anchor bolt development lenght

8/13/2019 Anchor Bolt Development Lenght

http://slidepdf.com/reader/full/anchor-bolt-development-lenght 1/70

~ c = c ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ = = = = = = = ~ ~ ~ ~ _ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ .. ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~

DEVELOPMENT LENGTH FOR NCHOR BOLTS

y

John E Breen

Final Report for ProjectNo

3-5-63-53Development Length for Anchor Bolts

Conducted for

TEXAS HIGHW Y DEP RTMENTInteragency Contract No 4413 - 830

and

CENTER FO t R A N S P O R T A T ~ RESE RCH LrIl Y

IIIUIIII



PREF CE

On June 1 1963 a research program to study design cri ter i for the proper

lengths for anchor bolt embedment into footings was undertaken by the University

of Texas Center for ighway Research in cooperation with the Texas ighwayDepartment and the Bureau of Public Roads. The program included a series of

tests on thir ty-six anchor bolts with varying embedment conditions. The anchor

bolts were embedded near the edge of a square footing specimen and were under

a combination of flexural tension bond and spl i t t ing conditions closely

approximating those in the prototype footing specimens. In these tests loaded

end slip behavior was determined at nominal working loads t f i r s t yielding

and in many cases at ultimate capacity. This report presents the resul ts of

these tes ts and a survey of the trends indicated thereby.



ACKNOWLEDGEMENTS

The tes ts described herein were conducted as a part of the over-all

research program of the University of Texas Center for Highway Research under

the administrat ive direction of Dean John J McKetta. The work was sponsored

joint ly by the Texas Highway Department and the Bureau of Public Roads under

an interasency contract between the Universi ty of Texas and the Texas Highway

Department. Liaison with the Texas Highway Department was maintained through

the advisory committee for this project consisting of Mr Wayne Henneberger

Mr Larry G Walker and Mr De Leon Hawkins the contact representative. The

study was directed by John E Breen Assistant Professor of Civil Engineering.



-------------------------------------------------------------------------................ ..

Object and Scope

CH PTER 1

INTRODUCTION

The general objective of th is invest igat ion was the establishment of the

development length or the embedment length required to develop the tens i le

yield capacity of STM A-7 anchor bolts with a yield strength of 33 ksi . At

the same time i t Was to explore the effect of various types of end anchorages

and certa in other variables . A substant ia l amount of recent t e s t data has

been compiled on the proper development length for the STM A305 deformed

reinforcing bars. Very l i t t l e material is avai lable concerning the length

required to develop the strength of large size anchor bolts . The main infor

mation exis t ing is a series of concentric pul lout t es t s reported by Duff

1Abrams which involved plain bars anchored with both nuts and nuts with washers.

In th is type t es t the concrete around the bolt is under compression. s has

been shown by Ferguson Thompson and Turpin2

the resu l ts of such concentric

pullout t es t s do not ref lec t the bond-slip behavior of elements which are in

the tension zone of beam or footing members such as the prototype sign support

footings. In addit ion the only s l ips which were measured in the Abrams ser ies

were those at the free end of the embedded bolt . Tests on anchorage lengths



r

rh I rhREACTION BEAM

II

II.

~ ~ L O A D CELL

1 4 YOKE , , , ~ R A M yOKE .......BEARING BLOCK ...

L J

t

STEEL BEAM

..J:

t:

Jd

;;JI

......

a )

:lJ----------i ~ T S T

r-I::= _ _ _ __ :.:1:1 BOLT

II

PLATE

L U

t

2



3

only two bolts were used. The to ta l number of bolts involved was thir ty s ix.

The anchor bolt sizes ranged from 1 1/4 to 3 with embedment lengths of 10

and 15 diameters. End anchorages consisted of ei ther a standard nut or a

combination of a standard nut and washer.



Specimens

CHAPTER 2

TEST PROGRAM

All of the footing specimens used in this program were of the same general

type. When 1 1/4 through 2 diameter bolts were used, four anchor bolts were

embedded in each footing specimen. However, only two bolts were used in each

specimen with the 2 1/2 and 3 diameter anchor bolts . The individual anchor

bolt specimens are designated by a four part code symbol. The f i rs t number

represents the nominal bolt diameter. This is followed by the l e t te r N for a

specimen with a nut anchorage only, W for a specimen with a nut and washer anchor

age and S for the series with special type of end anchorage. This l e t te r is

followed by a number which is the embedment length in terms of number of bar

diameters. The final l e t te r indicates the designation of individual companion

specimens. For example, the designation 1-1/4 W lOa designates a specimen in

which a 1-1/4 bol t with a nut and washer anchorage was embedded for 10 diameters

in a footing specimen. The a indicates that i t was the f i r s t of this type tested.

