Vessel Design Ref

5.0

Support for vertical vessels

t

(A)

Tall Cylindrical Process Columns

\

t

• Supported on cylindrical or conical shells (ski s)

• The support skirts are directly welded to the v ssel bottoms head or shell

• The skirt base is stiffened by a continuous sti ening ring, which consists of top and bottom annulus plates with intermediate verti al stiffeners, to reduce localized bending stresses.

• They are designed as cantilever beams

(B)

Small and medium sized vertical v ssels

• Supported on legs or lugs (brackets)

• Provision of good access to the bottom dished end and any nozzles located there • Minimum thermal stresses arising from shell- upport temperature gradient

Minimum diameter = 6" Maximum HID =5" Maximum LID = 2" Number of legs:

N = 3 for D < 3' 6"

N = 4 (or more) for D > 3' 6" Maximum operating temperature = 6500p

,

,

'D ~--7

,

, , H

• Supported on uniformly spaced leg supports • Four legs are usually used

• Legs are normally fabricated from equal leg gle and T section shapes. They are welded to the cylindrical shell wall, often usi g a reinforcing pad.

• Some manufacturers prefer to use supports m de from pipe that is then welded to the dished end.

5.1 Leg Supports

I I I I

'

-

'

-

'

-'-r-

'

-

'

-

'

-'

I I I I I I I .-._.- .-r·-.-.- .- . I I I I I I I '-'-'-'-i'-'-'-'-' I I I

Two possible ways of welding the angled beam and I-beam to the vessel. The choice is between "easy to weld" and "offering more flexural rigidity". Besides cold-formed beams, sometimes a round pipe may be used as leg column, which has equal strength in all direction and has a high bending rigidity.

Loads on the vessel

1. The wind load (Pw) is horizontal and acts at t e centroid of the projected exposed surface

2. The earthquake load (Pe) acts horizontally on he center of gravity of the vessel

3. Piping or other equipment loads are not consi ered

Stress Analysis (to determine design di

ensions)

• Support-leg columns

• Base plate

• Leg-to-shell weld size • Leg-to-plate weld size

• Stresses in the vessel shell at supports • Size of anchor bolts

Support-Leg Columns In the case of 4-leg support

Over-turning moment (MD) at the base is about the diametral axis A-A Vertical reaction (due to dead load)

=

WIN

MD R

~

1

Vertical reaction (R) due to turning moment = Mg/D,

(Db = Base diameter)

In the case of 3-leg support

Db Db' 3 MD =R(-+-sm30)=-RDb 224 =>

J~(

~...::..b __

-7)1

R '"",,---'-

,

"

, " , 1200 .--. -~--/i I I I I

Maximum load on the leeward side (compreSSioj side) is:

Co

=

Wo +

4Mb (Operating condition)

N NDb

C =WT

o N (Test condition, no win loading)

Maximum axial load on the windward side (upli

TWo 4Mb

=--+--o N ND

b

(operating condi ion)

T =_We

+

4Mb

o N ND

b

(empty)

The eccentric loads P1 and P2 at the column top are:

11

= ~

+ ~

(operating condition)

11

= WT (test condition)

N

D2 -__

Wo +

4M a ( . dition)

1- operatmg con rtion

N ND

P2 = - We

+

4M a (empty)

N ND

Lateral load (F) per column

It is derived based on equal deflection at the top edge of the leg support .

=>

For 4-column leg support, IIi =2Ix

+

2Iy

•

.

.

.

...

·

·

··

•

·

··

•

·

Base plate

C

.

C

ompresslOn stress =

-ab

B d· 11(d/2)·(a/2) 11(d/2) en mg stress

=

-

.

=

2

ba

~/

12

ba

1

6

Compressive stress must be always greater than e bending stress.

d

b

1<

d

)1

1<

Weld Size

Shear stress = PI 1(2L1

+

h)

Bending moment on the weld joint

= C (d/2)

+

F(L/4)

Size of Anchor bolt

S

·A

b

= (

4Mb -

WJ

a NDb N

Anchor bolts are designed to resist the uplift force.

a

)1

h

LIr:::-lLI

_I

ge;~etry

I

If W > 4Mb, no uplift exists and the minimum bolt size %to 1 inch.

5.2

Bracket Supports (or Support Lug)

For vessels with small to medium diameters

«

1 ft.) and height-to-diameter ratio 2.5 When the bracket or leg support is attached to th cylindrical shell, a longitudinal moment arises.

In

each case the vessel wall is su ject to the extemallongitudinal moment of 'Fd' where 'F' is the maximum react" on at the support and 'd' is the distance from the centerline of the support to the shell outside surface.

The stresses in the shell induced by the

Bracket can be found by the local load method. Vessel

Wall ~

I

(d ₎

Forces and stresses on the bracket

: Top bar I ~---~I---~~

t

a

T

R h Gusset Vplate ---Base plate d F F/2 ( ) ( b ) a ( ) F . d =R . d sina 2 R= F 2sina

Maximum compressive stress (Sg) on the gusset (regarded as an eccentrically loaded plated)

I

I I

i

I

I

I

where the force eccentricity, e=(d - b)sina !

2

s

=

R

+

6Re

g (bsina).tg £t.tg

Force and stress on the Top Bar

The top bar is assumed to be a simply-supported earn with uniformly distributed load Fd/ha

Maximum moment occurs at the mid-span

M =(Fd).~=Fda max ha 8 8h d" 6M Ben mg stress = 2 fa .C (2":::;c:::;8ta)

6.0

Saddle Supported Cylindrical Vess Is

The code design of saddle-supported horizontal L.P. Zick (1951, The Welding Journal Research beam and ring analysis so that the mathematical results he had available.

ylindrical vessel follows the work of upplement) who used a modified

odel agrees with the experimental

Most recent work has indicated that Zick's appr ch gives reasonable agreement only when a flexible saddle support is employed. Wh n the saddle is rigid the simple Zick's analysis significantly underestimate the peak str s in the vessel by a factor as much as 50%.

6.1 General considerations

(a) Saddle supports should be located to cause minimum stress in shell and without additional reinforcement

(b) Most vessels are supported on two saddle supports. The saddles have an embracing angle between 120 and 150 degrees. Any relative settlement of the supports does not change the support reactions, therefore, the stresses in the shell remain quite the same.

(c) In the case of very long vessel that rested on more than two supports, the

support reactions are calculated based on continuous beam theory and increased by 20 to 50% as a safety factor for relative support settlements.

(d) The support reactions are highly concentrated and they induce highly localized stresses at the support regions. To reduce highly localized stresses, the saddle must be designed to provide flexibility at the support-shell junctions. Extended plates or wear plates may be used to provide a gradual transition of structural rigidity between the support and the vessel's shell.

(e) One of the saddles should be designed at the base to provide free horizontal movement, thereby avoiding restraint due to thermal expansion.

The Mathematical Model

4

H I ~ I I

.:

I

-.

I Rorr I I

+--

I

~-j

.-.-.-.-.-.-.-.-.-.-.-.-.-.-.~.-.-.-.-. _._._._._._.-.-._' I - I -I , I , , I

<

A )

<

A ) ~ L , <, _-' I WR2/4 2Hw/3 ~ ~/8 2Hw/3

)

W

R

2

/

4

3H/8 ---. ~ w Q _Q A _A

)1

_1<

<

) L

>

<

Shear force diagram

Q

(1) The support reaction is Q - the total weigh is 2Q

(2) The dished head is replaced by equivalent ylindrical segment of length 2HJ3.

