DIAPHRAGM SEISMIC DESIGN METHODOLOGY

UA – LU – UCSD

DSDM Task Group First Meeting – Detailed Minutes

August 7, 2003

PCI Headquarters, Chicago, IL

Participants:

Roger Becker (Spancrete)Paul Johal (PCI)

Ned Cleland (Blue Ridge Design)Clay Naito (Lehigh University)

Tom D’Arcy (PCI)Suzanne Nakaki (Nakaki Bashaw Group)

Robert Fleischman (U. Arizona)Jose Restrepo (UCSD)

S.K. Ghosh (S. K. Ghosh Associates)Mario Rodriguez (UNMR)

Neil Hawkins (UIUC)Richard Sause (Lehigh University)

Phillip Iverson (PCI)Douglas Sutton (Purdue University)

Comments via email: Joe Maffei (Rutherford-Chekene)

1. Introductory Remarks

A. Discussion of Original RFP (D’Arcy)

Objective: to develop an industry endorsed design methodology for precast diaphragms:

  • the forces and displacements for which the diaphragm should be designed, including chord and collector units and connections;
  • the connections and details that can provide this performance; and
  • guidance on the relative stiffness of the diaphragm to the lateral force resisting system.

Scope: Floor diaphragms in precast concrete construction including

  • Topped and Untopped Construction
  • Hollow Core and Double Tee construction

B. Background on need for roughness guidelines (Becker)

ACI 318G attempted to clarify the surface roughness required to achieve the shear stress in the ACI Code. On balloting, 318H (Seismic) raised issues – a combination of the provisions in ACI Ch. 11, 17 and 21 was proposed but no research has been performed. The ACI approach allows either noncomposite or composite but for the latter the process required to develop reliable bond is not established – opinions vary from the need for no special preparation to provide ¼” roughness and horizontal shear reinforcement.

Group Consensus

  • Goal of program should be to answer the basic questions… What force do we design the diaphragm for and how do you design the diaphragm?
  • The portion of research pertaining to topped systems should focus on composite (rather than non-composite) systems. There is no need to directly investigate systems associated with typical CA construction techniques (i.e., cast-in-place topping w/out mechanical connectors).
  • The roughness required for reliable composite action between topping and precast elements, and the techniques needed to attain this roughness should be key components of the research.
  • Information on required roughness can be addressed either in the Lehigh baseline tests or in the UCSD quasi-static tests. (Naito)
  • The research should also have as a specific goals to describe the demand on the connections in these composite topped systems.
  • Pilot tests with and without topping – topping alone, connection alone, and then topping and connection together are likely required (Fleischman).

2. Project Overview (Fleischman)

Overall Integration and Flow of Research Activities was presented. The program involves significant individual efforts and significant integration. A concern raised by several DSDM members was that in attempting to perform the research, focus on the end product - procedures that can be used by precasters and designers - may become diverted. The research team believes that the proposed research approach is needed for meaningful findings but is keenly aware of the task group’s concern.

Group Consensus

  • The research team will develop mechanisms by which the Task Group can regularly evaluate the research direction.
  • The Task Group is expected to take a leadership role in keeping the research focused on the applied outcome.

Proposed Risk Mitigation Technique– At each DSDM research meeting, the Task Group will be provided with updates of work at each individual research sites and the integration of these research components, in order to evaluate the current direction and overall vision of the research. Meeting time will then be allotted for the Task Group to advise the research team on where research scope or direction requires modification; or where the research team’s efforts need to be refocused. In later research meetings, technology transfer/code development mechanisms will be developed.

3. Overall Design Approach (Fleischman)

A. Role of Elastic Diaphragm Design and the need for Ductile Detailing

Various overstrength factors have been proposed: up to 4.5 (); the team proposes targeting multiple seismic levels; elastic behavior at the design level seismic event seems tenable. For a severe seismic event, research indicates an inverse relationship between force level (overstrength) and possible ductility demand. It was clarified that the need for diaphragm ductility is not for energy dissipation purposes; rather the ability to accommodate plastic deformation without loss of load carrying capacity. Questions were posed to the task group as to their interpretation of acceptable design force level and acceptable damage.

Group Consensus

  • The desired research approach should be to determine the required overstrength factors to meet performance requirements established by the group.
  • Elastic Behavior at the design level EQ is an appropriate basic goal of the research.
  • For the MCE, avoidance of loss of life should be the goal – i.e. maintain gravity carrying integrity of diaphragm; any level of local damage (losing a few connectors, etc.) is acceptable provided the structure can remain stable. It is also not necessary to address the issue of repair since the lateral system will be heavily damaged.
  • An alternative to the term ductility may be required to avoid confusion with detailing for energy dissipation (Iverson).

B. Web Reinforcement Design Approach

Designs have promoted tension compliance in the web reinforcement. Tension stiffness combined with tension ductility was proposed (Fleischman) to allow for a more monolithic diaphragm and to lower amount of chord steel. This approach was not viewed favorably due to issues of serviceability (shrinkage, temperature) and the need to transfer high tension forces outside of the chords.