In a few cases this code designation is followed by a -2. This indicates a

retest to fai lure of a specimen which was taken up only to yield in the in i t ia l



B 1

. . , . - - - - - ~0 0 ,

/ ,/ \

b 0\

/ \I I

0/\ /\ /

0. . , II

E=3

ANCHOR BOLT

, - - _ { 4 - # 7X51

-S

4 -#7 X41

-0

1 . #2 SPI RAL

(PITCH Sll)

SPECIMENS B

ALL IY4 , IY2 223/4

ALL 134 , 211 24y411

ALL 2 Y2 32

ALL 3 32

1' ..{\J

{\J

II

m

E

3

4

4

4

BI

l

---

/ 0 d'-,

///0 0 \

8 31. 8 3 1. \78 ,8 I

I\\ ~ 0/, \ /

/ 'o /,:;...-

E=4

ANCHOR BOLT

L....-_{4-#9 x51

-S

4-#9x41-011

'---___ #2 SPIRAL

(PITCH S )

LONG BARS

8 - 7

8 - 9

8 - 9

6-#11

1--- B-

1.....- -..... .....

/, -

/ '-' '-'/ \

\v \{\J

\, ( \Jm ,

1-' 24 1'f l )-- - II

\ m\\\

/

) // '

.....

'-I E =411

ANCHOR BOLT

S -# l l x 51

-S

#2 SPIRAL

(PITCH S )

FIG. 3 SPECIMEN DETAILS

0 \



E ACE OF CONCRETE

SA/ _ I _ ' ~ ' ¢ 1 ~ ~ m ~ \ ~ ~ ~ ~ ~ ~ ~ \ \ I - - - - - - - - - - - - ~ I - JI 6 » 1. 17.5 » 1

243A

I 11

.1

FIG. 4 DETAILS OF SPECIAL

ANCHORAGE

SERIES 5

SS

SC

)1111 75 )10 I

So .

l \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ ~ I11

.111 17.5 /--.j



8

size was 1 . The cement factor was five sacks per cubic yard and the slump

range was 3 to 4 . In general, the water-cement ra t io was seven gallons per

sack. The water used was Austin ci ty water with no admixtures added. When

ready mixed concrete was used inspectors were provided at the plant to insure

proper batching and mixing procedures.

Anchor Bolts. The anchor bolts used were specified to be manufactured from

ASTM A7 steel with a minimum yield strength of 33 ksi . In a l l cases the bolts

were fabricated in local commercial machine shops from round bar stock. In one

set of 1-3/4 bolts , the companion specimen indicated that a high carbon or other

alloy steel Was substi tuted and the yield point was substantial ly above that

required. Appendix Figs. A1 and A2 shaw typical stress s t ra in curves for the

bolts used. The threaded sections were machined to conform to USS thread

specifications. The anchor bolts were not galvanized.

Reinforcing Steel . The reinforcing steel used was deformed ASTM A305)

bars meeting ASTM A15 intermediate grade specifications. The spirals were

wound and furnished by a commercial supplier using number 2 (not deformed) bars

of ASTM A15 intermediate grade s teel

Forms. All the footing specimens were cast in wooden forms as shown in

Fig. 5. The bolts were cast in a vertical posit ion at the top of the forms.

In al l specimens the height of the footing specimen was maintained at 6 feet .

This insured that the water gain effect would remain relat ively constant in



10

concrete bucket and an overhead traveling crane. The concrete was vibrated

into place with a large internal vibrator. Standard compression tes t cylinders

and modulus of rupture beams Were cast from each of the mix batches. In

general the concrete samples were taken from the portion of the batch which

was placed in the top half of the footing specimen. This was the portion of

the concrete surrounding the test bolts. Because of the slumps used good

compaction was obtained.

Curing. The specimens were trowled to a smooth finish and then l e f t over

night in the forms in the laboratory without any special protection other than

a piece of moist burlap placed over the open end of the form. In general this

burlap covering was kept in place unt i l the sixth day. Then the curing cover

and wooden side forms were removed and the specimen was moved into posit ion and

prepared for test ing. Wire rope slings were uti l ized for handling the specimens.

At no time was the specimen l i f ted by using the embedded anchor bolts.

Test Procedure. While the general procedure used in testing al l of the

bolts was identical, i t was found necessary to increase the capacities of

some of the elements of the tes t ing r ig to handle the 2-1/2 and 3 diameter

bolts. However, this was mainly a scaling operation. o dist inct ion will

be made between the two sets of equipment in the text or figures other than

to note that no monitor load cel l could be used in the 2 1/2 and 3 diameter

bolt tes ts since the ranges of loads applied were beyond the capacities of



Fig 6a Loading system



3

Fig. 6e. Test arrangement large bolts) as viewed from above specimen



14

of the compression centroid.