The weight of dished end is therefore = 2 w/3, acting at a distance 3HJ8 from

end of parallel

(3) The total length is =L

+

4HJ3

(4) The uniform load has a intensity w = 2Q/( +4H/3)

(5) The hydrostatic pressure that acts normal t the dished end creates a couple given by WR2/4

Points to note:

6.2 Longitudinal bending stresses

in

the she I (a) The bending moment at the mid-span '

2 (3H L) wR2 wL L L M} =--Hw

-+-

+---(-)+Q(--A) 3 82 4 24 2 =QL

[1

+

2(R2 - H2)/ L2 _ 4A] 4 1+(4HI3L) L

=K{~L)

Bending stress at the mid-span:

The above expression assumes that the full vessel section is available in resisting bending stresses and the cross section remains circular. For very thin vessels it is found that the cross section does not remain circular especially so during filling with liquid. Nevertheless, the expression gives satisfactory design dimensions for vessels with D/t ratio up to 1250.

(b) The bending moment at the saddle-support

2 (3H ) wR2 wA2

M2

=:":

S+A

The top portion ofshell above the saddle support would feform under load and is deemed ineffective in resisting longitudinal moment. So the moment of inerti atthis cross section is reduced to that of a ring with its top portion removed.

The effective arc is assumed to be: 2i1

=

2(8/2

+ ~

1 )

I

yJ:---<J

0o::

t,

NA

...

...- ~+

~C_~ Effective

portion

The position of the neutral axis, N.A. and the second moment of area I about this axis can be found.

_ rsini1 y= ~ (sini1

J

Cl

=

r ~ - cos A , [ . 2

A]

3. sm D. I =r t i1

+

smi1cos i1 - 2 ~

Longitudinal bending stresses atthe highest and lowest point of the effective cross section are:

(Highest point - tension)

. M)

SI =--- ·C)

I - (Lowest point - compression)

Allowable stress limits

The tensile stress combined with the pressure stress (pr/2t) should not exceed the allowable tensile stress for the shell material.

6.3

Shear stresses in the plane of the saddle

[

The distribution and magnitude of the shear stresses in the shell in the plane of the saddle depend a great deal on how the shell is reinforced.

! !

The inner shear force, V=Q-w(A+ 2H)= ~(L-2A) 3

f

+

4H 13

i

The outer shear force,

v

=

w(A

+

2H)

=

2Q( i

+

2H 13)

3 L 4H /3

Note: the inner shear force is greater then outer shear force when

Q(L - 2A) 2Q(A

+

2H 13)

=-:'---~ < or

L+4H13 L+4H13 L> 4A +4H13

(A) Shell is stiffened by a ring at support region

If the shell is made rigid, the whole section is effective in resisting load-induced shear stresses.

The shear flow (shear force per unit arc length) is: (~ is measured from the top of the cross section)

v .

d,

qo =-SIlly

nr

The maximum stress flow is when ~ =90 degree.

=> Max. shear stress = qo =

Q [

L - 2A ]

t nrt L

+

4H 13

(B)

Shell not stiffened by ring

When the shell is free to deform above the saddle, it is considered that the shear stress acts on a reduced cross section. The upper portion of the shell is considered

19 ( ())

2~

=

2(e/2

+

jJ /20)

= -

J(

--20 2

i

I

As a result, the shears in the effective portion will be increased by a factor:

Factor C = qo(unstiffened shell) =

r

sin2 ¢d¢ = J(

, qo(stiffenedshell) fsin2¢d¢ J(-a+sinacosa

(Vsin~) The shear flow, qo =C nr

"qo (Vsin¢)

The shear stress, S2 = -

=

c

---t srt

The maximum shear is now at the tip of the saddle, i.e. ~=a

(C)

Shell stiffened by heads (A <Rl2)

If the saddle is close to the end closure the shell is stiffened on the side of the head. It is assumed that the shell above the horn (tip of saddle) is stiffened by the end closure. The shear distribution in this upper region is therefore similar to that for a stiffened region.

For the lower portion - in the saddle region, (a::;!~::; n) ,the shear distribution can be

I

found by summing the shears to one side of the saddle. The sum of vertical shear force for the upper portion is equal to the sum of verticrl shear force in the lower portion.

I

I

---J

---I

I

Shear force near the enr closure

---1 ---I I I , I I

~r--r-

---~--_, ,

The vertical shear force,

V

=

21iL(sin~)t(sin~)rd~

=

Q

(a - sinacosa)

otirt n

The shear flow is assumed to be the same as that for the unstiffened shell, that is:

The shear flow, qo

=

c(Vsin~)

= ~.

Q (a - sin o.cosc.Isin

e

nr nr n

qo Q a - smacosa

The shear stress, S = - = - . . .sin ~

t tcrt n - a

+

smacosa

The maximum shear occurs at ~=a

Allowable stress limits:

I

6.4

Ring compression in the shell over the saddle

Assuming that the surface of the shell and saddle are in frictionless contact without attachment. Ring compression is caused by shear forces.

The ring compression in the region a:s; ¢ :s;J[

The shear flow is:

v .

d

.

(

1[ )

q3

=

-smlf' .

Jrr J[ - a

+

smacosa

The total shear force at any point on the shell arc above that point.

¢ Q(cos¢ - cosa)

Total shear force = fq3rd¢

=

.

J[ -

a +

smacosa a

The contact pressure between the saddle and shell would induce a tangential compression force similar to the above. That is:

Tangential compression force due to saddle force = Q( cos¢ - cos fJ) J[ - fJ

+

sin

j3

cos

j3

Therefore, the maximum tangential force is= _ Q(1

+

cos fJ) J[ - fJ

+

sin f3 cos f3 The contact pressure can be deduced from the tangential force:

1 Q(cos d. - cos

/3)

Contact pressure = _. --='---'----'If'__ ~-'----r J[ - fJ

+

sin fJ cos fJ

The maximum contact pressure occurs at

¢

=

7r

Max. contact force (at ¢ =7r )

1 Q(l

+

cosb)

- - - . --

The width of shell that resists this force was considered by Zick to be '5t' on each side plus the width of the saddle, i.e. width =b

+ -

.

In

a follow-up paper, he suggested width =

b +

1.

1

6-Jrl

I

The tangential stress can be calculated, S5=shea~force / width

The stress S5 is important when concrete saddle

J

used. It should be checked for large

diameter vessel.

I

Recent experimental and theoretical work on sad les welded to the vessel have found that this tangential stress is very small, about 111 of that predicted by Zick's

approach. However, for the saddle not welded to he shell, the Zick's approach gives the correct order of stresses.

The ring compression may be reduced by attaching a wea plate somewhat larger than the saddle surface area directly over the saddle.

Allowable stress limits

The compressive stress S5 should not exceed 1/2 of S,and is not additive to the pressure stress. If wear plate is used, the combined thickness of wear plate and the shell can be used to calculate S5, provided the wear plate extends r/I 0 inch beyond the horn.

Despite the limitations of Zick's approach it does provide a workable design method that has been used extensively over many years. However, the very high

circumferential stresses known to exist at the saddle horn region when the vessel is supported on a rigid saddle at not predicted adequately by the analysis. Although these peak stresses do exist, they are very local to the saddle horn and are unlikely to cause plastic collapse of the support. However, their existence does cause concern when the vessel is subject to high cyclic stressing.

Local stresses in shell due to loads on attachment

I

Types of attachments: Nozzles, supporting lugs,ilifting brackets, etc.

!

Main concerns - High concentrated stresses at tHe attachment due to combined

internal pressure and external loads applied through the attachment can be a source of failure if proper reinforcement is not supplied.

