Temperature Stress Behavior of Axially Compressed Restrained Steel Pipe Columns Under Fire Exposure

Literature Overview

This study by Qu Lijun and colleagues from the Chinese People's Armed Police Force Academy, published in the Journal of Civil Engineering in 2012, presents one of the most comprehensive experimental investigations into the thermal stress response of restrained steel pipe columns subjected to fire. The research was supported by the Ministry of Public Security Applied Innovation Program (2008YYCXWJXY117) and involved testing 215 Q345 steel pipe column specimens using a self-developed test rig that applied constant axial load with continuous temperature elevation. The experimental matrix included 13 levels of restraint stiffness, 3 levels of initial stress, and 6 slenderness ratios, which is remarkably thorough and provides a solid foundation for fire-resistance design and assessment of steel structures.

Core Technical Findings

The fundamental conclusion drawn from this extensive test program is that thermal stress constitutes the most critical load effect governing the fire behavior of restrained steel members. This finding carries profound implications for fire-resistance design codes, which traditionally focus on strength degradation at elevated temperatures while often neglecting the progressive accumulation of thermal stresses in restrained conditions. The test results demonstrate that thermal stress dominates failure initiation, and by the time thermal stress begins to decrease during the later stages of heating, the specimen has already sustained catastrophic failure.

The four-stage evolution of the thermal stress-temperature curve is particularly instructive for engineers:

The interplay between restraint stiffness and slenderness ratio is another critical finding. Higher restraint stiffness and larger slenderness ratios both produce greater thermal stresses and lower failure temperatures. Notably, when restraint stiffness is high and slenderness ratio is small, the plastic plateau region of the stress-temperature curve is extended, indicating a more ductile response before ultimate failure. This has direct implications for the design of braced steel pipe columns in buildings where fire compartmentation and structural continuity impose significant restraint conditions.

Slenderness Ratio and Failure Mode Analysis

The study clearly distinguishes between two failure modes governed by slenderness ratio. When the slenderness ratio is large, failure occurs through buckling instability, characterized by significant lateral residual deformation and relatively small axial deformation after collapse. When the slenderness ratio is small, failure is governed by material strength exhaustion, with large axial deformation and comparatively minor lateral displacement. This distinction mirrors the cold-load behavior classification between flexural buckling and yielding failure but extends it into the elevated-temperature regime where material properties are continuously degrading.

The effect of initial stress level provides another layer of engineering insight. At the early stage of heating, specimens with different initial stress levels follow essentially the same stress-temperature curve with identical slope, indicating that the elastic thermal stress development is independent of the pre-existing mechanical stress. However, specimens with lower initial stress sustain higher thermal stresses and separate from the initial common curve at a later temperature, while those with higher initial stress diverge earlier. This behavior suggests that the available thermal stress capacity is inversely related to the initial mechanical stress level, a finding that should be incorporated into fire assessment methodologies for existing structures with unknown load histories.

Engineering Practice Implications

From a practical standpoint, this research reinforces several design principles for steel pipe columns in fire-exposed conditions. First, the design of fire protection systems must account for the restraint conditions imposed by connected structural elements, not merely the thermal degradation of the pipe column itself. Second, the selection of slenderness ratio in fire-critical structural zones should consider the amplification of thermal stresses that occurs with higher slenderness values. Third, for existing structures undergoing fire assessment, the initial stress state must be evaluated because it directly influences the thermal stress capacity and the onset temperature of plastic yielding.

The experimental methodology adopted here—constant load with continuous temperature elevation—represents the most realistic simulation of actual fire exposure conditions for restrained members. This contrasts with isothermal testing, which cannot capture the progressive accumulation of thermal stresses that occur during the heating phase of a fire. Engineers involved in fire engineering design should advocate for test protocols that replicate realistic loading and heating sequences when evaluating the fire performance of steel pipe structural components.

Key Reflections and Study Insights

Having reviewed this work in depth, I find it particularly valuable that the authors systematically varied restraint stiffness across 13 levels rather than limiting the study to a few discrete cases. This granularity allows for the development of more precise predictive models for thermal stress estimation in practical design situations. The observation that the plastic plateau lengthens with higher restraint stiffness and lower slenderness ratio is especially significant for the seismic-fire interaction analysis of steel pipe braced frames, where the member's ductility under combined thermal and mechanical loading determines the overall structural response.

One area that warrants further investigation, as suggested by this study, is the influence of pipe wall thickness and cross-sectional geometry on the thermal stress distribution. The current tests appear to focus on uniform heating conditions, whereas real fires produce non-uniform temperature fields that could induce additional bending stresses in pipe columns. Future research should combine the restraint stiffness methodology presented here with non-uniform heating scenarios to provide more complete design guidance for practical fire engineering applications.