The instrumentation uti l ized was relat ively simple. In addition to the

pressure gage to measure the hydraulic pressure applied to the ram, and a load

cell to check the applied load, measurements were made of deflections and loaded

end sl ip. Dial gages were used as indicated at H, I and J in Fig. 6b to

measure the deflection at the two reaction straps and in the footing specimen

direct ly under the applied load. In addition an optical micrometer mounted on

the theodolite shown at K in Fig. 6c was used to measure the relat ive sl ipbetween the anchor bolt and the concrete face. In order to make this measure

ment, three targets were instal led on the specimen. Two of these targets were

mounted (one 1 above the bolt , the other 1 below the bolt) on steel rods that

were embedded in the concrete during casting. The third target was mounted

(using a strong permanent magnet on the bolt one half inch from the face of

the concrete. (L). The optical micrometer was used to read the relat ive dis

placement between the target mounted on the bolt and the targets which were

fixed to the concrete. This micrometer attachment has a sensi t ivi ty of 0.001 .

This system proved to be a reasonably sat isfactory method of taking the slip

measurements in an otherwise remote and fa i r ly inaccessible region.

The loads were applied in increments and held unti l the deflection dials

and the optical micrometer could be read and any cracking marked and recorded.

Loading for a part icular test would require approximat,ely one to two hours.

Because of the fact that each footing specimen contained a multiple number of



with 2-1/2 and 3 diameter anchor bolts only contained two bolts in each

footing, these tests were taken to failure without retest ing.

Specimen Behavior and Failure

15

Loaded end sl ip star ted almost immediately upon in i t i a l loading and pro

gressed with gradually increasing increments. In a few cases spl i t t ing of the

cover s tar ted over the bolt at the loaded end and progressed towards theanchored end. In most cases very defini te flexural cracks opened up across

the face of the beam. Usually the largest flexural crack occurred at the

vicini ty of the anchorage of the bolt (which represents the cutoff point ) .

The final fai lure of those specimens which were taken to ultimate fel l into

three different classes;

1. Specimens in which very defini te and pronounced spli t t ing appeared

along the ful l length of the bolt a t ultimate load, as shown in Figs. 7a and

7b. I t can be seen in Fig. b that a very wide spl i t t ing crack extends from

bolt 1-1/2 N lOa to the face of the specimen. I t can also be seen that the

concrete cover over this bolt is breaking off into two dist inct wedges. This

i s similar to fai lures found in pullout tests of deformed bars.

2. Specimens in which very l i t t l e longitudinal spl i t t ing appeared prior

to or a t failure but in which very intensive spalling and crushing of the

concrete in the vicini ty of the anchorage of the bolt appeared. Most of the



16

Fig. 7a. Spli t t ing fai lure



8

Fig 9a Fracture of bolt inside concrete



paz

Variables

CH PTER 3

N LYSIS OF D T

20

The primary variables in this investigation were three in number, namely

bolt s ize bolt length and type of end anchorage. Accompanying the variat ionin bolt s ize was the corresponding variat ion in the size of the standard nuts

and washers used for the end anchorages. A complete tabulation of the bolt

nut and washer dimensions i s given in Table 1.

In addition three secondary variables were introduced due to the nature

of the test specimens used and to the construction pract ices. The f i r s t of

these is the rat io of the clear cover over a bolt to the bol t diameter. Follow

ing the standard practices of the Texas Highway Department bolts in a footing

specimen of a given depth were located with a fixed center to center dimension.

Inherent in this method of detail ing is a decreasing clear cover for larger

bolt sizes. This can be represented both as an absolute decrease in cover and

as a decreasing rat io of cover to bolt diameter. These values are also tabu

lated and are given in Table 1.

Secondly the concrete strength varied somewhat during the tes t program.

While i t was desired that al l concrete should have the same f due to the

-



/)

ZIJJ

UIJJa../)

LL0

z

1 2 ~ ~ ~AVERAGE

1 -

8 ~

I

6 ~

4 ~

I



/)

5 0 ~ i ~

40 -------

en 30/)

UJ

Q:-

/)

- I~ 20

t;

1

TENSION TESTS

I VI , BARS

A SMOOTH BAR

- - - B THREADED BAR(ROOT AREA)

C THREADED BAR

(ASTM "MEAN AREA")

O ~ - - - - - - - - ~ - - - - - - ~ - - - - - - ~ - - - - - - ~ ~ - - - - - - ~ - - - - - - ~ - - - - - - - - ~ - - - - ~o 0.001 0.002 0.003 0.004 0.005 0.006 0.007 0.008

STRAIN in/in

FIG. II EFFECT O METHOD O CALCULATION OF STRESS IN A THREADED

SECTION

.

N.p.



25

concrete.

The plot of this s l ip against either load, bol t s t ress or bond s tress

gives the same curve shape for any given specimen since a combination of given

bolt size and length establishes f ixed ra t ios between load, tensile s t ress

and average bond stress.

