ENHANCEMENT OF SEISMIC PERFORMANCE AND DESIGN OF PARTIALLY
GROUTED REINFORCED MASONRY BUILDINGS Erik Avila Home Institution:
California State University, Los Angeles REU Site: University of
California, San Diego PI: Benson Shing, Ph.D Graduate Student
Adviser: Andreas Koutras August 22, 2014 i Abstract
PartiallygroutedstructuresconstructioniscommonintheMidwestern,Eastern,and
Northwestern United States. Seismic design provisions for partially
grouted reinforced masonry are based solely on research performed
on fully grouted shear walls. This joint
projectbetweenDrexelUniversity,UniversityofMinnesota,andtheUniversityof
California,SanDiegoaimstounderstandthebehaviorofpartiallygroutedmasonry
structures and to develop improved design details to enhance their
seismic performance.
AttheUniversityofCalifornia,SanDiegosNetworkforEarthquakeEngineering
Simulation(NEES@UCSD)site,thesystemlevelperformancewasstudiedby
constructing two full-scale, single-story, partially grouted
reinforced masonry buildings
andtestingthemattheEnglekirkStructuralEngineeringCentersLargeHigh
Performance Outdoor Shake Table (LHPOST). The first structure
represented common practice while the second structure had added
enhancements. The applied enhancements,
whichincludeddoublegroutedverticalcells,bondbeams,andladdertypejoint
reinforcement,enabledSpecimen2toperformbetterthanSpecimen1.Throughthe
shake tests, the seismic performance of the structures was better
understood. ii Table of Contents 1.Introduction
......................................................................................................................
1 2. Literature
Review............................................................................................................
1 3.
Methods...........................................................................................................................
3 3.1 Testing Overview
..................................................................................................................
3 3.2 Specimen One
........................................................................................................................
4 3.3 Specimen 2
............................................................................................................................
5 4. Student Contributions
.....................................................................................................
7 4.1 GoPro camera documentation
...............................................................................................
7 4.2 Use of MATLAB to Inspect Sensors
.....................................................................................
8 4.3 Structural inspection after motion
.........................................................................................
9 4.4 Strain Gage Yield Sequence
..................................................................................................
9 4.5 MATLAB Video and Data Synchronization
.......................................................................
11 4.6 Proposed Shear Friction Formula Problems
........................................................................
12 Results
...............................................................................................................................
12 Conclusions
.......................................................................................................................
15 Acknowledgements
...........................................................................................................
15 Further Information
...........................................................................................................
16 References
.........................................................................................................................
16 iii Table of Figures Figure 1: Prototype Building for
Testing...........................................................................
3 Figure 2: Shake Table Test Structure
..................................................................................
4 Figure 3: South Elevation View of Specimen
1..................................................................
5 Figure 4: East Elevation View of Specimen 1
....................................................................
5 Figure 5: South Elevation View of Specimen
2..................................................................
6 Figure 6: East Elevation View of Specimen
2...................................................................
6 Figure 7: Specimen 2 View 1 Captured with GoPro Camera
............................................. 7 Figure 8: Specimen
2 View 2 Captured with GoPro Camera
............................................. 7 Figure 9:
Malfunctioning Accelerometer Analyzed Using MATLAB
............................... 8 Figure 10: Functioning
Accelerometer Reading Analyzed Using MATLAB .................... 8
Figure 11: Crack Occurrence - Marking of Structure After Each Test
............................... 9 Figure 12: Strain Gage Data Using
Average Yield With MATLAB.First Yielding is Circled
..................................................................................
10 Figure 13: South view of wall elevation use to display strain
gage yield sequence.The colors represent the time at which the
yielding occurred. ....................... 11 Figure 14: MATLAB
Video and Data Synchronization
................................................... 12 Figure 15:
Specimen 1 after testing
..................................................................................
