Final Project Report Avila (2014)

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).