Vehicle-Bridge Coupling Vibration Response of Concrete-Filled Steel Tubular Tied Arch Bridges - A Shop Floor Perspective
What the Paper Really Tells Us
This study by Zhang Yao and Sun Zengshou from Tongji and Zhengzhou Universities addresses a question that every bridge fabricator should care about: what actually happens to the steel tube arch ribs and hangers when a loaded truck rolls across the deck at speed. The authors developed a vehicle-bridge coupling vibration analysis program using ANSYS secondary development, then systematically varied vehicle speed, vehicle weight, and bridge deck roughness grade to see how each parameter affects the dynamic response of the arch ribs, the end hangers, and the mid-span of the deck.
The key findings are worth restating for anyone who thinks dynamic analysis is just academic noise. First, as vehicle speed, vehicle weight, and deck roughness grade decrease, the mid-span displacement response peak increases continuously, and the maximum stress at the arch springing, quarter-arch, and crown sections increases while the minimum stress decreases. This is a classic fatigue loading pattern - a stress range that widens with heavier trucks and rougher decks. Second, the crown section stress response peak is less sensitive to vehicle weight, speed, and deck roughness than the arch springing and quarter-arch sections. Third, the mid-span acceleration response peak and the end hanger stress amplitude are most sensitive to a vehicle weight of 20 tonnes, and the end hanger stress amplitude actually decreases with increasing vehicle speed. Fourth, when the deck roughness grade reaches D level, the end hanger stress amplitude hits 186 MPa.
That last number deserves attention. For a hanger rod or cable, 186 MPa stress amplitude under repeated loading is not trivial. If the hanger is a steel rod with a yield strength of 400 MPa, that amplitude is approaching half the yield stress. If it is a strand or cable, the fatigue life at that amplitude could be significantly reduced depending on the detail category and the number of load cycles. The paper's conclusion that lower deck roughness grade and higher vehicle speed cause larger vertical acceleration of the vehicle body, which in turn causes larger vertical displacement of the bridge structure, reinforces the point that deck surface condition is not just a pavement issue - it is a structural fatigue issue.
Steel Pipe Fabrication Implications for Arch Ribs and Hangers
Now let me translate this into shop floor language. The arch ribs of a concrete-filled steel tubular (CFST) bridge are typically fabricated from seamless or welded steel pipe. For a tied arch bridge, the rib sections are usually large-diameter pipes, often in the range of 600 mm to 1200 mm outside diameter, with wall thicknesses from 12 mm to 30 mm depending on the span and the loading. The steel grade is typically Q345, Q370, or Q420 in Chinese practice, equivalent to ASTM A572 Grade 50 or Grade 60.
The fabrication of these ribs involves several critical steps that directly affect the dynamic response characteristics. First, the pipe sections must be cold-formed or hot-formed to the arch curvature. If the radius of curvature is tight, cold forming can introduce significant residual stresses and work hardening, which alter the local material properties and can affect the fatigue response under dynamic loading. For tight radii, hot forming is preferred, but this introduces its own challenges - controlling the heating temperature and cooling rate to avoid grain growth and degradation of mechanical properties.
Second, the longitudinal seams of welded pipe - whether LSAW, spiral welded, or ERW - have weld seams that are metallurgically different from the base metal. The heat-affected zone (HAZ) has different hardness, toughness, and residual stress characteristics. Under dynamic loading, these weld seams are potential fatigue initiation sites. The paper's finding that the arch springing and quarter-arch sections have higher stress sensitivity than the crown section means that the welds in those regions need extra attention during NDT. I would specify 100% UT and MT or PT on all butt welds in the arch rib, particularly in the springing and quarter-arch regions.
Third, the hangers are typically high-strength steel rods or cables. The paper's finding that end hanger stress amplitude reaches 186 MPa at D-level deck roughness is a direct callout to the hanger fabrication and installation quality. If the hangers are steel rods with threaded connections, the thread detail is a classic fatigue notch. The thread rolling or cutting process, the thread profile, and the fit between the rod and the anchor nut all affect the stress concentration factor at the thread root. For dynamically loaded hangers, I would specify rolled threads rather than cut threads, and I would require a minimum thread engagement length and proper lubrication during installation to avoid galling.