13 Figure 16: Specimen 1 after testing, Side View
............................................................... 13
Figure 17: Specimen 2 after testing, Front View
.............................................................. 14
Figure 18: Specimen 2 after testing, Side View
............................................................... 14
1 1.Introduction
Reinforcedmasonryconstitutesabout10%ofalllow-riseconstructionintheUnited
States.TheTMS-402/ACI-530/ASCE-5(MSJ
C,2013)buildingcoderequirementsfor masonry structures developed by
the Masonry Standards J oint Committee (MSJ C) and ASCE 7 (ASCE,
2010) classify reinforced masonry shear walls into three types:
special,
intermediate,orordinary.Thesewallscanbeeitherfullygroutedorpartiallygrouted
(Shing et al. 2013). Fully grouted masonry walls have grout placed
in every cell. Partially grouted ordinary walls are used outside of
the West Coast and have grout in the location of the
reinforcement.The construction of partially grouted reinforced
masonry buildings
iscommonpracticeintheareaswithintheMidwestern,Eastern,andNorthwestern
United
States.Partiallygroutedstructuresaredesignedusingshearstrengthandseismicdesign
provisions specified by TMS-402/ACI-530/ASCE-5(MSJ C, 2013). The
shear strength and seismic design provisions are based solely on
experimental results of fully grouted masonry shear walls (Shing et
al. 1990). Therefore, the motivation for this project derives from
the shear strength formula overestimating shear capacity. Not much
research has
beendoneonpartiallygroutedreinforcedmasonrybuildingstoimprovetheshear
strength formula and design provisions. This is a joint project
between Drexel University, University of Minnesota, and the
University of California, San Diego to understand the behavior of
partially grouted masonry structures and to develop improved design
details to enhance their seismic performance.
AttheUniversityofCalifornia,SanDiegosNetworkforEarthquakeEngineering
Simulation(NEES@UCSD)sitethesystemlevelperformancewasstudiedby
constructing two full-scale, single-story, partially grouted
reinforced masonry buildings. These structures underwent shake
tests at the Englekirk Structural Engineering Centers Large High
Performance Outdoor Shake Table (LHPOST), the worlds first and
largest
outdoorshaketable.Throughtheseshaketabletests,theseismicperformanceofthe
structureswasbetteranalyzed.Thefirststructuresconstructionreflectedthecurrent
practice, while the second had improved details. 2. Literature
Review
Theexpressionforshearstrengthinfullygroutedreinforcedmasonryissharedwith
partiallygroutedreinforcedmasonry,thereforeresultsmaynotbeaccurate.The
expression used in the Masonry Standard J oint Committee to
determine shear strength
hasbeentestedforaccuracythroughanexperimentoffourpartiallygroutedspecial
reinforced masonry shear walls by Minaie et al. (2010). The four
partially grouted special
reinforcedmasonrywallsweresubjectedtoreversedcyclicdisplacementinorderto
experimentally establish their in-plane behavior and assess the
appropriateness of current seismic design provisions for partially
grouted shear walls. Minaie et al. (2010) arrived at
theconclusionthattheshearstrengthexpressionforreinforcedmasonryshearwalls
provided by the MSJ C appears unconservative for partially grouted
masonry shear walls. The expression may not be appropriate because
it is based exclusively on tests with fully 2
groutedmasonryshearwalls,whichdisplayedfailuremodesthataredifferentthan
partially grouted
systems.Anexperimentalstudyoftheseismicperformanceofpartiallygroutednominally
reinforcedconcrete-masonrystructuralwallswascompletedtoevaluatethisasan
alternativetofullygroutedreinforcedmasonrystructuralwalls.Fivereduced-scale
structural walls were constructed and tested under scaled versions
of the 1940 El Centro,
Californiaearthquake,usingitsnorth-southcomponentrecordwithaconstantaxial
compressionloadthatrepresentedasingle-storybuilding(Kaspariketal.2014).
