METHODOLOGY FOR MITIGATION OF EARTHQUAKE HAZARDS IN UNREINFORCED

BRICK MASONRY BUILDINGS

Kariotis, J.C., Ewing, R.D., Johnson, A.W. (Co-Principal Investigators), and Adham, S.A.;

ABK, A Joint Venture, El Segundo, California, USA

ABSTRACT

Seismic hazards in existing unreinforced masonry buildings were investigated in order to provide a methodology to strengthen these buildings to appropriate resistance levels. The testing program was comprised of static and dynamic testing of walls and diaphragms, both in-plane and out-of-plane, and of anchorages between walls and diaphragms. In these guidelines for the analysis of existing buildings, there were several s ignificant departures from the code provisions for new construction. Results can be used as retrofit guideline s in accordance with the three seismic hazard levels of the 1978 ATC provisions based on effec­tive peak accelerations of 0.1, 0.2, and 0.4 g.

INTRODUCTION

Building construction using unreinforced masonry predates the development of seismic criteria that guide the design and construc tion of present-day buildings. A substantial number of these buildings are still being used in seismically active areas, even though investigations of earthquake damage have confirmed that this type of building has been a major contributor to loss of life during earthquakes. It has become imperative that a system of analysis methods and procedures--a methodology--be devised to determine realistic hazard mitigation requirements that will lead to cost-effective methods of retrofit of these buildings. In this way, the choice will not remain limited to either an enormous investment to make existing buildings conform to present standards for new construction or an economic loss resulting from the demolition of these buildings. Such a methodology can help meet seismic hazard mitigation goals of cities squeezed between threats to life safety and economic constraints. This paper describes the results derived from an extensive research program and gives guidelines for its application.

The research program resulted in the publ icati on of eight reports on various phases of the analysis and testing. The final volume, entitled The Methodology, provides in "gu ideline" form the procedures for investigating existing unrein­forced masonry buildings for the purpose of strengthening them to resistance levels that correspond to three levels of ground shaking intensity that describe the seismic hazard of the entire United States. This paper provides a summary of the procedure discussed in the Methodology volume.

BASIS OF THE METHODOLOGY

A review of research work on masonry showed that most of the effort has been directed toward determining the characteristics and response of reinforced masonry components to in-plane forces; and little or no effort was devoted to typical unreinforced masonry building response and the dynamic interaction among the building components. Accordingly, a research program was initiated that included several types of tests:

1339

• Dynamic testing of full-scale walls, out-of-plane

• Static and dynamic testing of full-scale diaphragms, in-plane

• Static and dynamic testing of walls, in-plane

• Anchorage between walls and diaphragms

As a result of these experiments, it was determined that elastic or equivalent static procedures are not completely satisfactory to define the dynamic and highly nonlinear response of unreinforced masonry buildings.

The experimental data were then used in conjunction with analytical models for four related component responses and their interactions:

• In-plane motions of endwalls and crosswalls induced by the earthquake ground motion

• Roof and floor diaphragms subjected to in-plane motions induced by the endwalls and crosswalls

• Walls subjected to out-of-plane motions induced by ground motion at the foundation level and by the diaphragm motion or by a pair of diaphragms

• Anchorage between the walls and diaphragms

IN-PLANE RESPONSE OF WALLS During an earthquake, the ground motion is transmitted from the building/ foundation interface through the endwalls (in-plane response) to the floor and/or roof diaphragms that drive the walls in the out-of-plane direction. Masonry shear walls can be considered rigid relative to the diaphragm stiffness and can be modeled as a rigid block resting on a soil. Analyses performed over a realistic range of building aspect ratios and soil stiffnesses showed that the ground motion is transmitted through the endwalls with little amplification. However, in-plane relative displacement of the shear walls (e.g., pier rocking and/or diagonal compression failure) will modify the transmitted ground motion significantly.