!

i

Design consideration:

• Opening in vessel shell must be reinforced or operating pressure

• Reinforcement is usually a rectangular or s uare pad welded to the shell • Over-reinforcement may create 'hard spot' on the vessel and induce large

secondary stresses

• Reinforcement material should be close to he opening for effectiveness, of which 2/3 of the required material should b. within a distance d/4 from the opening, where d is the diameter of the opening.

• Sharp junctions should be avoided; fillets should be incorporated to reduce the magnitudes of stresses at the junctions.

"The best arrangement is the so-called balanced reinforcement, which consists of about 35-40% of the area on the inside and about 60-65% on the outside. On many designs, however, it is difficult to place reinforcement on the inside. Balanced reinforcement is often used at manway and inspection opening where no nozzle is attached"

Area Replacement for Nozzles

This method formed the basic design method in many design codes. The origin of the area replacement idea is not entirely clear. Simply expressed one replaces the area cut away by the cross section of the hole in the shell and relocates it around the hole close to the cutout. Notice it is an area replacement rather than a volume replacement.

The disposition of the replaced area is important. To be effective it needs to be close to the edge of the opening where the stress field is increased. The extent of the

reinforcement is preferably equal the die-out distance of the peak stresses at the edge of the opening. That is why in some codes the extent of reinforcement is expressed as a function of

Jrl,

the characteristic length parameter for the die-out distance of the discontinuity stress. In any case one simply obeys the rules as stipulated and no

explanation is given. It should be noted that the distance for reinforcement is generally quite shout.

Cylindrical vessel with local loads on a rectangular attachment

Assumptions:

• Attachments are rectangular or square w th two edges parallel to the circular profile

• The radial force produces uniform press re over the attachment area • The moment loading produces a triangul r pressure distribution External loads

(a) Radialload, P

(b) Longitudinal moment, ML

(c) Tangential (or circumferential) moment, (d) Torque, T

(e) Shears VL, Vt

The shear stress in shell due to the torque Tis:

-r---

ri-,

T

- 2Jrrot - 2Jrr;t

Maximum shear due to VL or Vt is:

v

V

r'=_t_ or -r,=_L_

Jrrot Jrrj

The shear stresses rand t' are usually small enough to be disregarded.

Parameters for cylindrical shell:

Shell parameter: y =R I t or R I(t

+

tp)

Square attachment; ~=cl R where 'c' is the half-length of the loaded area

Cylindrical attachment: ~=O.875ro /R

Rectangular attachment: it can be converted into equivalent square loaded area. For small side ratios with a I b ::;1.5 , the equivalent c =-Jab /2

General expression for stress in the shell

Circumferential (tangential) stress:

i

For different loadings, the circumferential and tlhe longitudinal stresses are expressed in different parametric forms as follows:

I

(1) Radial load, P

0.

=

(

;'XcP~;)y

+

6;

.

]

i

=Cp(Plt2) (outward force) =C~(PIt2) (inward force)

u,

{Nep

6Mep}

O"ep

=

t2 R~ (Mt IR2~)'Y

+

(M{ IR~) =

C

t

(M

t 1(2R2B )

(2) Circumferential (tangential) moment, M,

(3) Longitudinal moment, ML &

Design considerations

(a) lfthe maximum stress at the attachment is too high, the shell must be reinforced by a reinforcing pad or the thickness of the reinforcing pad required for internal pressure must be increased. The width of the pad is such that stresses at the edges of the pad are below the allowable stress.

(b) If two local loads are too close to each another, i.e. within the stress die-out distance, then their influence on each other must be considered.

Note: The analysis presented above for local loads applied on cylindrical shell is too simplistic. More detailed and accurate analyses for different types of attachments are available in the literature and recommended by design codes, specifically for loads on the nozzle, and openings. For example, the Welding Research Council Bulletins 107 & 297 (WRC 107 & 297).

Design by analysis

Essentially Design by Analysis is based on the dea that if a proper stress analysis can be conducted then a better, less conservative, a sessment of the design can be made compared to the usual approach of Design by le. The philosophy was originated in the 1960's in the US. The motivation was drive by the sophisticated design work in the nuclear industry. There were many design £ atures that were not covered directly by the existing Design by Rule methods.

In the early years, all design by analysis ideas ere developed based on thin shell analysis and in particular the analysis of discon inuity effects including thermal discontinuities.

It was suggested that different types of stress h d different degrees of importance and this led to the idea of categorization of stress. T e stresses are cast in the form of

'stress intensities' to reflect the Tresca yield cri eria and then compared with specified stress limits that are set at different levels for the different stress categories. This methodology was first incorporated in ASME PV code Section III and Section VIII Division 2 in 1968 and later into BS 5500 as Appendix A. Many countries have now adopted the same basic approach.

Multiaxial Stress States

In real world, all stresses are three-dimensional. It is the simplifying assumptions that reduce the 3-D stresses into 2-D and I-D. Yielding in the presence of multiaxial stress states is not governed by the individual component but by some combination of all the stress components. The two commonly used yield criteria are the Von-Mises criterion and the Tresca criterion.

Von Mises criterion (distortion energy theory) states that yielding will take place when;

Tresca criterion (maximum shear stress theory);

Although it is generally agreed that the Mises criterion is better for common pressure steel, ASME chose to use the Tresca criterion as a framework for the Design by

Analysis procedure. The reason is that Tresca is the more conservative and it is easier to apply. The later is longer true now since computer can perform complex

calculations at ease.

In order to avoid the unfamiliar (and unnecessary) operation of dividing both

calculated and yield stress by two, a new term called 'stress intensity' was defined.

The stress differences of the principal stresses are as follows:

The

STRESS INTENSITY,

S is the maximum absolute value of the stress difference.

That is:

So the Tresca criterion reduces to: S=(J' y

Throughout Design by Analysis procedure stress intensities are to be used.

Stress Categories

Certain types of stresses are more important than others and that these should be assigned to different categories with different levels of importance having different

stress limits. ASME chooses the following categories: (A) Primary Stress

(i) General Primary Membrane Stress, Pm (ii) Local Primary Membrane Stress, PL

(iii) Primary Bending Stress, P,

(B) Secondary Stress,

Q

(C) Peak Stress, F

Primary stress is a stress developed by the imposed loading that is necessary to satisfy the law of equilibrium between external and internal forces and moments. The basic characteristic of a primary stress is that it is not self-limiting.

Secondary stress is stress developed by the self-constraint of a structure. It must satisfy an imposed strain pattern rather than being in equilibrium with an external load. The secondary stress is self-limiting, its may cause local yielding and minor distortion resulting from discontinuity condition or thermal expansion.

Peak stress is the highest stress in the region under consideration. The basic characteristic of a peak stress is that it causes no significant distortion and is objectionable mostly as a possible source of fatigue failure.

Failure modes

1. Excessive elastic deformation incl ding elastic instability 2. Excessive plastic deformation

3. Brittle fracture

4. Stress rupture and creep deformati n 5. Plastic instability - incremental co lapse 6. High strain - low cycle fatigue

7. Stress corrosion 8. Corrosion fatigue

In setting the stress limits, however, attention is concentrated in 3 areas. They are: (a) Avoidance of gross distortion or bursting, Pill' PL and

P,

(b) Avoidance of ratcheting, PL

+ P,

(c) Avoidance of fatigue, P+

Q

Relationship between stress limits to the various categories

Stress Intensity Allowable Stress Equivalent Yield

General primary membrane, Pm Sm

2S

3 y

Local primary membrane, PL 1.5 Sm Sy

Primary membrane +bending, 1.5 Sm Sy

(Pm+Pb) or (PL +Pb)

Primary +secondary 3Sm 2Sy

(PL+P,+Q) or (Pm+P,+Q)

Fatigue, 2Sa

-(PL +Pb +Q+F) or (Pm+Pb +Q+F) (allowable fatigue stress range)

The above limits are not always applicable; there are a number of special cases. In the case of nuclear vessels the service loadings are classified into normal, upset,

emergency and faulted conditions. This is formalized in ASME with k-factors applied to the limits. For example, for earthquake loading, k = 1.2, for hydraulic test k =1.25, etc.