Welding Quality and Fatigue Considerations Under Dynamic Loading
The paper's focus on dynamic response means that weld quality is not just a static strength issue - it is a fatigue life issue. For CFST arch ribs, the main welds are the longitudinal seams of the pipe sections, the circumferential butt welds between pipe segments, and the welds connecting the rib to the deck and the hanger anchorages.
The longitudinal seams of welded pipe are particularly critical. For LSAW pipe, the weld is a double-sided submerged arc weld, and the weld geometry - the reinforcement height, the width, and the transition angle to the base metal - directly affects the stress concentration factor at the weld toe. A weld with high reinforcement and a sharp transition angle can have a stress concentration factor of 2.0 or more, which means that the local stress at the weld toe is twice the nominal stress. Under cyclic loading, this is a fatigue crack initiation site.
For the circumferential butt welds between pipe segments, the fit-up gap, the root gap, and the weld profile all matter. If the misalignment between two pipe segments exceeds a certain limit, the eccentricity creates a bending moment at the joint, which increases the local stress. I have seen cases where a 2 mm misalignment on a 20 mm wall pipe resulted in a 15% increase in local stress at the weld. For dynamically loaded structures, I would specify a maximum misalignment of 1 mm or 10% of the wall thickness, whichever is smaller.
The weld quality requirements should follow the relevant standard - for bridge structures in China, this is typically GB/T 50661 or the design code for steel bridges. The NDT requirements should be 100% UT for full-penetration butt welds, plus MT or PT on the weld surface. For the hanger anchorages, the welds should be 100% UT and MT, and I would add a requirement for weld toe grinding or TIG dressing to improve the fatigue detail category.
Practical Lessons for Fabrication and Site Erection
Let me share a few lessons from my own experience that this paper reinforces.
First, deck roughness is not just a pavement contractor's problem. The paper shows that deck roughness grade D causes the end hanger stress amplitude to reach 186 MPa. In practical terms, this means that the deck surface tolerance must be specified and enforced during construction. If the deck is rough, the hangers will see higher stress ranges, and their fatigue life will be reduced. I have seen bridge projects where the deck surface was left rough to save money on the asphalt overlay, and within five years, the hangers showed signs of fatigue cracking at the anchorages.
Second, the vehicle weight sensitivity at 20 tonnes is a reminder that the design vehicle load matters. If the bridge is designed for a 20-tonne axle load, but the actual traffic includes overloaded trucks, the fatigue life will be reduced. This is a traffic management issue as much as a structural issue. For bridges in mining or industrial areas, where overloaded trucks are common, I would recommend a higher design vehicle weight or a more conservative fatigue detail category.
Third, the paper's finding that the crown section is less sensitive to vehicle parameters than the springing and quarter-arch sections suggests that the critical sections for fatigue are not where you might first think. The arch springing and quarter-arch sections are where the bending moments are highest, and these are the sections where the steel pipe is most highly stressed. For fabrication, this means that the welds in these regions should receive extra attention, and the pipe sections in these regions should have higher quality requirements - tighter dimensional tolerances, better surface finish, and more rigorous NDT.
Fourth, the paper's comparison of vehicle-bridge vibration response shows that lower deck roughness grade and higher vehicle speed cause larger vertical acceleration of the vehicle body, which in turn causes larger vertical displacement of the bridge structure. This is a coupled effect that is easy to underestimate. If the deck is rough and the traffic is fast, the bridge will see larger dynamic amplification factors, which means higher stresses in the steel tubes and hangers. For the fabricator, this means that the dynamic amplification factor used in the design should be checked against the actual deck condition and traffic speed. If the design assumed a smooth deck and moderate speed, but the actual condition is rough and fast, the steel tubes may be under-designed for fatigue.
In summary, this paper is a useful reminder that dynamic analysis is not just an academic exercise. The findings on vehicle speed, vehicle weight, and deck roughness have direct implications for steel pipe fabrication, weld quality, and hanger design. For the fabricator, the key takeaways are: pay attention to the weld quality in the arch springing and quarter-arch sections, specify rolled threads for hanger rods, enforce deck surface tolerance during construction, and check the design assumptions against actual traffic conditions. These are the lessons that separate a bridge that lasts fifty years from one that needs hanger replacement in ten.
Zhuojin Pipe Fitting Co., Ltd