K1.1 Scope
Section K1 describes requirements for prequalification of beam-to-column connections in special moment frames (SMF), intermediate moment frames (IMF), composite special moment frames (C-SMF), and composite intermediate moment frames (C-IMF) and of link-to-column connections in eccentrically braced frames (EBF). The concept of prequalified beam-to-column connections for SMF and IMF, as used in the Provisions, was originally adopted from FEMA 350 (FEMA, 2000a) and was subsequently extended to include prequalified link-to-column connections for EBF. In the 2016 edition of the Provisions, the prequalification of beam-tocolumn connections was further extended to include C-SMF and C-IMF.
Following observations of moment connection damage in the 1994 Northridge earthquake, these Provisions adopted the philosophy that the performance of beam-to-column and link-to-column connections should be verified by realistic-scale cyclic testing. This philosophy is based on the view that the behavior of connections under severe cyclic loading, particularly in regard to the initiation and propagation of fracture, cannot be reliably predicted by analytical means alone. Consequently, the satisfactory performance of connections must be confirmed by laboratory testing conducted in accordance with Section K2. In order to meet this requirement, designers fundamentally have two options. The first option is to provide substantiating test data, either from project-specific tests or from tests reported in the literature, on connections matching project conditions within the limits specified in Section K2. The second option available to designers is to use a prequalified connection.
The option to use prequalified connections in the Provisions does not alter the fundamental view that the performance of beam-to-column and link-to-column connections should be confirmed by testing. However, it is recognized that requiring designers to provide substantiating test data for each new project is unnecessarily burdensome, particularly when the same connections are used on a repeated basis that have already received extensive testing, evaluation, and review.
It is the intent of the Provisions that designers be permitted to use prequalified connections without the need to present laboratory test data, as long as the connection design, detailing, and quality assurance measures conform to the limits and requirements of the prequalification. The use of prequalified connections is intended to simplify the design and design approval process by removing the burden on designers to present test data and by removing the burden on the authority having jurisdiction to review and interpret test data. The use of prequalified connections is not intended as a guarantee against damage to, or failure of, connections in major earthquakes. The engineer of record in responsible charge of the building, based upon an understanding of and familiarity with the connection performance, behavior, and limitations, is responsible for selecting appropriate connection types suited to the application and implementing designs, either directly or by delegated responsibility.
K1.2 General Requirements
K1.2a Basis for Prequalification
In general terms, a prequalified connection is one that has undergone sufficient testing, analysis, evaluation, and review so that a high level of confidence exists that the connection can fulfill the performance requirements specified in Section E3.6b for SMF, Section E2.6b for IMF, Section F3.6e for EBF, Section G3.6b for C-SMF, and Section G2.6b for C-IMF. Prequalification should be based primarily on laboratory test data but supported by analytical studies of connection performance and by the development of detailed design criteria and design procedures. The behavior and expected performance of a prequalified connection should be well understood and predictable. Further, a sufficient body of test data should be available to ensure that a prequalified connection will perform as intended on a consistent and reliable basis.
Further guidance on prequalification of connections is provided by the commentary for FEMA 350, which indicates that the following four criteria should be satisfied for a prequalified connection:
- (1) There is sufficient experimental and analytical data on the connection performance to establish the likely yield mechanisms and failure modes for the connection.
- (2) Rational models for predicting the resistance associated with each mechanism and failure mode have been developed.
- (3) Given the material properties and geometry of the connection, a rational procedure can be used to estimate which mode and mechanism controls the behavior and deformation capacity (i.e., story drift angle) that can be attained for the controlling conditions.
- (4) Given the models and procedures, the existing database is adequate to permit assessment of the statistical reliability of the connection.
