Thermal Effects Due to Failed Expansion Joints

[CivStruck]

I have recently been investigating the implications of failed expansion joints, particularly in steel bridges (superstructure), and the effects of a uniform temperature gradient. The applicable code is AASHTO LRFD Bridge Design Specicfications. Of particular interest is the load combinations and factors listed in Table 3.4.1-1. From what I gather the following load factors should be applied to the force effect due to uniform temperature gradient (TU):
-Strength Combinations: 0.50*TU
-Service Combinations: 1.00*TU
-Displacements: 1.20*TU
-Steel Substructures: 1.00*TU

My primary question is why is a 0.50 factor used for strength and 1.00 used for service? The code justifies the 1.20 for displacements so that expansion joints aren't undersized.

Again, focusing primarily on steel, it was proposed to me that 1.0 was used for service before the steel yields and 0.5 would be used for strength after the steel yields and strain is essentially relieved. This sort of makes sense to me, but I then question how the yielding process works; does yielding occur or do anything to relieve stress if the beam is still constrained by fixed ends? I have more questions related to this, but it really stems from a lack of explanation for AASHTO (looked up some of the sources, but none answered my questions). I would love to see any explanation of the rational used.

If anyone has any other ideas or resources related to the force effects that occur in a constrained steel member and the process of failure, I would be grateful for any help!

 

[ishvaaag]

Traditionally, the rheological effects have been being added even to limit states at service level; i.e., you might have 1.5 for DL, 1.6 for LL and then temperature effects factored by 1. So at least is clear the rheological effects have not been thought to require safety factors akin to those of the weights, or wind action. How much of the rheological effects was required to consider concomitant with the other forces in a per hypothesis base has been a quality of every code. See that service and steel substructure factors are 1, and 1.2 for displacements (gap control and envelope of movement), hence the code still wants that the actual movement be considered for particular cases where either local ruin (gap control), bad service life (envelope of displacements) or underdesign (consider full temperature effects in what will affect substructure) are paramount.

Respect how to, essentially you need a program able to map the initial field of temperatures to the structure, and then relax it to equilibrium (solve for the resulting stresses). Most likely for most bridge situations SAP 2000 or similar is such a program.

I once did a pair of Mathcad worksheets based in the article of the paper collection

Book:
Puentes Mixtos
Estado Actual de su tecnología y análisis
Comunicaciones y mesas redondas de las primeras jornadas internacionales
Barcelona 23-27 noviembre 1992
Dirección y edición: Julio Martínez Calzón
CICCP. Madrid, C/ Almagro, 42, 28010 Madrid

Article:
ANÁLISIS TÉRMICO DE LOS PUENTES MIXTOS, GRADIENTES EQUIVALENTES, FORMULACIONES Y VALORES PRÁCTICOS
Enrique Mirambell Arrizabalaga
Dr. Ing. CICCP
Departamento de Ingeniería de la Construcción, Universidad Politécnica de Barcelona
Escuela Técnica Superior de Ingenieros de Caminos de Barcelona

that I may or not have ported well to the worksheet examples that I made, and I post here if something just to remember that the self-equilibrated thermal stresses will be present in bridges. It is brought in the comments to the paper that such longitudinal stresses use to be in the range of 15 to 20 MPa (further quoting that such is also the case of warping stresses for the structural type they seem to be talking about).