For attachments and supports the limits are:

The membrane stress intensity :S1.2 Sm(0.8 Sy) Membrane + bending stress intensity

s

Sm(1.33 Sy)

I

For nozzles and openings:

Membrane + bending stress intensity

s

i .25 Sm(1.5 Sy)

Some cautionary words are necessary for the u wary. The manner in which the symbolism is used can lead to confusion. For e ample a stress limit on some combination of stress categories denoted as CPL+P, +Q) needs to be clearly

understood. It is the stress intensity evaluated om the principal stresses after the stresses for each category have been added tog ther in the appropriate way. It should not be interpreted as the combination of stress ~ntensity from each category.

I

In summary: ONLY add stresses, DO NOT add stress intensities.

A trivial example of the wrong way of summing the stresses in given below:

Stresses Pm Q Pm+Q

Sx= S] 10 25 35

Sy= S2 10 -5 5

Sz= S3 -2 0 -2

The maximum stress intensity for Pm= 12

The maximum stress intensity for

Q

=30

The maximum stress intensity for (Pm+ Q) =37 (from [Pm+ Q] column)

It is wrong to add stress intensities of Pmand Q, that would give (12 + 30) =42

When we add stresses, of course, they need to be in the same directions and at the appropriate locations for the identified combination of loads. The approach is to evaluate all the stresses for the different types of loading. These should be assigned to categories as necessary. Then the stresses in the various categories should be summed and finally the stress intensities calculated for the particular combination of categories required.

FE Analysis for Pressure Vessel Desig

The Design by Analysis is closely rooted in thi shell discontinuity analyses. When FE (finite element) method is used, some diffi ulties in stress categorization occur. The FE gives accurate stress information for c mplex geometries. These stresses may vary nonlinearly through the thickness. For ass ssment purposes it is necessary to linearize the stress distribution and separate membrane and bending effects. In a simple case the procedure would be straightforward and membrane, bending and peak elements of the stress could be identifies. Unfo unately things are not always so simple. Firstly the Linearization procedure is it elf subject to a number of

uncertainties. Secondly the bending componen in general may include primary bending as well as secondary bending.

In practice it tends to assume the membrane str ss intensity as primary and the bending stress intensity as secondary (which m y not be conservative). In critical situation the designer may wish to impose his dwn conservatism at this point.

Until today no entirely satisfactory solution has been found for the stress linearization. However, alterative methods may be forthcoming that would by-pass the

categorization problem or at least simplify its interpretation. The Standards allow the design to be based on limit load analysis with a suitable factor where the factor has to be the same as the main shell (i.e. 1.5). Design may proceed directly with a factor on load without detailed consideration of the stresses. The approach seems promising if it can be extended to complex loading situations it could provide a relatively simple alternative to the current classification route.

ASME identifies 8 modes of failure "which confront the pressure vessel designer." The evaluation of failure modes requires the computation of membrane and bending stresses and their classification into certain categories primary, secondary and peak -to which different design allowable stresses applied.

The original techniques for evaluating the stress limits were based on shell theory by which membrane and bending stresses are determined directly - so there is no much confusion in the classification of stresses.

With the advent of finite element (FE) techniques, the transition from the stress distribution to the failure mode requires a different path.

The results of axisymmetric or 3-D solid FE analysis are not immediately in a form suitable for the extraction of shell type membrane and bending stresses. Difficulties are associated with linearization procedure used to obtain membrane and bending stresses.

Unless we are dealing with well established ca es, as listed or referenced in codes, there has always been a problem with the cate orization of stresses into primary and secondary.

The problems of assessing primary and second ry stress failure modes and their relationship to stress results from axisymmetri and 3D geometries were first addresses by Hechmer and Hollinger in 1986.

"3D stress criteria - aweak link invessel design and a lysis", PVP Vo1.109, A Symposium on ASME Codes and Recent Advances in PVP and Valve Technology including aSurvey of Operational Research Methods inEngineering, July 19 6, ASME, New York, NY.

Three approaches for determining the membra e and bending stresses were discussed: (i) stress-at-a-point

(ii) stress-along-a-line, and (iii) stress-on-a-plane.

A quantitative comparison of the three approaches was presented in:

3D stress criteria - application of Code rules, "PVP Vo1.120, Design and Analysis of Piping, Pressure Vessels, and Components, July 1987, ASME, New York, NY.

The study shows that the 3 approaches can give substantially different results. The most complex of the three approaches is stress-on-a-plane. The definition of the plane for 3D geometries is subjective and the resultant stresses and conclusions are merely engineering judgement.

Some issues:

It should be emphasized that these issues actually arise from the nature of the Code rules, rather from any deficiency in the finite element solution.

In 2D axisymmetric analyses, the bending stress can be calculated using component normal and/or shear stresses or principal stresses. The distinction is that for a given set of geometric reference axes, component stress directions remain constant with location and load application whereas principal stress directions vary with location and load application. This distinction is important when considering the pros and cons of using component versus principal stresses.

Which stresses are consistent with bending theory? Code implies that bending is applicable only to normal stress components, because the Code links bending to

For 3D geometries, the issue is evaluation of s

I

esses along lines versus on planes. The code implies evaluation along a line. H01ever, the code does not preclude the use of planes.

Two PVRC grants were established to investigate and document the need to update the ASME B&PV and Piping Code criteria and re~uirements for relating 3D stress

distributions to failure criteria. The findings ar1presented in the following paper.

J.,L. Hechmer and G.L. Hollinger, 3D Stress Criteria, JvP-Vol. 210-2, Codes and Standards and Applications for Design and Analysis of Pressure vess~l and Piping Components, ASME 1991.

Recommendations

I

• The stresses for Pmcan and should be calculated by simple equilibrium equations. The same is true for Pb if Pmis small (for example, the plate structures). Stresses for Pmneed only be evaluated in basic structural elements. Designer should apply his ingenuity to calculate equilibrium stresses, not to extract stresses from a general FE model.

• It is appropriate to calculate PL stresses in the vicinity of all discontinuities. There are discontinuities where PL stress exists, but need not be evaluated. Because code rule reinforcing rules ensure that PL limit is met.

o Nozzle-shell junctions o Formed heads to shells

o Cones to shells

o Tapered cylinders to shells

• Linearization algorithm calculates the net force distribution on the cross section. The average net force can be calculated from the total net force. The average net force is then subtracted from the net force distribution that is used to calculate the bending moment. The bending moment is computed relative to the neutral axis. • Calculate (PL

+

Pb) and (P

+

Q) in the basic structural elements (and not in the

transition elements). The reason is that plastic collapse and gross strain

concentration will not occur in the stiff transition elements. They will occur in the more flexible shell element.

If a fatigue analysis is to be performed within a transition elements due to high stress concentration, it may be appropriate to consider the (P

+

Q) in the adjacent structural elements.