I t has been mentioned that in the tes t procedure many of the specimens

were loaded i n i t i a l ly only unt i l apparent yielding. Then the specimens were

unloaded and a corresponding set of zero load readings were taken. Upon

completion of the test ing of the other bolts in the footing specimen, many

of the bolts were reloaded unti l fa i lure became evident . After examination

of the data observed in these t es t s i t was found tha t a l l of the reloaded

specimens seem to have a common behavior. As indicated in Fig. 12 the s teel

s t ress-s l ip curve OA indicates the results of the in i t i a l test ing of specimen1-1/4 W lOb. Testing was discontinued a t A and the load released. The to ta l

permanent s l ip upon removal of the load was observed as indicated by the

descending branch AB. After test ing of several of the bolt specimens, tes t

1-1/4 W 10b-2 was run. In th is tes t the s tee l s t ress-s l ip curve BCD was

observed. On the same figure is i l lus t ra ted the resu l t s of a general ly

similar tes t specimen in which the bolt was i n i t i a l ly tes ted to a much higher

sl ip level . This s teel s t ress-s l ip curve (1-1/2 N lOb) is shown by curve

I t should be noted that the l a s t observation agrees with an observed value

1



H

/)

60

50

40

/)

/)

wa

/)

30w

/)

Z

«w

E 20

Een

10

0..2

A .-

1 /4 W O b ~ /fI

/

c

---IY4NIOb

J /-1 14 W Ob-2

.02

I tJ /

I

J,

t

1

B

.04 .06 .08 10

GROSS END SLI P in

D

- - = IY 4W IO b - 2

.12 14 16

FIG. 12 EFFECT OF INTERRUPTED LOADING ON STEEL STRESS-SLIP RELATION

N

0 \



1-4

oo

ooooI.IJ

a:l - /)

..JI.IJI.IJl - /)

Z<tI.IJ

::?

60

50,

/.

I ~ ~ ~ U ; {40

30

20

10

-------;:::... ------ '-=----------------. ,.. . . .- I W lOb

~ I Y N I O b

IY4 N 150

/

•

•.

.

I

0

0.02 0.04

o0.06

0.02

0.08

0.04

0.10

0.06

GROSS SLIP - in.

0.12

0.08

0.14

0.10

FIG. 13 STEEL STRESS-SLIP CURVES FOR IY4 BOLTS

0.16

0.12

'.



60

50

tot

(/) 40

( / )

/)

UJ

0:l-

(/) 30..JUJ

UJ

t;Z

~ 20::

10

00.02

FIG. 15

~ ..- ..- - - - - I I ' ' ' ~ - ---7 . .... - - - - - - - - --- , - - ---- - - - - -

~ ---

13/4 N 100

(HI-STRENGTH BOLT)

. . . . . / .

/ / ; A/ / . /

/ , ,4---13/4 N 150

" • i 13/4 W 100 (HI-STRENGTH BOLT)

. I ~ W 15 a

_ . 1 1 0 / 4 : ~ ~ ~ . ~ O R A G E

0.04 0.06 0.08 0.10 0.12 0.14 0.16

0 0.02 0.04 0.06 0.08 0.10 0.12

GROSS SLI P - in.

II

STEEL STRESS- SLIP CURVES FOR 10/4 BOLTS

N\0



60

50

.....(J) 40

.,..(J)

(J)

LLJ

aI -(J)

30J

LLJ

LLJI -(J)

Z<:(

20LJ

:E

00.02

FIG. 16

__ ------ r WIOo. ..,.,. 1_

. . . . , . -- , 1 2 N ' O b

- 2 N 100 - - - - - - - - ~ ........ .. .-: ....... := == =. . . . . - . - - - - - - - - ~ - - - - - ~ t '. ~ . . ; . . . . - . . - --_;; -- - - -

, . , , - -

0.04

0

... ....,.--1 I f I I I I I t ~ .

••

/2 N 150

- , 2WI50

0.06 0.08 0.10

0.02 0.04 0.06

GROSS SLIP - in.

0.12

0.08

STEEL STRESS-SLIP CURVES FOR 2 BOLTS

0.14

0.10

0.16

0.12

w0



o10

0

0

1

0

~ z

J

V rt

(

J

IS~ SS3~lS l33 S N~3~

00OJ0dto0dV0

0

d

10

Z

(

J

0d

o

o



50

H~ 40

en

enI.IJ

a

en 30J

I.IJ

I.IJ

t;

Z

~ 20:E

10

y4 SC 1 45°)

. , / ~

,- - . .

i,..., 13/4 S8 1 30°)

.... 13/4 SA 1 SAME AS 13/4 N 10 SERIES)

__ .£10/4 SO ~ ~ ~ ~ ~ : ~ - -

O ~ · - - - - - - - - ~ - - - - - - ~ - - - - - - - - ~ - - - - - - ~ ~ - - - - - - ~ - - - - - - - - ~ - - - - - - - -0.02 0 04 0 06 0.08 0.10 0.12

GROSS SLIP - in.