ROOF AND FLOOR DIAPHRAGMS SUBJECTED TO IN - PLANE MOTIONS The dynamic response of diaphragms shows a nonlinear hysteretic behavior for ground motions of moderate and higher intensities. The analytical model de­veloped for this type of diaphragm requires only two parameters to define the force-deformation envelope (i.e., the ultimate force capacity and the initial stiffness) and one parameter to define the degrading, unloading and reloading stiffness. For typical unreinforced masonry buildings, the diaphragm stiffness is modeled by nonlinear, hysteretic shear springs, and the sidewall mass (the walls are assumed to crack) and tributary diaphragm mass are lumped at the nodes. Peak velocities at the top and bottom of the walls can be obtained from the model, as well as relative deformation s between the top and bottom of the walls .

WALLS SUBJECTED TO OUT-OF-PLANE MOTIONS The dynamic stability of fully anchored unreinforced walls subjected to out-of­plane motion s was determined from full-scale testing. The parameters that affect stability are:

1340

• Velocities imparted by the diaphragms to the ends of the walls

• Ratio of weight of wall in the stories above the story under consideration to the weight of the wall in the story under consideration

• Height/thickness (H/t) ratio of the wall in the story under consideration

ANCHORAGE BETWEEN WALLS ANO OIAPHRAGM Adequate anchorage of the walls to the diaphragm is an essential part of achiev­ing hazard mitigation in unreinforced masonry buildings. Anchorage forces have been developed for use in the methodology that are based on tests and nonlinear, dynamic analyses of the diaphragms. Although not a new concept, the paramount consideration of the methodology is life safety. This is obtained by limiting building damage and by minimizing the probability of the separation of the walls and parapets from the floors and roof. The collapse of parts of the gravity load-carrying system that are sensitive to rotational or relative displacement is investigated.

NEW CONCEPTS As stated earlier, the guidelines proposed for the analysis of are not the same as the code provisions for new construction. departures are:

existing buildings Si gnifi cant

• Oue to the sensitivity of earthquake hazard mitigation recommendations to the intensity of ground shaking, the use of state-of-the-art documents for seismic hazard zoning is recommended.

• Oue to the nonlinear, dynamic response of unreinforced masonry buildings, the procedures for each seismic hazard zone are separately defined rather than using a factored coefficient for each seismic hazard zone.

• Input ground motions for earthquake hazard zones are taken from the updated-­although still tentative--seismic design provisions of the Applied Tech­nology Council (2). However, the ground motions are defined by mean values rather than upper bounds of motions.

• The seismic response model for the buildings is modeled as a rigid block on flexible soils. This basic in-plane response model is modified for walls with a limited interstory shear capacity and ductile-like behavior.

• The velocity amplifications and the relative displacement response imparted to the walls, out-of-plane, are based on nonlinear, dynamic analyses that have been correlated with full-scale diaphragm tests.

• Oynamic stability concepts for unreinforced wall elements subjected to out­of-plane motions are utilized in lieu of requirements for an elastic resis­tance capacity.

• Materials resistance capacities are based on inelastic behavior of materials.

• All existing materials and elements in the building that are distorted by relative horizontal or interstory displacement are considered in the response model and the structural resistance model.

1341

FIELO SURVEYS ANO PRELIMINARY OESIGN

The methodology for mitigating seismic hazards in unreinforced masonry buildings is presented for the three seismic hazard zones described by the ATC 3-06 pro­visional guidelines (2). These seismic hazard zones are defined by Effective Peak Accelerations (EPA) of 0.1 g, 0.2 g, and 0.4 g.

The procedure for using the methodology begins with a five step fie.ldsurvey that is the same for all seismic zones: to prepare preliminary framing plans for roofs and floors, prepare preliminary elevations of all walls, investigate anchorage of walls, investigate wall materials, and test existing materials of the anchorage systems . Once these preliminary steps have been accomplished, the analysis procedure begins . This is done in the following steps. For 0.1 9 and 0.2 9 seismic zones:

• Identify all hazardous building elements on framing plans. floor plans, and wall elevations.