Diaphragm  factors were discussed. A capacity design approach was proposed in which collectors are designed with an overstrength, and the chords are designed with a higher  factor than the shear reinforcement to promote ductile modes (Fleischman). It was indicated, however, that chords in one direction are collectors in the other, raising difficulties with using the approach in all but direction-preferential cases (Nakaki).

It was pointed out that existing code factors are "judgment" based and may not fit well with a rational diaphragm analysis. Therefore, it was suggested that the project produce technically sound design criteria that can be rationally applied to code-specified phi and omega factors. A phi factor of unity was proposed with the research simply determining appropriate overstrength factors. (Maffei).

Group Consensus

  • Web reinforcement should be designed with tension compliance.

C. Applicability of Construction Methods

The Task Group reiterated their position on pursuing topped composite (rather than noncomposite). The Task Group also expressed a desire to extend the applicability of untopped diaphragms into high seismic zones. The untopped research should examine systems with pour strips (particularly for high seismic) rather than fully dry systems; very little is known about the performance of continuous connectors (Cleland). It was noted that while a fairly clear understanding exists as to the nature of requirements in high and low seismic zones, there is not a clear notion of the level of detailing required for moderate zones (Cleland).

Group Consensus

  • Develop design procedures for untopped diaphragms in high seismic zones if feasible.
  • Research scope is to provide methodologies suitable for the entire country.
D. Diaphragm classifications for different seismic zones

A framework for the design methodology was proposed by the research team that incorporates multiple diaphragm classifications through choices of different overstrength factors, different ductility detaling requirements, and diaphragm span limitations. Three classifications were proposed:

  • Ordinary – (Non seismic) Elastic demands on diaphragm connections (A, B)
  • Intermediate – Low levels of inelastic demands on diaphragm connections (C)
  • Special – Maintain integrity over large deformation demands to prevent loss of gravity carrying capacity (D,E)
  • A discussion arose on how many diaphragm design classifications are necessary. Consensus was not reached; some Task Group members thought two classifications are adequate; others thought three are necessary.
  • A concern was raised that using the terms Ordinary, Intermediate, and Special is that it will lead to confusion among code writing bodies. Diaphragm classification terms “Elastic Design” and “Limited Ductility” were suggested (Maffei).
  • It was suggested that the designer have a choice in type of diaphragm design that does not have any arbitrary restrictions pertaining to seismic zone (Maffei). This approach will allow designer expertise to determine whether it’s more economical or appropriate to design for higher forces. Also, a need was expressed that the ductile design procedures must be straightforward to limit the amount of additional engineering effort, thereby avoiding use of the elastic design solely because it is simpler to employ.

The use of a ductility factor was proposed to relate local ductility of connectors to global diaphragm displacement ductility capacity (Maffei). The research team plans to make this relationship transparent to the designer, and calibrated based on the tested displacement capacity of the connections determined in the research and the specific diaphragm configuration as suggested.

Group Consensus

  • A standard should be developed that serves the needs of both seismic and non-seismic design practices. Thus, the methodology developed should be representative of both elastic and ductile design philosophies. Both techniques are being used in practice, and favoring one may isolate a large group of designers.

4. Design Philosophies (Restrepo)

(A)Alternative Detailing Philosophies

The simple beam analogy used in current design may not be the most effective technique nor does it seem to be applicable for irregular floor plans. The question was also raised as to whether the boundary element approaches place too much ductility demand on the diaphragm. Alternative detailing philosophies were presented and discussed:

  • Strut and tie; stringer-panel: The stringer panel method, used in the past for the design of wooden airplane wings, is attractive in that it eliminates force concentrations at nodes. It was mentioned that those successful designs depended on out-of-plane flexing of the panels. However, it was thought that in combination, the strut and tie and stringer panel methods could produce a powerful approach (Hawkins).
  • A capacity design method was presented that promoted strong joints and allowed distributed small cracks along the topping above the panel. However, it was indicated that durability of the system is the overriding factor, to the point that the volume change of the diaphragm associated with temperature and shrinkage is typically accommodated by pre-cracking and sealing joints. Therefore, a crack width of even three hundredths is viewed as unacceptable (D’Arcy). Shifting the failure to mid panel could lead to cracking of the precast element which will compromise the resistance to weathering.

Group Consensus:

  • Explore the use of alternative design approaches for irregular floor plans.
  • It is imperative to limit cracking at the panel joints to preserve system durability as a major selling point of precast construction.
  • It was suggested that precast could be promoted as a system possessing superior durability and designed with a more rational approach (Iverson).