• For assessing the membrane stress limits (Pm

+

Pb), all the stress components (3 normal

+

3 shear components) should be included. The average principal stresses must be computed from the average stress components through the thickness and

44 Page 2 of24

Conventional bolts are usually made to the specific project requirements by steel fabricators or they may be purchased in standard sizes (diameters and lengths) from steel suppliers. The availability and cost of conventional bolts are generally based on demand and fabrication requirements. The types of conventional anchor bolts most often used are discussed below.

Headed Bolts. Square or hex-headed ASTM A 307 bolts are frequently used as anchor bolts due to their

wide availability and relatively low cost (see Figure 1). Higher strength bolts, such as ASTM A 325 bolts,are available and can be used, but are more expensive. A washer placed against the bolt head is often used with the intention of increasing the bearing area and thus increasing the anchor strength. However, the actual strength increase obtained by adding a washer is small, if any, and under certain conditions (small edge distances), may actually decrease the tensile strength.

(::])

A) HEX-HEAD

Bent Bar Anchors. Bent bar anchors, frequently used in masonry construction, are usually made in "J" or "L" shapes (see Fig. 2). Even though the "J" and "L" shapes are the more popular, a variety of shapes (see Fig. 3) is available since there currently is no standard governing the geometric properties of bent bar anchors. These anchors are usually made from ASTM A 36 bar stock and are shop-threaded.

Headed Bolts

44 Page 3 of24

T

o v ~ .,-_,,/_~ 1h TO 1% D ~

~ ,_~

~IaI~

0

~

_I

T

A) Bl" BOLT B) "J" BOLT

"L" and "J" Bent Bar Anchors

FIG.2

A) EYE BOLT

S) "U" BOLT

44 Page 40f24

Other Bent Bar Anchors FIG.3

Plate Anchors. Plate anchors are usually made by welding a square of circular steel plate perpendicular to the axis of a steel bar that is threaded on the opposite end (see Fig.4).There are no standards governing the dimensions (length, width or diameter) of the plate.The American Institute of Steel Construction does limit the fillet weld size based on the plate thickness (see Table 1).Both the plate and bar are usually made from ASTM A 36 steel.

~~\

~ ~) ~'-- -"'Im A) CIRCUlAR PLATE ANCHOR B) SQUARE PLATEANCHOR Plate Anchors FIG. 4

Through Bolts. As the name implies, through bolts extend completely through the thickness of the masonry and are composed of a threaded rod or bar with a bearing plate located on the surface opposite the attachment (see Fig.5).In the early 1900's, through bolts were usedin loadbearing masonry structures to tie floor andwall systems together. Often decorative cast bearing plates were used since through bolts were visible on the exterior masonry surfaces (see Fig.6). Today, through bolts are primarily used in industrial construction where aesthetics are not a principal concern, or in retrofitting existing structures.

Through bolt rods are usually made from ASTM A 307 threaded rod orthreaded ASTM A 36 bar stock.

44

Through Bolt

FIG. 5

Decorative Through Bolt Bearing Plate

FIG.6

44

Proprietary Anchor Bolts

,

---_._--_.- ._--- --- .._--- - - .- -

-*American Institute ofSteel Construction

Page 60f24

Proprietary anchors are available through a number of manufacturers under numerous brand names.

Although the style and physical appearance of the anchors differ between manufacturers, the basic theories behind the anchors are very similar. For this reason, proprietary anchors can be divided into two generic categories: expansion-type anchors and adhesive or chemical-type anchors.

Expansion Anchors. Two different types of expansion anchors are generally recommended by their manufacturers for use in brick masonry: the wedge anchor and the sleeve anchor (see Fig 7). These anchors develop their strength by means of expansion into the base material. Wedge anchors develop their hold by means of a wedge or wedges thatare forced into the base material when the bolt is tightened. The

wedges create large point bearing stresses within the hole;therefore, this anchor requires a solid base material to develop its full capacity. For this reason, voids formed by brick cores and partially filled mortar joints in some brick masonry may make the construction unsuitable for wedge anchor installation.

44

FIG. 7

Page 7 of24

Sleeve anchors develop their strength by the expansion of a cylindrical metal sleeve or shield into the base material as the bolt is tightened. The expansion of the sleeve along the length of the anchor provides a larger bearing surface than the wedge anchor, and is less affected by irregularities and voids in the base material than is the wedge anchor. For this reason, sleeve tnChors are recommended by their

manufacturers for use in brick masonry more often than we ge anchors.

Drop-in and self-drilling anchors (see Fig. 8) are two other t. pes of expansion anchors available, but are typically not recommended by their manufacturers for use ii masonry. The reason for this is due to the

embedment and setting characteristics of the two anchors. Both anchors are produced to allow shallow embedment depths and are expanded or set by an impact setting tool. The combination of shallow embedment and high stresses imparted by the expansion tend to cause cracking or splitting in masonry.

Depending on the extent of cracking or splitting, the anchor could experience a reduction in load-carrying capacity or undergo complete failure during installation.

A)SELFORILllNG ANCHOR B) DROP4N ANCHOR.

Other Proprietary Expansion Anchors FIG. 8

There are several considerations that should be examined when contemplating the use of expansion-type anchors in brick masonry. These are: 1) Expansion anchors should not be used to resist vibratory loads. Vibratory loads tend to loosen expansion anchors. 2) Specific torques are required to set expansion anchors. Excessive torque can reduce anchor strength or may lead to failure as excessive torque is applied. 3) Expansion anchors require solid, hard embedment material to develop their maximum

capacities. Some brick construction may not provide a good embedment material due to voids formed by brick cores and partially filled mortarjoints.

Adhesive Anchors. Two basic types of adhesive anchors are currently available. The major difference between the two is that one anchor is manufactured as a pre-mixed, self-contained system, whereas the second type requires measurement and mixing of the epoxy materials at the time of installation. The more popular self-contained types use a double glass vial system (see Fig 9) tocontain the epoxy. The outer vial contains a resin and the inner vial contains a hardener and aggregate The glass vial is placed in a

pre-44 Page 13 of24

In hollow brick construction, the units are laid so that the cells are aligned and provide continuous channels for reinforcing steel placement and for grouting. Depending on the design, every cell or intermittent cells may be reinforced and grouted (see Technical Notes 41 Revised). The anchor embedment detail will depend on the reinforcing pattern used in the construction. Figure 15 shows typical embedment details for conventional anchors embedded between reinforcing cells. The anchor should be solidly surrounded vertically and horizontally by grout for a minimum distance of twice the embedment depth (1b) (Figs. 14 and 15) for full tension cone development. The tension cone theory is discussed in following sections. This may require that some cells be partially grouted. A wire mesh screen can be placed in the bed joint across cells that are to be partially grouted to restrict the grout flow beyond a certain point. Figure 16 shows typical embedment details for conventional anchors embedded in reinforced cells. In this detail, the anchor may be tied with wire to the reinforcing to secure the anchor during the grouting process Again, the anchor should be solidly surrounded by grout to a minimum distance of tWife the actual anchor embedment depth, both

vertically and horizontally. I

i f

P;; "L" socr

s,

",I"SOLT CI HEA.OED;BOLT

IF

;

:

:

::~

~·*

I;>

O)PlA15 ANCHOR

DMIN.

Conventional Anchors in Reinforced Hollow Brick FIG.15

44 Page 14of24

0) P~JE P.fl-cHOR

Conventional Anchors in Partially Grouted Hollow Brick

FIG.16

Two typical embedment details for conventionally embedded anchor bolts installed in composite brick and concrete block construction are shown in Fig. 17. As shown, anchor bolts may be placed in the collar joint between the brick and block wythes or placed into cells in the concrete block wythe and grouted into place. In details similar to Fig. 17(a),the anchor bolt type and diameter may be controlled by the width of the collar joint. Collar joints should bea minimum of 1in. (25 mm) wide when fine grout isused, or a minimum of2 in.