FIG. 19 STEEL STRESS - SLIP CURVES FOR 10/411

BOLTS

WITH SPECIAL ANCHORS

0.14

ww



I ll



0 030

2 KSI



36

include the 2 diameter bolts since only one of these shows a sl ip in excess

of this l imit . Both 2-1/2 bolts with nut and washer anchorages meet the

cr i te r ia although those with nut anchorages do not . However, the 3 bolts

have an exactly opposite behavior. Bolts with nuts developed much less sl ip

than those with nuts and washers. This is apparently due to the extremely

shallow cover (1-1/4 clear cover) over the washer in this series which led

to early cracking. I f the cr i te r ia of 0.02 is accepted, a l l of the bolts

would meet the cr i te r ia

Figure 2 i l lust ra tes the observed s l ip a t the nominal yield point of the

bolts , that is a mean area steel s t ress of 33 ksi While the allowable s l ip

cri ter ia for th is case cannot be easi ly formulated, i t i s in terest ing to note

that al l specimens except three of the 3 bolts indicate less than 0.03 end

sl ip at the nominal yield. In the case of the 3 bolts , the specimens with10 D embedment length had failed (as indicated) a t less than 33 ksi , while

the 15 D specimen with nut and washer anchorage had a sl ip of approximately

0.046 a t this stress level . I t can again be noted that generally the 15 D

specimens with washers show the least amount of s l ip a t th is stress level . In

the 10 D specimens, s l ip does not seem to vary consistently with the type of

end anchorage. In some cases, specimens with nuts show less sl ip than specimens with nuts and washers. The opposite .is true in other cases.

In order to provide a quali tat ive indication of the relat ive relat ion



9

Fig 22 nd movement of nut

-------- .. ... . . . , . . , ~ ~ ~ . ~ . , ~ ~ . - - - , . ~ . .... . , ~ - ~ . - . - .. - - - . - ·-··-··--··-·-,· .·-·· ·· _ . M' . ..• ,_ ., ....



5 I-L15 ANCHOR

o 10 N• 10 W

[ ] 10 SO

4 l- I I 15 N

• 15 W

f) •.. 3

jf)

0

•f)

•0

141

a:l - • •f)

J

I l 0 •LL I 2 8 i • 8

LL I

I -0 •f)

0 I I •] NO ANCHOR)

10

O L I - - - - - - - - - - - - - - - - - - - - ~ - - ~ - - ~ - - ~ - - - - - - ~ - - - - - - ~ - - - - - - - - - - - - - - - - - - - - - - -I Y4 IY I 3/4 2 Y 3

BOLT DIAMETER t in

FIG. 24 STEEL STRESS AT A SLIP OF 0.01 in

P-o



t

41

bearing capacity prior to the development of thei r ful l yield point s t ress .

Unless this yield point can be developed, the s t ructural design wil l not havethe desired factor of safety. Accordingly, even though i t i s real ized that

the extreme amounts of sl ip occurring at high loads would be undesirable, a

study has been made of the ultimate strength character is t ics of the specimens.

Fig. 25 shows the ult imate s teel s t ress f ) as a function of the boltsu

diameter, D. All specimens tested to fai lure except the 10 D 3 bolts developed

the nominal yield point of 33 ksi . In general the 15 D embedment was somewhatmore effect ive than the 10 D embedment. However, this effectiveness was

defini te ly much less than the 50 increase in embedment length. Upon study

of the data i t was fe l t that one variable in the tes t was not ref lected in

this figure. This was the amount of clear cover over the bars. As has been

discussed previously in the section on specimen behavior and fai lure, the

larger bolt specimens with less cover fai led due to crushing of the concrete

over the anchor. After an attempt to re la te ult imate steel s t ress to clear

cover proved unsatisfactory i t was decided to invest igate the re la t ionship

between ult imate s teel s t ress and the ra t io clear cover divided by bolt

diameter. This re la t ionship is i l lus t ra ted in Fig. 26. Comparison with

Fig. 25 indicates clear cover as a function of bolt size seems to be a

signif icant parameter. Two dis t inct curves are vis ib le , one for the 10 D and

the other for the 15 D embedment length. In addition, the nature of the



.Jo

o ]

<

•0

<

•0

<30

••

oI-

•<]0

o

o

0

0

0

~

1

V

~

S'S1S1331S 31WI1ns

.



•;

j

43

stresses under such a t r iaxia l stress condition. However, even considering

this , such magnitudes of stress are surprising.

In order to check the general magnitude of the bearing stresses developed

by the anchorage devices, a simplified set of computations were made. Since

the ultimate bond characterist ics of the 1-3/4 bolt without an anchorage

device were comparatively small, i t was assumed that at fai lure the total

tension force in the bolt was taken by the anchor device. This certainly

was true in the cases of the two ultimate tension failures occurring at the

anchors inside the specimen. I t was probably true in the majority of other

specimens. Since only the general magnitude of the bearing stress i s sought,

no further precision is just i f ied Equating the tensi le force in the bol t

to the bearing stresses on the nut or anchor we see that

A f = A fs su c c

where A is the mean area of the bolt ,s

f is the steel s t ress on the mean

area as tabulated in Appendix A,su

f is the concrete bearingc

is the area of the concrete in direct bearing. Therefore,

A= @ f

A suc

stress , and Ac



D

Nominal

BoltDiameter

in.

1 1/4

1 1/2

1 3/4

2

2 1/2

As

Mean

StressArea,

Sq in.

