Semester Award Granted

Spring 2025

Submission Date

May 2025

Document Type

Thesis

Degree Name

Master of Science (MS)

Thesis/Dissertation Advisor [Chair]

Barry T. Rosson

Abstract

This paper investigates the derivation and performance of new stiffness coefficients. The estimated coefficients aimed to improve the geometric stiffness matrix representation and their application in elastic second-order analysis for steel frames. The newly developed (C1-C4) coefficients incorporate non-linear effects and reduce computational efforts to efficiently enhance the accuracy of second-order analysis. These coefficients are particularly beneficial for braced structures where they allow more refined analysis using fewer elements per member, especially as the load applied to the frame approaches the critical buckling load. However, for cases of unbraced frames, using these approximated coefficients showed no significant advantages in comparison to the conventional formulations.

The research uses a simplified method for calculating amplification factors through a single element per member of moment frames. Three computing methods analyzed include the predictor-corrector method, eigenvalue buckling analysis, and the ��2 multiplier method. Such methods provide approximations of amplification factors which incorporate the nonlinear second-order effects. An analysis of numerous moment frames under different load scenarios was performed in MASTAN2 for evaluation purposes. Linear regression analysis enabled researchers to derive a design equation as the amplification factors showed an established linear association with load increment numbers up to 1% relative error. The paper outlines specific suggestions about both precision and workflow and the operational boundaries of these techniques.

Download

Share

COinS