(50 mm) wide when coarse grout is used (see Technical Notes 7A Revised). When thecollar joint dimension is in the 1 in.(25 mm) range, it may become difficult to position anchor bolts in the collar joint and maintain the recommended clear distance between the masonry and the anchor (Fig. 17).The practice of using soaps to accommodate anchors larger than the collar joint is not recommended because the reduction in the brick masonry thickness around the anchor could lead to strength reductions. If the anchor

dimensions required are larger than the collar joint, a detail similar to that shown inFig 17(b) should be considered.

44

B) ANCHOR IN BLOCK WYTHE

Page 15 of24

GROUT

STOP

Conventional Anchors in Composite Brick/Block Masonry FIG.17

Through bolts are typically installed after construction and grouting by drilling through the completed masonry work. When through bolts are to be installed after construction in reinforced brick masonry, care should be taken during installation to avoid cutting or damaging reinforcement while drilling the through bolt holes. Reinforcing bar locations can be identified by specially tooled joints or other marks made during construction.

Proprietary Anchors

Proprietary expansion and adhesive anchors typically require special installation procedures and equipment. The manufacturer should be contacted to determine the appropriate anchor for a particular application, the correct installation procedure and if any special installation equipment is required. Improper application and installation of proprietary anchors may lead to less than satisfactory structural performance.

Typical proprietary anchor details are shown in Fig.18.Itis suggested that proprietary anchors be embedded in head joints when facing or building brick are used. This reduces the possibility of placing anchors in brick cores that occur within the thickness of the brick and adjacent to the bed joint surfaces.

Anchors setin grouted hollow brick should be placed in holes drilled in the bed joints so that they intersect grouted cells, or should be placed in holes drilled through the faces of the units into the grouted cells. As with conventional anchors, proprietary anchors should be solidly surrounded vertically and horizontally by grout for a minimum distance of twice their embedment depth.

44

]

n

"/ 11

r

u

'~It-::ll>

J

f'(

l

u

'V IJ A)GROUTED OOLL~R ,JOINT CONSTRUCnON 'b

J

~

!

~

l

.,

[

II

~;;J

~

II

-.'v

II

Typical Proprietary Anchor Details

FIG.18

ANCHOR BOLT DESIGN

Page 16 of24

Anchor bolts are used as a means of tying structural elements together in construction and therefore, provide continuity in the overall structure. In virtually all applications, anchor bolts are required to resist a combination of tension and shear loads acting simultaneously due to combinations of imposed dead loads, live loads, wind loads, seismic loads,thermal loads and impact loads. For this reason, and also to insure safety, anchor bolt details should receive the same design considerations as would any other structural connection. However, due to a lack of available research and design guides, anchor bolt designs are based largely on past experience with very little engineering backup. This situation may lead to conservative,

uneconomical designs at one extreme, or nonconservative designs at the other.

Recently, however, research investigating thestrength of conventional and proprietary anchors in masonry has been completed. Reports have been issued thatevaluate anchor performance and suggest equations to predict ultimate anchor strengths. Bycombining the research findings with design practices currently

used in concrete design, equations for allowable tension, shear and combined tension/shear loads for plate anchors, headed bolts and bent bar anchors are under consideration for adoption in the proposed "Building Code Requirements for Masonry Structures" (ACIIASCE 530). These equations are outlined below.

Tension

The tensile capacity of an anchor is governed either by the strength of the masonry or by the strength of the anchor material. For example, if the embedded depth of an anchor is small relative to its diameter, a tension cone failure of the masonry is likely to occur. However, if the embedded depth of the anchor is large relative

44 a

~~st°MIN.

OJ THROUGH BOLT --V-:;"Cr~.r-r-r;!;~:},-",...r-r-~"TT"" D MIN. B) .~••BOLT o ~~.~~~ ....~.~ C) HEADEOBOLT

Conventional Anchors in Grouted Collar Joints FIG.12

Page 11of24

Typical embedment details of conventional anchors in multi-wythe brick construction are shown in Fig. 13. A brick,or portion of a brick, is left out of the inner wythe to form a cell for the embedded anchor (Fig. 14). After the anchor is placed, the cell is filled with mortar or grout prior to placement of the next course.

44

---

-Page 10 of24

aThe manufacturer should always beconsulted when adhesive anchors areto beused inareas where contact with chemicals islikely.

~

'''

]

W .J ~ V) Z W ~ 50

"~rl '1"""1t4_._-r--r-.-rl 'IT,'1-'--'---'-rf "--.",,,,, ,..r"Tf""f·T'"f~r"1'I,I_._i " T,'I"-'-"--'--'-'-TI "'"""!

100 150 200 25G

50

TEMPERATURE, 'F

Effect of Temperature on Ultimate Tensile Capacity

FIG.11

INSTALLATION DETAILS

Conventional Anchor Bolts

Typical embedment details for each type of conventional anchor used in grouted collar joint construction are shown in Fig_ 12_The conventional embedded anchors (headed bolts, bent bar and plate anchors) are usually placed at the intersection of a head joint and bed joint. By using this location, the brick units adjacent to the anchor can be chipped or cut to accept the anchor without altering the joint thickness.

44 Page 12 of24 Q .1.. f" D) PLATE ANCHOR B) '~" BOLI E)

1HROUGH

sOLT

··

Conventional Anchors in Multi-Wythe Brick Masonry FIG.13

44

MIX

Page 9 of24

There are special requirements and limitations. that should be considered when contemplating the use of adhesive anchors in brick masonry. They are: 1) Specially designed mixing and/or setting equipment may be required 2) Dust and debris must be removed from the pre-drilled holes to insure proper bond between the adhesive and base material. 3) The adhesive mixture tends to fill small voids and irregularities in the base material. 4) Large voids (due to brick cores, intentional air spaces and partially filled joints) may cause reductions in anchor capacities. This is especially true with the self-contained adhesive anchors since a limited volume of epoxy is available to fill the voids and provide a bond to the anchor. 5) The adhesive bond strength is reduced at elevated temperatures and may also be adversely affected by some chemicals (see Table 2 and Fig. 11).

H ARD!-ENER RESIN

PLACE

Site-Mixed Adhesive Anchor FIG. '10

44 Page 8 of24

and mixing the adhesive components. The other type of adhesive anchor requires that the epoxy components be hand-measured and mixed before the epoxy is placed into a pre-drilled hole. A threaded rod or bar is then set into the epoxy mixture, as shown in Fig. 10. Adhesive epoxies usually vary slightly between manufacturers, but the steel rods or bars are typically ASTM A 307 or ASTM A 325 threaded rod, or ASTM A 36 shop-threaded bar.

A) EPOXY CAPSULE

a) THREADED ROD

C)INSTALLED ANCHOR

Self-Contained Adhesive Anchor FIG.9

44

TABLE 4

Allowable Shear on Anchor Bolts -l-rcrn use

1985 Edition"

Page 22 of24

(a) ,A.LLOV\,I,ABLE SHE.A,R C)~\J.A.f",JCHOF.'BOLTS1 FOR CLAY .A.f",JD

COf\JCF.'ETE t"llASOf\JRY

Tot<lI Allowable

Diameter Elllbedmenf Shear3

(inches) (inches) fibs}

1/4 4 270 :3/8 ₄ 41D 1(2 4 550 5/B 4, ~~I

-

3/4 ~Ir- 1100 7/8 f3 1500 "7 18504 I 1-1fa 0 22504 u

'P',n anchor bo~ is eIbolt that h;",,,eIright elngle extension of elt leelst three dierneters.

