures developed by the surfaceburst.

2-2.3 Air Blast Induced Ground Motion

The air blast induced soil stresses can be groupedinto two categories depending on the relative time ofarrival between ground shock and air blast. Super seis-mic conditions occur when the air blast velocity exceedsthe dilatational wave velocity of the soil. Under theseconditions the air blast arrives before the blast inducedground motion. The outrunning case occurs when the airblast velocity is less than the dilatational wave veloc-ity of the soil.

The characteristics of the ground motion producedunder super seismic conditions differs considerably fromthose produced under outrunning conditions. However, thedetailed difference is only significant to shock isola-tion design. As far as the buried tank design is concern-ed, the vertical blast pressure induced soil stresses, Pe,at the crown of the tank are of primary importance, andthese stresses are identical under both of the ground mo-tion regimes.

2-2.4 Attenuation of Dynamic Stress with Depth

The pressure to be used in the design is the verti-cal free field stress, Pz, at a depth corresponding tothe crown of the tank (see Figure 2-1). The magnitude ofPz is a function of the air blast characteristics at theground surface above the point of interest and of thestress-strain properties of the soil. At the ground sur--face the vertical soil stress will have the same magni-tude and time variation as the air blast overpressure. Asthe pressure wave propagates down through the soil, atten-uation of peak pressure takes place along with a lengthen-ing of the rise time and positive phase duration. Theseeffects, which are shown schematically in Figure 2-5, arecaused by the non-linear stress-strain behavior of thesoil under initial loading and by the energy absorbed inpermanent strain deformation.

2-7

GROUND SURFACEt j '

Figure 2-5 CHANGE IN PRESSURE- TIME CURVESWITH DEPTH

2-8

- = = . . . . .

The vertical free-field soil stress at depth, z,can be expressed as a function of the peak surface over-pressure and a depth attenuation factor, a in the fol-lowing manner (3,9)

= P (2-1)

in which

a for p0 < 1000 psi (2-2a)z i+0

Lw

or

____ 27.eaz 1(-.)7o 5 ) for p> 1000 psi (2-2b)z 1+ zw48)'

and

L =230 ft. (100 pSiV11 (_.L) 1/3 (C ) (2-3)\ PO 1MT (622 fps)

where:

C = propagation velocity of the peak stress wave5

Lw = depth factor

W = weapon yield

z = depth below ground surface

The effective propagation velocity of the peak soilstress during the initial loading cycle is given by:

2-9

I

C - (2-4)S

where:

Ps= mass density of the soil

E = dynamic secant modulus of the soil in5 one dimensional compression

The secant modulus is determined at stress levelscommensurate with the dynamic overpressure expected. Ifthe dynamic secant modulus is unknown, Cs may be takenas one-half the seismic velocity for surface soils abovethe water table. (10)

2-2.5 Pressure Amplification in Saturated Soils

The presence of the ground water table in the vi-cinity of the buried tank will alter the attenuation ofblast pressure. If the water table is at ground surface,no attenuation of the air blast pressure should be con-sidered (i.e., az = 1.0). (10) If the water table liesbetween the ground surface and the structure as shown inFigure 2-6, the attenuation factor, az, for the dry soil,from ground surface down to the water table at depth, Zw,should be computed by Equation 2-1. At the water table,the vertical pressure transmitted through the soil-waterinterface will be:

P2 C2P 2[2)1 cz po (2-5)zwP cl + P2C2-

where:

= seismic velocity of the soil layerabove the water table

C2 = seismic velocity below the watertable, (not less than 5100 ft/sec.)

p = unit mass of soil above the watertable

p = unit mass of saturated soil belowthe water table

2-10

.*. ... .. R DWATER TABLE

..............................

21-1

Below the water table the transmitted pressure should notbe attenuated with depth.

Due to possible pressure reflection at the watertable interface it would be conservative to consider thesoil pressure immediately above the water table to beequal to the pressure below the water table.

The presence of a high water table presents severaldesign problems in addition to the pressure amplification.Excavation and construction will require expensive dewa-tering equipment. When the tank is finally placed andbackfilled, provisions must be made to keep it in theground under buoyant uplift forces. There are three al-ternatives:

1. Provide permanent dewatering.

2. Provide massive hold down anchors.

3. Use a shallow excavation and mound the soilas shown schematically in Figure 2-7.

From a practical design standpoint, every effort shouldbe made to keep the tank above the water table.

2-2.6 Soil Arching

A steel tank buried in a homogeneous soil behavesas an inclusion, producing a discontinuity in the soilfield. If the soil is stiffer than the tank, under ver-tical loads the soil will transfer the load around thetank by arching action as shown in Figure 2-8. If onthe other hand, the tank is stiffer than the surroundingsoil, vertical loads in the soil will be transferred tothe tank, thus increasing the effective pressure actingon the crown of the tank above the normal free-fieldpressure.

By definition, arching is that fraction of the free-field soil pressure at the crown elevation of the tankwhich is transferred to or away from the tank by shear inthe soil. Or in equation form:

A ( - (2-6)

2-12

MOUNDED BACKFILL SLOPE IS A FUNCTION; OF OVERPRESSURE

kD

:' " ' : ' ' :':: ' :' .T ... .'.. ..

SD .......... ..

v-HIGH WATER TABLE

Figure 2- 7 SHALLOW EXCAVATIONWITH MOUNDED BACKFILL

2-13

... .. ... ..

Po,

INITIAL 'GROUND

STIFF RIGID TANKI NCLU SWO1'

SFT SOIL

Figure 2- 8 SOIL ARCHING

2-14l

where:

Pe = the vertical soil pressure at interfaceof tank crown

p = uniform pressure in free-field at eleva-tion of crown

The effective pressure at the crown of-the tank dueto soil arching is:

P = (- A) pv (2-7)

For tanks that are stiffer than the surrounding soil thearching factor, A, will be negative, thus producing higher

design forces than if the tank were at the ground surface.

The following empirical arching equation has been es-tablished by Gill and True (7) for buried cylinders in drydense granular soils:

e-nD (2-8)\A

where:

M= A c (2-9)

g Pe

SdA 0 (2-10)g AsD

and

A = arching factor

Aon = experimentally determined constants(Ao = 0.87 and n = 0.135 for sharpgrained sandblaster's sand)

A = geometry factorg ,

e = Naperian constant

2-15

M - confined compression (secant) modulusc of soil = E /0.5775

pe = effective interface pressure at crownof tank = (1-A) P

6 -- relative deflection between soil andtank

S = plan perimeter of tank

do depth of cover over crown

As plan area of tank

D = mean diameter of tank

For horizontally buried tanks of length L, the geometryfactor is:

4d ( D - 2D + 2L (2-11)Ag D iD2 + 4DL -4D 2

The relative deflection between tank and surrounding soilis:

= v c D -( M D (2-12)

where:A = horizontal expansion of tank diameterx

Ay = vertical expansion of tank diametery

A x= average vertical strain in the tankD

Pv

average vertical strain in the soilc

2-16

"aL

)

Substituting Equation 2-7, 2-11 and 2-12 into Equation2-9 yields the following expression for D:

4d° (rD - 2D + 2L) A Mc l) 1-0 ( c 1 1 (2-13)

(wD2 + 4DL - 4D2 ) D Pv (12-3

For convenience, Equation 2-13 may be expressed as:

1-A

where:

I4do(1D -2D + 2L) A M(0 =_x . _c 1) (2-14)

(nD2 + 4DL - 4D2 ) D Pv

Then Equation 2-8 may be rewritten as:

(1-A)( o = e (2-15)

Experimental evidence indicates that for dry .granularsoils the maximum value of active arching is:

A Tan o (2-16)

where, o is the angle of internal friction of the soil.

From elastic theory and tests data on buried cylin-ders (2) the deflection ratio in Equation 2-14 is:

A .c _ 2(1- e0024a2) (2-17)

D PV

2-17

Z<

where: ______

Ln~c

and

Et = elastic modulus of the tank

I = Moment of inertia of the tanksection = t3/12

t = thickness of tank

thus:

4d (D - 2D + 2L) e0.024a2 (2-18)0(1 - e - " 2 a ) ( -

(nD2 + 4DL - 4D2 )

To assist the designer in solving the arching Equa-tion 2-15 for the arching factor, A, a graphical plot ofthe function has been prepared in Figure 2-9.

2-2.7 Dynamic Amplification

Nuclear air blast is a short duration high intensityform of loading which excites dynamic inertia forces ina soil-tank system. Dynamic amplification of verticalsoil pressure at the crown of -1tank can be approximatedby the following relationship 3)

- 1

DF = (2-19)Ttt + 2pd 1 d Tt

td

in which the tank-soil system is idealized as a single-degree-of-freedom oscillator, having an elasto-plasticrestoring force, and excited dynamically by a triangularimpulse of instantaneous rise time.

2-18

.... .............

I f FT-i t

Ap 0.87

Ae 0.80 T- T F

03

A.F 0.70ARCHING FUNCTION-

1 1! If A I-AA,= 0.60:1 1 1 1 1 1' -4 =e-

Q6 -r

if

Ae 0.50 T_ ft _FT I

-t-t--777I f i I It I

Q 4 I, III fit I I, I Trill ;iI IIIt if tf 11 it I

f

z t t 11 If 1 11 jjj I i T

LAJ

LL TF '17Q2

1 11

I f I I t I it I I I I I I I I I I I I

I I I I I I I I I 1 111 . t I I I I I I I I I i I I I if I I I

It -it I _fTT

0 4h H1 i- ad 1.4r l i 6. -T-'-t'-H+ t t I I I 4+14+ 9.9 t 4

- 7

j ARCHING FUNCTION, 7-7 -n#-r= fill

..02 1 Tf i l l i t l I ? f j t f ' I f i t -1 1f i l l 1 1 f i t I t

7-1- i i I , i I 1 47It 11 F- F it -rl I I I Ae i.oI'll -1 1 1 f i-L If rt i l l 1 1 1 1 1 f 1 : ..1 1 1 ,J_!_L' LL I Ft I I I t i

Z ZT:4 Tp if + T

-Q4 444-1-: 51' 5-MIL

Figure 2-9 GRAPH OF ARCHING FUNCTION

2-19

The term, td, is the posityive time duration of theair blast impulse which is defined as an equivalent tri-angular impulse of durations ti, t50 , and t. in Section2-2. For design purposes select the triangular pulsewhich produces the largest dynamic amplification.

The ductility factor, p, is a ratio of the totalplastic strain in the tank under peak loading to theelastic strain at yield. Since the tanks must be design-ed to survive repeated blast loading, strains in the tankcannot exceed the yield limit. Thus, if p is set equalto one, the response of the tank to dynamic loading willbe elastic.

The term, Tt, in Equation 2-19 is the natural periodof vibration for the tank in its fundamental compressionalmode. This is the primary mode of oscillation in whichthe tank will respond to dynamic vertical soil pressures.There are ozher modes which will be excited as well, suchas the flexural modes shown in Figure 2-10. These modesare generally not significant for tanks with length todiameter ratios of two.

The tank's natural period in the compressional mode

1cD

Tt - D (2-20)

where, Ct is the compressional wave velocity in the tankmaterial which may be computed by the relationship:

, (1 - v) Et

C = ( + v) (1t- 2v) (2-21)

where:

v = Poisson's ratio

Et = elastic modulus of the tank

Pt = mass density of the tank

For steel, Ct = 18,470 ft/sec.

2-20

I+

/

COMPRESSIONAL FIRST FLEXURAL

MODE MODE

2-2

\ \.

SECOND FLEXURAL THIRD FLEXURAL

MODE MODE

Figure 2-10 NATURAL MODES OF VIBRATIONFOR TANK CROSS SECTION

2-22.

2-3 Yield Load Capacity of Tank

The critical pressure distribution on the tank isproduced by uniform pressure as shown in Figure 2-11.This pressure distribution is a very close approximationfor tanks buried below the water table. For tanks buriedin dry granular soils, the lateral pressures will normallybe smaller than the vertical pressures. Under this imbal-ance of loading a certain amount of lateral expansionwill take place until the passive soil pressure againstthe sides of the tank equalizes the vertical pressure. Atthis point the pressure on the tank may be approximatedby a uniform distribution.

Secondary bending stresses will be produced in thecylinder as it ovalates. The magnitude of these stresseswill not be significant in the design, provided the tankis buried at least one-half diameter and preferably onediameter below the surface.

The membrane stresses in the cylinder portion of thetank under uniform pressure are:

Pcr D

ah - (2-22)h 2t

and

Per D

S- (2-23)m 4t

for the hoop and meridional directions respectively.

Assuming failure criterion to be the yielding of thetank and assuming the stresses in Equations 2-22 and 2-23to be the principal stresses, then by von Mises yieldcriterion:

ah2 m+a 2 = y (2-24)-h Oh~m m y

where a is the yield str 'is in compression.

y

2-22

-Pc r Pcr

PRESSURE DISTRIBUTION

Pcr Pc

I.m(Tin T "h crh

CRITICAL MEMBRANE STRESSES

Figure 2- 11 PRESSURE DISTRIBUTION FORDESIGN CAPACITY OF TANK

2-23

By substituting Equations 2-22 and 2-23 into Equa-tion 2-24, the yield criteria reduces to a simple rela-tionship for the critical design pressure as a functionof yield stress, tank thickness, and diameter:

c 2.31t (2-25)Pcr D Gy

2-4 Buckling

There are no analytical relationships for the buck-ling capacity of horizontally buried cylindrical tankswith spherical end caps. However, experimental resultsfrom static tests (6) on internally pressurized tanks in-dicate that buckling will not occur as long as the prin-cipal membrane stresses remain below the yield limit ofthe tank material.

Chelapati (4)has derived an expression for the crit-ical buckling pressure of buried cylinder which has beenshown to agree w.-th experimental results (1) For cylin-ders of large length-to-diameter ratios, large diameter-to-wall-thickness ratios, and large values of foundationcoefficient - conditions existing in the tests (11 - thecritical buckling pressure can be expressed as:

Pcr - 0.612 (r/t 4.5 McD (2-26)(nrt )3 D3 (-6

where:

Pcr - critical buckling pressure

t = cylinder thickness

r = cylinder radius

Et = elastic modulus of the cylinder material

E = elastic modulus of the soil = 0.577 Ms c

2-24

M = one-dimensional compression secant modulusc at the surface pressures of interest

I = t3/12 = moment of inertia of the tank sec-tion

which applies to the case with all-around soil support.As a conservative check on the buckling capacity of thetank, Equation 2-26 may be used.

2-5 Internal Pressurization

The primary benefit achieved through internal pres-surization is a thinner tank thickness with resultingsavings in material, fabrication, and construction costs.Internal pressurization pre-tensions the tank shell topartially offset the compressive stresses produced in thetank under nuclear air blast.

Since the tank will be internally pressurized almostcontinuously throughout its life, except for times whenit is inspected and cleaned, it is suggested that a safetyfactor of 2.0 be applied to this design loading.

2-6 Dead Load of Soil

In the high overpressure range from 200 psi to 3000psi, the influence of soil dead load on the tank designis negligible. However, in the overpressure range under200 psi soil dead load should be considered.

2-7 Design Procedure

2-7.1 Proportioning of Tank Thickness

In this section, a design procedure for selection ofthe tank wall thickness is outlined in a step-by-stepmanner. In section 2-8, two design examples are presentedto illustrate this procedure. One example deals with thedesign of a tank in dry granular soils, the other coversthe design case in saturated granular soils where the wa-ter table is above the tank.

The design procedure presented here is limited tohorizontally oriented cylindrical tanks with sphericalend caps.

2-25

The following parameters must be known or assumed in order to complete the design:

D = diameter of the tank

L = length of tank

d = depth of burial, from ground surface0 to crown

zw = depth of water table below ground surface

O = yield stress of tank material

Et = elastic modulus of the tank

v = Poisson's ratio for the tank

Yt = unit weight of the tank material

E = elastic secant modulus of the soil

s Ys = unit weight of unsaturated soil :

Ysw = saturated unit weight of the soil

o = angle of internal friction of the soil

P0 = peak overpressure level

W = weapon yield

A = arching factor (assumed)

Steps in the design procedure are as follows:

1. Determine the positive phase duration for theequivalent triangular air blast impulse fromFigure 2-4, as suggested in Section 2-2.1.

2-26

2. Determine the tank's natural period of vibra-tion in the compressional mode by Equation 2-20:

Tt =TDCt

3. Calculate the dynamic amplification factor usingEquation 2-19:

1

DF -Tt 2 iI21+

Itd 1+ - iTtt1 + .7 --

Substitute ti, t50 and t. into the equation fortd and select the largest dynamic amplificationfactor. For an elastic system, p = 1, and themaximum dynamic amplification value which couldpossibly be obtained is two.

4. Calculate the pressure attenuation factor, a , byEquation 2-2a or 2-2b:

S l+- Lfor p < 1000 psi

= (i + z)( 8) for p0 > 1000 psi

If the crown level is above the water table calou-late az at the crown level of the tank. If thecrown level is below the water table, calculateaz at the water table.

2-27

5. If more than 3/4 of the tank diameter is belowthe water table, calculate the pressure ampli-fication factor by Equation 2-5:

zw [2 P C + P2 C2 ) az o

If less than 3/4 of the tank diameter is belowthe water table no pressure amplification fac-tor needs to be considered.

6. Compute the arching factor, A, from Equation 2-7.

a. Guess at a tank thickness or use Figure 1-2tc get an approximate value.

b. Compute the expotential term in Equation2-17:

1.5 Ma = Ln c

EtI/D3

c. Compute the arching exponent in Equation2-18:

4do(iD - 2D + 2L). 2- 0 (1 - e 0 2 4 )

(nD2 + 4DL - 4D2 )

d. Compute the right hand side of the archingfunction in Equation 2-15:

e-n = e 0 .13 5€

e. Compute the active arching factor, Ao, fromEquation 2-16:

Ao = Tan o

2-28

1,

f. Enter Figure 2-9 with the value for e-nand A and read out, A.

7. Compute the dead load pressure acting on thecrown of the tank, including the archingeffects -

With no water table present:

PDL = (1- A)ysd °

For the case with water table:

PDL = (1- A)YsZw + (1 - A)(ysw - yw)(d - Zw)

+ y (d - z)-w 0 w

where, yw is the unit weight of water.

8. Calculate the total effective pressure acting atthe crown of the tank due to all external loads.Neglect internal pressure.