0.969

1.405

1.899

2.498

3 999

T BLE 2

BE RING RE S

A A Acn cw- - L

Net Net Acn

Nut WasherArea, Area,

Sq in. Sq. in.

1.82 5.84 0.533

2.61 7.84 0.538

4.13 10.14 0.460

6.00 14.59 0.417

7.29 14.73 0.548

44

A A As cr- - L

A Critical Acw cr

Stress

Area

Sq in.

0.166 27.0 0.036

0.180 26.5 0.053

0 187 47.9 0.040

0.172 47.0 0.053

0 272 45.3 0.088



70

H 60

~UJ

FAI LURE TYPE::

50

I T = BOLT TEN SION

S = SPLITTINGJ

UJ C = COVER CRUSHING-

UJ 40

:i

~ ANCHORj 30l XsS CS-C 0 10 N

•lOW

II 15 N15 W

20CLEAR COVER

833 1 1 1.5 1 79 1 9BOLT DIAMETER

BOLT SIZE in. 3 V IY2 2 10/4 IY4

.j::VI

FIG 26 ULTIMATE STEEL STRESS AS A FUNCTION OF CLEAR COVER



46

washers were used, i t can be concluded that the to ta l area of the washer was

not ful ly effect ive in developing bearing. Examination of the specimens af ter

failure indicated that the washers were general ly bent back over the nuts.

This would indicate a diminished effectiveness in bearing.

Bond Stresses

Any single curve for s teel s tress vs. s l ip by a change in ver t ical

scale, can be turned into a bond s tress vs. s l ip curve. However, the change

in scale is different for each bol t size and for each embedment length.

Because of the way in which the average bond s t ress has been defined, speci

mens of identical bol t size but different embedment lengths and with almost

identical steel s tress vs. sl ip curves would yield quite different bond s tress

vs. sl ip curves. For instance, in Fig. 17 specimens 2-1/2 W 10 and 2-1/2 W 15

have pract ical ly identical s teel s t ress-s l ip curves up to about 3 ksi . Yet,

i f the bond stress was computed at any given s l ip the bond s t ress for

specimen 2-1/2 W 15 would be only 2/3 that for 2-1/2 W 10. This is probably

not ref lected physically since i t i s possible that the local bond stresses

a t a part icular point might be almost ident ica l . s a matter of fact near

ultimate, local bond stresses for the two cases are probably close to zero

since t is the end anchorage which seems important. Therefore, i t was fe l t



7

Fig 27a Cover sp lling



49

By matching the actual location of the anchor and the cracking patterns shown

in Fig. 27c with those of Fig. 27b the location of the nut and washer have

been indicated. t can be seen that the crushing seems to radiate from the

anchor in a pyramidal pattern a t angles somewhere between 30 and 45 degrees.

Postulating that this same type st ress dis t r ibut ion might be taking place in

three dimensions leads to the idea of a t runcated cone of st ress radiating

out from the head.

The following hypothesis is advanced: Assume that the cr i t ica l s tress

area for bearing will occur not at the end anchor but on the base of the

cone defined where this cone of st resses intersects the surface of the speci-

men This is i l lus t ra ted in Fig. 28. This area which we will cal l the

cr i t ica l st ress area can be computed as

where is the to ta l cover measured from the center l ine of the bolt and D

is the diameter of the bolt . The values of this cr i t ica l st ress area are

given for the various bolt size and cover combinations in Table 2. By

recomputing the bearing stresses on this f ict ious s tress area a new index

can be obtained which will include cover. Since



51

7

0 IONto t e lOWf)

6 6 15N

A 15W

~ I ~6S c FAI LURE TYPE

II

' A S C T = BOLT TENSIONc J

5AC S = SPLITTING

C = COVER CRUSHI NGf)

e S Cf)

D = DISCONTINUEDJ 6Ca::: AT YIELD

OS Cf)

C) 4 e cza:::« OS t ~J

D

J

: CtI§u

C ~ S S ~ t:::u C

f



90

8

70

60

40

II S C

~ S C

' .S-C

~ CllC

A i ~ C e

'

o ION

e lOW

II 15 N

~ 15W

FAI LURE TYPE

T = BOLT TENSION

S = SPLITTING

52

C = COVER CRUSHING

o = 01 SCONTINUEo

AT YIELD

~ S

~ ~

.

.'



53

should possibly ref lec t the general varia t ion of tensi le strength with com-

pressive strength. Since many investigations in bond, shear, diagonal

tension, and tensi le capacity of concrete have shown that these values tend

to vary with the square root of f I ra ther than as a l inear function of f I

C c

th is correction was used as an index of the varia t ion of concrete strength.

In Fig. 30, the same data are re-expressed as a function of the ra t io f /r -.j.l.c

and the parameter clear cover/bolt diameter. Comparison of Figs. 29 and 30

wil l indicate that much of the scat ter has diminished. This is especially

t rue in the low cover-diameter rat ios . The 3 bolt specimens for 10 D and

15 D now plot almost together reducing a great deal of the scat ter of the

previous figure.