K1.2b Authority for Prequalification
While the general basis for prequalification is outlined in Section K1.2a, it is not possible to provide highly detailed and specific criteria for prequalification, considering the wide variety of possible connection configurations and considering the continually changing state-of-the-art in the understanding of connection performance. It is also recognized that decisions on whether or not a particular connection should be prequalified, and decisions on establishing limits on prequalification, will ultimately entail a considerable degree of professional engineering judgment. Consequently, a fundamental premise of these Provisions is that prequalification can only be established based on an evaluation of the connection by a panel of knowledgeable individuals. Thus, the Provisions call for the establishment of a connection prequalification review panel (CPRP). Such a panel should consist of individuals with a high degree of experience, knowledge, and expertise in connection behavior, design, and construction. It is the responsibility of the CPRP to review all available data on a connection and then determine if the connection warrants prequalification and determine the associated limits of prequalification, in accordance with Section K1. It is the intent of the Provisions that only a single, nationally recognized CPRP be established. To that end, AISC established the AISC CPRP and developed ANSI/AISC 358, Prequalified Connections for Special and Intermediate Steel Moment Frames for Seismic Applications (AISC, 2022b).
Use of connections reviewed by connection review panels other than the AISC CPRP, as permitted in Section K1.2b, and determined suitable for prequalification status in accordance with the Provisions, is subject to approval of the authority having jurisdiction.
K1.3 Testing Requirements
It is the intent of the Provisions that laboratory test data form the primary basis of prequalification and that the connection testing conforms to the requirements of Section K2. FEMA 350 specifies the minimum number of tests on nonidentical specimens needed to establish prequalification of a connection or, subsequently, to change the limits of prequalification. However, in the Provisions, the number of tests needed to support prequalification or to support changes in prequalification limits is not specified. The number of tests and range of testing variables needed to support prequalification decisions will be highly dependent on the particular features of the connection and on the availability of other supporting data. Consequently, this section requires that the CPRP determine whether the number and type of tests conducted on a connection are sufficient to warrant prequalification or to warrant a change in prequalification limits. Both FEMA 350 and the Provisions refer to “nonidentical” test specimens, indicating that a broad range of variables potentially affecting connection performance should be investigated in a prequalification test program. It may also be desirable to test replicas of nominally identical specimens in order to investigate repeatability of performance prior to and after failure and to demonstrate consistency of the failure mechanism. Individuals planning a test program to support prequalification of a connection are encouraged to consult with the CPRP, in advance, for a preliminary assessment of the planned testing program.
Tests used to support prequalification are required to comply with Section K2. That section requires test specimens be loaded at least to a story drift angle as specified in Sections E3.6b and G3.6b for SMF and C-SMF and Sections E2.6b and G2.6b for IMF and C-IMF or a link rotation angle as specified in Section F3.4a for EBF. These provisions do not include the additional requirement for connection rotation capacity at failure, as recommended in FEMA 350. For purposes of prequalification, however, it is desirable to load specimens to larger deformation levels in order to reveal the ultimate controlling failure modes. Prequalification of a connection requires a clear understanding of the controlling failure modes for a connection; in other words, the failure modes that control the strength and deformation capacity of the connection. Consequently, test data must be available to support connection behavior models over the full range of loading, from the initial elastic response to the inelastic range of behavior and finally through to the ultimate failure of the connection.
The story drift angle developed by a moment connection test specimen is the primary acceptance criterion for a beam-to-column moment connection in a moment frame. In an actual building, the story drift angle is determined as the story displacement divided by the story height and includes both elastic and inelastic components of deformation. For a test specimen, story drift angle can usually be determined in a straightforward manner from displacement measurements on the test specimen. Guidelines for determining the story drift angle of a connection test specimen are provided by SAC (1997).
When a connection is being considered for prequalification by the CPRP, all test data for that connection must be available for review by the CPRP. This includes data on unsuccessful tests of connections that represent or are otherwise relevant to the final connection. Testing performed on a preliminary connection configuration that is not relevant to the final design need not be submitted. However, parametric studies on weak and strong panel zones of a connection that otherwise match the final connection are examples of developmental tests that should be submitted. Individuals seeking prequalification of a connection are obliged to present the entire known database of tests for the connection. Such data are essential for an assessment of the reliability of a connection. Note that unsuccessful tests do not necessarily preclude prequalification, particularly if the reasons for unsuccessful performance have been identified and addressed in the connection design procedures. For example, if 10 tests are conducted on varying sized members and one test is unsuccessful, the cause for the “failure” should be determined. If possible, the connection design procedure should be adjusted in such a way to preclude the failure and not invalidate the other nine tests. Subsequent tests should then be performed to validate the final proposed design procedure.