• Calculate recommended anchorage force at each floor above the building base and at the roof level.

• Verify capacity of existing wall anchors.

• Oesign retrofitted wall anchorage systems.

• Oesign bracing systems for parapets and appendages extending above the roof anchorage level.

In addition, special investigations may be required by the methods recommended for 0.4 9 seismic zones for the following conditions:

• If wall H/t ratios are in excess of historic standards or building height ­plan dimension ratio exceeds 3, and the structure is founded in soft soils.

• If diaphragm discontinuities exist adjacent to an unreinforced masonry wall.

• If the building survey has determined that parts of the vertical load­carrying system may act as a tie to a shear wall, and horizontal displace­ment of that part of the vertical load-carrying system relative to the shear wall will cause loss of bearing capacity.

• If the building survey has determined that major elements of the vertical load-carrying system are supported on masonry piers. and if there is a probability that significant relative rotation of beams on bearing surfaces will oCcur.

ANALYSIS GUIOELINES

The analysis procedures for seismic hazard zone EPA = 0.4g are more extensive. and in the following sections of this paper these procedures will be summarized in the form of guidelines.

The guidelines describe a probable response of existing building elements that is correlated to element displacements that extend into the inelastic range. Capacities of existing materials are given as yield capacities. Yield capacities of structural elements are used for design of retrofitted systems.

1342

1. Anchorage of Wall Elements. Calculate the recommended anchorage force at each floor above the building base and at the roof level, where the anchorage force is equal to 1.0 times tributary wall weight. This procedure includes the design of the bracing system for parapets and appendages extending above the roof anchorage level. If existing wall anchors are to be used as part of the wall anchorage system, verify capacity of the embedded ends of the existing wall anchors by nondestructive testing. Qualify nondestructive testing by limited destructive testing. Determine by analysis the capacity of the connection to the diaphragm of the existing anchors.

For analysis of the anchorage of the walls to the diaphragm, a C is recommended that is an upper bound of acceleration amplification . This uppeP bound of amplification is appropriate for diaphragms that have near-elastic response. This near-elastic response can be due to diaphragm stiffness and strength or to a small span/depth ratio.

2. Stability of Anchored Wall Elements. Allowable H/t ratios of walls with minimum quality mortar for several types of buildings are given in Table 1. Even so, these H/t ratios are dependent on the presence of crosswalls and on diaphragm demand/capacity ratio and span length. Crosswalls, as defined here, are existing walls constructed of materials other than unreinforced masonry or retrofitted structural elements that extend between all diaphragms at all levels of the building. The definition of crosswalls includes all lateral load resist­ing systems with a defined nondegrading load-displacement behavior. Their spacing and capacity are defined in Table 2. Buildings with diaphragms conforming to the requirements of Table 3 qualify as "buildings with crosswal1s."

TABLE 1. ALLOWABLE HEIGHT/THICKNESS RATIO OF UNREINFORCED MASONRY WALLS WITH MINIMUM QUALITY MORTAR

Walls of one-story buildings First-story walls of multistory buildings Walls in top story of multistory buildings All other walls

*Crosswalls

Crosswalls*

20 20 14 20

All Other Buildings

14 20

9 15

TABLE 2. MINIMUM CAPACITY OF CROSSWALLS ANO SPANS (L) OF DIAPHRAGM1,2,3 Diaphragm Span (L)

in Meters

90 or more 54 minimum

Demand/Capacity of Diaphragm

1.0 or less 2.0 or more

Minimum Capacity of Crosswalls as a Percentage of Diaphragm Capacity

30 30

1. Minimum demand/capacity ratios may be interpolated for diaphragms with spans between 54 and 90 meters.

2. Maximum spacing of crosswalls is 12 meters measured in the direction of the span.

3. Not applicable for steel decking detailed for lateral load resistance, concrete filled steel decks, and concrete framed floors.

1343

TABLE 3. MINIMUM OEMANO/CAPACITY RATIO ANO SPAN OF OIAPHRAGMS BETWEEN UNREINFORCEO MASONRY SHEAR WALLS FOR QUALIFICATION AS "WITH CROSSWALLS"*

Horizontal Span Between Shear Walls

18 M or less

Minimum Oemand/Capacity Ratio

2.5 3.0 54 M maximum

*Minimum demand/capacity ratios may be interpolated for diaphragms with spans between 18 and 54 meters.

The methodology uses the criteria of predicted dynamic stability shown in Figure 1. The parameters that affect stability, developed from full-size testing , are:

• Input velocities imparted by the diaphragms to the ends of the walls

• Ratio of weight of wall in the stories above the story under consideration to the weight of the wall in the story under consideration

• H/t ratio of the wall in the story under consideration

1. 75 ....-----,.....------,----,-------,---...,.------,

1.50 I-----t-~....>..c_~-----!------t----t------i

1. 25 I------.p..,,---~--.....:~---"'-~~----t----t------i

5.0

~ 1. 00 1-----+----"-.::---i-""'-.::::---+--~:--t-...::-..=_-+-4.:..:... 0,,-----., ~ 3· 0

VI VI

'" VI

>

2 . 0

1.0 o. 751-----+---..::......,jc-----=::.......::--+----=-t-=---~--___i 0.5

0.1 0.0

0.501-----t------i-----!--"-----t----+-:0:-;:/W-:----{

V, SRSS - SQUARE ROOT DF SUM DF PEAK VElOC1T1ES AT 0.251---- TOP ANO BOHOM DF WAll SQUAREO(m/s) +--- - -j

H/T - HE1GHT TO TH1CKNESS RAT10 DF WAll

O/W - OVERBUROEN WE1GHT TO WAll WE1GHT RAT10

I O L-__ ~ __ ~~ __ ~ __ -J ___ ~ __ ~

5 10 . 15 20 25 30 35

H/T

FIGURE 1. UNREINFORCEO MASONRY WALL STABILITY CRITERIA, 98% PROBABILITY OF SURVIVAL

1344

Crosswalls conforming to the mlnlmUm requirements of Table 2 may be introduced into the building to increase the acceptable H/ t ratio of walls; or, walls that exceed the recommended H/ t ratio may be braced by suppl~nental members panning between diaphragm levels.

Recommendations for design and i nstallation of the supplemental bracing members are:

• Design brac ing members for 0.4 times the tributary wall weight.

• Deflection of the bracing member, calculated using recommended forces, should not exceed 0.15 times the wall thickness.

• Horizontal spacing of the vertical bracing members should not exceed one­half the unsupported height of the wall or 10 ft. maximum.

• The vertical bracing members should be anchored to the floor or roof framing independently of the recommended wall anchorage system.

3. Computation of Earthquake Response Force. Recommendations for computation of earthquake response forces are:

• Calculate weight of building as a lumped weight at each floor, mezzanine, and roof level. For convenience, tabulate the weight computations as in­plane wall weight (WW) and weight tributary to diaphragms (W D), at each level, for each axis of analysis of the building.

• For analysis of the connection of the ends of diaphragms to select Cp from Table 4. However, the shear used for design at the ehd of the diaphragm need not exceed (v u • D). Yield vu, of diaphragms are given in Tab l e 5.

the walls, of connection capacities,

• The restoring shear capacity, VR, of any shear wall composed of piers need not exceed 0.2 W + Ln 0.2 W , and the diaphragm shear at the shear wa ll at any level need n~t exteed th~ yield capacity (vu ' D) of the diaphragm at that level.

• For analysis of in-plane sheRr in each shear wall when determined to be critical, use V = 0.4 Ww + LI 0.4 WO. However, the diaphragm shear at the shear wall at any level need not exceed the yie ld capacity (v u ' D) of the diaphragm at that level.

