Kurt H. Gerstle

Effect of connections on frames

Abstract The ways in which rotational flexibility of beam-column connections affect the behavior of steel frames are outlined. Then, observed connection behavior is summarized and possible formulations for analysis are described. This connection behavior is incorporated in linear and non-linear analyses of steel frames, and predictions are presented for the behavior of flexibly connected frames. Some of these predictions are validated by comparison with test results. Lastly, a computerized design method which includes realistic connection behavior is outlined.

Elastic properties of two coals

Abstract The elastic finite-element analysis of coal mine openings requires knowledge of the mechanical behavior of the coal, which is best supplied in the form of material stiffness matrices, or their inverse, the compliance matrices. The elastic properties of two different coal types are presented in the form of compliance matrices in this paper. Coal samples from two sources, one in the Pittsburgh seam of Pennsylvania, the other in the Herrin No. 6 seam of Illinois, were tested by two different laboratory methods to determine complete compliance matrices for three-dimensional stress states. The data are analyzed statistically, results obtained for the two coals by different test methods are compared, and possible idealizations for analysis are discussed. It appears that modeling of coal as an orthotropic material is a reasonable engineering approximation. Although the principal compliances are sufficiently similar to suggest isotropic behavior, the shear compliances are much larger than those obtained from the relations for an isotropic elastic material. Evidence is also presented which casts doubt on the applicability of the elastic model for other than monotonic load histories.

Analysis of time-dependent deformations of openings in salt media

Abstract An analysis of the time-dependent deformations of openings in salt media is described. The procedure involves (1) deriving three-dimensional constitutive equations for a Carlsbad potash material by means of laboratory tests; (2) developing a method of analysis capable of handling these constitutive equations and the geometries encountered in mining structures; (3) application of the method to particular field problems for which data are available. From results of uniaxial compression tests of potash specimens carried out in this investigation, and data obtained by other investigators from multiaxial testing of a similar material (rock salt), a three-dimensional viscoelastic-viscoplastic model is postulated. The finite-element method is chosen to handle the constitutive equations of this material model in the analysis of plane strain and axisymmetric problems. Good comparison with existing closed-form solutions is demonstrated for this approximate finite-element analysis, which is then used to calculate the deformations around a horizontal circular opening in an underground mine. The predictions are compared with field measurements.

Strength of flexibly connected steel frames

Abstract An analytical study of the strength of flexibly-connected steel frames is presented. A computer program which includes the effects of material and geometric nonlinearities of members and connections is developed, and used in a parametric study which indicates that in most cases, increased connection stiffness leads to an increase of frame strength, but in exceptional cases, it may result in a slight loss of strength.

Behavior and analysis of bracing connections for steel frames

Abstract This paper presents the second part of the results of a study devoted to the analysis of heavy steel bracing connections, and to the effects of those connections on the behavior of braced frames subjected to static loads. The paper deals initially with the finite element analysis of one type of bracing connection, in which the structural fasteners such as bolted clip angles and fillet welds are modeled using the two-dimensional nodal interface element developed in a previous companion paper. Next, two series of full-scale tests of bracing connections subjected to tensile loading are described and the responses measured during the tests compared with the predictions. Good agreement was found between tests and predictions when the overall performance of the bracing connection is not strongly influenced by slip in the bolted fasteners. Lastly, one series of tests on beam-to-column connections made up of friction-bolted double clip angles is presented to validate the finite element model for this type of connection. Once validated, the finite element analysis is used to derive fundamental bracing connection flexibilities and the relative strength interaction domain to be used in a companion paper.

Analysis of flexibly connected braced steel frames

Abstract This paper presents the third part of the results of a study devoted to the analysis of heavy steel bracing connection, and to the effects of those connections on the behavior of braced frames subjected to static loads. The paper presents initially a model compatible with standard frame analysis computer programs and capable of capturing the basic behavior of bracing connections. This is accomplished by representing the bracing connection by an equivalent three-member truss whose behavior can simulate relevant stiffness and strength of the prototype connection. This equivalent truss model is applied to the analysis of a single-story braced frame under service loads, and the sensitivity of the frame to different bracing connection stiffnesses, beam-to-column connection rotational rigidities and brace eccentricities is explored. Lastly, a more complex eight-story braced structure is analyzed under factored loads by yielding of connections and members to demonstrate the inelastic capabilities of the approach.

Reliability of Flexibly-Connected Steel Frames in Sway

An investigation of the influence of connection stiffness on the reliability of flexibly-connected steel frames in sway is carried out. The emphasis is on consideration of connection behavior and variability. Currently available reliability methods neglect these effects. First-order, second-moment probabilistic theory is used in computing the frame reliability index. The results of analysis for a typical portal structure demonstrate that the inclusion of connection stiffness in linearly-elastic steel frame reliability analysis is conceptually and computationally simple. It is demonstrated that the frame reliability may be strongly affected by connection stiffness. The effects of beam-to-column stiffness ratio and beam-to-connection stiffness ratio on frame reliability are also investigated.