Most moment connection tests are quasi-static, meaning that the specimens are subjected to a slowly applied displacement protocol. Testing that applies real-time earthquake loading to full-scale moment connections are challenging because (a) real-time testing requires specialized testing facilities that are not common, (b) quasi-static testing allows mid-test evaluation of behavior and associated adjustments, and (c) buildings can have a wide range of periods, so a considerable number of tests might be required to capture the full range of strain rates. According to the categorization in Barsom and Rolfe (1999), earthquake loading is considered an intermediate loading rate as compared to dynamic loading rates typically associated with blast or slow loading rates, which are representative of quasi-static testing of connections and coupons.
Tests conducted on steel material at intermediate loading rates show increased yield stress (approximately 5 to 10%) and some reduction in ductility and toughness (Bruneau et al., 2011; Barsom and Rolfe, 1987). While toughness and transition temperature has been shown to typically reduce with intermediate loading rates as compared to slow loading rates (Barsom and Rolfe, 1987; Kaufmann and Fisher, 1995), in less constrained plane stress conditions, toughness may be increased with intermediate loading rates. Furthermore, for well-detailed connections, yielding leads to a rise in temperature that causes increased toughness that may offset strain-rate related reduction.
Tests conducted at real-time earthquake loading rates have been compared to tests conducted at quasi-static loading rates, such as Uang and Bondad (1996), Nakashima et al. (1998), Suita et al. (1998), Dusicka et al. (2007), Lamarche and Tremblay (2011), and Tremblay et al. (1997). Tests on pre-Northridge connections with lower toughness weld material, which experienced weld fracture without significant yielding (or associated temperature rise), suggested that specimens were more prone to fracture with earthquake strain rates (Uang and Bondad, 1996). Conversely, tests on well-detailed connections with yielding showed that earthquake loading rates do not adversely affect ductility (Tremblay et al., 1997) and, instead, that rises in temperature can cause increased ductility and an increase in moment strength of the entire connection (Nakashima et al., 1998; Suita et al., 1998). The effect of loading rate was evaluated both analytically and experimentally during the FEMA/SAC project (FEMA, 2000e) in response to the 1994 Northridge earthquake, and it was determined that real-time testing was not necessary to understand the behavior of moment connections that follow post-Northridge seismic provisions.
K1.4 Prequalification Variables
This section provides a list of variables that can affect connection performance and that should be considered in the prequalification of connections. The CPRP should consider the possible effects of each variable on connection performance and establish limits of application for each variable. Laboratory tests or analytical studies investigating the full range of all variables listed in this section are not required and would not be practical. Connection testing and/or analytical studies investigating the effects of these variables are only required where deemed necessary by the CPRP. However, regardless of which variables are explicitly considered in testing or analytical studies, the CPRP should still consider the possible effects of all variables listed in this section and assign appropriate limits.
K1.5 Design Procedure
In order to prequalify a connection, a detailed and comprehensive design procedure consistent with the test results and addressing all pertinent limit states must be available for the connection. This design procedure must be included as part of the prequalification record, as required in Section K1.6. Examples of the format and typical content of such design procedures can be found in FEMA 350.
K1.6 Prequalification Record
A written prequalification record is required for a prequalified connection. As a minimum, the prequalification record must include the information listed in this section. The prequalification record should provide a comprehensive listing of all information needed by a designer to determine the applicability and limitations of the connection and information needed to design the connection. The prequalification record need not include detailed records of laboratory tests or analytical studies. However, a list of references should be included for all test reports, research reports, and other publications used as a basis of prequalification. These references should, to the extent possible, be available in the public domain to permit independent review of the data and to maintain the integrity and credibility of the prequalification process. FEMA 350 provides an example of the type and formatting of information needed for a prequalified connection.
For connections prequalified by AISC CPRP, ANSI/AISC 358 serves as the prequalification record.