For the case without water table above the tank:

= DF (1-A)z + DL

For the case with water table above the tank:

Pe DF L2 C + P C 2)] (1-A)ap 0 + pDL

2-29

(

9. Calculate the internal pressure, using a safetyfactor of 2.0:

1Pi= 3 Pe

10. Determine the critical yield pressure on thetank:

cr e -

11. Determine the required tank thickness to resist,Pcr' from Equation 2-25:

Pcr D (Pe - pi ) D

2.54a -Y y

12. If the computed tank thickness is the same asthe assumed value used in calculating the arch-ing factor, A, no further work is required.If not, use computed tank thickness and repeatsteps 6,7,8,9,10 and 11 as many times as re-quired to make the solution converge.

13. Check the critical b'uckling pressure of thetank using Equation 2-26:

p 0.621 = 4.5cr (r/t) 3 c D3

14. If the tank has been designed for an over-pressure level under 200 psi, the final. selec-tion of thickness may be controlled by a com-bination of non-blast induced loads. Checkthe combined influence of dead load, construc-tion equipment loads, normal vehicular loads,and any other surface loading which might bepresent.

2-30

To further simplify the design process, a computerprogram has been written which carries out each of thesteps, 1 through 12. It also checks the buoyant upliftforce produced on the empty tank by high water table andgives the required hold-down forces. A general flow di-agram of the program along with example solutions are pre-sented in Appendix A. This program is available throughthe U.S. Naval Civil Engineering Laboratory, Port Hueneme,California.

2-7.2 Discussion of Design Procedure

The design procedure, as presented, is based on sev-eral conservative assumptions. First, it is assumed thatthe maximum resistance of the tank section is controlledby the yield stress of the material. In actual fact thetank can resist external loads which produce plasticyielding and even buckling of the cylinder section, with-out rupturing. (6) Yield stress was selected as the designlimit in order to insure no loss of function under repeatednuclear air blast.

Secondly, dynamic increase in the strength character-istic of the tank was neglected. For mild grade steelsthe increase in yield strength can range from 20% to 40%.For high strength steels, which are strain hardened, theincrease generally is negligible.

Thirdly, only partial advantage has been made of in-ternal pressurization. The safety factor of 2.0, limitsthe internal pressure to 1/3 the total effective external

pressure acting on the tank. By reducing the safety fac-tor to 1.5, internal pressurization could be increased,and the required tank thickness would be reduced by about20%.

Dynamic forces produced by the sloshing action of liq-uid fuel stored in the tank have not been included in thedesign procedure. Experimental evidence and analyticalverification indicate that the natural period for liquidsloshing is very long in comparison with the fundamentalperiod of the tank in its compressional mode. Thus, peakresponse of the tank to blast pressure will occur long be-fore internal loads due to liquid sloshing are developed.In addition, the inertia loads generated by the sloshingliquid are small in magnitude compared to the externallyapplied soil pressures.

2-31

The design procedure outlined in this section isbased on two loading conditions, the dynamic externalblast pressure and the static internal air pressure. Forthe -donamic- loading--no s afety factor is considered otherthan the reserve capacity of the tank section above yieldstress. For internal pressure, a safety factor of 2.0has been suggested. This margin of safety is on the lowside when compared to the 2.5 which is normally specifiedbetween allowable design stresses and yield stress in theASME Boiler and Pressure Vessel Code - Section VIII - Di-vision 1. The fact that the tanks are buried tends tominimize the hazards resulting from accidental ruptureunder static internal pressure. In addition, the primarypurpose for the installation will be that of fulfilling amilitary function. These two factors tend to justify thelower safety factor suggested. However, the designershould exercise his own judgment in establishing a safetyfactor which he feels is consistent with the unique fea-tures and conditions at the site under design.

The design and fabrication of the vessel should fol-low the ASME Code recommendations as closely as possible.Particular attention should be given to the design ofpipe connections and manholes to the tank, so that properpressure vessel details are used in the nozzle geometryand weld procedures.

2-7.3 Backfill Requirements

The design procedure which has been presented is val-id only if granular material is used as backfill aroundthe tank. It is recommended that construction specifica-tions require backfill with material placed at a dry den-sity corresponding to 95% of the standard Proctor maximumdry density.

The specified thickness of dense granular backfillshould depend on the characteristics of the adjacent soilfield and should be specified by a competent soils engineer.It is recommended that a minimum thickness equal to 20% ofthe tank diameter be used all around the tank.

2-7.4 Design for Special Soil Conditions

Because of the erratic nature of measured forces intanks and cylinders buried in clayey or silty soil, it isrecommended that any tank installation in such materialbe backfilled with dry dense granular soil. The backfillmust extend below the bottom of the tank to be effective.

2-32

I!

For rock sites, over cut the excavation in length,width, and depth to allow for a generous layer of densegr nuir a i, whi c should be placed around theentire tank. Provide a concrete cap slab as shown inFigure 1-4 to prevent the tank from rebounding out ofthe ground during nuclear blast loading.

High water table is another site condition whichwill require special design considerations, as noted inSection 2-2.5. Dewatering for excavation will be a dif-ficult, if not impossible task in coarse grained materi-als because of the large potential flows. In fine siltysands dewatering might be economically feasible with theuse of well points. As noted earlier, every effortshould be taken to place the buried tank above the watertable. Under high water table conditions, the use ofseveral small diameter tanks may be more economical thanone or two large diameter tanks if the smaller tanks canbe placed above the water table.

For sites which combine high water table conditionswith loosely compacted granular soils, the potential ofsoil liquefaction due to ground motion generated by nu-clear blast or earthquake should be carefully evaluated.If the tanks were to be placed below the water table un-der these site conditions, positive measures should betaken to preconsolidate the soils through the use ofvibro-.flotation techniques. It is recommended that apositive hold-down force be included in the design ofthe tank. This could be provided by concrete encasementof the tank or by rock anchor techniques.

2-8 Design Examples

Two design examples are presented in this section todemonstrate the procedure outlined in Section 2-7. Bothexamples deal with a 16 foot diameter tank buried one di-ameter in granular soil. The weapon yield is 5 M.T. andthe overpressure level is 1500 psi. The distinguishingdifference between the two cases is the location of thewater table. In Example No. 1 it is 100 below the groundsurface, whereas in Example No. 2 it is 5 feet below thesurface.

2-33

Design Example No. 1 p

D = 16 ft. I.D. ELEV. 0

L = 2D

d = D = depth of burial doTANK

z, = 100 ft.waI = 90,000 psi, static

Et = 30 X 10 6 psi D Pi

v = 0.25

Yt = 490 lbs/cu.-ft.

E = 33,000 psis

YS = 110 lbs/cu.-ft. ELEV. -100 WATER TABLE 7 -

Ysw= 130 lbs/cu.-ft.

S = 3700

PO = 1500 psi

W = 5 M.T.

1. Determine ti , t50 , and t. from Figure 2-4 -

S0.37 sec. 00 psi] 1/2 [5 MT] /3 = 0.164 sec.ti = .7e. 1500 psi] [ -- MT

t5o 0.0195 sec. [5-M] / 0.033 sec.

2-34

t = 0.01 sec. 5 MT 1/3 0.017 sec.

2. Determine period of tank vibration in compressionalmode -

Ct = 18,470 fps

T, (16 ft.) 0.0027 sec.r 18,470 fps

3. Calculate dynamic amplification factor -

Let = , linear-elastic system

For T Tt t 0.0027Sd t. 0.164

d 1

DF = = 2.0.0027 + 1/2T .17)0027

1+ .7

This is the maximum value which DF can have in anelastic system so there is no need to consider t5 0or t0.

4. Calculate the pressure attenuation factor, az , atthe crown -

Cs = s Ys

Ca= '(32.2 ft/sec2)(3Opc psi)(144 in2 /ft2 )

Cs=V (110 pcf)

CS = 1180 fps

2-35

L = 230 , 100psi) 1/2 ( 2 f/3

Lw = 193 ft.

For p0 greater than 1000 psi use Equation 2-2b

__ 1 27.5 1 . 27.5

__ PcO . 19.3 ft) (1500)0.48

a = 0.759z

5. Calculate pressure amplification factor due to watertable. In this case the water table is 100 feet be-low ground surface. Therefore no stress amplifica-tion will occur.

6. Compute arching factor, A -

From Figure 1-2 select a trial tank thickness of1.00 inch.

a = Ln~ c(E \I/D /

Ln 1.5(33,000 psi)(16 ft x 12in/ft) 3a Ln: 12.40

(30 x 106 psi)(l in3/12)(0.577)

4d (gD - 2D + 2L) -0 024a 2

(gD2 + 4DL - 4D2 )

for L = 2D

4d° (t + 2) -0. 024a 2

= D(7 + 4) ( - e )

2-36

4(16 ft)(i + 2) 12 ' 2L_= (16 ft},ir ,. s (i - eO-t(2.) )

-€ = 2.81

-no -0.135(2.81)e e : 0.684

A = Tan 370 0.7535

Enter Figure 2-9 with these values for e- and Aand read, A = 0.315. o

Thus, (1-A) = 0.685

7. Calculate dead load of soil -

PDL (1- A) y d

- 0.685 (110 pcf)(16 ft)

144 in2/ft2

PDL 8.37 psi

8. Calculate the total effective external pressure -

Pe = DF (1-A)a Zp° + PDL

= 2.0(0.685)(0.759)(1500 psi) + 8.37 psi

Pe = 1562 psi + 8.37 psi = 1570 psi

9. Determine internal pressure -

Pi 1570 psi 523.5 psi

2-37

10. Calculate critical yield pressure -

Pcr = 1570 psi - 523.5 psi 1056 psi

11. Determine tank thickness-

t PcrD = (1056 psi)(16ft x 12in/ft)2.31 cT y 2.31 (90,000 psi) .967 in.

12. The closest rolled plate thickness is 1 inch, so use1 inch plate. There is no need to go through anothercycle of iteration since the arching factor is basedon an assumed tank thickness of 1 inch.

13. Check for elastic buckling -

EPcr = 0.621 v (r/t) 3

= 0.621 A/(33,00psi)(3OxlO6psi)(l in)3 = 657 psi

V (8 ft x 12in/ft) 3 6

Thus, the critical pressure for elastic buckling'isless than the net external pressure of 1056 psi cal-culated in Step #10. Based on these results, thetank should buckle elastically. However, the buck-ling equation is very conservative in this case. Itwas derived for long thin-walled cylinders withoutstiffening end caps. The actual tank under analysishas a short cylindrical section and is stiffened bythe two spherical end caps. Static tests on buriedmodel tanks, internally pressurized (6)indicate thebuckling resistance of the tank to be three times aslarge as the value given by Equation 2-26.

Using these test results as a guide, the criticalbuckling pressure would be predicted as-

Pcr = 3 (657 psi) = 1970 psi

which is well above the design pressure of 1056 psi.

2-38

Design Example No. 2D 16f.WMRID = 16 ft. _ f

L 2DdDd o 16'd= D

' 0

z = 5 ft. "1

y = 90,000 psiyD 16 'P

Et = 30 x 10 6 psi

v = 0.25

= 490 ibs/cu.-ft.t--OAULAR SOIL

Es = 95,000 psi

's = L10 lbs/cu.-ft.

y, 130 lbs/cu.-ft.

0 =410

PO 1500 psi

w = 5 M.T.

1. Determine ti, t 5o and t. from Figure 2-4

t = 0.164 sec.

t50 = 0.033 sec.

t =0.017 sec.

2-39

2. Determine period of tank vibration in compressional

mode -

Ct = 18,470 fps

T v(i6 ft) - 0.0027 sec.t 18,470 fps

3. Calculate dynamic amplification factor - this willbe the same as in Example No. 1

DF = 2.0

4. Calculate the pressure attenuation factor, az' atthe water table-

C(32.2 ft/sec2)(95,000 psi)(144 in2/ft2)'

s V (110 pcf)

Cs = 2000 fps (for soil above the water table)

= 230 ft. 100 psi) '/2 (5 MT) 1/3 2ooo fps)Lw 20 f.(1500 psi) 1 MT 622 fps)

Lw = 327 ft.

a = 1 ft 27.51+ •_ " - 0. 810+ 327 ft (1500)0'48

This is the pressure attenuation down to the watertable.

5. Determine the pressure amplification factor at thewater table -

P 2C1 + p 2 C2 Y1C1 + Y2C2

2-40

2c10(o130 pcf)(510 fps)

2pcf)(2000 fps) + (130 pcf) (5100 fps)]

=1.5

6. Compute arching factor, A-

Assume a plate thickness of 1.5 inches

a = Ln

Ln 1.5(95,000 psi)(16 ft x 124/ft)3 (12)[ (30 x16psi)(.282 in)3(0 .577) Ja =Ln 2.48 X 10 12.42

=4d Or +2) (1 e-0024a2

D(ir + 4)

4 ~(16 ft)(ff + 2) (1 - e-0024(12.4+2) 2)(16 ft)(ir + -)

* = 2.81

e-n = e- 0 .135(2 "81) = 0.685

A = Tan 41 = 0.870

Enter Figure 2-9 with this value and read, A = .401

Thus, (1-A) = 0.599

2-41

7. Calculate dead load of soil -

PDL = (1 - A)y szw + (1 - A)(y - yw )(d - zW )

+ yw (d0 - zW )

(0.599)(110 pcf)(5 ft)

(144 in2/ft 2 )

+ (0.599)(130 pcf - 62.4 pcf)(16ft - 5ft)

(144 in 2/ft2 )

+ (62.4 pcf)(16 ft - 5 ft)

(144 in 2/ft2 )

PDL = 2.30 psi + 3.12 psi + 4.74 psi = 10.16 psi

8. Calculate the total effective external pressure -

Pe = DF 2 [PlCl + P2C2)] (1 - A)azPo + PDL

= (2.0)(1.5)(0.599)(0.810)(1500 psi) + 10.16 psi

Pe = 2192 + 10.16 = 2202 psi

9. Determine internal pressure -

S 2202 = 734 psip-3

2-412

10. Calculate critical yield pressure -

cr = 2202 psi - 7 34 psi = 1468 psi

11. Determine tank thickness -

= PcrD (1468 psi)(16 ft. x 12 in.)2.31 2.31 (90,000 psi)Y

t = 1.36 inches

12. The closest rolled ,plate thickness is 1.375 inches,so use this size. The assumed plate thickness of1.5 inches used in computing the arching coefficientis not sufficiently different from the final calcu-lated thickness to warrant a further round of anal-ysis.

13. Check for elastic buckling-

EE

Pcr = 0.621 (r/t)3

0.621 L/95

000 psi)(30 x 06 psi)(.375 in)3

V (8 ft x 12 in/ft) 3

Pcr = 1,797 psi

The critical pressure for elastic buckling is greater

than the net external pressure of 1468 psi, calculated

in Step #10. Thus, buckling will not be critical inthis case.

2-43

a.4. Check buoyant uplift force-

Buoyant uplift on empty tank,

10,0 3 Y

B = 24 w

1oT (16 ft) 3 (62.4 lbs/cu ft)24

B = 334,567 lbs.

Dead weight of tank = 114,940 lbs.

Dead weight of soil = 1,370,000 lbs.

Total dead load = 1,484,940 lbs.

Safety factor against buoyant uplift,

1,484,940 44S.F. = 334,567 5= "

Thus, buoyancy will not be a problem once the tankis covered with 16 feet of backfill. If soil in-stability due to the high water table is a potentialproblem, positive measures should be taken to tiethe tank down. Otherwise it will float to the sur-face as soil liquefaction occurs due to dynamicloading from nuclear air blast or ground shock.

In comparing the design thickness for the tankwhich was calculated in Example 1 with that from Example2, the primary cause for increased thickness was due tothe presence of high water table. The resulting pressureamplification of 1.5 forced the design thickness to in-crease 50% over the size in Example 1. The stiffer soilcondition in Example 2 produced a slightly larger archingfactor, A, which reduced the effective pressure on thetank. This reduction in design pressure partially off-set the influence of high water table and resulted in anet increase in tank thickness of 37.5%.

2-44

2-9 Corrosion Protection

For internally pressurized steel tanks, corrosionprotection is a very important design consideration.Failure of an internally pressurized tank due to corro-

*+ sion represents a greater hazard than the corrosion ofunpressurized fuel tanks. The potential of stress cor-rosion is also enchanced by internal pressurization whichplaces the tank in a state of moderately high workingstress throughout its life.

There are many factors which affect the corrosivenature of the soil in which the tanks are buried. An un-derstanding of some of the more important factors willhelp the designer select the proper backfill material.The following are the primary factors affecting soilcorrosivity (5)

1. Soil Texture - Fine, even-textured soils areless corrosive than coarse soils. Soils ofuniform composition, such as sand or thesandy or silty loams, are less corrosive thana mixture of soils, such as sand and clay.

2. Soil Acidity - Highly acid soils are corro-sive, while mildly acid soils are not nec-essarily corrosive.

3. Organic Matter - Soils low in organic matter,such as clean sands, are usually less corro-sive than soils high in organic matter.

4. Bacterial Content - Soils with anaerobic bac-teria (usually heavy, water-logged soils,such as clay) are very corrosive.

5. Electrical Resistivity - Low resistivitysoils (0 to 3000 Ohm-cm) are corrosivebecause their high conductivity allows aneasy path for the flow of current from an-odic areas to cathodic areas on the struc-ture.

6. Moisture Content - Wet soils are usuallymore corrosive than dry soils.

2-45

Dry, clean sand without solub,.le salt content is one

of the best non-corrosive soil conditions and forms anexcellent backfill material. It is recommended that apositive form of corrosion protection be provided inaddition to a non-corrosive soil. The most economicalmethod is the use of a good quality commerical coatingon the outside of the tank to reduce the exposed metalarea to a minimum and thus isolate the metal from thepotentially corrosive environment of the soil. In con-junction with the coating, cathodic .protection should beprovided to prevent corrosion at discontinuities orfaults in the coating material.

Every effort should be made to avoid dissimilarmetal couples, such as ,L brass or copper pipe couplingin combination with steel. All piping should be coatedand wrapped.

If the corrosive environment is removed from theexterior surface of the tank by providing a coating bar-rier and cathodic protection, stress corrosion will haveeffectively been prevented. There are additional stepswhich may be taken to insure against stress corrosion.A full anneal and slow cool of the fabricated tank is adependable means of preventing stress-corrosion cracking.For high strength, cold worked steels, a full anneal isnot always permissible because it lowers the strengthproperties of the metal.

2-10 Mechanical Consideration for Internal Pressurization

The use of internal pressurization in a buried fueltank presents mechanical design situations which would

not normally be considered in the case of unpressurizedtanks. Allowance must be made for depressurization ofthe tank at times of refueling, cleaning and inspection.During operation the high internal pressure will forcefuel out of the tank down the fuel supply line to theengines. Thus, a pumping system will hot be required.However, in its place a compressor must be substitutedwhich will provide the internal pressure. Pressure re-duction valves must be installed on the fuel supply linesto drop the fuel pressure to atmospheric level before itis supplied to the engine.

These and other mechanical considerations are pre-sented in this section of the report.

2-46

2-10.1 Compressor

To provide internal pressure to the fuel tanks atpressures up to 1000 psi, an Ingersoll-Rand compressor(Model 15T2X) or equivalent will serve the purpose eco-nomically. Because of possible down time due to repairsand maintenance, two of these compressors should be pro-vided at each site. The cost of a single compressor withhigh operating reliability which would insure continuousinternal pressurization is greater than the cost of twocompressors of the type recommended.

In order to insure resistance of the fuel storagetanks to repeated nuclear air blast, the compressor sys-tem must survive nuclear attack in an operable manner.Thus, the compressors should be placed in the hardenedshelter which the fuel tanks supply. Each compressoroccupies a space 58 inches in length, 28 inches in width,and 33 inches in depth.

2-10.2 Piping

Nuclear blast pressures will cause large relativedisplacements in the soil as the pressure front movesover the soil. Prediction equations are available (8)forcalculating the probable maximum relative displacementwhich would be expected between adjacent tanks for anygiven soil condition, overpressure level, and weaponyield. With this information the designer can provideflexibility in the pipe system to meet the relative dis-placement criteria.

Particular attention must be given to the design ofpipe attachments at the pressurized tank. Due to rela-tive motion of the soil adjacent to the tank under arch-ing action, large shear forces are developed at the tank-to-soil interface along the side walls of the vessel. Anypipe connection made in this zone will experience thesehigh shearing forces and should be isolated from the soilthrough use of a crushable wrapping or isolation sleeve. (15,16)

If flexible piping is used between tanks, the stiff-

ening influence of internal pressure should be accountedfor in the design of the flex-hose.

2-10.3 Valves

Three separate valves must be provided in the pres-surized tank system.

72-47

In the fuel supply line to the engines a. pressurereduction valve will be required to reduce the high in-ternal pressure in the tank system to atmospheric levelor slightly higher depending on the head losses in theline.

Because of potential rupture of a single line orpremature failure of a tank due to undetected weaknessesin material a system of pressure closure valves should beprovided at each tank-to-pipe connection. These couldbequick closing solenoid valves, activated by a pressureswitch in the pipeline which respond to a sudden pressuredrop. Thus, failure at one point in the tank system wouldnot weaken the entire system to subsequent nuclear attack.

A system of pressure relief valves will be requiredin order to depressurize a portion of the tank system forrefueling, cleaning, and inspection.

2-10.4 Access

A special manhole will be required in the top ofeach tank with access passages to the surface. The man-hole must be pressure tight and be easily opened fromoutside the tank.

2-11 Site Investigation

Throughout this report every effort has been madeto emphasize the important part which the soil mediumsurrounding the tank plays in its design. The tank de-rives much of its strength and protection from a favor-able soil environment. Thus, it is only appropriatethat attention be given to the particular soil proper-ties needed by the designer to properly evaluate thesite-soil conditions and the improvements which may berequired to upgrade the soil immediately surroundingthe tank.

General information on the site-soil conditionssuch as thickness of overburden, uniformity of soillayers, depth to bedrock and water table is basic tothe design. For construction purposes informationabout soil stability during excavation, the availabil-ity of select backfill, and the anticipated settlementis required. For sites with high water table, data onthe permeability and draw down characteristics must beestablished in order to design dewatering procedures.

2-48

Specific material properties should be determinedfor each s6il type encountered at the site or used asbackfill. The following list contains the propertiesmost useful to the designer:

1. Elastic secant modulus, E (static and dynamic)

2. Shear modulus (dynamic)

3. Confined and triaxial stress-strain relationships,(static and dynamic)

1. Poisson's Ration, v

5. Propagational or compressional wave velocity, Cs

6. In situ unit weight

7. Saturated unit weight

8. Void ratio

9. Relative density

10. Bearing capacity

11. Shear strength

12. Angle of internal friction, ko

13. Electric resistivity

14. Corrosioi potential

Items 1, 4, 5, 6, 7, and 12 are directly applicableto the design procedure outlined in Section 2-7. If a morerefined analysis should be required using dynamic finiteelement techniques, items 2 and 3 may also be required.

The bearing capacity and shear strength of the soilis of direct use in designing for construction loads.This information is also applicable to the foundationdesign for support facilities which may be located atthe site.

The void ratio and relative density are field measur-able properties used to control the placing of backfillaround the tanks.

2-49

Appendix C has been prepared as a guide in specify-ing and reviewing soil investigation-procedures. Theappendix deals with siting studies, geological explora-tion, sub-surface exploratory drilling and sampling, geo-physical surveys, laboratory testing, and final reportrecommendations.

2-50

~1 I

I ' I'.

Section 3

OPTIMIZATION STUDY

3-1 Introduction

An economic study was made to determine the optimumtank volume at various overpressure levels. Parametricvariables in the study included the depth to water table,internal pressure, and yield strength of the tank steel.In order to optimize the tank size on an economical ba-sis, realistic cost figures were obtained from tank fab-ricators and major construction contractors who were lo-cated near the area of proposed construction.

A computer program was written to carry out a seriesof analyses for each parametric case. Results are pre-sented in graphical form for direct interpretation andapplication by the designer.

3-2 Cost Factors

The major costs of a buried fuel storage tank are inthe tank material, excavation and backfill. Each of thesecosts is directly dependent on the tank diameter or volumeas will be shown. Additional costs result from internalpressurization, piping and maintenance. These are minorin nature and are directly dependent on the total. storagevolume at the site rather than the volume of a given tank.

3-2.1 Tank Costs

Tank costs may be broken down into the following ba-sic items:

1. Material

2. Fabrication

3. Transportation

4. Placing

5. Corrosion Protection

All of these costs except transportation are direct-ly dependent on the volume of the tank. Transportation

3-1

costs are primarily dependent on the route of travel noton the size of the tank. Thus for cost comparison pur-poses, the transportation costs from fabricator to jobsite, have been excluded from the total tank costs.

There is a limitation on the maximum size of shopfabricated tanks which can be hauled by truck or train.By truck the maximum diameter is 10 feet and by rail itis 12 feet. Tanks of greater diameter must be trans-ported in segments and field fabricated at the site.

To establish realistic values for the total cost ofa fabricated tank, a survey was conducted of the majorfabricators. Each was presented with information regard-ing the range of possible tank sizes, the material type(high strength steel with a yield stress of 90 to 100 ksi)and fabrication criteria (ASME Pressure Vessel Code).Both field and shop fabrication were considered. Resultsfrom the survey were plotted and two cost equations wereestablished (See Figure 3-1). The equations were delib-erately fitted to the high side of the data in order tobe conservative.

As can be seen from the results, the minimum costoccurs at plate thickness of 1 inches. Plates of thisthickness are ideal for bending, require the minimumamount of jigging and are readily welded. For thinneror thicker plates the fabrication costs increase due tothe above factors. The cost differential between shopand field fabrication is roughly 25 cents per pound.

3-2.2 Excavation Costs

Excavation costs can range widely due to the varia-tion in anticipated soil conditions which range from hard-rock to nearly impregnable swamps. In order to establishan estimate of the probable cost for excavation and shor-ing, three soil conditions were considered, rock, sand,and clay. The cost of dewatering was not included in theexcavation costs.

For rock the most economical excavation is a verti-cal pit. Estimated excavation costs range from $15 percuoic yard to $45 per cubic yard. In sand and clay soilsvertical excavation using driven piles and sheathing tohold back the soil appears to be the most economicalmethod. The cost of excavation and shoring will varywith depth of total excavation as shown by the cost equa-of Figure 3-2. Costs increase at a rate of 15 cents per

3-2

}1

1.60

a FIELD FABRICATED

* SHOP FABRICATED

1.40'

FIELD FABRICATED

1.20-

100-'-J

z

1l'2 ---SHOP FABRICATED

0 30-

.60 - 4 4

MINIMUM YIELD STRENGTH OF 90 ksi

.400 2 3 4 5

PLATE THICKNESS, t (INCHES)

Figure 3-1 TOTAL COST PER POUND OF STEELPLATE- INCLUDING MATERIAL,FABRICATION, PLACING, ANDCORROSION PROTECTION

3-3

16 'NOTE: COSTS INCLUDE EXCAVATION ANDSHORING, FOR DRY SANDS AND CLAYS.

14-

Z 2

0"

0 80a

o 6-

> 43

x

00 10 20 30 40 50 60DEPTH OF EXCAVATION H, (FEET)

Figure 3-2 EXCAVATION COSTS VS, DEPTHOF TOTAL EXCAVATION

3-4

I

foot for excavations up to 30 feet in depth. Below this

depth, costs increase at a rate of 25 cents per foot.

At sites with high water table, dewatering will benecessary in order to excavate. The cost is directlydependent on the depth of water table, depth of excava-tion, and soil permeability. For gravelly soils thepermeability is so high that dewatering -may not be eco-nomically feasible; in clayey soils dewatering will nor-mally be a simple process. Because of the potentiallywide range in dewatering costs from site to site, thesecosts have been omitted from the optimization study.

3-2.3 Backfill Costs

Select backfill material will normally be placedaround the entire tank (See Figure 3-3). Washed ordredged sand is ideal for this purpose because of itslow corrosion potential and uniform material propertieswhen properly compacted. Washed sand costs $4.50 percubic yard, dredged sand is less. Both would probablyhave to be hauled to the site at a rate of $1.30 percubic yard per hour of turn around time. Hand or ma-chine compaction will range in price from $1.50 percubic yard to $3.00 per cubic yard. For cost optimiza-tion purposes the select backfill in place was estimatedat $7.00 per cubic yard.

The remaining portion of the excavation above thetank and select backfill (See Figure 3-3) is~filled inwith ordinary backfill obtained at the site. This ma-terial may be a coarse sand with some rock fragments.It is placed under careful compaction controls similarto the select backfill, however, its cost is minimal.Ordinary backfill normally costs $1.50 per cubic yardin place. For the optimization study it was assumed tocost $3.00 per cubic yard due to the strict placing re-quirements.

3-2.4 Pressurization Costs

A 15 horsepower compressor, which is capable ofpressurization to 1000 psig, will range in cost from$3,000 to $5,000. In Section 2-10.1 it was recommendedthat two compressors be purchased in order to have oneas a backup. Thus, the total compressor cost could beas high as $10,000. This is negligible in comparisonto the cost of the tanks alone.

3-5

ORDINARY

BACKFILL

m TAN KI. I

0....J .. ..

C D C

W: WIDTH OFEXCAVATION

NOTE: FOR OPTIMIZATION STUDY,C =E = .21)--* 3.0Ofeet

Figure 3-3 EXCAVATION AND BACKFILL

3-6

Operation and maintenance of the compressor shouldnot exceed $15 per month. This cost is directly propor-tional to the total storage volume at the site, not theindividual tank volume. Thus, for economic comparisonpurposes between tanks, all compressor costs have beenexcluded from the study.

Miscellaneous costs of valves, piping, etc. have

also been neglected in the economic comparison.

3-3 Parameters Considered in Study

The following parameters represent the more signif-icant variables which could be considered in an optimiza-tion study:

1. Costs

a. Tank

b. Excavation

c. Backfill

2. Tank Geometry and Properties

a. Shape

b. Diameter to Length Ratio

c. Orientation in the Ground

d. Yield Strength of the Material

3. Site Factors

a. Depth of Burial

b. Depth to Water Table

c. In situ Soil Properties

d. Backfill Properties

4. Nuclear Weapon Characteristics

a. Weapon Yield

b. Overpressure Level

3-7

5. Internal Pressurization

It is evident from the list of possible parametersthat an exhaustive study of each one would involve a ma-jor effort. For the purposes of this report certain keyparameters were selected for study. The others were heldconstant or allowed to vary in a prescribed manner.

The variable parameters are:

1. Tank volume or diameter which varies in 2 footincrements from 8 feet to 30 feet.

2. Overpressure level - 3000, 1500, 1000, 600, 200,and 10 psi.

3. Depth of water table - 0, 5, 13, 30, and 100feet.

4. Internal pressurization - with pressure andwithout pressure.

5. Yield strength of tank steel - 90 ksi versus36 ksi yield stresses.

6. Excavation costs for rock site versus sand orclay site.

The tank diameter and overpressure level were vari-ables in each of the four cases studied. The remainingparameters were varied individually within each case.

Tank costs are based on shop and field fabricationcurves of Figure 3-1. For tanks 12 feet in diameter orunder, shop fabrication is used, while tanks over 12 feetare field fabricated. The cost curves of Figure 3-1 arebased on high yield strength steel such as T-I or equiv-alent with a minimum yield strength of 90 ksi. For mildsteel, such as A-36 with a yield strength of 36 ksi, thecost equations of Figure 3-1 are reduced by 50% to re-flect the lower material and fabrication costs.

Excavation costs in sand and clay are based on thecost function of Figure 3-2, while excavation in rock is$28 per cubic yard and is independent of depth. Thecross-sectional area of excavation is shown in Figures3-3 and 3-4 along with the specified areas of select and

3-8

Po(t) (5 MT WEAPON)

z0 ***ORDINARY

BACKFILL

40- W1"' - " ! BACKFILL

w" -- '' INTERNAL TANKSPRESSUREK

~ Pi+ I/5 Pe ,3 e

wa

N io

SOIL MODULUS =33,000 psi

WATER TABLE

C =E =.2D 3 FEET

VARIABLE PARAMETERS:

TANK DIAMETER, D =8 FEET TO 30 FEETOVERPRESSURE, Por 3000, 1500, 1000, 600,

200 AND 100 psiDEPTH TO WATER TABLE, = 0, 5,13, 30 AND 100 FEETINTERNAL PRESSURE, Po= 0 AND P /3YIELD STRESS OF TANK, f y= 90ksi AND 36ksi

Figure 3-4 PARAMETERS OF TANKOPTIMIZATION STUDY

3-9

ordinary backfill. A clearance between excavation andtank along its sides, bottom and ends is 20% of the tankdiameter with a minimum of 3 feet. This minimum is basedon working space requirementcs.

Cost for select backfill is $7.00 per cubic yard andfor ordinary backfill $3.00 per cubic yard.

The tank diameter to length ratio is held constantwith the length equal to two diameters as shown in Fig-ure 3-4. It is oriented horizontally in the ground. Highstrength steel with a 90 ksi yield is considered in allbut one of the cases studied. For this case A-36 steelwith a yield stress of 36 ksi was evaluated,.

The tank is buried at a depth of one diameter belowground surface in sandy soil. The elastic secant modu-lus for the soil was selected at 33,000 psi, which wasfound to be a good average for sandy soils.

A nuclear weapon yield of 5 MT was selected for allcases studied.

Internal pressure within the tank is calculated ineach analysis to equal one third the effective externalpressure acting on the tank.

Figure 3-4 summarizes the key parameters, both vari-able and non-variable, used to analytically model theburied fuel tank in the following optimization studies.

3-4 Results from Case Studies

3-4.1 Case 1 - Optimum Costs as a Function of Over-

pressure and Tank Volume

For this case the water table is located 100 feetbelow ground surface, giving a solution for dry soil con-ditions. Both overpressure and tank diameter are vari-ables in the analyses.

The plot of plate thickness in Figure 1-2 was devel-oped from this case study. It reflects the theoreticalvalue for plate thickness. Due to fabrication consider-ation the minimum plate thickness for shop fabricatedtanks is inch and for field fabrication it is inch.In addition, available plate sizes come in incrementalthicknesses which become coarser as the total platethickness increases. These practical limitations have

3-10

been included in the economic evaluation of the tanksand have resulted in certain discontinuites in the the-oretically smooth optimization curves, as will be seenin many of the following figures.

Total tank costs for a single tank including mate-rials, fabrication, installation and corrosion protec-tion, excluding excavation and backfill, are given inFigure 3-5. There is a distinct discontinuity in costsbetween shop fabricated and field fabricated tanks dueto the material and fabrication cost differentials. Itshould be noted that for low overpressures ranging from200 psi on down to 10 psi, minimum plate thickness con-trols the design. In fact even at 600 psi, the designof field fabricated tanks up to 20 feet in diameter iscontrolled by the minimum workable plate thickness ofinch.

Total cost for a single tank including tank cost,excavation, and backfill is given in Figure 3-6. In thehigh overpressure range excavation and backfill accountfor roughly 12% of the total cost. While in the lowoverpressure range this figure increases to 50% of thetotal cost for large diameter tanks. Note also that forany given tank diameter the maximum range in total tankcost with overpressure can be as much as 500%.

The tank size which yields the minimum cost for en-closing a given volume is normally considered the opti-mum design. In this study the total tank cost to enclosethe volume equivalent to that contained in a 30 foot di-ameter tank was computed for- each of the tank sizes con-sidered. The tank costs were then normalized by dividingeach tank cost by the minimum tank cost, for any givenoverpressure level. The resulting ratio of total tankcost to minimum tank cost was plotted as a function oftank diameter and a smooth curve was passed through thevalues to produce the resulting optimization curve shownin Figure 3-7.

From this figure it is obvious that the 10 to 12foot diameter shop fabricated tank is the most economicalsolution. Field fabrication of the tank will cost 30 to50% more depending on the tank volume and overpressure lev-el. For field fabrication in the high overpressure rangeof 3000 psi there appears to be an optimum tank size be-tween 18 and 2" feet in diameter. At 200 psi overpressure,minimum plate thickness for field fabrication produces asharp discontinuity at the 12 foot diameter tank size.

3-11

SHOP FIELDFABRICATED FABRICATED

1,000

NOTE-C-ST-I-CLUuES MATERIAL,

0

0

10 -7

I IOft.:'

8 10 120 1 1 50 2 0 02 4 2 20 250

CAPACITY OF TANK (1000 GALS)

Figure 3-5 TOTAL TANK COST AS A FUNCTION OFPEAK OVERPRESSURE LEVEL ANDTANK DIAMETER -CASE I

3-12

iil

I,. SHOP _ FIELD

FABRICATED FABRICATED

,00

!o !NOT' COT ICLUES TANK COST,=

,EXCAVAT1ON a BACKFILL

4,-4

1,000

7--

0

_ __

I1 :I

8 10 12 14 16 18 20 22 24 26 28 30DIAMETER OF TANK (FEET)

I. I00

5 10 20 50 100 2-00 250

CAPACITY OF TANK (1000 GALS)

Figure 3 -6 TOTAL COST PER TANK INSTALLED AS AFUNCTION OF PEAK OVERPRESSURELEVEL AND TANK DIAMETER -CASE I

3-13

SHOP IFIELDFABRICATED-1 FABRICATED

1.70 I_ _ _ _ _ _ _ __ _ _ _ _ _ _ _ _ _

NOTE: TOTAL TANK COSTSINCLUDE TANK COST,

1-5\ EXCAVATION, AND BACKFILL

1.60-0

10'-0

____________ Po 3000psi

1,\0

0 z

0 0 1.40Y' Of

z-

0 %X

Z 0-tlz0 "*'

oo P0 1000 psi

1.20 L _ _

kI D

1.10. D1- pi_

WATER TABLE

1.00- 118 10 12 14 16 i8 20 22 24 26 28 30

DIAMETER OF TANK (FEET)

Figure 3-7 NORMALIZED TOTAL TANK COSTS AS AFUNCTION OF PEAK OVERPRESSURE LEVELAND TANK DIAMETER-CASE I

3-14~

This same plate thickness limitation forces the optimumfield fabricated tank size up into the large diameterrange between 26 and 28 feet. At 600 psi overpressurethe limitation on plate thickness does not influence theresults over as wide a range of tank sizes. However, itdoes alter the pattern of the optimization curve in the12 to 20 foot diameter range.

3-4.2 Case 2 - Influence of Water Table Depth on Costs

For this case study the depth of water table wasvaried, starting at ground surface with zw = 0 and mov-ing on down to 5, 13, 30 and 100 foot depths. Overpres-sure and tank diameter were varied as well.

Figure 3-8 shows the influence of water table onthe required design thickness of the tank. Since 6 inchplate is the maximum size available in high strengthsteel, large diameter tanks will not be feasible underhigh water table conditions and high overpressure levels.At 1500 psi overpressure level, the required plate thick-ness for the large diameter tanks is reduced to 3 incheswhich is a more workable size, see Figure 3-9.

The influence of water table on total tank cost isgraphically presented in Figure 3-10 and Figure 3-11 foroverpressure levels of 3000 psi and 1000 psi respectively.The range in tank cost can be as much as 300% between asite with no water table and one with water table at thesurface. These costs do not include dewatering or tankhold down costs. By comparing Figure 3-10 and Figure 3-11it is evident that the cost differential due to water ta-ble decreases with reduced overpressure level. At 200 psi

overpressure there is no further cost differential. Thusthe presence of water table at thi; low overpressure doesnot appear to influence the tank design to a point that itis economically significant.

The variation of total tank cost as a function ofoverpressure, holding the water table constant, is shownin Figure 3-12. By comparing the results at 600 psi over-pressure with those of Figure 3-5, the case without watertable, it is evident that the minimum plate thickness forfield fabricated tanks is no longer the controlling factorin the tank design. Instead, the design is dictated bythe blast pressures which are amplified as they passthrough the saturated soil.

3-15

SHOP L FIELD

FABRICATED FABRICATED

"-J

_ 4.

H7H

t

-.-

I.

.

0#

.J

8 10 12 14 16 18 20 22 24 26 28 30DIAMETER OF TANK (FEET)

I I I I I..-

5 10 20 50 too 200 250

CAPACITY OF TANK (1000 GALS)

Figure 3-8 PLATE THICKNESS AS A FUNCTION OFWATER TABLE DEPTH AND TANKDIAMETER -CASE 2. (P0 :3000 psi)

3-16

• ~ ~ ~ ~ ~ 4- H

_j ,

SHOP FIELDFABRICATED FABRICATED

IS t

w

w

w

-44

w

-Jt

4 08 10 12 14 16 i8 20 22 24 26 28 30

DIAMETER OF TANK (FEET)

5 10 20 50 100 200 250

CAPACITY OF TANK (1000 GALS)

Figure 3-9 PLATE 'THICKNESS AS A FUNCTION OFOVERPRESSURE LEVEL AND TANKDIAMETER - CASE 2 (FULL WATER TABLE)

3-17

SHOP FIELDFABRICATED FABRICATED

0

0

CL)

0

0 0

DIAMETERTE OTAK(ETBL

5 10 20 50 100 200 250

CAPACITY OF TANK (1000 GALS)

Figure 3-1O TOTAL COST PER TANK AS A FUNCTIONOF WATER TABLE DEPTH AND TANKDIAMETER - CASE 2 (P0 3000 psi)

3-18

SHOP FIELDFABRICATED FABRICATED

11000=00 4'WTR AL

-J

z

I.o

1-00

0

IO

58 10 12 14 16 18 20 22 24 26 28 30

DIAMETER OF TANK (FEET)

5 10 20 50 I00 200 250

CAPACITY OF TANK (1000 GALS)

Figure 3-11 TOTAL COST PER TANK AS A FUNCTIONOF WATER TABLE DEPTH AND TANKDIAMETER - CASE 2 (Po~ 000 psi)

3-19

i • • • • m

SHOP FIELDFABRICATED FABRICATED

1,000

x

.J-J

0

', z

w - 100

0

0ZI---

DIAMETER OFBL TAKFET

51020010 0 5

.J.

F- .

CAPACITY OF TANK (1000 GALS)

Figure 3-12 TOTAL COST PER TANK AS A FUNCTIONOF OVERPRESSURE LEVEL AND TANKDIAMETER - CASE 2 (Z 0)

3-20IAAIYOAN IO AS

An optimization curve for total tank costs under var-ious water tab'le conditions is presented in Figure 3-13.

The ratio of total tank cost to minimum tank costhas been normalized using the minimum cost for a tank in-stalled at a dry site as the divisor. Thus, the plottedcurves represent the relative cost between various watertable depths as well as the relative cost between tanksizes at any given water depth. The optimum tank diame-ter is between 10 and 12 feet for all water table condi-tions. Thus, the presence of water table does not alterthe optimum tank size, however, it is worth noting thatat 1000 psi overpressure, full water table increases thetotal cost for shop fabricated tanks by 50% and for largediameter, field fabricated, tanks by 100% over the tankcosts in dry soil conditions.

3-4.3 Case 3 - Influence of Internal Pressure onTank Costs

To investigate the influence of internal pressuriza-tion on optimum tank design two cases were considered.One with normal internal pressure conditions and the oth-er without internal pressure.

The resulting plate thicknesses calculated for thetwo cases are shown in Figure 3-14. It is evident thatinternal pressurization significantly reduces the requir-ed plate thickness. This is particularly true in themedium to high overpressure range.

A comparison of total tank costs is presented inFigure 3-15, for various overpressure levels. Internalpressuiization reduces the total tank cost by as much as50% in the high overpressure range. In the low overpres-sure range, under 200 psi, fabrication requirements dic-tate the minimum plate thickness rather than the externalpressure loads. Similarly, the influence of internalpressurization does not effect the determination of platethickness in this overpressure range.

For the case without internal pressure, the optimumtank size is in the 10 to 12 foot diameter range, whichis the same as for the cases studied previously.

3-4.4 Case 4 - Inf-luence of Tank Yield Strength

From the results of Case 1 it is apparent that inthe low overpressure range, tank costs are controlled by

3-21

2.40 . 0-

Pz1000 psi

2.20. Z [

WATER TABLE v

2.00

0

z 4

8 10 12NG 14OFB 2 2 4 2 8 3

DIAMKIETROTAK(E)

"8i ,:::: 10 12 1 I I'"b 22 2 6 3

Figure 3 -13 NORMALIZED TOTAL TANK COSTS AS AFUNCTION OF WATER TABLE DEPTH ANDTANK DIAMETER - CASE 2 (Po 1000 psi)

3-22

SHOP FIELD_ _

FABRICATED- FABRICATED

& -

6. 4±1 4'-r'LLLI-t'

5.i "-" 4.

C-

08 10 12 14 (6 18 20 22 24 26 28 30

DIAMETER OF TANK (FEET)

5 10 20 50 1OO 200 250

CAPACITY OF TANK (1000 GALS)

Figure 3-14 PLATE THICKNESS AS A FUNCTION OFINTERNAL PRESSURE, OVERPRESSUREAND TANK DIAMETER -CASE 3

3-23

II-

SHOP FIELDFABRICATED FABRICATED

-J

ME o=E - -

01

80 107-: H2AI 6 18 2 2 24 2 8 3

5 10 120 50 16 11002 4 2 20 305

CAPACITY OF TANK (1000 GALS)

Figure 3-15 TOTAL COST PER TANK AS A FUNCTIONOF INTERNAL PRESSURE, OVERPRESSUREAND TANK DIAMETER -CASE 3

3-244

J

minimum plate thickness requirements rather than strength.

Thus, the economic advantages of using a low yield

strength steel, such as A-36 with a yield stress of 36ksi, were investigated.

Figure 3-16 presents the calculated plate thicknessrequirements based on design pressure criteria. At 200psi overpressure, the calculated plate size for A-36steel closely parallels the shop and field fabricationrequirements of inch and inch respectively. It isinteresting to note that for the 3000 psi overpressurelevel, the design plate thickness for A-36 steel exceeds6 inches. Fabrication of plates this size is costly, ascan be seen by the total tank costs shown in Figure 3-17.These results further point out the fact that high yieldstreagth steels are the most economical solution in themedium to high overpressure range. While in the low over-pressure range under 200 psi, normal mild steel with ayield strength of 33 to 36 ksi is more economical.

The plot of normalized total tank costs in Figure3-18 reveals that the optimum size for tanks shop fabri-cated from A-36 steel is 12 feet in diameter and forfield fabrication it is 20 feet in diameter. These re-sults apply only to the low overpressure range of 200psi and under.

3-4.5 Case 5 - Influence of Soil Conditions onExcavation Costs

An economic study was made of the excavation costsin rock versus sand or clay soils. The variable wastank diameter or excavation volume. Water table was setwell below the bottom of the excavation to exclude dewa-tering costs from the comparison.

The resulting excavation costs are presented in Fig-ure 3-19. As anticipated, rock excavation costs exceedsand and clay soil excavation costs by a margin of 600%in the small tank diameter range and by 100% in the largetank diameter range.

Backfill costs have been added to the excavationcosts to provide an economical comparison between the twocosts. Backfill costs are a very small percentage of thetotal excavation cost for large diameter tanks. They doincrease to significant proportions for small diametertanks under sandy or clayey soil conditions.

3-25

SHOP owl FIELDFABRICATED FABRICATED

4.

I

XT , t

±0±......- - - -

w+

4. . .

-- J

.

w =Os p20s+f # Ll

CAPACTY OFTANK 1000 GAs)Fiur ....PAT.TICNES.S.. FNCIO.O

YILDSRESOERRESRELEEAN TAN DIAETE -i CASE

4-2

=9L.5 s H4 00ps

SHOP FIELDFABRICATED FABRICATED

0 O - - - ......

z

00

0

0.

8 10 12 14 16 18 20 22 24 26 28 30DIAMETER OF TANK (FEET)

5 10 20 50 100 200 250

CAPACITY OF TANK (1000 GALS)

Figure 3-17 TOTAL COST PER TANK AS A FUNCTIONOF YIELD STRESS, OVERPRESSURELEVEL AND TANK DIAMETER - CASE 4

3-27

SHOP ___ FIELDFABRICATED- FABRICATED

1.70 ----

1.60- - __

1.50 - __

U .40 - _

z 4

0 z30000,

00

0 O

1 0 0 - -

6 T E

f y 3 6 ksi

P.0 10 0 -

s_

8 10 12 14 16 18 20 2 2 2 4 26 28 30DIAMETER OF TANK (FEET)

Figure 3 -18 NORMALIZED TOTAL TANK COSTS AS AFUNCTION OF PEAK OVERPRESSURELEVEL AND TANK DIAMETER -CASE 4

3-28

SHOP FIELDFABRICATED FABRICATED

21,000~

C,)

-Vj

I--

U

Q~I-00

5

8 10 12 14 16 I8 20 22 24 26 28 30DIAMETER OF TANK (FEET)

5 10 20 50 100 200 250

CAPACITY OF TANK (1000 GALS)

Figure 3-19 EXCAVATION AND BACKFILL COSTS FORROCK SITE VERSUS SAND OR CLAYSOIL CONDITIONS -CASE 5

3-29

3-5 List of Major Findings and Conclusions

From the five cases studied the following conclu-sions can be derived:

1. From an economical standpoint, the optimumtank size is the 10 to 12 foot diametershop fabricated tank.

2. High water table conditions may increasethe material and fabrication costs for atank by 300%.

3. High water table in combination with 3000psi overpressure produces design require-ments which exceed the available platesize for high strength steels.

4. Internal pressurization provides a signif-icant savings in tank costs. This savingsfar exceeds the initial costs of compressorand special mechanical equipment used topressurize the tanks.

5. For overpressure levels over 200 psi, highstrength steels such as U.S. Steel's T-l,or equivalent, with a 90 ksi yield strength,are more economical to use than mild steelswith a yield strength of 36 ksi. Under 200psi overpressure, fabrication restrictionson minimum plate thickness make mild steelsthe economic choice.

6. Rock site excavation costs exceed soft soilsite costs by a minimum of 100%.

3-30

1.i

REFERENCES

Allgood, J.R., Ciani, J.B., and Lew, T.K., Influenceoj Soil Modulus on the Behavior% of Cylinders Buried

in Sand, Technical Report R-582, U.S. Naval CivilEngineering Laboratory, June 1968.

2 Allgood, J.R. and Herrmann, H.G., Static Behavior o6

Butied Reinforced Concrete Model Cylinders, TechnicalReport R-606, U.S. Naval Civil Engineering Laboratory,January 1969.

3 AMF Advanced Systems Laboratory, Sanguine FacilitySystem Study (U), Naval Civil Engineering LaboratoryContract Report CR 69.020, Port Hueneme, California,

December 1969, (Classifibd).

4 Chelapati, C.V., Citicalt Pressure 6o% RadiallySupported Cylinders, Technical Note N-773, for the

U.S.-Naval Civil Engineering Laboratory, January1966, (Contract NBY-32280) (AD 627082).

5 Corrosion Prevention and ContrLol, NAVDOCKS M0-306,

Department of the Navy, Bureau of Yards and Docks,June 1964.

6 Gates, W.E. and Takahashi, S.K., Static and DynamicModel Test.s o6 Pressurized, Underground Fuel StorageContainers, Comparison o6 Experimental Resuts withFinite Element Analysi4, NCEL Technical Report(currently in preparation).

7 Gill, H.L. and True, D.G., Active Arching o6 SandDuring Static Loading, Technical Note N-759,U.S. Naval Civil Engineering Laboratory, November1966.

8 Haltiwanger, J.D. et al, Attachments and Connectionsto Buried Structures, Naval Civil Engineering Labora-tory Contract Report - Contract No. NBY 32279, October1965.

9 Hendron, A.J. and Auld, H.E., The Effects o6 SoilProperties on the Attenuation o6 Airblast-nduacedGround MotionA, Proceeding of the InternationalSymposium on Wave Propagation and Dynamic Propertiesof Earth Materials, University of New Mexico, August1967. R-1

10 Hendron, A.J., Gamble, W.L., Haltiwanger, J.D., andNewmark, N.M., Design of Cylindrical ReinforcedConcrete Tunnel Lineu to Resist Ait Overpreurues,Report to the U.S. Naval Civil Engineering Laboratoryby N.M. Newmark, Consulting Engineering Services,Urbana, Illinois, June 1968.

11 Murphy, H.L., Groand Motion Predictions 6o NuclearAttack Area Studies, Stanford Research Institute,May 1967.

12 Newmark, N.M. and Hall, W.J., Preliminary DesignMethods 6o% Underground Protective Structute6,AFSWC-TDR-62-6, June 1962.

13 Newmark, N.M. and Haltiwanger, J.D., Principles andPractices 6o Design o6 HardenedStructutes, AirForce Design Manual, AFSWC-TDR-62-138, December 1962.

14 Sauer, F.M., Clark, G.B., and Anderson, D.C., NucleatGeoplosics, Part IV, EmpiticaZ Analysis o6 GroundMotion and Cratering, Stanford Research Institute,May 1964.

15 True, D.G., Design Criteria for Flexible Utility Con-nections, Naval Civil Engineering Laboratory, Tech-nical Report R608, December 1968.

16 True, D.G., Exploratory High-Explosive Field Tests o6Flexible Utility Connections, Naval Civil EngineeringLaboratory, Technical Report R638, September 1969.

17 Wilson, E., et al, Dynamic Analysis o6 Two-DiensionalStructure - Media Systems with Initial St4ess and Non-homogeneous Damping, Naval Civil Engineering LaboratoryContract Report, CR69.019, January 1970.

R-2

'I1.

APPENDIX A

COMPUTER PROGRAM FOR

DESIGN AND COST OPTIMIZATION

A-I Introduction

The design procedure for horizontally buried, inter-nally pressurized fuel storage tanks, as outlined in Sec-tion 2-7, was coded for analysis on the IBM 1130 computer.To generate the cost optimization curves of Section 3, aseries of subroutines were added to the program which cal-culated material quantities and costs. Thus the computerprogram which is presented in this Appendix will calculateboth tank thickness and costs. The cost functions includ-ed in the program are those given in Section 3-2.

A-2 Computer Program

Figure A-I is a flow diagram of the program logic.Following the figure is the program listing on Pages A-4to A-15. All input data has been defined and units notedon the first sheet. Definition of terms used in the pro-gram follows the input information.

The program is dimensioned to take 15 different over-pressures and 15 different tank diameters in a single run.This number may be increased by modification of the pro-gram dimension statement.

A-3 Sample Problems

The two design examples of Section 2-8 are presentedas sample input-output to the computer program. DesignExample No. 1 is on pages A-3 to A-4 while Design ExampleNo. 2 is on pages A-5 to A-Il.

A-1

START

GENERATE CURVE FORARCHING FUNCTION vs. AWHERE -3.0 6A cA°0

/,z'-READ AND PRINT SOIL,

' TANK AND WEAPON CHARACTERISTICS

COMPUTE COMPRESSION WAVEVELOCITIES OF TANK AND SOIL

E READ NUMBER OF OVERPRESSURES AND

T ANK DIAMETERS TO BE INVESTIGATED

4 READ IN ONE OVERPRESSURE VALUE

COMPUTE THE MAXIMUM DYNAMICAMPLIFICATION FACTOR

SET THE FIRST TRIAL VALUE OFTHE ARCHING FACTOR TO ZERO,ASSUME THE SOLUTION ASTHE NEXT TRIAL VALUE

/.

GENERAL LOGIC DIAGRAM FOR DESIGN AND COST OPTIMIZATIa

COMPUTE THE TOTAL EFFECTIVEPRESSURE AT THE TANK CROWN

COMPUTE THE INTERNAL PRESSURE AS APERCENT OF THE EFFECTIVE PRESSURE

COMPUTE THE TANK THICKNESS FORTHE NET EXTERNAL PRESSURE

INTERPRET A NEW ARCHING FACTORFOR THE COMPUTED THICKNESS.

[CALL INTER]

~YES

ADJUST PLATE THICKNESSFOR AVAILABLE PLATE SIZE.

[CALL PLTHK]

COMPUTE COST OF TANKMATERIAL AND FABRICATION.

[CALL TANCO]

IF WATER IS PRESENT,COMPUTE THE BUOYANCY FORCE

ON THE TANK,

12

)N PROGRAM (TKOPT)

COMPUTE VOLUME OFEXCAVATION AND COST.

[CALL EXCOS]

COMPUTE VOLUMES OF SPECIAL ANDORDINARY BACKFULL AND THERE

RESPECTIVE COSTS.[CALL BAKCO3

COMPUTE TOTAL COST AND

STORE ALL COSTS FOR SUMMARY

NO COMPLETE

NO FOR ALL TANK

DIAMETERS

YES

PRINT SUMMARY OF COSTS FOR EACHTANK DIAMETER FOR THE PARTICULAR

OVERPRESSURE

4PRESSURES BEEN

YES

FIGURE A-1

A-2

€A

00

iIi

oz

4 00

00

0 U. IL 0

wi

v u V I w(L C

* *11 WI *1 *d .

*j 0.D -1-t0 a _ A x *0 0 * )l *

ieu~ 0. * *440 .00 000 I i LzO -4 * q .. * *

0 U * *. * 0I ** 0.4

0. < X0 0 eiXf1N0~t~N- w -* *-

-4 ZI **0* *

* -C 0 - * 0 *LL I- *; *C I, Z

0* <0 0Z 0 z < I* 0 * -V

W- * -* DUU -. z - *z Z* o WIJ Z*4

I.: *., 1.- 1n 0 * 4*.->>* 4*tj.W

0 0* WJ U W LL 0 =)L- (Z < . * I-* 4* *U)- 0 0* _*00 0 JO- IW t flI. LL ) * 3c* if)*z

4*0.* * ) on u < - x u WI ) z z*04W w* * 0 -)0 a - cr.4 I-- X CL* 0* z z 0*4ze- <* X*XOZWJOX WJ3-0 :I * U* 00* U* -

I.-U x* tD~w: ) tow V) 10* * .- - * -W* *- I.-- -- )-LLZD 1 )0* 0* V) 0 -1 4 1.-

z 0 l* 0* WWVUtnuuWz OW WuI ** W* Vi) V) < zNI-* Z* X3:-- .- 30Z 44 j (X 1I-* WW* Z*W)

0* 0 - 0I-L01- 1)0 a 01 0* w* 0* Ix zz ~ .. M V) LL V ) .- - Of 1 * 0.* 0. 0.* w W 0

WOO* (I < itt Z 0. z w w of .Z0 M* x* x~ 1-*0.

0* -*)W )-W Mt - 0 1 * U* VU* .- W

A- 3

uI 1 *

I *0 * u

--I *0*

0n0. -o 0I

Id) I Id*

0. **Itl 0 10 * 0

<f 0. 0 I* 0 * 4

NN~~Ix 0 I .

UJII

rI II 0ILL

V) Il 0 *JIW - UI * b* -

>- 00 *(x

0I zE zk Z N D

0 d w J I *:uu-4* of

,4 0 Z I * *r-zi0W1QI < ~ I--

coU D It I W i01 < I 1 * I u4> W0 41

OZw Eli 1.-I<V 0%oou- le v *- *

OW d 01I I1 *04*v o-

010 0 01 0-I*I WI0

0 IL ELI * <o- I "Irl1 WI 0 * * 0U) V*zo o U 4 -4 1 * * N

I- m 10 00% 0 a wtw*0* w up WI 0I ZE* *o J-UN1 <co

o 0 01 -, .)-) I- O * * 0400 0 * z 4 LI Wl 1.3 * *00 0 * 0i M*'0 - 1 1mW *> v In

.0 * - tt <WW E0C 4 to4.-I *X * I0*0N1 ' 1- Ill0 *-.4- 'It 000 *w <

* 4n 01 .15 0000* ** ) ILL WN t- . - U

-WWWW m~1 Z *CW C In 4 E - * 'WUWUI~~zI t 0 D Ido0 t'. iI ItW0.l~liuu o IL on .1O'A '- I 4 - *

* 0 l'- ' -).V) ->In I toI0 * *t N0 0 0 0 0 CZ -400 VVv. W-.1 0000 0*u ow z -- V 0 L < * *

'-44~~~0 ID '. I . *. v U) W L - q 0LV< I wN

* IiX W to0 w -<t X0 I 00) *<orL -W* CL Cu-Z 0: l- . V I . Ifl. * I- *=O Q-* FZN <<0 . lz W I :)-4 < )0 ON-L. W) a)Z L .LLU. I in-'-* -4

11 11 Ills01 <- It I-- 0U I MIII.- . l :)-40:1. P- O' 0 u d - >*il J I *04* 0

or2 0-w* < Z9-VW W -3 V (D)-4 I D :*)4 D4 :w ew<zI - * Z 4- <-M j - 1- -I 1< I ,-Z Z-*Z.1W>- .4< ><ILD V -- Z I<.a*<-*

00tjccw* IL - W . V< - I dtW*V

4-Z * - U) ZWVW EI A-4

N

w 0

CaLl)

w

It

00

CNI

* ~ ~ ~ ~ ~ I *1(-1) ~ S

*w t u0 u u U) IWWWWIJWWWWWWWW*W I0.Q. *0 x *NNw ww ww w P - W W W * lI.L*

ocr * * U)).In (2.CIn Id Ill w U. w w w w w U. -L ILU. * *= * a. co(a ) IIU Ulto F-I alI-I *~ 0V)* -. * 1 40. 00. 0.~ E11:lL ILI 000000000000* .

w to* * EU000000000000* 440 w* IN 00000000000 < 10 . 0 00 0 0 00 0* o0- *0

U.If~ U * w *0*0000 .00000 I G 0ONm- a o0 C0N4 0 GO* *0 4 EnL0.* *N c* 00In0( 0 @ C 4NNN~* *'~

0z * * f0-400 04n ;,glN.4 0 *4 *444N NN* *.-z4)- * -m *.4-1000SI * * *

* *00 *. I4,M 0 40lI.- L-* * *l U 0 0 * I

0 - * * * JL * .-

CL Z .11 * Id4 - U* l'*

z 0 l'- * * U I-L *UU*

0 1- * * 0 00 I

D--. * *0 I.- 0 oa <S *Z- *I-. * * * O I) l-C 0 0~

-J MiInE O It5-. v *- * 0* *

Ill - ILIz l -i w wujCLUJ * J&I-- LL 03 x 0 ) > l'

z~~~~~~~A 5)U . - - *WZI-- ) IL4/) 0 u < >> I

0- 1.-

I %I

0L0 f1 01( -I

4L.m 0z I z

I i )Io

CI ; ;0 1

If 0 1 If 0

)z N I

0 1 1.- 0 I

*LL 0 r I I*0 I X.D 0 ) vi I D 0 o ad

*J N 0 0j 4f *d N IliIL m- 1' - 01 *L X a V

0. < *L 1 I< u*L I.a In u ~ 0 In L II V) * 0I*i to Wd *. t-NI *j >U W**40t0

*4 .4 <U)U 0 02 0 < 1UU Z4 < ;: 0.40 <*4 I- 300 Z4 i-I * 2- 30.0 -44 0 1* 0 W~ * 0 - N

LI* ' tn 04 0 . Il *: u- I.u~r -

WWLIJW f-I *- *O?.4 I. .- WWW t- I 0'U)WUODW 0- 0OI-n 1 ofWUW 0. 0lf.001

* 0, <..4-> - Is 0 U.,e -> --

0000* Z-40 coVI 0 0000 Z .00 UUL

00000* 0 01-- IO O 0 1~--U

- O N N I IL .... Oil NU -II-O~A* I "'L M4N 10 0 . l :1 I-L 4Mg 09 -

00 .- w m CNI 4 O-N-4 * t to m 4 4 I

o 10 1f 40- In~ * 4-1.- 0*100 9 1 1.- 0-Iz r.- in 0 LU Li

I La. P.- I.- I. * I I.- LC2 I D ;In*w L 0 ' w IL 00r I V) Wa~ ILU I L 1

*. 04 -0 LL I.A L . MW W* EIL rI- 0 UM M N-.00000* -. 02 N4 It L 11 .0, q4 ' L

00 0 * l-0 < LU - > I * 20 O ix- -.t . 0 -* UN.;- -;I <<; I i C! I- C* UN.h: L 0M If<Z c 10 %n L<- N 2> -Z I 0 .t* - I-~ U I II I

In 113 <. n3U - < -. I in z* IA C :-3) - i<V- I

I-Z 0* - U) ItUL0L - 0 U) LU w ZLU <X Z 4 14 0.0 -<a)0~ IL <I 4 -to. N I.0.<(~a2IL* 0 0E4 4 -Ox 0 z uU)0I I 1~Al2-U £t0 M 3Z M-iw.LZuc 0<D I

-00 0.4 inLU0<0o I I IO I.4 ~-001-0 V) 0WU) ~ 1. 0aLUD <4U Z< mLILIL. 1. 0) LiLUD <4U ZIx ILL A I

3: 0w< N~U0 <0 -0 I 4U) ll z* IfW0 I.)-0 0 ILUL a.S U. . n<II U. IU D- U- I.-I2 U I

< -z4 * -l- 31 Vl0 5Z u uLUL I "t-X * .-- "0 V -ZLxv UL0LU-< x Z-1 0. a - I.I< U 0 zE I 0.LJI<x. -. OU)4- -C4 ZZ LXCLk(~z L .- l-w 0 1 x m* .L lU 1-- 1 It U. -D - IZ1 .- Il C 04x f I.--U)02WZ(t2) 0D 1.- 1::) 0 i-* -VILI< -0.N4-0w 4 I D 0Z(L-* Z-UW < - M 4 W I

1/*MWO<* <--4WoUj <0>-0000 I >- 0f<* )-<DLULj4<0>00001I

004LD3~* 02-Ij.0.U<UCM>>>> I 004LD3* 1I0 UUm> >I

A-6

I I

1011a.mII M 0.

4~1 NND0Ir- 40

NI 0 1

CMo1 N 0

in I

w xII I

Il X I-- W Lw- .1 1

(X(LU I iN'N JAII N0. 0-3 >N 0 ZcI. t- > V 0 : 11-C ;: . ILI 0 <I *.-4 0 Il 0 < I

N 4 0 - Q1 1 O z z 0 111 IlIL .I-I LLu U 1. .I U

Il -1 01 1(ll I-01 I

0 1 i

0 I 0 I

of Ix I'. IILII U z: iI

v)04 O 0 )I 0- 01 : 01I- 3 0 . -1I I I. K0 j-1 0

0 tI NI 1 0 ( L I0.- U0 0 40lI . ( In 0 I

(n 4 LU in a,~l LU91-0 1-0- -ItI D:'V)~I ILI IL I l -a0. 1 O W * 0 IL w r- 0 (w0 IM .L t L U VW ~- -- I MU. WUu U. 3 z -4N4- - I

0. LU 10. 4 LU .1

44- LUO0 z1 CM4- Lu uu

.-0:-- owL eii -4 1- I0 ~ ~ 1UU VIm -tNi . 101 0nUU 4 oVIInl-- i01IL .-r%14 u ULL . - 1 0 l-NILU (-'I 1LIN z -C D .- _ NI In < LD _j. t'I

-LU4 I-- n' U-LU -NL 0) '0. ('1 0I LL 4AL

Z 4 40 C I u < I 14 0 XNIu < I0NM 0 ) <M I 1 04-0 n z<t* 0 1. 0 0c I I I * 0 0I If 4 c I

I. 00< LU l " > I l' 0 LU <4 - IX A 4!W I -- -. I * < 4 4 UWW Za0- _j

V I. u' 1.- *: le <I--I- u -a 4* I-. 1- 0. * . <

< * LU: W e'O' * < LU w 300vv u 0*- Q --- U) (4t0 x w o I Oz 0* 4 I-.1 In 0(0w t-t-*j 4K x -1 m0 LI I.- LNX-Mw <II <X * & 40(0 ZU3<WX&M<d-4

X*c -PII 0 4- 30.0. - . w 0 . -5 * I- 3.0 .zI .-44 .1V 0 '3D . * 0- << NI W n * 0- < ( l .

VW Q * <D'- -0040 0000 VI 0 * B'- <0fli00001I i--- * I - CC<- 1uO~ 1- i---s- I< - -I-u I

I.-ULLiLU 0 L t0.-1h- I Zw ww aL ULULxULU'-* 0V - cS U LU1t-0 WW4 1M W w X 0 < ox XXXX >-I < * -0 V) .1 - T. <X)T.-Xl

0000 w0* _j .0 - 0__ - .1V I W10000m0* < .- 0 NU.1 OVI-1Lj .1_ I00000<* 0 -< w---VIo 0 00003* 0 -< WW---VIo~

0n -jc * oI~ o~uc > > 1~ 4 l ml * Z 4um u 4.-. I>

* ~ IL(0 -. A. I-7 L( I

r-%0 0 tI CI % 4 1(nO - 05 z O- 034L o- .I L 4 0-

(4 LA I t- L I

0)) 0 11)U0 0 1

U)W0~~~ LUIU110 L

czoA D - I ".UV. W LO wS a.4 LU

w JLm :). X) LUQ. mU D)

0 wU LL V u I ocu . V)u ( I-i >LU 0 ty I- I -jLU> US- X

.- LU %D 4 -43 0-LU 5-4 <U,~ 0 45 0a)

,4 Z N. L Z z zU* 0 k U 4 it 0 0 X 4 V

*L 0 LU I. I. v 0 U.4. w 44 5- U. w * 44 I

* 44 I- I I- J -4* q a. 05 * -l -0I

Go 0 * I

* ~ ~~~~ H1 * ' I* e~N co LU 004ft'I* U* O )

0 0 0'L * 1D

* 0 4.0 )))>-- 1 0 4.40>--) 100000*N ; o n 0 a It 0-V O O O 0 0 --- I

I.-ON.0.n* I- I NL m 4 I a,.0 I-- mn* I - I-LU 4nO5I5 40-4U).4 * mU. ) 0 %0 44 l0-4U).- * UZ F- 44 a5) -'A * w w 1 o D 4- I'A Cq0itw * Z 4 4.- I

*LL .U L L n 1 m LLU 4 L . *L W LL.U L-LI I.: 0 tC* IL.4 z) 4L 1 *u U>. Z-) 1 0I 3 r~q1 tt

00 . * zS-00 w o4 i .0 0 (1T * 0r--o U)0, 5* 0tNC ) za

*, 4- ' 4 UN 12. - I a * 0 - 0 LU w - NS4(9 4 . uL U Ix N-40 -I1 4q 1*4 u OUWU (K N-4 I

Z . 5<~ D _->- I Z z. < U.D V -- i4 ~ LL LU . U) .- LUU.L 4VL* U. .. t) - V4 -

0U.I-I-L z 0.5O 0ZZ Ix -u <OS I z5--- U 00 C N uZZ <-.~ml If-0 0.4 -U 0. V) zU < ~ (a4 1- U40 0 LU W z4 Ia.0 -U.La, 04U ccL U L I I I I I.~, -l Cy, 0 .ImIL.0 L I

z (LI * -I-lD0 >-Z L zlI-z *"L U. 114-0 IV, >U LU< Li t0* Z-0 )4 *Z4< v I v.U-~ 1.OU4- 0 z u < u

.- z a* > -1 0 ft44Lu w00z I I- z '0* >-<~~~4 U 00 0004cm3 0-U.00UVMO> >> 004cfl3*W~xw 05I±O .04MO>> >

ix~~~~~~~A 8Lw0 c .wU .

-j ( - wx zw v -- w 0011

N

(Aa a-a (LA LI 0 0 4 n1

0. 0m -It a-M0I

C; .-I .- . -aL

(Gin Z I Gini

uJ II.-fr 0 IOOf 0

4~~~ 1AnO 0Z 4f. 0 ZI

uwO U.1II'02 0-wfO I'- W CeO.

o 4 00 NOI N) 010w w -- 0w 0:F- V

-i 04 In 0: j> V)a

a . L1 z a L

v- I. - <a Il u .I- LL P- 0 4

U.n I I Z IU) N. IU))0 0. OD OD ~ 0. t

a. - 1. a < w .

*-< ~ 1.iJ 1I .- <- F.O <I

'g~w0 vi 1 0rww vi

0- m 1 0%lAO I- f-:3c0 L - -JC 4 t-I 1.- -.wL<C -4 NaC 4 4I

N I 1 0 4lP* 0 r4 4 a *q I- a. , - 4 N II(1U

toII * L 0 4 1 V*wwIl 0 UJ m I

00 0 * z 4 14 ,- 10 >1u ll 0 00 * z 4IN Go u I* 0 0 I N 0 0 'r

1 1 * - 1- NI : 1* - 01O* a n '- 1- Il 0. 1* 1 *N&A I- I- - - 0*nz : I *( I

:) N - V) C I r* :) a

* oz 4z t--- I 00 . * o Ill.- )zlf.-f. *lI

* q 0 x *u< I z 0 CcoI fI* 1- N NO.In. 0z 4 to Is0'e Iel I0;:11- If U) 0-0f.4t to I I I -41- 0-1,0 0 a- OVA) o tcifl co I

'-.zJIJW f .- * 00OV < w- W w t.- a- *Ow w

a~u.. ~ <J.L w. U..i.1 N x I II N.ua.u a.. ~-r0000 -i IA* l 4- 0z l' <ufe>---- * w 4 V *40 -CC <X I

00000* LL XI - V .4,u < -lI 00000* U. -3:U ~ 4-VNU< I-VI100000* 0 0 ----- vzlooooo * 0 0 zuwz

OOO<X * 1- 0Il 4-Wxmw NI ILI <X * 1- 1W 1. LXMw41CON4NI- - l * & xu m 44ww l IO 4 m w @-- Ill a.aa z - ~ o'DO M-* x 11. 4C<-vIl In * 0 4.- l 40 L II

* IL L iJG . ) z w I W IL ii) 1 .w00 <L o . U U . I*n ooz 0u a < <:Il 1 C >O O tnr * <42Z :) w 1 <0-* ILIN0 N' U< Ix *C <O'0 Nf u< a C P

* 0- -a n z~m a *o- - O-N-a.O ZWW LL I4 11111f* X-0.04- c 1i o2 a F. <w - 00 _- o 'n 002Xx I

MM - * < 1 - V)a0 VAD- D D ) I IXXI * J l -V F-O. x< u W -. =)- Iw 1* Z -. UWW N.1-40 j jj_ I W* < U Z I4t 1, i-0j_ ji

>e t-.* ).-IN-1-J <N <4< I ) 1* U,--<wN1-J <N <4<<~ W* I-3in .- U <V.->>I <- W* U. -mU -vX <U->>

1-Z 0* .- 4 in ZZ UW ZI -Z 0* - iA-9U 0Z

e.o.

-. 5 i

in I5

0.. f I CL I

to IIL 0000 I 41t 4 n 03

CL .; 0 , - < I

en0 14 InI 0.. *- NLU U I t I

N D0 o C 0 N 0 ZI

of 0*, 0 I 0*- 0 4

1- e *ILL I-N *L ai

NO N0 NO N 0It IfIll Ill

(AU ) kInU C . UII

W K- WI Il - <w- 0.II- .

1*. 0< 0. <III 5 V) u I a )v 0

0 L:U 5 vL v I

UJI 1WN 0. I 0I

<L a III I

u~ici 0 5 n1c

LUW l U) W,, LW -I U) I40aOD0. fn t - 'A0. #. 4 4

0.1 31LU:.50. 4 LU .540

0-40LO 4 04-. 0 41

I-- ULU us---0 .5 CIDLU u-- .1

*:1- .LU2<n s-0 ml I-- .InUN CD(nInN4LII*0, 0~ 0 n .4w X. NI0'~m

*1 WWWW 0 U nn 0I .1I4 I V*U WWU 0 W -t 4 IL 0L LUL LL L W 0 LU jLL . -4tOD(4 I

o en 4 - I- *- - 4 - I*M- I>00 0 * N W 4 N rU'l 0 0 * N: z 0 C4I%' uuUE

* C 1 : 2 0 r4 -1- .I - i 1 * 2 0 .

0 N- .1 0 X- 1 * (v tn N0I

LL In IL-4 in LLL in LU *@4'O I IL LU **. 40

Z 4 4Un i Z0Wcc( b flI 4 <inU 0U3WW I-of* I- CY - 0. 0-.4z1.-.-o' 1 0 - 30 0. 0 CO I-40* 00 0 1,*f V o I IIII -1 0 l'014n 0c

111- 010o -0 .<I X W 5-1- ON4 I-. 0 " 0

U)LULULULUI-* 0 L~n ( 0 LI < < ) U UL W 0z L:I-- w V N. I-<Iz.LLLL LL .- 4N0Nu > Z; Z z C.LLLL VL >.4-W~I--- I

00000* 0 u a4-LI0o0 0' 0w -C * - A0 J I<Ll x 0 . m <I w -( x * :- 03 SIN 0 , xa

0W l- F W Mt zW -t* W. s-0 N1'OMW tOIW* M- I L 4K X I - L ,0O.N4:9- 4 CL V 44 Io I DO 0 * U_ .< Ml 54 Il

41 .Iln o < - V) z . .. t.h )I aWC JWA * ZL iL 4f.-. I

V)X OK* IL ) O- <0U0 0 V)~ 0 * IL D ) -1 ~ <I -0II toIZ I.- L) -. 4ZU ~(* F.-I0 N 0< w *- < < ,-_Zwww 0WC( II 0<

:) N 0 U) I- I *j It--D0 I-z w IIM5 11 -Ws- N- <) IOXWx I <M5 * -- C-I- I in OW

<Z 0.-* Z IL4-< V ->-.ZJJ I ujzx0.* Z-UIl4l V -- >-Z- I>-eLW IL *-3i .-<OW - . U<V-O I < LU* IL -31) -.141110 -.0o-Z 0c* - U ELULUZI s-Z * Q-wm4 U ZZLUULUOZ

LU~~~~~~~~A 10*4N.L2s-IX.4LU4 0L -<L0<

41

0 ~ 0

m 0 1 N 0 0.1 0

14 1 0.

zI I

l-I

0 0

rc0 0 11~n 1 I-Ia

W.J i 01 O N0w0 1

UjL f )I I.

I-I Z L D-4 ZL

0 zu z I 01I

00 D 4 1 4

< - 1- 0: 11<I.

1 - 1

o Iu~ t-I 0 I~J0I- Oiuoa r i- .01 l'- 4CO A I N - 0IN

r; $ 01 4 -wT C-i 4.8n )4on o'0W 0 10. t- 0l , 0. 481

00 0 * 0 I q N rU V 4 0 * Z 1 0(J0 4 UULU.I* 0 Oil 0. -; *. 0; L I I* co 40 1-U I- I-II I I 0 O t * I- I- 3- -j1* N~ cm 0 v *Y 1 In 0n 10iIu*n 4 I- 0I * z <D

* ) U.0 - - 1I -1 : -L 0)LLo*..l I *)U I

n*f)z II :/'O u I i 0 ZUjlZ 01 OD 0 C*0 of L) I * 0 i 0 0 NILK v 0 I* 4\ 3- V) 0.<0 o I *4 1-.-.-. *IAU V 0'In I

* - 30.0 V) 0 o NII I I 3I - 0 .0 11 on INo* l< 0 - >- I 1. 010<Q - ) I< vI- OIIf 1.- . .1 In I * < - 0IU'I \ z 11 U'.. a I

1z V-1- I4 -1 1-IO .4 NI<< 1-3--- U :- M le 11 < < 4 (-WiJW v~- > - Z < u I. >-WW~J -1 Z 0I'

tNIA * U Vo ) 4411 U l .'- * Q U 0 44v I

IIIi <. :; i . CIA IIn IW <x1 .xao

Z0'wOOa* <D'O U4-O OI~ ~ < z000 i <0-0 0* l' 0- W <4 I I.- U I I W M'C.0 < n <4I ofI

I.-0 1 - I LA 0a w 1 1 1 * - . 0 1 - II I Ii-004 iii m >-* I0'4jLi0. L < Y *0VW Z x~-0 <* x ;F.LL xZ .1140V j x<7 . I

cl 7. ID U.4-I0 wI a< x 0 - z.-3- U 44<u0>-Dn Iw < - l* - ULL<~ _j - >'Z j I W<Z M a* z- QU L< a _U --- Z _ I>-WD<* U.. .- wtz)-< vDtJ-O1> < D * tL .-.3In -WV4U- OL

~0 a . * >-t a-W&O <V0->> 1, 0i I0 0.4 w-W40 3:U

Li IU~n 41) L. .LLLL I I~J~* 4A-11 LLILL

* . In C1 In CI 0 t- r- t- (. a, It in1o~ * o 0 M0C1 - 4 .t in 4.4 in 4 0

*0 in N cm .0 N '10 Ca% a- C' 0 0 0

j in C m -nN0 N 0 co co 4 n 0 0 04-o* r C1 -4 C'- 0 0 CM in An 't 4 In

*1 4 N CI 4 in t- 0%

Z *

CY *- a, 0% a4 CL N IN 0 t-~ In 0 N 10

> 0 - In 0 m -1 -1 In (n- I 10 0

0* 0* .4 N N 0 -1 N 0q In 10 Go 4 N

L>* N (IN I 0 CI (n 4 q4 In C0 .0 in 0*44 0 4 4 n C OD .0 0M t- 4V 0"

x 0* c;C; I C

4 v r- CM 0 in CC N It O 0) CS It CI 10* - %n4 In 00 C C C 0

*Z * 4 CCn 0 0 (n N coI . -40 m IN 0

-. 1 01 4 CIq 4 I 10 -4

* co m (1 A 4 t-%*- 11 -

w* I

(* * 0 I 0 N1 Cci a)I 0 N 0 CI N) 0

Z * -* .4 - In It (n Nl 04 4 Go C4 0 0n

-4* 4 .0 4C 4n In CI t N C- 0 in In>- 1 CCv 01 In o, -4 0 m f- %4 CID C-

*03 -4 Nl CI In C 0 N In % 01 CCI0)1 00 4 u4 -4 .-

ILII

X * * I- p t n c \.,rU *< ) -4 NC 0- N 0' 0 01 N 0 0 - inV) t N0 i 0IN N n I Nl -n 0 0 In CV

a,* In 10 N0 01 4 I 0 0 CC 0 N1 O 10 C41 44* < -4 m N 0 O 0 0 co aI 0 C a, 4 0

0000 S * 1- 04 4 n 0 ' N N r1 4 C -

*> LI N4 N-4 n 01 4 N cI n 0

14it I 1 4 n 0 N- 0CI 4o '0 in .Cc

-.1 00* *l -41 0 r D

< 4 01 N It .- I 4- C in - 1 0

I.- tN 4L In N 0 4 Cv (V1

0. M A - *

wi~J* <

w/4C* l'- 0 0 0 0 0 0 0 0 0 0 0 0

a.~. * * 01.0 N 4 z z0 0 ; 0 z z 4 '0m .1 0- v V- NI N N N N Cr1

0 _jI-*I

A-12

APPENDIX B

COMPARISON OF EXPERIMENTAL RESULTS

WITH EMPIRICAL EQUATIONS

B-I Test Description

A series of static and dynamic tests were conductedon small scale stainless steel tanks, s hown schematicallyin Figure B-1. The tanks consisted of a cylindrical sec-tion and two hemispherical end caps. The models wereburied one diameter below the soil surface in a horizon-tal position. Internal pressurization was used to eval-uate improvement in buckling resistance of the thin-walled structures.

Two static tests and two dynamic tests were con-ducted on identical models. The first of the static anddynamic test series were conducted at pressure levels lowenough to prevent yielding and buckling of the tank. Tankstresses from these two tests (Static Test #8 and DynamicTest #2) are presented in this appendix.

In the second static test, external pressures wereincreased until yield strains were reached and the tankbuckled. The pressure was increased beyond the bucklingpoint without further increase in load to the tank dueto full transfer of load away from the tank through soilarching. Buckling of the tank did not produce a rupture.

Complete details of the entire test program are givenin Reference 6.

B-2 Tank Stresses

For Static Test #8 an internal pressure of 100 psiwas used to counterbalance an external soil surface pres-sure of.155 psi. The tank strains were monitored at elev-en gages located arounid the perimeter of the cylinder atits centerline, see Figure B-1. Stresses computed fromthe test data are plotted in Figure B-2 on polar graph pa-per. The right half of .he graph represents hoop stresseswhile the left half indicates meridional stresses. Twoexperimental plots are shown on each half. One, markedPi (Exp.), is the plot of stresses induced under internal

B-I

Po 2

COOKS BAYOU SANDdo= D E-' A E5 33,000 psi

I0. 024"

D1"p. 100 psi - -STEEL TANK

Es= 29.3 x lpsi

( I__ ____i f y =37,080 psi

L AL =2 D

00

HOOP STRAINS MEASUREDAT GAGES - I,3,9,1I,13a15

2700 13 eA5 900

II 1 12J MERIDIONAL STRAINSMEASURED AT GAGES

9 10 2,4,10,12 Ek14180*

SECTION AA

Figure B-1 MODEL TEST OF INTERNALLYPRESSURIZED BURIED FUEL TANKS- STATIC BDYNAMIC TESTS

B-2

STATIC TEST #8Px155 psi 0

i OXPs

100

pl-~B 3(M

pressure loading (i.e. no external pressure). The secondplot, marked Pi - Po, (Exp.), represents the state ofstress under combined external and internal pressureloading. A dashed line has been used to connect the re-- d --. points- -for purposes -P -readability Te

actual magnitude of stresses between monitored valuesis unknown.

For comparison purposes, hoop and meridional stresseswere computed using the empirical design formulas ofSection 2. These results have been superimposed on thesame graph in Figure B-2. By comparing the experimentaland empirical plots it is apparent that the tank behavesas a pressure vessel under uniform internal pressure withthe soil providing little or no resistance. Results fromthe model tests and the empirical formulas are virtuallyidentical. The invert hoop stress measured experimentallyappears to be a questionable record.

The empirical design equations of Section 2 arebased on the assumption that the actual distribution ofexternal pressures on the tank may be approximated by auniform distribution. This assumption gives reasonableresults in approximating the principle hoop stress whichoccurs at the spring line of the tank, see Figure B-2.The uniform pressure distribution assumption is an upperbound on all hoop compressive stresses. However, in themeridional direction this assumption underestimates thecompressive stress in local regions near the crown andinvert. It was in these local zones, at the junction ofcylinder with hemispherical end cap that buckling of thetank occurred during the second cycle of static loading -

to failure.

Results from the dynamic test are given in FigureB-3. A rectangular pressure pulse with nearly instan-taneous rise time was used to generate the dynamic sur-face pressure, p , which had a peak value of 150 psi.The tank was intgrnally pressurized to 100 psi. For com-parison purposes, peak stresses were computed in the hoopand meridional directions using the empirical relation-ships of Section 2.

From a comparison of Figure B-2 and B-3, it is ap-parent that the stress patterns for static and dynamicloading are nearly identical. The only variation betweenthe two cases is in the magnitude of the stresses. Sinceinternal pressures and peak external surface pressuresfor both tests were virtually the same in magnitude,

B-4

DYNAMIC TEST #2P = I5Opsi 00

E PEENL -

E PiR - -MP +2 0P--oE

00

DYNAIC TSEUTSADEPRIA OML

B-5

the basic difference in the resulting stresses is due toamplification resulting from dynamically applied loads.The dynamic amplification factor, DF, in -this case is 1.5.

As seen in Figure B-3, for the dynamic case, the em-pirical equations conservatively bound the experimentalvalues of peak hoop stress. In the meridional direction,however, the empirical equations underestimate the dynam-ic test results by 100%. From a design standpoint, thisis not as serious as it may appear, since hoop stresseswill normally be larger in magnitude than meridionalstresses. Thus, sizing of the plate thickness will bedictated primarily by the hoop stress.

B-3 Arching

From the static test results arching factors werecomputed which ranged from 0.32 to 0.68. For the firstcycle of external pressure loading, the arching factorranged from 0.32 to 0.39 while in the second cycle itranged from 0.52 to 0.68. Thus, the first cycle ofloading compacted the soil, so that subsequent cyclesloaded a .significantly stiffer soil medium. As noted inSection 2-2.6, arching is a function of the relative stiff-ness between the tank and the surrounding soil. As thesoil increased in stiffness, more of the load was trans-ferred away from the tank to adjacent soil by shear.Arching factors from the dynamic tests ranged in valuefrom 0.34 to 0.37.

For comparison purposes, the arching factor was com-puted using the empirical equations of Section 2-2.6.The resulting arching factor, A, was 0.41, which comparesquite well with the experimental values from the firstcycle of static loading and both of the dynamic testcases.

B-4 Buckling

The empirically determined buckling capacity of thetank as given by Chelapati's (4)formula from Section 2-4.is

Est A/(33,000 psi)(29.3x10 6 psi)p = o.6121/ - 0.612

P(r/t)3 V (6 in/0.024 in)3

cr = 152 psi

B-6

Lwhile actual buckling occurred under a net external com-

pressive pressure of 498 psi. The tank had an internalpressure of -only 13 psig at the time it buckled.

From these and other test results on vertical tanks(6)

it is apparent that the Chelapati buckling equation givesconservative results for internally pressurized tanks ofthe general geometry tested.

* I.-

' B-7

APPENDIX C

RECOMMENDED SOILS INVESTIGATION

C-1 Introduction

This appendix presents a series of recommended soilinvestigations which are applicable to the design of un-derground fuel storage tanks. These recommendations wereprepared by Mr. H. Klehn, a member of the firm of Damesand Moore, consulting engineers in the applied earth sci-ences, Los Angeles, California.

C-2 Design Requirements

The proposed facilities will consist of steel stor-age tanks placed below ground surface. The tanks will bedesigned to resist hig' nuclear blast pressures. Theymay range in diameter from 10 to 30 feet, and in lengthfrom 20 to 60 feet and be buried either in a horizontalor vertical orientation. It is estimated that the bottomof the tank will be from 20 feet to 65 feet below the ex-isting ground surface. Subsurface materials may includedry granular soils, saturated granular soils, or relative-ly shallow bedrock..

The method of placement would be in open excavationswith required dewatering. The soil conditions will nor-mally be improved by overexcavation and backfilling withhigh quality compacted fill soil.

A comprehensive series of studies will be requiredto provide criteria for design of the vessels to resistthe nuclear blast pressures. Under these loading condi-tions it will be necessary to know the properties of thesubsurface soil and rock materials when subjected to dy-namic forces at relatively high pressures. Of particularinterest are the dynamic secant modulus, compressionalvelocity, and Poisson's Ratio.

C-3 Recommended Soils Engineering and EngineeringGeology Investigation

C-3.1 General

Because of the close relationship of the soils en-gineering and engineering geology studies, these studies

C-i

3

should be done concurrently, as part of one investigationperformed-tya single c-onsultLng--8roup. The inhvestiga-

tion would include the following elements:

1. Siting studies.

2. Geological mapping survey.

3. Subsurface exploratory drilling andsampling.

4. Geophysical survey.

5. Laboratory testing.

6. Engineering analysis and preparation ofreport outlining specific recommendationsfor use by the design engineer.

The final report of the investigation should be pre-pared jointly by the soils engineer and the geologist,and should contain specific recommendations along withsupporting data. The results of all tests should be in-cluded in the final report.

C-3.2 Site Selection

The siting team should include the soils'engineer andengineering geologist. Prior to making the site selections,all available geological and soils data should be reviewed.This information would form the basis of preliminary opin-ions concerning the probable soil and rock conditions.

During the siting team's tour of the project area, thesoils engineer and the engineering geologist should advisethe rest of the team concerning the following:

1. Probable thickness of overburden soils.

2. Probable depth to ground water.

3. Type of soil and bedrock.

4. Engineering characteristics of the

subsurface materials.

C-2

It is anticipated that the information concerningthe subsurface conditions will have a substantial influ-

* ence on final site selection.

C-3.3 Geological Survey

A detailed geological survey of each of the siteswill be required. The engineering geologist should haveavailable to him topographic maps at the approximatescale of 1 inch equals 200 feet. The contour intervalwill depend on the site terrain; howeverthe intervalshould not exceed 5 feet in any case.

It is recommended that both black and white, andcolor aerial photographs be obtained specifically forthis investigation. This would be stereoscopic coverageextending at least one mile beyond the site limits. Thescale of the photographs will depend on the terrain andsite geology; however, the direct photographic printsshould be no smaller than a scale of about 1 inch equals800 feet.

The engineering geologist should make a complete in-spection of all erosional features, rock outcrops, exca-vation and road cuts. In addition, test pits should beexcavated as required to further define the geologic con-ditions at the site.

The engineering geology map of each site should in-clude the boundaries of the soil and rock types, theorientation of joints, bedding and shear zones in therock, and the extent of all landslides, serious erosionor other areas of past or potential instability.

Following the initial geological mapping, the geol-ogist should assist the soils engineer in selecting loca-tions for the subsurface exploratory borings. In addition,the geophysical survey should be planned to further definethe subsurface conditions. Sufficient subsurface explora-tion should be completed to define any suspected or knownsubsurface geological structure.

It should be the responsibility of the engineeringgeologist to prepare subsurface sections detailing thesite conditions.

The geological studies should include a review of allground water information available for the area. This de-

scription should include anticipated seasonal ground water

C-3

fluctuations, as well as the conductivity of the groundwater, and presence of corrosive chemicals. Sufficientsubsurface exploration should be performed to-complete.lydescribe the local ground water conditions.

The engineering geologist should present his opin-ions concerning the suitability of the site for construc-tion of the proposed facilities, along with his evalua-tion of the engineering properties of the rock and soilmaterials.

C-3 .4 Drilling and Sampling Program

The purpose of the drilling and sampling program willbe to explore the subsurface conditions. During the pro-gram, sufficient core samples should be obtained for lab-oratory testing and classification of all soils and rock.

The soils engineer and engineering geologist shouldlay out the locations of all borings. During drilling,any adjustments or additions to the drilling programshould be made in the field.

The depths of the borings will depend to some extenton the materials encountered. However, for planning pur-poses, it is expected that the borings will be in the,range of 50 feet to 100 feet deep.

There are many types of drilling equipment whichcould be used. It should be the responsibility of thesoils engineer and the engineering geologist to jointlyselect the proper type of drilling equipment to accom-plish the work at each site.

Each piece of drilling equipment should be under thecontinuous supervision of either the soils engineer or theengineering geologist. This professional would be respon-sible for the drilling and sampling operations and wouldmaintain an accurate log of each boring.

Several alternate methods of sampling should be avail-able. The actual method used will depend on the type ofsample desired and the soil or rock encountered.

Representative, undisturbed samples of relativelydense granular materials and firm to stiff cohesive mate-rials can be obtained using a sampler of the type shownin Figure C-1. When a sampler of this type is driven in-to the soil, the number of blows required to cause the

C-4

DRIVNG ORPUSR4O SOIL SAMPLER TYPE U -MECHNISMFOR SOILS DIFFICULT TO RETAIN IN SAMPLER

U. S. PATENT NO. 2,318,O42COUPLING ~ -,

WATER OUTLETS/

NOTCHES FOR It

ENGAGINGFISHING TOOL

NEOPRENE GASKET - CHECK VALVES

HEAD

SPACE ToRECEIVEDISTURBED SOIL

N TEIALTERNATE ATTACHMENTSEENHGSCSEEEAD EXTENSION' CAN

HEAD* ARE SPLIT SAMREL~

COE.ETAtHER

SPLOT BARREL -..ITO FACILITATE REMOVAL

of CORE SANPLE)

SPLIT BARREL,

WINTCMG DEVIC

Figue C I OIL AMPERSIT CRE.REAI0-5

sampler to penetrate each foot when driven with a calibra-ted weight should be recorded. This driving resistancecan be used to correlate samples between borings. In ad-dition, the blow count provides a means for estimatingthe relative density of the soils, soil strength and othersoil properties.

Where soft or sensitive soils are encountered, apiston or Shelby tube sampler, of the type shown on FigureC-2, should be utilized. This type of sampler has the ad-vantage of a bit with a thin cutting edge which results inless sample disturbance during the sampling operations.This type of sampler also should be used when attemptingto obtain accurate density samples of loose granular mate-rials for the purpose of evaluating the relative density.

Depending on the type of soils, it may be desirableto perform field tests during the drilling program. Fieldtests provide a means of evaluating the in situ physicalproperties of the soil. This is particularly importantwhen suitable undisturbed samples of the soil materialscannot be obtained.

At a minimum, standard penetration tests (AmericanSociety for Testing and Materials, Designation D1586-63T)should be performed in each boring at intervals of notgreater than 10 feet when drilling in granular soils. Inthis test, a sampler similar to that shown on Figure C-1is driven with a hammer weighing 140 pounds, dropping froma height of 30 inches. The number of blows required tocause the sampler to penetrate 1 foot can be used to esti-mate relative density of sandy and gravelly soils.

Another field test which may be considered is thestatic penetrometer test, similar to the Dutch cone pene-tration test. This test procedure is described in a paperby Plantema in the Ptoceedings o6 the Second InternationaZConereence on Soit Mechanics and Foundation Engineering,1948. This test is a measure of the static resistance topenetration of the undisturbed soils at the bottom of theboring. A penetrometer is provided with a cone-shaped tip,and the force required to cause penetration is recorded.This test is particularly useful in cohesive soils and hasbeen correlated with various soil properties, including thesoil strength.

A third test which may be useful in measuring the insitu soil properties is the pressure meter test. This test

C-6

REACTION MEMBER

COUPLING

PISTON SAMPLER

HEAD "

THE DAMES & MOORE PISTON SAMPLER

HAS BEEN DEVELOPED TO OBTAIN SAM-

PLES OF SOFT SOILS WITH A MINIMUM OFCYLINDER DISTURBANCE. THE MOST SIGNIFICANT

FEATURES ARE THE SEALING PISTONWHICH CONFINES THE SOIL DURING

TIE ROD

SAMPLING AND THE SAMPLE TUBEWHICH HAS A WALl THICKNESS OF ONLY0.042 INCHES.

-PISTON CUPS

DRIVING PISTON AT THE START OF THE SAMPLING, THE

LOWER END OF THE SAMPLE TUBE IS

ADJACENT TO THE SEALING PISTON AlTHE BOTTOM OF AN EXPLORATION

TEST BORING. THE SEALING PISTON,CYLINDER, HEAD, AND REACTION MEM-BER REMAIN STATIONARY DURINGSAMPLING. COMPRESSED AIR, COM-PRESSED NITROGEN, OR WASH WATER

ARE FORCED INTO THE CYLINDERTHROUGH THE SAMPLING RODS FROMSAMPLE TUBETHE DRILLING EQUIPMENT. THE DRIV-

ING PISTON MOVES THE SAMPLE TUBE

SIS DOWNWARD INTO THE SOIL.SEALING PISTON

Figure C- 2 PISTON SAMPLER

C-7

utilizes an inflatable tube or cylinder which is introduc-ed into the boring. As the tube is expanded against thewalls of the boring, the pressure required to cause a unitstrain or deflection of the soils forming the walls ismeasured. This stress-strain relationship is a functionof soil strength and compressibility and has been corre-lated with these properties.

If low strength silts or clays are encountered, theshear strength and stress-strain characteristics of thesesoils can be measured in situ by the vane shear test. Forthis test, a steel vane is pushed into the soil at the bot-tom of the boring, and the torque required to rotate thevane is recorded.

Where bedrock is encountered, it is recommended thatsamples be obtained by using NX coring tools. NX coringtools provide a clean rock core, 2-1/8 inches in diameter.This has become a standard core size, and provides rocksamples which can be conveniently tested in the laboratory.A great deal of information is available from tests on NXcore samples for correlation of physical properties. Uti-lizing cores of other diameters would influence the testvalues and make it difficult to access rock propertiesbased on information available from previous work.

The rock should be continuously cored, and an effortshould be made to obtain 100 percent core recovery. Anaccurate log of all bedrock cores should be maintained,and the cores should be photographed immediately upon ex-traction from the boring. The log should include the num-ber, inclination and spacing of joints or fracture planesobserved in the bedrock cores.

The RQD index should be recorded as described byProfessor D.U. Deere of the University of Illinois. Pro-fessor Deere has developed this index value as a measureof the overall massiveness and competency of bedrock ma-terials. Utilizing his method, the cumulative length ofindividual core pieces longer than 4 inches is reportedas a percent of the total length of core run. This isreferred to as the RQD index.

All soil samples and rock cores should be packagedin watertight and vapor-tight containers, and stored in aprotected place prior to laboratory testing. Since manyof the physical properties of soil and rock are dependenton the moisture content of the materials, every effortshould be made to maintain the samples at the field mois-ture conditions.

C-8

-' '

During the drilling operations, it may be desirable

to perform pumping tests to evaluate the permeability andhydraulic characteristics of the subsurface materials.This would be particularly important where dewatering ofdeep excavations is anticipated during construction ofthe proposed facilities. The method and duration of thepump tests should be reviewed by the design engineer pri-or to initiation of the tests. At least two observationwells should be placed in the vicinity of the pumped wellto permit accurate measurement of the drawdown of thewater level during test pumping.

The field log maintained for each boring should in-clude the following information:

1. The date, type of drilling equipment, andsupervising engineer or geologist.

2. The drilling contractor's name and address.

3. Boring locations in terms of distances orcoordinates to the nearest foot in plandimension.

4. Elevation of the ground surface at eachboring location to an accuracy of 0.10foot.

5. Type of sampling, depth, length of samplee.r core atteripted, the percent recovery,and the RQD index for rock cores.

6. Field classification of materials drilledand sampled.

7. A visual field evaluation of the relativefirmness or density of the sampled materials.

8. The loss of drilling fluid circulation ifrotary-wash drilling methods are utilized.

9. Observations regarding ground water flowinto holes through fractures or permeablesoils.

10. Initial and static water level measurementsobtained daily for a period of at least twoweeks following completion of each boring.

C-9

11. Results of in situ tests.

0-3.5 Geophysical Survey

The geophysical survey will provide additional in-formation concerning the depth and type of soil and'rockbeneath the sites. In addition, the velocities withwhich shock waves pass through the subsurface soil androck are measured as part of the geophysical explorationprogram.

It is important that the uniformity of subsurfaceconditions be known for quite a distance surrounding theactual site to be developed. This information is neededin the finite element analysis and to estimate the ap-proach of outrunning shock waves through the subsurfacematerials.

It is recommended that the geophysical program in-clude a minimum of two seismic refraction profiles orlines at each site. Generally, these lines should be atright angles to each other, centered on the approximatecenter of the proposed structure. The lines should ex-tend approximately 1,000 feet beyond the site limits.

For the finite element analysis, the seismic refrac-tion survey should include measurement of both the com-pression wave and the shear wave velocities of the soiland rock present in the subsurface profile to a minimumof three times the depth of the bottom of the structures.

As part of the geophysical program, measurements ofuphole velocities should be made. These measurements areuseful in evaluating the dynamic properties of the sub-soils in the immediate vicinity of the proposed structures.It is recommended that the uphole velocity be measured inthe exploration borings which are drilled to a minimumdepth of 50 feet below the proposed lowest point of exca-vation.

The results of the geophysical explorations shouldbe correlated with the engineering geology survey andwith the results of the drilling program. Any structuralirregularities or other poorly defined subsurface condi-tions should be explored by drilling.

C-10

C-3.6 Laboratory Testing_ _Static Tests: Conventional soils engineering tests

are used to define the density, moisture content,strength, compressibility and permeability of the sub-surface soils. When known, these basic soil propertiespermit evaluation of the performance of subsurface soiland rock materials when subjected to static loads.

It is recommended that the moisture content and drydensity be measured for all samples. This soil propertyis easily obtained and is useful in approximating soilstrength and compressibility.

For granular samples, it is important to know therelative density of the soils. It is recommended, there-fore, that the minimum and maximum densities of repre-sentative samples of granular soils be measured. Knowingthese two values, and the in-place density of the soils,it is possible to calculate the relative density of thein situ soils. From the relative density, the compress-ibility of the soils under load can be estimated. It ispossible also to evaluate the static subgrade modulus ofthese soils and the probable soil strength from the rel-ative density. In addition, the susceptibility of satu-rated granular soils to liquefy under shock loading canbe estimated, based on the relative density of the soil.

Depending on the type of soil, i.e. granular soilor cohesive soil, one or more types of strength testsshould be performed. For cohesive soils it is recpmmend-ed that triaxial compression tests or unconfined compres-sion tests be performed as described in Figure C-3. Ifthe soils are saturated, the pore water pressures shouldbe measured during tests.

In the case of granular samples, triaxial compres-sion tests could be utilized. Ordinarily, however, thestatic shearing strength can be adequately measured us-ing direct shear tests. The method of direct shear testsis described in Figure C-4.

In addition to performing pumping tests in the fieldto measure the permeability and hydraulic properties ofthe soils, it is sometimes advisable to perform perme-ability tests in the laboratory. The method of perform-ing these tests is described in Figure C-5. Basically,this test is a measure of the relative ease or difficultywith which water passes through the sample. The results of

C-11

METHODS OF PERFORMING UNCONFINED COMPRESSION AND TRIAXIAL COMPRESSION TESTS

THE SHEARING STRENGTHS OF SOILS ARE DETERMINEDFROM TIlE RESULTS OF UNCONFINED COMPRESSION ANDTRIAXIAL COMPRESSION TESTS. IN TRIAXIAL COMPRES-SION TESTS THE TEST METHOD AND THE MAGNITUDE OFTHE CONFINING PRESSURE ARE CIIOSEN TO SIMULATEANTICIPATED FIELD CONDITIONS.

UNCONFINED COMPRESSION AND TRIAXIAL COMPRESSIONTESTS ARE PERFORMED ON UNDISTURBED OR REMOLDEDSAMPLES OF SOIL APPROXIMATELY SIX INCHES IN LENGTHAND TWO AND ONE-IIALF INCHES IN DIAMETER. THE TESTSARE RUN EITHER STRAIN-CONTROLLED OR STRESS-CONTROLLED. IN A STRAIN-CONTROLLED TEST THESAMPLE IS SUBJECTED TO A CONSTANT RATE OF DEFLEC-TION AND THE RESULTING STRESSES ARE RECORDED. INA STRESS-CONTROLLED TEST THE SAMPLE IS SUBJECTEDTO EQUAL INCREMENTS OF LOAD WITH EACH INCREMENTBEING MAINTAINED UNTIL AN EQUILIBRIUM CONDITIONWITH RESPECT TO STRAIN IS ACHIEVED.

YIELD, PEAK, OR ULTIMATE STRESSES ARE DETERMINED TRIAXIAL COMPRESSION TEST UNIT

FROM THE STRESS-STRAIN PLOT FOR EACH SAMPLE ANDTHE PRINCIPAL STRESSES ARE EVALUATED. THE PRINCIPAL STRESSES ARE PLOTTED ON A MOHR'SCIRCLE DIAGRAM TO DETERMINE THE SHEARING STRENGTH OF THE SOIL TYPE BEING TESTED.

UNCONFINED COMPRESSION TESTS CAN BE PERFORMED ONLY ON'SAMPLES WITH SUFFICIENT COHE-SION SO THAT THE SOIL WILL STAND AS AN UNSUPPORTED CYLINDER. THESE TESTS MAY BE RUN ATNATURAL MOISTURE CONTENT OR ON ARTIFICIALLY SATURATED SOILS.

IN A TRIAXIAL COMPRESSION TEST THE SAMPLE IS ENCASED IN A RUBBER MEMBRANE, PLACED IN ATEST CHAMBER, AND SUBJECTED TO A CONZ"NING PRESSURE THROUGHOUT THE DURATION OF THETEST. NORMALLY, THIS CONFINING PRESSURE IS MAINTAINED AT A CONSTANT LEVEL, ALTHOUGH FORSPECIAL TESTS IT MAY BE VARIED IN RELATION TO THE MEASURED STRESSES. TRIAXIAL COMPRES-SION TESTS MAY BE RUN ON SOILS AT FIELD MOISTURE CONTENT OR ON ARTIFICIALLY SATURATEDSAMPLES. THE TESTS ARE PERFORMED IN ONE OF iHE FOLLOWING WAYS:

UNCONSOLIDATED-UNDRAINED: THE CONFINING PRESSURE IS IMPOSED ON THE SAMPLEAT THE START OF TIlE TEST. NO DRAINAGE IS PERMITTED AND THE STRESSES WHICHARE MEASURED REPRESENT THE SUM OF THE INTERGRANULAR STRESSES AND POREWATER PRESSURES.

CONSOLIDATED-UNDRAINED: THE SAMPLE IS ALLOWED TO CONSOLIDATE FULLY UNDERTHE APPLIED CONFINING PRESSURE PRIOR TO THE START OF THE TEST. THE VOLUMECHANGE IS DETERMINED BY MEASURING THE WATER AND/OR AIR EXPELLED DURINGCONSOLIDATION. NO DRAINAGE IS PERMITTED DURING THE TEST AND THE STRESSESWH.CH ARE MEASURED ARE TIlE SAME AS FOR THE UNCONSOLIDATED-UNDRAINED TEST.

DRAINED: THE INTERGRANULAR STRESSES IN A SAMPLE MAY BE MEASURED BY PER-,FORMING A DRAINEE, OR SLOW, TEST. IN THIS TEST THE SAMPLE IS FULLY SATURATEDAND CONSOLIDATED PRIOR TO THE START OF THE TEST. DURING THE TEST, DRAINAGEIS PERMITTED AND TIlE TEST IS PERFORMED AT A SLOW ENOUGH RATE TO PREVENTTIlE BUILDUP OF PORE WATER PRESSURES. THE RESULTING STRESSES WHICH ARE MEAS-URED REPRESENT ONLY TIlE INTERGRANULAR STRESSES. THESE TESTS ARE USUALLYPERFORMED ON SAMPLES OF GENERALLY NON-COHESIVE SOILS, ALTHOUGH THE TESTPROCEDURE IS APPLICABLE TO COHESIVE SOILS IF A SUFFICIENTLY SLOW TEST RATEIS USED.

AN ALTERNATE MEANS OF OBTAINING THE DATA RESULTING FROM THE DRAINED TEST IS TO PER-FORM AN UNDRAINED TEST IN WHICH SPECIAL EQUIPMENT IS USED TO MEASURE THE PORE WATERPRESSURES. THE DIFFERENCES BETWEEN THE TOTAL STRESSES AND THE PORE WATER PRESSURESMEASURED ARE TIlE INTERGRANULAR STRESSES.