,f!.,standard machine bo~ i:::oacceptable.

"Of the total required embedment, a minimum of five bolt diameters must be perpendicular to the masonrv surface.

1--10reduction in value" required for uninspected mesonrv.

",f!.,pplicable for unit:; ha"iing a net area strength of 2500 psi or more.

(b) ,A.LLOVVABLE ~:HE.A.R IJr\j EiOLn:; FOR D ...1F'1F.~ICALLY DE::::IC;r\JED MA,Sm',JF~Y EXCEF'T Ur'"JBURf\JED cu~.,,( U~',JlTS

Solid Grouted

Diameter Embedment' Masonry Masonry

Bolt (inches) (Shear in (Shear in

(inches) Pounds! Pounds)

112 4 350 550 5/8 4 500 750 3/4 5 750 1100

na

f3 1000 1500 7 1250 113502 1-1/8

e

1500 22502

',f!.,nadditional 2 inches of embedment shall be provided for anchor bolts located in the top of columns: for buildinqs located in Seis:mic Zones: ~,los:.2, 3, and 4. 2Pennitted onl'i with not less than 2500 pounds per sq in. units

*Reproduced from the Uniform Building Code, 1985 Edition, Copyright 1985 with permission of the publisher, The International Conference of Building Officials."

44 Page 17 of24

on the smaller of the two loads calculated for the masonry and anchor material. Thus, the allowable load in tension is the lesser of:

(Eq.l )

or

(Eq. 2) where: TA= Allowable tensile load, Ib,

Ap= Projected area of the masonry tension cone, in2, fm = Masonry prism compression strength (In composite construction, when themasonry coneintersects different materials, fm should be based on the weaker material), psi,

AB=Anchor gross cross-sectional area, in2, fy=Anchor steel yield strength, psi.

The value of Apin Eq.1 is the area of a circle formed by a failure surface (masonry cone) assumed to radiate at an angle of 45° (see Fig. 19)from the anchor base. When an anchor is embedded close to a free edge, as shown in Fig 20,a full masonry cone cannot be developed and the area Apmust be reduced so as not to over-estimate the masonry capacity. Thus, the area Ap,in Eq. 1 will be the lesser of:

(Eq 3)

or

(Eq. 4) where: Ap= Projected area of the masonry tension cone, in.2,

1b= Effective embedded anchor length, in., 1be = Distance to a free edge, in.

44

Full Masonry Tension Cone

FIG. 19 . '~" G', ~..----"~~ A) PROJECTED CONE Page 18 of24

44 Page 19 of24

Reduced Masonry Tension Cone FIG.20a

B) PROJECTED AREA

Reduced Masonry Tension Cone FIG.20b

The effective anchor embedded length (1b) is the length of embedment measured perpendicular from the surface of the masonry to the plate or head for plate anchors or headed bolts. The effective embedded length of bent bar bolts (1b) is the length of embedment measured perpendicular from the surface of the masonry to the bearing surface of the bent end minus one bolt diameter. Where the projected areas of adjacent anchors overlap, Ap of each bolt is reduced by one-half of the overlap area. Also, any portion of the projected cone falling across an opening in the masonry (i.e., holes for pipes or conduits) should be deducted from the value of Ap calculated in Eqs.3 or 4.

Shear

The allowable shear load is based on the same logic as the allowable tension load.That is,the anchor capacity is governed by either the masonry strength or the anchor material strength. The distance between an anchor and a free masonry edge has an effect on the masonry shear capacity. Calculations have shown that for edge distances less than twelve times the anchor diameter, the masonry shear strength controls the anchor capacity. (C.I ations based on masonry with f'm = 1000 psi and anchor steel yield strength with f . = 60 ksi. Therefore, where the edge IS ance u or exceeds 12 anc or diameters. the allowable shear

-44 Page 20 of24

or

(Eq. G)

where: VA = Allowable shear load, lb.

When anchors are located less than 12 anchor diameters fro! a free edge, the allowable shear load is determined by linear interpolation from a value of VA obtained in Eq. 5 at an edge distance of 12 anchor diameters to an assumed value of zero at an edge distance 0 1 in. (25 mm). This takes into consideration the reduction in the masonry shear capacity due to the edge d,istance.

Combined Tension and Shear

Allowable combinations of tensile and shear loads are based on a linear interaction equation between the allowable pure tension and pure shear loads calculated in Eqs. 1,2,5 and 6.Anchors subjected to combinations of tension and shear are designed to satisfy the following equation:

T / TA+V / VA~ 1.0 (Eq.7)

where: T = Applied tensile load, lb..

V=Applied shear load, lb.

Proprietary Anchor Bolts

The allowable load equations previously presented are intended for use with plate anchors, headed bolts and bent bar anchors and have been proposed to the ACIIASCE 530 Committee on Masonry Structures. However, when the allowables from these equations are compared to test results for proprietary anchors,

they appear to produce acceptable safety factors.

Allowable Loads. Average factors of safety are 4.0 for tensile tests and 5.0 for shear tests on proprietary anchors. The combined tension/shear interaction equation produced an average safety factor of 7.0 when compared to test results on proprietary anchors. Therefore, based on comparison to test results, the allowable load equations proposed in this Technical Notes are suggested for use in the design of proprietary anchors in brick masonry. The embedment depth used to calculate the allowable load values should be equal to the embedded depth of the proprietary anchor.

Edge Distance. Edge distance is of particular concern when expansion anchors are used in brick masonry, due to lateral expansion forces produced when the anchors are tightened. These forces are often large enough to cause cracking or spelling of the brick when edge distances become small. To date, no research has been conducted in this area. Therefore, due to the lack of information, it is suggested that a minimum edge distance of 12 in.(300 mm) be maintained when expansion anchors are installed in brick masonry.

44 Page 21 0[24

Through Bolts

There are no known published reports available addressing the strength characteristics of through bolts in brick masonry. However, based on the conservatism in the allowables for bent bar anchors and proprietary anchors, the allowable load equations should provide acceptable allowable load values for through bolts used in brick masonry. The embedment depth used to calculate the allowable load values should be taken as equal to the actual thickness of the masonry.

Current Codes and Standards

i

At the present time, one model code and one design standard contain provisions for anchor bolt design in brick masonry. The BIA Standard, Building Code Requirements for Engineered Brick Masonry, and the

Uniform Building Code cover design allowables and embedment depths for anchors loaded in shear. There are no provisions for axial tensile loads or combined tension/shear loads in these documents. Tables 3 and 4 show the allowable shear loads and minimum embedment depths from the two documents. The values in Table 4(a) are based on rational analysis and in Table 4(b) on empirical analysis. As can be seen, the tables are very similar and are generally more conservative than the allowable shear loads obtained from

Eqs.5 and 6for the same embedment depths (Table 5).

!xl ! L:J

•From Building Code Requirements forEngineered Brick Masonry,Brick Instituteof America,August 1969.

'In determining thestresses on brickmasonry,the eccentricityduetoloaded

boltsand anchorsshallbe considered.

280ltsand anchors shall be solidlyembedded in mortar orgrout

3Noengineering orarchitectural inspectionofconstructionandworkmanship.

4Constructionand workmanship inspected byengineer,architector competent representative.

44 _{Page 23 of24}

'American ConcreteInstitute/American SocietyOfCivil Engineers Committee 530 on

MasonryStructures.