The seismic response factors, C , of the diaphragms are given in Table 4. These factors equal or exceed the sei~mic zone EPA to account for amplification of input motions that are applied to the ends of the diaphragms. However, the upper bound of response shear that can be coupled with the shear walls is the yield capacity of the diaphragm.

TABLE 4. RESPONSE FACTOR, C , FOR SHEAR CONNECTION OF HORIZONTAL DIAPHRAGM p

Single layer of boards with applied roofing 0.45 Double layer of boards or blocked plywood 0.8 Steel decking not detailed for lateral load resistance 0.6 Concrete filled steel decks or concrete framed systems with span/depth ratio of 2 or less 0.4

1345

TABLE 5. YIELD CAPACITIES, v , OF EXISTING ROOF ANO FLOOR CONSTRUCTION u

Description of Existing Construction

Straight sheathing with roofing applied on the sheathing or a single layer of tongue and groove sheathing without roofing

Straight sheathing with plywood overlay

Unblocked plywood sheathing with roofing applied on the sheathing

Diagonal sheathing with roofing applied on the sheathing

Plywood sheathed floors or roofs with blocking at panel edges

Double board system with finish flooring laid over diagonal sheath­ing or multiple board systems with board edges offset

Metal roof deck system designed for minimal lateral load capacity

Metal roof deck systems designed for lateral load capacity

Concrete filled steel decks

Concrete framed floors

Yield Capacity of Materials Vu in lb/ft, shear

300

650

400

750

2-1/2 x shear values listed in design codes such as UBC Table 25-J or SBCC Supplement to Chap­ter XVII

1800

1800

3000

As determined by static yield capacity testing

Concept of v is not applicable u

The response at a shear wall is calculated as the hazard zone EPA times the weight of the shear wall and the weight that can be dynamically coupled with the shear wall. The effective coupling of the weight tributary to a diaphragm is calculated as EPA times the calculated weight. but the coupling response is limited to the yield capacity of the diaphragm(s) at any level. This procedure is not intended to give an arithmatical summation of peak element response, but is a probability of combinations that is similar to seismic design procedure for new buildings.

4. Distribution of Response Forces. For buildings with shear walls that exceed height/length ratios used in response studies and that are founded on soft soi1s, redistribute the response forces in accordance with the ATC pro­visions (2). Redistribution of response forces is not retommended except for these special cases.

1346

5. Analysis of Horizontal Displacement Control Elements. Recommended analysis procedures for horizontal displacement control elements are based on dynamic testing and modeling. The procedure is as foll ows:

• For diaphragms without crosswalls:

Calculate demand/capacity ratio WD 2v u D

where WD= Total weight tributary to diaphragm

Vu Yield capacity of diaphragm (see Table 5)

D Diaphragm depth

From Figure 2, using the appropriate diaphram yield capacity and span length, determine adequacy of existing diaphragm. If the existing diaphragm does not meet the span limitations, the diaphragm must be retrofitted to increase vu' or crosswalls may be added to limit relative horizontal displacement.

150

125

100

75

50

25

O L-__ ~ __ -L ____ L-__ ~ __ ~ __ ~

O

FIGURE 2 ACCEPTABLE SPAN FOR DIAPHRAGMS

(BASED ON DISPLACEMENT CONTROL CONCEPTS)

1347

• For diaphragms with crosswalls:

Calculate demand/capacity ratio: WD 2v • D + L V

u c

LV total yield capacity of crosswalls that are spaced not to exceed that speEified in Table 2.

• If the spacing of existing crosswalls is that specified in Table 2 and the capacity LV

C exceeds 20% of WD' the span of the diaphragm shall be unlimited.

• For multistory buildings, Vc utilized for diaphragm analysis at any upper story shall be added to the WD of the story below for analysis of that story.