Figure C -3 UNCONFINED COMPRESSION

AND TRIAXIAL COMPRESSION TESTS

C-12

-~,-!

METHOD OF PERFORMING DIRECT SHEAR AND FRICTION TESTS

DIRECT SHEAR TESTS ARE PERFORMED TO DETERMINE

THE SHEARING STRENGTHS OF SOILS. FRICTION TESTS

ARE PERFORMED TO DETERMINE THE FRICTIONAL RE-

SISTANCES BETWEEN SOILS AND VARIOUS OTHER MATE-

RIALS SUCH AS WOOD, STEEL, OR CONCRETE. THE TESTSF)

ARE PERFORMED IN THE LABORATORY TO SIMULATE

ANTICIPATED FIELD CONDITIONS.

EACH SAMPLE IS TESTED WITHIN THREE BRASS RINGS,DIRECT SHEAR TESTING

TWO AND ONE-HALF INCHES IN DIAMETER AND ONE INCH~& RECORDING APPARATUS

IN LENGTH. UNDISTURBED SAMPLES OF IN-PLACE SOILS

ARE TESTED IN RINGS TAKEN FROM THE SAMPLING

DEVICE IN WHICH THE SAMPLES WERE OBTAINED. LOOSE SAMPLES OF SOILS TO BE USED IN CON-

STRUCTING EARTH FILLS ARE COMPACTED IN RINGS TO PREDETERMINED CONDITIONS AND TESTED.

DIRECT SHEAR TESTS

A THREE-INCH LENGTH OF THE SAMPLE IS TESTED IN DIRECT DOUBLE SHEAR. A CONSTANT PRES-

SURE, APPROPRIATE TO THE CONDITIONS OF THE PROBLEM FOR WHICH THE TEST IS BEING PER.

FORMED, IS APPLIED NORMAL TO THE ENDS OF THE SAMPLE THROUGH POROUS STONES. A SHEARING

FAILURE OF THE SAMPLE IS CAUSED BY MOVING THE CENTER RING IN A DIRECTION PERPENDICULAR

TO THE AXIS OF THE SAMPLE. TRANSVERSE MOVEMENT OF THE OUTER RINGS IS PREVENTED.

THE SHEARING FAILURE MAY BE ACCOMPLISHED BY APPLYING TO THE CENTER RING EITHER A

CONSTANT RATE OF LOAD, A CONSTANT RATE OF DEFLECTION, OR INCREMENTS OF LOAD OR DE-

FLECTION. IN EACH CASE, THE SHEARING LOAD AND THE DEFLECTIONS IN BOTH 'tHE AXIAL AND

TRANSVERSE DIRECTIONS ARE RECORDED AND PLOTTED. THE SHEARING STRENGTH OF THE SOIL

IS DETERMINED FROM THE RESULTING LOAD-DEFLECTION CURVES.

FRICTION TESTS

IN ORDER TO DETERMINE THE FRICTIONAL RESISTANCE BETWEEN SOILAND THE SURFACESOF VARIOUS

MATERIALS, THE CENTER RING OF SOIL IN THE DIRECT SHEAR TEST IS REPLACED BY A DISK OF THE

MATERIAL TO BE TESTED. THE TEST IS THEN PERFORMED IN THE SAME MANNER AS THE DIRECT

SHEAR TEST BY FORCING THE DISK OF MATERIAL FROM THE SOIL SURFACES.

Figure C-4 DIRECT SHEAR AND FRICTION TESTS

C-13

METHOD OF PERFORMING PERCOLATION IESTS

The quantity yan the velocity

of flow of water which will es- "

Cape through an earth structure

or percolate through soil are

dependent upon the permeability * 5 t ; •

e

of the earth structure or soil.

The permeability of soil has • ioften been calculated by empir-

Ical formulas but is best de-

termined by laboratory tests, its ,

especially in the case of con-

petted soils.

A one-inch length of the

core sample is sealed in the

percolation apparatus, placed

under a confining load, or sur-

charge pressure, and subjected

to the pressure of a known head +

" ,

of wattr. The percolation rate

is computed from the measure-

ments of the volume of water

which flows through the sample . ... . ..

in a series of time intervals.

These rts are usually ex-

pressed as the velocity of flowAPPARATUS FOR PERFORMING PERCOLATIONS TESTS

In feet per year under a hy- Shows tests in progress on eight samples simultaneously.

dreulic gradient of one and at

a temperature of,20 degrees Centigrade. The rate so expressed may be adjusted for any set of conditions involving

the same soil by employing established physical laws. Generally, the percolation rate varies over a wide range at

the beginning of the test and gradually approaches equilibrium as the test progresses.

Vuring the performance of the test, continuous rea~irgs of the deflection of the sample are taken by means of

mircmeter dial gauges. The mocunt of cospreasion or expansion, expressed as a percentage of the original length

of the sample, is a valuable indication of the comprossloa of the soil which will occur under the action of load or

the expansion of the soil as saturation takes plane.

Figure C-5 PERCOLATION TEST

C-14

. . .. _+

•

rU

the permeability tests are useful in evaluating the de-watering requirements for excavations during construction.In addition, the potential for the soils to liquefy undershock loading is a function of-permeability.

Consolidation tests are a measure of the one-dimensional compressibility of the soil when completelyconfined. This test is performed as described in FigureC-6. Typically, this test permits evaluation of the set-tlement or consolidation of the soil when subjected tostate loads. However, the test can be performed where-in -c-lic loads are applied. In this case, it is a mea-sure of the compressibility of the soil under confinedconditions when subjected to momentary loads.

IThe electrical resistivity and corrosion potentialof the samples of soil and rock materials should be mea-sured. These tests are useful in anticipating the riskof deterioration of buried steel and concrete structures.

Dynamic Tests: Because of the requirement for thevessels to withstand shock loading, several unique labo-ratory tests will be required. Of primary interest arethe stress-strain properties of the soils and rock underthe range of dynamic loads anticipated. As stated ear-lier, shock loading may be as high as 3,000 pounds persquare inch in the form of a single shock wave passingover the site surface.

Ideally, the dynamic tests would be performed underconditions which closely simulate anticipated field load-ing conditions. Pressures of 3,000 pounds per squareinch, or pressures approaching this, are unusual in nor-mal soil testing. It is believed, however, that suitabletriaxial compression testing equipment could be developedto test samples at confining pressures on the order of3,000 pounds per square inch or less, if necessary.

However, before these high pressure tests are per-formed, the more conventional dynamic testing equipmentshould be utilized. Tests would be performed in exist-ing dynamic triaxial equipment which has a capabilityfor testing in the range of 100 to 200 pounds per squareinch confining pressures. The type of equipment usedfor these tests is shown in Figure C-7.

The test involves taking an undisturbed sample or alaboratory prepared sample of the on-site soil and apply-ing an all around confining pressure. Typically, the

C-15

M ETHOD OF P ERFORMING CONSOLIDATION TESTS

CONSOLIDATION TESTS ARE PERFORMED TO EVALUATE TIlE VOLUME CHANGES OF SOILS SUBJECTED

TO INCREASED LOADS. TIME-CONSOLIDATION AND PRESSURE-CONSOLIDATION CURVES MAY BE PLOT-

TED FROM TIlE DATA OBTAINED IN TIlE TESTS. ENGINEERING ANALYSES BASED ON THESE CURVES

PERMIT ESTIMATES TO BE MADE OF TIlE PROBABLE MAGNITUDE AND RATE OF SETTLEMENT OF THE

TESTED SOILS UNDER APPLIED LOADS.

EACH SAMPLE IS TESTED WITHIN BRASS RINGS TWO AND ONE-

HALF INCIIES IN DIAMETER AND ONE INCHl IN LENGTH. UNDIS-

TURBED SAMPLES OF IN-PLACE SOILS ARE TESTED IN RINGS UTAKEN FROM THlE SAMPLING DEVICE IN WHICH THE SAMPLES

WERE OBTAINED. LOOSE SAMPLES OF SOILS TO BE USED IN

CONSTRUCTING EARTH FILLS ARE COMPACTED IN RINGS TO

PREDETERMINED CONDITIONS AND TESTED.

IN TESTING, TIlE SAMPLE IS RIGIDLY CONFINED LATERALLY

IDEAD LOAD-PNEUMATIC

BY TIlE BRASS RING. AXIAL LOADS ARE TRANSMITTED TO THE CONSOLIDOMETER

ENDS OF TIlE SAMPLE BY POROUS DISKS. TIlE DISKS ALI.OW

DRAINAGE OF TIlE LOADED SAMPLE. TIlE AXIAL COMPRESSION OR EXPANSION OF THE SAMPLE IS

MEASURED BY A MICROMETER DIAL INDICATOR AT APPROPRIATE TIME INTERVALS AFTER EACH

LOAD INCREMENT IS APPLIED. EACH LOAD IS ORDINARILY TWICE THE PRECEDING LOAD. THE IN-

CREMENTS ARE SELECTED TO OBTAIN CONSOLIDATION DATA REPRESENTING THE FIELD LOADING

CONDITIONS FOR WIIICII TlIE TEST IS BEING PERFORMED. EACH LOAD INCREMENT IS ALLOWED TO

ACT OVER AN INTERVAL OF TIME DEPENDENT ON TIlE TYPE AND EXTENT OF THE SOIL IN THE

FIELD.

Figure C-6 CONSOLIDATION TESTS

C-16

• • • • • i • • • • •l, .

METHODS OF PERFORMING PULSATING LOAD TRIAXIAL TESTS

I

PULSATING AXIAL LOAD TESTS ARE PERFORMED TO EVALUATE THE DYNAMIC PROPERTIES AND

THE LIQUEFACTION POTENTIAL OF THE SOILS UNDER SIMULATED ANTICIPATED FIELD I.OADING

CONDITIONS.

PULSATING LOAD TESTS ARE STRESS CONTROLLED AND ARE PERFORMED ON UNDISTURBED OR

RECONSTITUTED SAMPLES OF SOIL APPROXIMATELY 6 INCHES IN LENGTH AND 2V2 INCHES IN

DIAMETER. THE SAMPLES ARE ENCASED IN A RUBBER MEMBRANE, PLACED IN A TEST CHAMBER AND

SUBJECTED TO CONIINING PRESSURE THROUGHOUT THE DURATION OF THE TEST. THE TESTS MAY

BE RUN ON SOILS AT FIELD MOISTURE CONTENT OR ON ARTIFICIALLY SATURATED SAMPLES. THE

TRIAXIAL EQUIPMENT ACTING THROUGH1 A BELLOFRAM SYSTEM APPLIES A PULSATING AXIAL LOAD.

THE CYCLING SPEED OF THE LOAD CAN BE VARIED BETWEEN V: TO 5 CYCLES PER SECOND TO

SIMULATE THE FIELD LOADING FREQUENCY.

DYNAMIC PROPERTIES DETERMINATION

To EVALUATE THE DYNAMIC PARAMETERS, THE SOIL SAMPLE IS LOADED IN CYCLIC COMPRESSION.

THE LOAD AND DEFLECTION ARE RECORDED ON TWO CHANNELS OF A RECORDING OSCILLOGRAPK.

BY TAPPING THE OUTPUT OF THE LOAD AND DEFLECTION TRANSDUCERS AND APPLYING THESE TO

VERTICAL AND HORIZONTAL PLATES, RESPECTIVELY, OF A CATHODE RAY OSCILLOSCOPE, A

HYSTERESIS LOOP IS PRODUCED. THIS LOOP IS PHOTOGRAPHED, AND THE PHOTOGRAPH IS USED TO

EVALUATE THC DAMPING VALUE PRESENT. THE PROCEDURE IS REPEATED AT VARIOUS STRAIN

AMPLITUDES TO EVALUATE THE DYNAMIC PROPERTIES IN THE RANGE OF INTEREST ON A PARTICU-

LAR SAMPLE. THE LOAD AND DEFLECTION VALUES OBTAINED FROM THE OSCILLOGRAPH ARE USED

TO EVALUATE THE DYNAMIC MODULI OF ELASTICITY.

LIQUEFACTION POTENTIAL

TO EVALUATE THE LIQUEFACTION POTENTIAL, THE SOIL SAMPLE IS SUBJECTED TO AXIAL CYCLIC

LOADING. THE MAGNITUDE, FREQUENCY, DURATION AND SEQUENCE OF LOADING IS DETERMINED ON

THE BASIS OF PAST EARTHQUAKE RECORDS. TIlE LOAD. DEFLECTION, AND PORE PRESSURE ARE

RECORDED ON THREE CHANNELS OF A RECORDING OSCILLOGRAPH. THESE RECORDS ARE USED TO

EVALUATE THE LIQUEFACTION POTENTIAL FOR TIAT PARTICULAR SOIL TYPE UNDER THE TEST

CONDITIONS.

Figure C -7 DYNAMIC TRIAXIAL TESTC-17

confining pressure is equal to the existing static over-burden pressure. While confined by this all around pres-sure, an axial deviator stress (increased axial loadingbeyond the all around confining pressure) is applied cy-clically. The strain and increased pore pressure result-ing from this instantaneous loading is measured. If de-sired, the loading cycles can be continued and the re-sponse of the sample observed for many cycles.

For the conditions anticipated at the site, it isbelieved that the strain resulting from the initial loadapplications would be useful in evaluating the dynamicstress-strain modulus of the materials. It is this par-ticular item which must be evaluated for use in calcula-tion of the depth attenuation factor, az, in Equation2-1 and in dynamic finite element analysis should it berequired.

C-4 Engineering Analysis and Report Preparation

During the progress of the soils engineering andengineering geology investigation, frequent meetingsshould be held to discuss the field and laboratory testresults as they become available and to make the neces-sary adjustments to the program.

The results of the field exploration and laboratorytesting program will provide the basis for the engineer-ing analyses. When the analyses are complete, a finalreport should be prepared which includes the followinginformation:

1. Complete description of all soils andgeologic work performed at the site.

2. Plot Plan, on which are shown the lo-cations of all borings, field tests,and geophysical profiles.

3. Specific information for each boring,including the number of samples, depthsof sampling, depth of ground water, andsamples selected for laboratory testing.

4. A general description of the site condi-tions.

C-18

5. Engineering geologic map and a descrip-tion of the engineering properties ofthe soil and rock strata.

6. Ground water conditions at the site.

7. A tabulation of all field and labora-tory test data, including the resultsof geophysical exploration program andlaboratory testing program.

8. Results of the engineering analyses andspecific recommendations concerning:

a. Relative density, type and extentof materials to be encountered ateach site.

b. Excavation problems and recommenda-tions for slope angles during con-struction.

c. Extent of overexcavation required.

d. Stability of proposed excavationand of fill and backfill slopes.

e. Bearing capacity and settlementcharacteristics of the soils sub-jected to static loads.

f. Suitability of the materials avail-able for use as fill, backfill andsubgrade support.

g. Compaction characteristics of theindividual soil types relative totheir use as fill or backfill, in-cluding the properties of the fillsoils following compaction.

h. Need and method of ground water con-trol, as well as the presence of soilconditions and water levels which ne-cessitate dewatering with well pointsor other dewatering systems.

C-19

i. Substances in the ground water andsoils deleterious to concrete andsteel.

J. Magnitude and rate of settlementunder backfill loads.

k. Design values for active and .pas-sive soil pressures at varyingdepths, including the effect ofthe water table.

1. Recommended values for staticand dynamic modulus of elasticityand Poisson's ratio for use in thefinite element analysis.

m. Compressional and shear wave ve-locity of all soils and rock pres-ent in the subsurface profiles.

The results of all tests and recommendations shouldbe organized in a form permitting interpretation by thedesign engineer. The method of presentation of test dataand results shall be by table, graphs or charts, depend-ing on the desire of the design engineer.

C-5 Services During Construction

The soils engineer and engineering geologist shouldbe retained to inspect construction operations. Thisinspection should be frequent enough to permit adjust-ments in the design if unusual or unanticipated subsur-face conditions are encountered.

C-20

I3

UnclassifiedSecurity Classification

DOCUMENT CONTROL DATA- R & D(Security classilication of title, body of abstract and indexing annotation .u..t. be entered when the overall report Is clAssified,)

1. ORIGINATING ACTIVITY (Corporate author) za. REPORT SECURITY CLASSIFICATION

T. Y. Lin & Associates14656 Oxnard Street 2b. GROUPlVan Nuys, California 91401

3. REPORT TITLE

DESIGN OF PRESSURIZED FUEL STORAGE TANKS TO RESIST AIR OVERPRESSURES

4. OESCRIPTIVE NOTES (Type of report and inclusive dates)

Final Report - 28 Dec 1968 - 28 May 19705. AUTHOR(S) (First name, middle initial, last name)

William E. Gates

. REPORT DATE 7s. TOTAL NO. OF PAGES 7b. NO. OF REFS

May 1970 27 17So. CONTRACT OR GRANT NO, 98. ORIGINATOR'S REPORT NUMBER(S)

N62399-69-0-0017b. ROECT NO. 6825351-009 PROJECT SANGUINE

C. 9b. OTHER REPORT NO(S) (Any other number that may be assignedthis report)

d. CR 69. 024

10. DISTRIBUTION STATE"MENT

Each transmittal of this document outside the agencies of the U. S.Government must have prior approval of the Naval Electronic SystemsCommand (NAVELEX) Washington, D.C. 20360 , Code PME 117.

II. SUPPLEMENTARY NOTES 12. SPONSORING MILITARY ACTIVITY

P Naval Civil Engineering LaboratoryPROJECT SANGUINE JPort Hueneme, California 93041

I3. ABSTRACT

'--4Ths report presents a design procedure and cost optimization curvesfor cylindrical steel tanks with spherical end caps that are buriedwith their axis parallel to the ground surface. The fuel storagetanks are designed to resist high overpressures associated withnuclear air blast. Internal pressurization .has&be n utilized tominimize material and fabrication costs and increase the bucklingcapacity of the tanks.

The design procedure is believed to be valid for overpressures range,from 10 psi P,3000psi. The overall reasonableness of the designprocedure hasezbeen confirmed by experimental tests on buried steelcylinders and pressurized tanks in granular materials. ( ).

OR 73 (PAGE I)F NOV 614 Ti__________,

0102.014.6600 S-cutity.s iiLti In

UnclassifliedSecurity Classification

14. LINK A LINK B LINK CKEY WORDS , - - -

ROLE WT ROLE WT ROLE WT

Buried fuel tanks

Internal pressurization

Nuclear blast resistant

Steel tanks

Shallow buried

Design procedure

Cost Optimization

Soil-structure interaction!F

hII

DD .R,..1473 (ACK) Unclassified(PAGE 2) Security Classification