I t was fe l t that the relat ion shown in Fig. 30 might prove helpful in

formulating some rough guides for design cr i te r ia . Since i t is desirable

that any such cr i te r ia be on the conservative side, a lower envelope Wasdrawn in the form of the st raight l ine AB indicated on Fig. 30.* This l ine

can be expressed by the equation

In order to have a clearer idea about the relationship proposed, i t is

necessary to investigate what happens a t the l imits and what the overall shape



100

80. .RANGE OF DATA

REPORTED

I \60

fer

Jif40

1

EXPRESSION

DERIVED

20

ABRAMS

I I

O 34 5 6

CLEAR COVER/BOLT DI METER

FIG. 31 PROBABLE GENERAL FORM OF CRITIC L BEARING STRESS COVER

RELATIONSHIP

\.IIp



55

However, in this par t icular specimen a greater amount of cover was used than

in the present ser ies . The clear cover over the bol t was 4-1/4 . This makes

the ra t io clear cover/diameter equal to 2.43. The computed f =

The concrete st rength was 5734 psi and hence the ra t io f / ~ =cr c

1640 psi .

21.5. This

point is plotted on Fig. 31. In addition, in Abrams' original set of pullout

specimens he tested some 3/4 bars with a nut and washer anchorage embedded

for 8 in a 8 diameter concrete specimen. The length of embedment is again

approximately 10 diameters. For these specimens the ra t io clear cover/diameter

equals 4.84. The specimens fai led by sp l i t t ing of the concrete. He reports

only average bond stresses but from these i t can be calculated that f = 349cr

psi for the specimens with nuts and 385 psi for the specimens with nuts and

washers. The cylinder strength would be approximately 1860 ps i hence the

ra t ios f ~ = 8.1 and 8.95 respect ively . These values are also shown incr c

Fig. 31 and tend to indicate the general t rend of the resu l t s . However, i t

must be emphasized that any conclusions as to the desirable amount of cover

deduced from these results are extremely approximate and should be considered

within the l imits of t es t data avai lable in developing these e x p r e s s i o n ~ As

in a l l empirical expressions a great danger exis ts in extrapolat ing l imited

tes t resu l t s . However, in th is case i t is fe l t tha t an attempt should be

made to give some guidance as to the magnitude of adjustment of cover in orderto increase anchor bolt performance. Further tes t ing should be done to

validate such changes.



56

For each standard bolt there is a fixed re la t ion between D and A Substi tutingsm

these in the previous equation we can get an expression which is a cubic function

of C. Solving this numerically and obtaining C to the nearest quarter inch we

obtain the following resu l t s

D C Calculated C Actual

1.25

1.5

1. 7

2.00

2.50

3.00

4.25

5.00

6.00

6.75

8.50

10.75

6 ~ : :

8

8

8

8

A comparison of the minimum desired cover and the cover uti l ized indicates

that the present cover is sa t i s fac tory except for the 2-1/2 and 3 bolt sizes.

This would be expected from the nature of the tes t resu l t s The specimens

which suffered from apparent lack of cover were the largest s izes. The results

would indicate that the 2 1/211 bolt cover should be increased s l igh t ly (The

cover from the center of the bol t to the outer edge i s equal to C/2.) This

indicates that the cover should be increased from 4 to 4-1/4 . However, for

the 3 bolt a more substant ia l change is indicated. The present 4

cover



57

used. SA was a standard 1-3/4 bol t 10 D embedment and a standard nut. Bolt

SB wasthe

sameas

SAexcept for the

3bend a t midpoint. Bolt

SC wassimilar

to bolt SB except that a 45

bend was put in the bol t . Bolt SD simply had

17-1/2 of plain unthreaded bar with no anchorage. The bent part of the bolts

were placed along a radius towards the center as i l lus t ra ted in Fig. 4. The

resul ts of this ser ies is shown in Fig. 19. I t has been previously mentioned

that the specimen with no end anchorage simply slipped out a t a very low s tress

level . Specimen SA behaved the same as the other specimens with s imilar embed-

ment. Specimens SB and SC indicated a sl ight ly s t i f fe r load-slip response.

However, when the values are compared to those shown in Fig. 15 for the similar

1-3/4 bol ts i t wil l be noted that they did not develop as high an ult imate

strength. Possibly th is was due to very severe internal sp l i t t ing . The higher

bearing stresses developed in the other specimens could not be obtained.

Another poss ib i l i ty is that the bend in the bar reduces the effect ive embedment

length and causes a s t ress ra i ser with an adverse effect . Unfortunately the

results of th is series must be regarded as somewhat inconclusive in that

while the bending of the bolt seems to give somewhat bet te r service load

character is t ics the ult imate strengths found were lower than in the unbent

anchor bolts .

P i lo t Tests-High Strength Bolts



The following conclusions were made:

1 The pullout tests indicated that A7 (33 ksi yie ld point)steel anchor bolts can be ful ly developed with a 15 dia

meter embedment length in a l l bolt s izes. (Steel st ress

computations are based on the mean diameter st ress area.)