For the special case of horizontal displacement control of an "open-front" building, the recommendation for diaphragms with shear walls at the diaphragm ends may be used (Fig. 2). To utilize Table 2, an equivalent LI is calculated . The wall weight, WW' at the open end is used to calculate LI:

L = 2 Ww • L + L I WD

Compare demand/capacity ratio of diaphragm with an acceptable span calculated as LI. If acceptable crosswalls exist, calculate

for entry to Figure 2. v • D + LV

u c

6. Analysis of Vertical Displacement Control Elements. For shear walls that are divided into piers by door and window openings, calculate the restoring shear capacity of the pier system as:

n P D L 0.9 ~ x I x

Where P = Axial load on pier x

In-plane depth of pier

Least height of pier if opening height on sides of pier varies

For computation of restoring shear, the stability moment of a fully cracked pier system is used. The restoring shear is computed from the weight on the pier, P, times 0.9 of the in-plane pier depth, D. The r ecommended 0.9 factor is based on in-plane testing of piers.

Compare this restoring shear capacity with the minimum recommended r estoring shear:

0.2W + w

n L 0.2W D 1

Compare calcu lated VR on each pier with in-place shear capacity V, where

v A a V = l.5

1348

where A = Gross area of pier, and va = Allowable shear = 3/4 (3/4v t + P/A)

= 20th percentile of in-plane test shear values reduced to equivalent shear at zero axial stress, and P = Axial load on pier

and for all piers VR <V, supplement restoring shear by materials designed at yield capacity. If for any pier VR >V, in-plane shear failure is probable and piers must be analyzed for shear capacity, using the following four steps:

• Distribute response shear V to pier system using stiffness as D/H.

• Calculate v = 1.5 V/A for stiffest pier.

• If v>va ' increase shear capacity of wall with consideration of relative stiffness of existing and new materials.

• For walls without openings and with height/length ratio S 0.5, calculate v = V/A.

7. Interconnection of Building Elements. A continuous load path for all calculated response forces should be provided. However, interconnection capacity of existing materials need not be analyzed. Two design steps must be under­taken: to design the tie system parallel to the shear wall for distribution of calculated response forces, and to design the diaphragm distribution tie system for retrofitted crosswalls or shear walls.

8. Review of Vertical Load-Carrying Elements. If the building survey has determined that major elements of the vertical load-carrying system are sup-ported on masonry piers, provide independent structural steel columns or equivalent at the face of the masonry pier. An independent foundation system is not required. An exception is if a shear wall is retrofitted into the line of bearing masonry piers, such as in the plane of an open front or a wall not continuous to the building base; then independent support columns are not required.

SUMMARY

A useful methodology for the mitigation of seismic hazards in existing unrein­forced masonry buildings has been established based on a research program that combined analytical and experimental investigations. Several new concepts were introduced that are significant departures from the current code provisions for new construction. The results, given here in "guideline" form for the highest of the three seismic hazard levels defined by the 1978 Applied Technology Coun­cil provisions, were originally presented in an eight-volume report produced by ABK, A Joint Venture, of which The Methodology is the final volume.

1349

ACKNOWLEDGEMENT

This research was conducted by ABK, A Joint Venture, for the National Science Foundation under Contract No. NSF-C-PFR-78-19200 and Grant No. CEE-8100532. The Joint Venture ABK consists of the three firms, Agbabian Associates, S.B. Barnes & Associates, and Kariotis & Associates, all in the Los Angeles area. The principal investigators for the three firms are R.D. Ewing, A.W. Johnson, and J.C. Kariotis. Dr. J.B. Scalzi served as Technical Director of this project for the National Science Foundation and maintained scientific and technical liaison with the joint venture throughout all phases of the research programo His contributions and support are greatly appreciated.

REFERENCES

(1) ABK, A Joint Venture. Methodologyfor Mitigation of Seismic Hazards in Existing Unreinforced Masonry Buildings, Vol. 8, "The r~ethodology," ABK-TR-08. El Segundo, CA: Agbabian Associates, Jn 1984.

(2) Applied Technology Council. Tentative Provisions for the Development of Seismic Regulations for Buildings, ATC 3-06. Palo Alto, CA: ATC, 1978.

1350