(B)Irregular Floor Plans

The design procedure must be applicable to all conceivable diaphragm layouts, including diaphragm openings, non-rectangular diaphragms, and multispan continuous diaphragms. It may be preferable to not use the term chord at all, and instead promote evaluation of shear and moment strength at critical sections. Further, computer modelling of diaphragms is much more common now, and with powerful software available, the project should develop computer modeling guidelines as part of the proposed design requirements. Development of illustrative examples would also be useful. (Maffei)

(C)Hollow core

Hollow core design methods were discussed including the profiled Italian system by Menegotto. A question arose regarding under what conditions the serrated profile is required (Becker). It was pointed out that Menegotto used undereinforced chords, and the research seemed to verify a preconceived notion. A topped diaphragm design that could serve as an effective solution for Hollow Core is a draped strand method in which prestressing strands at lower stress are placed to follow the contours of the stress trajectory. It was pointed out that this design need not be only applicable to HC, but could be used as a generic approach (Becker).

Group Consensus:

  • Parallel experiments to those in Italy may be needed (Hawkins).
  • Include the reinforcement crossing the panel at its ends when testing hollow core joints. Bars in the keyway (Raytech flatbar bend – Gleich?) would be required.

5. Shake Table Diaphragm Studies in Mexico (Rodriguez)

The results of the shake table research being conducted on precast diaphragms at the University of Mexico was presented. One phase of testing has been completed. Preliminary results indicate that most of the system damage occurred at the boundary of the diaphragm. Additional research is underway to examine strut-and-tie designs for non-composite topped diaphragms. A desire exists to coordinate efforts with the DSDM research.

Group Consensus

  • Attempt to synchronize the research efforts to allow direct comparisons; keep in contact to avoid duplication. Note: We might consider holding a session in a future ASCE Structure congress (2005?) or WCEE jointly with researcher from Mexico, New Zealand, and Illinois.

6. Diaphragm Reinforcing Detail (Naito)

(A)Mechanical Connectors

An overview on the use and general behavior of panel-to-panel connectors was presented. Available experimental results were reviewed. Some historical perspective and clarification was provided: mechanical connections were developed out of a need that was not necessarily seismic; the High Concrete stud connector is not proprietary; and other effective connector details exist, e.g., continuous bar with end plates (Cleland). Extensive discussion surrounded the approach to proprietary vs. common mechanical connectors. The panel viewed the use of proprietary connections in the testing program with misgivings based on past difficulties experienced in the PRESSS program. It was the task group’s recommendation to:

  • Test standard connectors that can be readily detailed.
  • Examine non-proprietary connectors and determine acceptance level for different performance targets. Classify details.
  • If topped panel tests without connectors (“CA design”) are needed for model calibration, the researchers should make the reason for these tests clear to the construction community to avoid the false impression that this type of diaphragm detailing is being endorsed.
  • Develop standards and acceptance criteria that can be used to validate (pre-qualify existing and qualification testing procedures for new) connectors. A set of standard criteria, protocols and data presentation formats are required.

Group Consensus

  • Avoid testing proprietary connectors as part of the study. Determine an approach for selecting connector types and classifying behaviors such that tests can be conducted and prequalification criteria developed for connectors without becoming locked into a specific proprietary connector.

(B)Issues Related to Construction Details

The Task Group indicated that experimental evaluation of the panel joints must include the effects certain construction details and techniques to accurately reproduce the actual behavior.

Anchorage: The connection to the vertical system is an important part of the load path, including their performance during relative motion of components, and should be addressed in the research. (Nakaki).Connections to inverted tee beams and spandrels failed in the Nisqually earthquake outlined in a Metro Report (Hawkins).

Other construction related issues included:

  • Welded wire fabric: Explore the use of regular steel wire to make the mesh (Sutton)
  • Composite action: develop an understanding the effects of roughness on composite action of diaphragms.
  • Load path integrity: Assess the integrity of the diaphragm through a study of the full load path such as the inverted tee to panel connection, etc.

7. Physical Scope (Sause)

(1)Prototype structure

Due to the finite budget and aggressive time line, it is not an efficient use of resources for the research team to design prototype structures. Additionally, the task group members bring design expertise, historical perspective, and knowledge of markets and current practice to the process the university team does not possess. Existing examples from previous research or past design guides may not directly meet the needs of the DSDM objectives. It was therefore proposed that the panel deliver a portfolio of prototype structures. The portfolio should include wall systems, frame systems and encompass the range of floor plans, lateral system layouts, and system redundancy found in practice. The following comments were advanced:

  • Parking structures should be included
  • Makes up approximately 60% of the market (D’Arcy)
  • Typically consists of 3 – 60ft bays with ramp in the middle
  • Hollow core is rarely used in exterior or corrosive situations; hollow core is often put on stiffer structures; tees often on frame type structures (Iverson)
  • Examining stadium diaphragms was discussed. These structures were considered too specialized for the research and should not be included.

The candidates for the prototype structures that emerged included:

  • Parking Structure
  • Multi-Family Residential
  • Bearing wall (hotel/motel)
  • Office Building
  • Low rise structures such as a warehouse or prison type building

Group Consensus

  • Need industry to provide a portfolio of structures by end of September. The research team will solicit, rapidly screen, and select among these prototype structures for discussion with the wider project advisory board later in the year to for broad input.

(2)Code Issues