According to Kasparik et al. (2014), the use of partially grouted
nominally reinforced
masonryalsoresultsinareducedcost,comparedwithtraditionalreinforcedmasonry
systems used in seismic zones that are typically fully grouted
within the plastic hinge zones and require higher reinforcement
ratios. Unfortunately, the capacity of the shake table did not
allow the testing of the walls to failure. This was an attempt to
understand partially grouted masonry systems. Key aspects of the
wall were examined pertaining to the wall yield capacities,
stiffness degradation, period shift, displacement ductility, and
seismic forcereduction factors. Although this experiment involved
nominally reinforced masonry and not fully reinforced masonry it
does seek to prove the cost effectiveness of using a partially
grouted system. Unreinforced masonry is economically competitive
but
couldperformpoorlyinearthquakes.Thereasonsforthispoorperformancearethe
inherent brittleness, lack of tensile strength, and lack of
ductility; that is, a lack of the propertiesgiven to reinforced
masonry by the steel reinforcing (Hess 2008). The cost effective
advantage of partially grouted system is evident nonetheless.Cost
effectiveness plays a major role in the design of structures but
seismic performance is even more crucial. Seismic resistance of
partially grouted masonry is yet to be verified and current
analysis and design methods must be validated (Shultz 1996).
Schultz (1996) attempted to study the seismic performance of
partially grouted masonry shear walls by testing six partially
grouted shear walls with bond beams. Bond beams serve to connect
and strengthen the walls. Outermost vertical cells of the walls
were reinforced vertically and grouted while a single bond beam
containing horizontal reinforcement was placed at mid-height of the
walls. Schultz (1996) arrived at the conclusion that partially
grouted masonry is a viable lateral-load resisting system for areas
of moderate and low seismic
risk.Shultz(1996)addedthattheresistancetothedrifthistoriesisstableandalso
featureshighinitialstiffnessandampleenergydissipation.Finally,verticalcracks
arising from stressconcentrations between ungrouted andgrouted
masonrydominated wall behavior and the sliding friction between
masonry panels and the concrete surface contributed to the seismic
resistance.
Thesepreviousexperimentswereattemptstounderstandseismicperformanceofa
partiallygroutedstructurewhetheritisfullyreinforcedornominallyreinforced.Itis
evident that not much research has been done in this topic and this
experiment will serve to be the first project with conclusive data
that may lead to a new successful enhanced design for these
partially grouted reinforced masonry systems.3 3. Methods3.1
Testing Overview
Twoone-storypartiallygroutedreinforcedmasonrystructuresweretestedatthe
Englekirk Structural Engineering Center at Camp Elliott, a site
located 15 km away from the main UCSD campus. The specimens were
built on the LHPOST, the worlds largest outdoor shake table with
dimensions of 12 m x 7.6 m (40 ft x 25 ft). The shake table is
uni-directionalwithmotionalongthelongdirection.AccordingtotheNEES@UCSD
website (UCSD, 2013), the shake tables peak acceleration when bare
is 4.2 g and 1.2 g when a 400 ton payload is present. The tables
peak velocity is 1.8 m/s. Finally, the table can resist an
overturning moment of 35 MN-m when bare, and 60 MN-m when a 400 ton
payloadisapplied.ItwasessentialthattheLHPOSTisabletoaccuratelyreproduce
near-fault ground motion effects (UCSD, 2013). Figure 1 depicts the
prototype building, the dark lines on the plan view represent walls
and the small openings represent windows
anddoors.Figure2showstheactualteststructure,whichisaselectionfromthe
prototype seen in Figure 1. Figure 1: Prototype building for
testing (Source: Andreas Koutras) 4 Figure 2: Shake table test
structure (Source: Andreas Koutras) 3.2 Specimen One This
experiment began with the construction of a one-story partially
grouted reinforced masonry structure that reflected common practice
on the East Coast. Location of vertical #4 rebar was determined
with the MSJ C 2013 building code and while bond beams were not
required, they were placed to see if they proved to be practical.
Bond beams are not usually used in East Coast construction.
Although most of the construction reflects East Coast practice,
since bond beams were added this was considered a modified East
Coast design. Figures 3 and 4 display two different views of
specimen 1. The two figures show
darkgrayareas,whicharewherethegroutisplacedandalsothelocationofthe
reinforcement. The specimen was monitored using 178 strain gages,
180 displacement transducers, and 39 accelerometers. Specimen one
was tested using the El Centro 1940 motion, which was of magnitude
6.9 Mw. 5 Figure 3: South elevation view of Specimen 1 (Source:
Andreas Koutras) Figure 4: East elevation view of Specimen 1
(Source: Andreas Koutras) 3.3 Specimen 2 After testing the first
specimen a second specimen was constructed with an enhanced design.
The new enhancements included vertical double reinforced cells with
#3 rebars,
whilemaintainingthesameverticalreinforcementratioasthefirstspecimen.Bond
beams were maintained with #4 rebar. Ladder type joint
reinforcement of size 3/16 in. wire was added in every course. This
follows the MSJ C building code recommendation
thatjointreinforcementshouldbeappliedineverycoursetobeconsideredas
reinforcement. The base was also roughened to prevent sliding and
to increase the shear
coefficient,anddowelswerealsoadded.Thetoppingthicknessoftheroofslabwas
increased from 4 inches to 12 inches. This can be seen in both
Figures 5 and 6. Figures 5
and6aretwodifferentviewsofthesecondspecimen.Thesetwofiguresshowthe
enhancements incorporated into the second specimen. 6 Figure 5:
South elevation view of Specimen 2 (Source: Andreas Koutras) Figure
6: East elevation view of Specimen 2 (Source: Andreas Koutras) 7 4.
Student Contributions 4.1 GoPro camera documentation
GoProcameraswerefirstinstalledbeforethesecondstructurewastestedonthe
LHPOST at the Einglekirk testing facility. A total of 15 GoPro
cameras were set up in and around the structure. Four cameras
documented the inside of the building while 11
camerasrecordedtheoutside.Thesecameraswerefirstsyncedtobecontrolledbya
single remote control and to capture the entire test from different
angles at the same time. Figures 7and 8 showexamples of GoPro video
images of the second specimen. The
GoProvideoswereusedtoexaminethestructuresperformanceduringthevarious
motions. Four videos were selected to be used in MATLAB. The MATLAB
program was to display the data and videos simultaneously. Figure
7: Specimen 2 View 1 captured with GoPro camera Figure 8: Specimen
2 View 2 captured with GoPro camera 8 4.2 Use of MATLAB to Inspect
Sensors Sensor feedback was provided in the form of MATLAB plots.
These plots were inspected for any problems in the sensors
performance.Approximately 40accelerometers were inspected using
MATLAB plots. Figure 9 displays the output of a flawed
accelerometer and Figure 10 displays a functioning accelerometer.
The plot titles display the node that belongs to the sensor for
reference. Reporting which sensors are not performing allows for
inspection before further testing. After inspecting the plots a
list of malfunctioning
sensorswascreatedandpresentedtothegraduatestudentsohecouldinspectthose
sensors. Figure 9: Malfunctioning accelerometer analyzed using
MATLAB Figure 10: Functioning accelerometer reading analyzed using
MATLAB 9 4.3 Structural inspection after motion During the testing
period from J une 20 to J une 25, 2014 Specimen 2 was subjected to
various table motions. After each motion the structure was
inspected and cracks were marked with a certain color representing
that specific motion. This assisted in finding any patterns and
tracking how the enhancement of the structure was performing after
each motion. Figure 11 shows the markings on the wall, which have
been labeled with the
nameofthemotionandthedate.Theundergraduatestudentparticipantwasonly
available to assist during the testing of the second specimen.
Figure 11: Crack occurrence - Marking of structure after each test
4.4 Strain Gage Yield Sequence
Approximately177MATLABplotsforthestraingageswereanalyzedforyielding.
MATLAB plots displayed the history of the strain gage and the
yielding was to be found by inspecting the plots and recording
under what motion the yielding occurred. A special script was used
when two strain gages were in the same location to obtain the
average yield to be used in the strain gage sequence. An example of
the MATLAB plot using the average can be seen below in Figure 12.
The top feedback response in blue is data from one strain gage and
the bottom feedback response is from the other strain gage at the
same location. The green feedback is the average of those two and
is what is used to determine when the sensor yields. For locations
with only one sensor a single feedback
wasusedtofindtheyield.Yieldingcanbeidentifiedintheplotwhenthefeedback
response, green for average or a blue for single feedback, crosses
either the top or bottom
redparalleldottedlines.Theaimofthistaskwastoidentifythemotionwhenthe
yieldingoccurredandtorecordthemotioninExcelforthelastfivemotionsofthe
10 second specimen and the last eight motions for the first
specimen. The first yield was
recordedandassignedaspecificcolor,thefirstyieldingiscircledinFigure12.This
inspection was done more than 300 times to create a yielding
sequence. From the two Excel files, the location of the strain gage
was marked on the specimen plans and color-coded depending on when
it yielded. The result was a plan with color coded strain gages
representing the yield sequence as shown in Figure 13. Figure 12:
Strain gage data using average yield with MATLAB. First yielding is
circled 11 Figure 13: South view of wall elevation used to display
strain gage yield sequence. The colors represent the time at which
the yielding occurred. 4.5 MATLAB Video and Data Synchronization
Following the collection of videos from the GoPro cameras, the
videos and the data was synchronized using MATLAB to run two
different GoPro views and display the data at the same time. Three
plots of base shear vs. drift ratio, drift ratio vs. time, and
input acceleration vs. time are displayed simultaneously with the
videos. The MATLAB script was changed numerous times to make sure
the data being displayed correlated with the
videosbeingplayed.TherewerefourMATLABscriptsintotal,twoforthefirst
specimenandtwoforthesecondspecimen.Thesevideoswereuploadedandare
availableattheNEES@UCSDwebsite(http://nees.ucsd.edu/projects/2014-partial-grouted-masonary)
12 Figure 14: MATLAB video and data synchronization 4.6 Proposed
Shear Friction Formula Problems The TMS 402-13 standard does not
have provisions to calculate the shear-friction strength of
reinforced masonry walls. Dr. Benson Shing is working on a proposed
formula for the nominal shear strength Vnfin a future edition of
MSJ C.Data for walls that showed significant sliding was provided.
The experimental shear strength was compared with a value
calculated using the proposed formula and tabulated in Table 1.
Shing 6 and Shing 8 are walls tested at UC San Diego while UT
PBS-03 and UT PBS-04 are cantilever walls tested at the University
of Texas, Austin. The formula used to find these theoretical values
will soon be implemented and available in the 2016 MSJ C. This will
aid in the design and construction masonry structures. Table 1:
Shear Friction Test Results and Proposed Shear Strength Wall
SampleExperimental Vnf, kipsCalculated Vnf, kipsPercent Error Shing
650535.66 Shing 849549.26 UT-PBS-0382853.53 UT-PBS-044747.30.6
Results The first specimen failed in a non-ductile manner with
diagonal shear cracks and horizontal sliding. Figures 15 and 16
display the first specimens extensive damage and concrete masonry
units can be seen on the floor. It is evident that the common East
Coast practice needs improvement.The enhancements in the second
specimen proved to be effective and this can be seen in Figures 17
and 18. The same damage that is seen in the 13 first specimen is
not visible after the testing of the second specimen. The testing
of these structures has been a great step forward in the field of
masonry construction. With a few enhancements the seismic
performance of partially grouted reinforced masonry structures has
really improved. Figure 15: Specimen 1 after testing Figure 16:
Specimen 1 after testing, Side view 14 Figure 17: Specimen 2 after
testing, Front view Figure 18: Specimen 2 after testing, Side view
15 Conclusions The experiment shows that Specimen 1 failed in a
non-ductile manner with diagonal shear cracks as well as horizontal
sliding. Sliding was observed at the base before the occurrence of
cracks on the superstructure and to try to prevent this, heavily
reinforced concrete stoppers were incorporated at the end of the
base of the main wall in specimen 1. Since sliding was observed in
Specimen 1 the base was roughened for the Specimen 2. It was
observed that the grouted cells did not have sufficient width to
prevent shear cracking. Specimen 1, which reflects common practice
outside of the West Coast, performed poorly. Specimen 2 involved
improved design details which attempted to enhance the ductility of
the structure. The bond beams proved to be effective in Specimen 1
and were retained in the Specimen 2. The placement of the joint
reinforcement weakened the mortar bed joints, resulting in early
cracking along the joints. The increased cross-section of the
grouted cells prevented shear failure of the grout and led to a
more flexural dominant behavior. Following the objective of the
project, which was to enhance the seismic performance of partially
grouted reinforced masonry buildings, the applied enhancements
which included the double grouted vertical cells, bond beams, and
ladder type joint reinforcement, proved to assist Specimen 2 to
perform better than Specimen 1. This was the first time a full
scale partially grouted structure was tested on a shake table. This
project will assist in the understanding of the seismic performance
of this type of structure while also providing an enhanced
alternative to this type of construction. The system level
performance was successfully analyzed and will serve as a great
contribution to the masonry structure community. Acknowledgements
This research is supported by the National Science Foundation under
the Network for Earthquake Engineering Simulation program with
Award No. CMMI-1208208. NSF REU grant EEC-1263155 and the NEES
grant CMMI-092718. Any opinions, findings, and conclusions or
recommendations expressed in this material are those of the
author(s) and do not necessarily reflect the views of the National
Science Foundation. Thank you to the principal investigator Dr.
Benson Shing for his support throughout the program and for
allowing me to be part of this project. Thank you to Andreas
Koutras for sharing his knowledge and for his extensive support
throughout this project. Thank you for providing figures for this
report. Undergraduate student assisted in post data analysis
towards the end of the project. 16 Further Information For further
information on this projectREU student Erik Avila:
[email protected] Graduate Student Advisor- Andreas Koutras:
[email protected] PI Dr. P. Benson Shing: [email protected]
References American Society of Civil Engineers. (2010). ASCE 7-10,
Minimum Design Loads for Buildings and Other Structures, Reston,
VA, 2010.Hess, R. L. (2008). The ShakeOut Scenario Supplemental
Study: Unreinforced Masonry (URM) Buildings, SPA Risk LLC. Denver,
CO. . Kasparik, T., Tait, M., and El-Dakhakhni, W. (2014). Seismic
Performance Assessment of Partially Grouted, Nominally Reinforced
Concrete-Masonry Structural Walls Using Shake Table Testing. J.
Perform. Constr. Facil., 28(2), 216227. Minaie, E., Mota, M., Moon,
F., and Hamid, A. (2010). In-Plane Behavior of Partially Grouted
Reinforced Concrete Masonry Shear Walls. J. Struct. Eng., 136(9),
10891097. MSJ C (Masonry Standards J oint Committee ) (2011), TMS
402-11/ACI-530-11/ASCE 5-11: Building Code Requirements for Masonry
Structures, The Masonry Society, American Concrete Institute, and
ASCE/Structural Engineering Institute.Schultz, A.E. (1996) Seismic
Performance of Partially-Grouted Masonry Shear Walls. Eleventh
World Conference on Earthquake Engineering, Acapulco, Mexico,
246-256. Shing, P., Ahmad, H., Moon, F., Shultz, A., and Koutras,
A. (2013) Improving Seismic Performance of Partially-Grouted
Reinforced Masonry Buildings. 12th Canadian Masonry Symposium,
Vancouver, British Columbia. Shing, P., Schuller, M., and Hoskere,
V. (1990). In-Plane Resistance of Reinforced Masonry Shear Walls.
J. Struct. Eng., 116(3), 619640. University of California, San
Diego. (2013). Shake Table Specifications. NEES@UCSD, . (J uly 28,
2014).