1Assuming fm =2,000 psi

ASTM A36 steel fy=36 ksi

Edge Distance = 12Bolt Diameters

SUMMARY

This Technical Notes is the first in a series on brick masonry anchors, fasteners and ties, It covers anchor bolt types, detailing and allowable loads for anchor bolts in brick masonry. Other Technical Notes in this series will address brick masonry fasteners and ties.

The information and suggestions contained in this Technical Notes are based on the available data and the

experience of the technical staff of the Brick Institute of America. The information and recommendations contained herein should be used along with good technical judgment and an understanding of the

properties of brick masonry. Final decisions on the use of the information discussed in this Technical Notes

arenotwithin the purview ofthe BrickInstitute of America andmustrestwith the project designer, owner or both

REFERENCES

1. Manual of Steel Construction, 8th Edition, American Institute of Steel Construction, Inc., Chicago,

Illinois, 1980.

PRF-44 Page 24 of24

3. Brown, R.H. and Dalrymple, GA, Performance of Retrofit Embedments in Brick Masonry, NSF Award No. CEE-8217638, "Static and Cyclic Behavior of Masonry Retrofit Embedments (Earthquake Engineering)", Report No.1, April 1985.

4. Hatzinikolas, M.; Lee, R.; Longworth, J. and Warwaruk, J., "Drilled-In Inserts in Masonry Construction", Alberta Masonry Institute, Edmonton, Alberta, Canada, October 1983.

5. Building Code Requirements for Engineered Brick Masonry, Brick Institute of America, McLean,

Virginia, August 1969.

6. Uniform Building Code, International Conference of Building Officials, Whittier, California, 1985.

7. Technical Notes on Brick Construction 17 Revised, "Reinforced Brick Masonry, Part I of IV", Brick Institute of America, McLean, Virginia, October 1981.

8. Technical Notes on Brick Construction 41 Revised, "Hollow Brick Masonry-Introduction", Brick Institute of America, McLean, Virginia, 1983.

9. Specification for the Design and Construction of Load-Bearing Concrete Masonry, National Concrete Masonry Association, McLean, Virginia, April 1971.

10. The BOCA Basic/National Building Code, 9th Edition, Building Officials and Code Administrators,

International, Country Club Hills, Illinois, 1984.

11.Standard Building Code, Southern Building Code Congress, International, Inc ..Birmingham,

Alabama, 1985.

12. Technical Notes on Brick Construction 7A Revised, "Water Resistance of Brick Masonry-Materials, Part 1\ of III", Brick Institute of America, Reston, Virginia, 1985.

-- 4Yi' -- 1 - 2x4 or 1-2x6 --- 5W -- 1- 3x4 or1-3x6 ---- 6" -- 2- 2x4or2. 2x6 8" -- 2-3x4or2 .3x6 ---- 8Yi -- 1-4x6or1-6x6 _10W __ 1-4x8orl-6x8 12Y4_ 1-4xl0 or 1-6x10

ATS-AB anchor bolts are pre-assembled anchor bolts that have been designed for usewith theATSsystem. Theyare available in 18",24"and 36"lengths and match the

strength andmaterial grade of the corresponding Strong-Rod connecting rods. The heavyhexnuts arepressed onto theboltto keepthem inplace.

Material: Standard (Model ABJ - ASTM A307, GradeA

High strength (Model AB_H) - ASTM A449 orASTM A193, Grade B7

Higher strength (Model AB H1_- 50) - ASTM A434, ClassBDorASTMA354, Claks B_i D Finish: None i Naming Scheme: ATS-AB5Hx24 ATS

=::J

T

L

Length Anchor Diameter

Bolt and Grade

*Units inVa"Increments

(Ex: 9='fa"or 1%")

Anchor Bolt Bolt Diameter Plate Washer Size 1, Component Color

Model No. - (in) (in) _i. (in) Cod'e

ATS-AB5 S/s 3fsxl'hxlY, 1v. Blue

ATS-AB7 1's 3fsx2v..x2.'A ... r 3reen·,> ...

ATS-AB9 1% 3/8x 23iAx2% PiA Orange

ATS-AB5H s/s 3/Sxl 'h x1'/2 _. 1'.4 Blue

ATS-AB7H 'Is 3,4,X2'.4x2V. 1'h Green

ATS-AS9H 11/8 3/ax2:v..x2%

.'.'

j3iA .•...Orange

....

ATS-AB9H150 1% 'hx3x3 1'/8 Orange

ATS-AB10H150 1% 1x3'hx3'h 2'h Purple

1. Anchor rodsareavailable in 18",24"and36"lengths.

2. Standard Anchor boltsare based onminimum Fu = 60,000 psi andFy=43,000 psi. 3. Highstrenqtn anchor bolts arebasedon minimum Fu=120,000 psi andFy=92,000 psi. 4. H150anchor bolts are based onminimum Fu=150,000 psi and Fy = 130,000 psi.

ANCHOR BOLT LOCATIONS

Anrnor bolts shall be specified bythe Designer. 1 -2x4or1-2x6=4Vi 1-3x4or1 -3x6 =5W 2 - 2x4or2-2x6=6' 2-3x4or2'3x6=8' 1-4x6 or1-6x6=8Y, 1 -4x8or1-6x8=lOW 1 -4x10 or1.6x10= 12'1. 1-2x4or1-2x6=4'fl 1-3x4or1-3x6=5'h" 2 -2x4 or 2-2x6=6' 2-3x4 or 2 -3x6=8' 1- 4x6or1 -6x6=8'h" 1-4x8or 1-6x8 =10Y, 1- 4x10 or1 -6xl0 = 12Y; ATS-AB Anchor Boll ~ COMPRESSION / MEMBERS Perpendicular-To-Wall Installation

ATS: Anchor Bolts Page 4 of 4 Anchor Rod f.'odel No. ATS·A85 ATS·,l"B7H ATS-,4B9H ATS·ABBH15G

1. IBC calculations are based on ACI 3113,Appendix D

2. For UBC and IBC wind design, embedment de, is based on the design strength of the anchor per AISC. Embedment and edge distances are calculated in order to attain a ductile steel failure mode.

3. For IBC seismic design, concrete strength is reduced by a factor of 0.75 per ACI 318, Section D.3.3.3. Steel strength is based on AISC calculations and does not include an 0.75 reduction factor. Embedment and edge distances meet the ductile requirements of ACI 318, Section D.3.3.4.

4. For UBC design anchor design for 2500 psi minimum concrete assumes no special inspection and a multiplier of 2.0 on the concrete per section 1923.3.2. For 3000 psi and 4500 psi concrete, special inspection is assumed and a multiplier of 1.3 is applied.

5. Plate washers have been designed for plate bending.

6. Alternate anchor bolt solutions may be provided by the Designer.

7. Foundation dimensions are for anchorage only. The Designer is responsible for the foundation size and reinforcement for all load conditions.

ReLated Catalog Pages (PDFs): ""top

C-ATS07 (AnchorTiedQwn$ysiem),page 16 (173k) Order frE?",_cat§Log~bymail

ATS: Anchor Bolts Page 3 of 4

Anchor Rod Model Nu.

•• See footnotes below

Wind and Seismic Design 97 USC without Supplementary Reinforcing:

Anchor Rod Model No. ATS-ABS ATS-AB7 ATS-'/\B9 Jl,TS-AB5H .w.TS-AB7H I\TS-AB9H ATS·AB9H1 SO

• See footnotes below

Seismic Design All IBC Codes:

Anchor Rod IWodel No.

A,TS-AB5H ATS-AB7H

•• Seefootnoles below Wind Design All IBC Codes: