BSSC Issue Team 4, Meeting 2, November 2 and 3, 2016
Meeting Notes
November 2
Attending: Jason Collins, Joe Ferzli, Dave Fields, S. K. Ghosh, Gino Kurama (2nd half), Dawn Lehman (1st half), Laura Lowes, Andy Taylor, Tom Xia
Remote: John Wallace (in part), JQ Yuan
S.K. Ghosh convened the meeting at 8:30 AM.
Today’s meeting will be mainly on the topic of concrete shear walls; tomorrow’s meeting will cover other types of shear walls
Today’s agenda:
1. Coupled shear walls – Dave Fields’ work. Ferzli commented that the definition of a coupled shear wall needs to include wall dimension along the span of the coupling beam.
2. ASCE 7-16 Table 12.2-1.
a. Is a FEMA P695 analysis required to create a wall type called coupled shear walls? Joe Maffei and Dawn Lehman are writing a justification for why a P695 analysis is not required.
b. Shear design of shear walls. Jack Moehle has assembled a database of 23 existing high-rise building. This database clearly demonstrates that design shears based on ELF are substantially less than actual shears calculated from NLRHA. Later, Ghosh will show Moehle’s slides, which will be attached to these notes.
c. John Wallace has a proposal to ACI 318 Sub-H for addressing shear amplification. Wallace also sent four papers for IT4 to consider.
d. Lehman and Lowes also have work to share
3. Detailing – we should examine detailing and decide if we need to make recommendations
Ferzli asked about the timeline for our work. Ghosh noted that the schedule was in the Meeting 1 summary. Ghosh reviewed that schedule. We must aim for completion of the bulk of our work by Fall of 2018.
Ghosh said that the work of Moehle and Wallace’s proposal are focused on tall, shear-wall-core, PBSD-designed buildings. Wallace disagreed about his proposal to Sub-H and said that it will address all types of shear walls.
Lowes asked: Why not use the degree of coupling as a definition of coupled walls?
Ferzli believes we need to use more than beam aspect ratio in the definition.
Ghosh reviewed the shear wall discussions from the Meeting 1 notes.
Fields gave a presentation, and provided a copy for attachment to these notes. The bottom line from the presentation was that coupled shear walls are defined as those walls with coupling beam aspect ratios greater than or equal to 3. Fields said this seems like a reasonable definition given the data he presented.
Ferzli believes beam type should play a role: diagonal reinforcement, moment frame reinforcement, fiber beam, steel beam, precast. He also suggested not using a single value of 3 as this creates a step function. He suggested a gradual transition.
Lehman was concerned about drift demands. She has observed compression failures at wall bases.
Xia agreed – we need an upper limit on aspect ratio of the coupling beams. This prevents the formation of very weakly coupled walls.
Lehman and Fields: were in favor of the Los Angeles approach for PBSD design of walls: design for service level and MCE events only – no code-based design.
Fields said that in the latest version of his presentation he provided additional information on modeling parameters. The definition he proposes is intended to provide a basis for designation of a new R-factor. There may be ways to improve the definition:
· How many levels of coupling are required?
· What about mixed aspect ratios?
· What about the aspect ratios of the walls that are being coupled?
He noted that for buildings greater than about 400 feet in height, wind may begin to govern shear wall design.
Ghosh encouraged all attendees to write down the thoughts they are sharing during the meeting. These are very valuable to others who may not be familiar with design of high-rise shear wall buildings. Ghosh asked fields if he would include degrees of coupling (according to the Canadian definition) for the walls in his data set.
Ghosh asked a general question: what differences are expected to be observed for types of coupling beams other than cast-in-place concrete? Lehman, Fields, and Ferzli said that they felt there would be very little difference in observed behavior.
Ferzli gave a presentation, a copy of which will be attached to these notes. The two main conclusions of his presentation were that
· Coupled walls perform very well
· Span/depth ratio is a good start for a definition, but we also need to include consideration of the stiffness of the walls that are being coupled.
Ghosh asked: What are the implications for the coupled wall definition concerning the following points:
· Geometry (e.g. length) of the coupled walls
· Should the definition contain an upper bound on aspect ratio?
· What about coupled walls with coupling beams that have mixed aspect ratios?
· How should the definition be related to the degree of wall coupling?
Lehman suggested also examining isolated walls as a point of comparison.
Ghosh said that Fields’ efforts will lead up to a definition being agreed to by ACI 318 Subcommittee H. Ghosh believes that symmetry of walls may also be a factor in the definition.
Lehman said that asymmetric walls, such as single T’s, C’s bending in the weak direction,
do not perform well.
There was a discussion of asymmetric walls. It was generally agreed that symmetric walls are preferable. Should symmetry be part of the definition of coupled shear walls? Some members felt possibly not.
Ghosh discussed Fields’ work. Will it be written up for publication? Fields said there are currently no plans to publish it. Ghosh felt it is important to document this work for the purpose of using it to justify a code change. He suggested writing the study up as a report, as if it is for a client. This could probably be published in Part 3 of the NEHRP Provisions. Ghosh also requested a summary of typical conditions actually encountered in shear wall design in practice. Finally, he noted, and it was agreed, that all code changes related directly to concrete should go through ACI 318.
Taylor (Chair of ACI 318 Sub-H) said that he is planning on balloting the simple definition of a coupled shear wall, presented today by Fields, in Sub-H, possibly by the end of November. This will have the benefit of eliciting comments from Sub-H members.
Xia noted that IT4 will likely need to work with IT2, chaired by Sandy Hoehner, which is responsible for the R-factor Table.
Lowes gave a presentation “Coupled Walls”, a copy of which will be attached to these notes.