2. Anchor bolts of A7 (33 ksi yield point) steel with dia

meters of 2-1/2 or less can be ful ly developed with a 10

diameter embedment length.

3. An important variable affecting the developable tensi le

strength is the amount of cover over the anchor. Results

of calculations based on a l imited empirical expression

indicate that the clear cover should be increased for the

2-1/2 and 3 diameter bolts over the 2-3/4 and 2-1/2

used in these tes ts .

4. Considering a 0.01 gross end s l ip at service loads as a

l imi t the following mean area steel stresses were

developed in 4650 psi concrete: (see Fig. 24)

D

in. 10 D Embedment 15 D Embedment

1 1/4 28 ksi 32 ksi

60

''-



61

1. The clear cover over the large bolts (2-1/2 and 3 diameter) should

be increased i f possible. The clear cover should be at leas t as

great as that used with the smaller (1-3/4 and 2 D bolts . A more

detailed recommendation for determining the cover required is given

in Chapter 3.

2. Additional specimens should be tested to investigate the effect of

clear cover, part icular ly with respect to the development of large

bolts . The data provided by this program is quite l imited in this

respect.

3. Consideration should be given to the effect of repeated loadings on

the anchorage characteris t ics . Literature searches have indicated

that very l i t t l e information is available in this area.

4. Further study should be given to the use of higher strength steel

bolts . The l imited pi lot tes ts indicate these may be more eff icient

a t service load levels than the larger bolts . t is possible that

some of the cover modifications associated with the very large dia

meter bolts could be par t ia l ly avoided by providing reduced diameter,

high strength bolts .



62

REFERENCES

1. Abrams, D. A., Tests of Bond Between Concrete and Steel, University of

Il l inois Engineering Experiment Station Bulletin No. 71, December, 1913,pp. 97-99.

2. Ferguson, P. M., Turpin, R. D., and Thompson, J . N., Minimum Bar Spacing

as a Function of Bond and Shear Strength,1I Journal of the AmericanConcrete Insti tute, Vol. 50, June 1954, pp. 869-888.

3. Ferguson, P. M., Breen, J . E., and Thompson, J . N., Comparative BondEfficiency of Large High-Strength Deformed Bars, report to the Committeeof Concrete Reinforcing Bar Producers, American Iron and Steel Ins t i tu te ,February 1962, pp. 66.

4. Texas Highway Department, Standard Specifications for Road and Bridge

Construction, January 2, 1962.

5. American Society for Testing Materials, Tentative Specification for Low

Carbon Steel Externally and Internally Threaded Standard Fasteners,ASTM Designation: A 307-58T, 1958 Standards, Part I p.,749.

6. Slaughter, E. M., Tests on Threaded Sections Show Exact StrengtheningEffect of Threads, Metal Progress, Vol. 23, No.3, March 1933, pp. 18-20.

7. American Standard ASA B18.2-l960 Square and Hexagon Bolts and Nuts,published by American Society of Mechanical Engineers, p. 17, p. 26.



6

PPENDIX



TABLE A (con t)

SUMM RY OF DATA

fs atSlip= Slip=

Slip a t

Specimen f ' f

kYoS1.

0.01 0.02 fs= 20 f = 33sksi ksi ksi in. in.

4.66 36.7 18 32

4.74 30.6 24 34

4.74 30.6

.011 .021

.006 .018

2N10a

2N1 b

2N10b-2

2W1 a

2W10a-2

2W1 b

2W10b-2

2N15a

2N15a-2

2W15a

2W15a-2

4.66 36.7 20 34 .010 .018

4.66 36.7

4.74 30.6 24 31 .006 .024

4.74 30.6

5.17 33.5 21 29

5.17 33.5

5.17 33.5 28 37

5.17 33.5

2-1/2N10a 4.63 30.7 19 34

.009 .027

.007 .016

.011 .019

65

Clear Cover

fsu Failure fcr ~ c piameterksi psi Ic '

42.5 D 2260 33.2

38.5 D 2045 29.7

49.0 C 2600 37.8

50.3 D 2670 39.3

48.3 C 2570 37.8

38.5 D 2045 29.856.8 S 3020 43.8

41.8 D 2220 43.2

56.2 C 2990 41.5

45.7 D 2430 33.6

64.0 C 3400 47.3

39.7 S 3510 51.5

1.5

1.1

~ ~ ; ' ~ f 1 t j F > ? 1



r

50

40

In 30

UlUlwan 2

1

001 002 003 004 005 006 007

STRAIN in.l in.)

FIG AI MILD STEEL ANCHOR BOLT STRESS STRAIN DIAGRAMS

008

\

\

•



f)

120

100

eo

6UJ

at- J )

4

20

.001

/

/

.002

/

/

/

/

.003

//

//

/

/

/

/

/

Q) 13/4

.004 .005

STRAIN in.l in.)

YIELD STRENGTH

AT 0.002 OFFSET

91 KSI

.006 .007 .008

0'>

J

anchor bolt development lenght

Documents