One of the main points of the presentation was that the compression pier eventually ends up resisting the entire shear imposed on a coupled shear wall system.
In response to the presentation, Fields said that typical behavior observed in design is relatively low drifts. The behavior he has highlighted in his study is for drifts in the 2.0 percent range. It would be interesting to know the distribution of shear between tension and compression piers at 0.5, 1.0, and 2.0 percent drift.
Xia noted that the walls discussed by Lowes are planar walls. He wondered what differences would be observed with flanged walls, such as core walls.
Fields said that shorter buildings are probably not usually governed by drift, but by shear wall strength. Taylor agreed.
Xia said that coupled walls are usually not found in low-rise buildings with planar walls. Ferzli did not completely agree. There are certain types of buildings like mid-rise hotels, where coupled planar shear walls may be used.
______
Lunch Break
______
Jack Moehle had made a presentation on wall shear and column shear at the recent 318-H meeting in Philadelphia. A copy of the presentation is attached to these notes.
Ghosh presented a summary of this presentation, using Jack’s slides.
With regard to columns, Fields wondered what axial load should be considered. From analysis? That is too low. From plastic hinging of beams? If so, how many beams yield? Sub-H should take this up.
Regarding walls: Moehle splits shear amplification into “dynamic amplification” and “overstrength amplification.” Neither has a clear relation to building height.
Lowes’ presentation on wall amplification factors references a 2011 paper by Rutenberg. Lowes will supply a copy of this paper. Rutenberg developed dynamic amplification factors including the influence of both 1st and 2nd modes, depending on which dominates behavior. This improved correlation with analytical studies.
Ghosh said this result tells us much about blade shear walls. How does this relate to coupled blade shear walls? Are any modifications required?
Wallace joined the meeting remotely to make a presentation. He showed ACI 318 Sub-H ballot CH19-009, which has been revised for a 2nd ballot. Wallace presented background materials. He is working on a code change proposal that would be an extension of the CH19-009 ballot. Ghosh requested that Wallace provide a copy of the paper by Rejec, Isakovic, and Fischinger so that it can be distributed to the committee.
Ghosh wondered if we need to look more closely at the capacity side of the equation. Short term, peak, dynamic loading may result in higher shear strength.
Taylor asked what previous research on rate effects are available. What results would be applicable to the question at hand? Wallace thought there is a need to replicate dynamic shear that is simultaneous with yielding in flexure. He knew of no existing studies that address this combination of factors.
Kurama asked if peak moment demands and shear demands occur simultaneously.
Lowes said that analytical work in Canada with continuum models studying high rate loading with simultaneous shear and moment demands has been conducted. It is not known if the results are applicable to the current discussion.
Xia said that the Canadian code already considers dynamic amplification. Are Canada’s capacities about the same as U.S. capacities? Ghosh said yes, the capacities are comparable.
Wallace asked if we are considering instantaneous shear, how should that be done? He prefers being conservative on the wall capacity side of the equation, i.e., don’t consider possible increase in strength due to short-term shear.
Xia asked if there have been problems with walls in past earthquakes?
Wallace said that we have not yet had a very good test, i.e., a strong earthquake, affecting modern shear wall structures. Ghosh agreed.
Wallace gave a second presentation, a copy of which will be attached to these notes. This was on reliability based design for wall shears in buildings. The essential question asked in this presentation is “What is the reliability of the current shear wall design criterion?”
Ferzli asked Wallace about shear shift to the compression pier. Wallace said that multiple testing programs have shown this shift. Wallace is concerned about biaxial loading on a core wall – what will happen at the corners?
Ghosh asked all presenters to send copies of their presentations.
______
Break
______
There was further discussion after the break
Kurama presented information on a study of high strength concrete and high strength reinforcement in shear walls. He looked exclusively at shear capacity. He concluded that VecTor2 analytical models were the best predictors of observed behavior. Gulec and Whittaker provided the best empirical closed form predictive method, with reduced scatter compared to other methods. ACI 318-14 overestimates lateral shear strength for walls without boundary regions but is conservative for wall with boundary regions.
Ghosh asked for Wallace’s thoughts on flexural vs. shear walls? Should we divide walls into these two types?
Wallace responded that he believes there are actually three classes of behavior: shear governed, transition, and flexure governed.
Ghosh asked how do we define the separation between these types. Perhaps based on shear demand? Based on height/length ratio and shear strength?
Wallace said there is definitely a transition region between shear governed and flexure governed walls. It is difficult to test in the transition region. Wallace showed a plot that was an attempt to classify walls based on curvature ductility.
Lowes showed several slides. There are three factors affecting drift capacity:
· Shear stress demand
· Cross section aspect ratio
· Axial load
All three of these factors seem inter-related. Lowes used ATENA for analysis of three tested walls.
Fields said that we would prefer ductile walls that are classified distinctly from non-ductile walls. PBSD is trying to provide ductile walls. Currently the walls themselves remain nearly elastic through the MCE. The only nonlinearity is in the coupling beams. Shear failure is protected against through design. He supports Lowes/Lehman proposal for extending the boundary element depth.
Lowes said there is very little inelastic flexural behavior in core shear wall systems, but in isolated walls she has seen problems with flexure.
Ghosh said this could be a height effect, with higher modes changing the force distribution in walls, possibly reducing flexural demand. If this is not the case we need to address the different behaviors of isolated walls and core walls.
Wallace said that creating a core wall system with sufficient shear strength creates higher stiffness and flexural capacity, leading to reduced tendency for flexural yielding. This is why flexural behavior is nearly elastic.
Fields agreed that we design for very high shear capacity, which indirectly leads to nearly elastic behavior in flexure.
Wallace said that a major challenge is predicting strength loss.