Traffic Signal Pole Fatigue in Wyoming
A Research Overview
J. A. Puckett, Ph.D., P.E., Professor
B. P. Collins, P.E., State Bridge Engineer
Two traffic signal structures recently collapsed in Wyoming. The collapse was the result of a fracture at the connection between the cantilever signal light support arm (mast arm) and the pole connected to the foundation. The purpose of this article is to advise the transportation engineering and maintenance community about this potential problem and to outline our experience with this issue, and finally, to outline research that is underway to address this problem. Because the problem is due to wind-induced fatigue, other transportation agencies have experienced such problems and may do so at an increasing rate with more cycles occurring on aging poles.
The Wyoming Department of Transportation (WYDOT) confirmed (through an inspection of the failed connections) that the traffic pole failed at the toe of the welds as a result of fatigue cracking.
These structures did not fail under an extreme-event wind. Vibrations at lower wind speeds most likely caused the fatigue initiation and subsequent crack growth. WYDOT visual inspection with dye penetrant of approximately 840 poles indicated that approximately 1/3 of the poles inspected have fatigue cracks ranging in length from 1/4 to 20 in. (6 to 500 mm) around the box connection between the pole and mast arm. Because inspections only indicate cracks that have propagated to the surface, the damage may be more significant than these numbers suggest. Analysis of the inspection data indicates that the pole directional orientation is not a strong factor predictor even though winds in Wyoming have a prevailing direction. A stronger predictor of cracking was the type of connection detail of the pole-to-mast-arm connection box. Connections that have continuous welds that perform better than connections that have welds discontinued at the corners. See figures.
Note open connection fatigue characteristics are quite poor - research at Lehigh University indicated a category similar to AASHTO E'. As illustrated in the bar chart, the field inspections qualify the record of the open-box detail.
A phenomenon called oil canning is likely adding to the fatigue damage accumulation.
Here, the rather stiff plate of the box connection deforms the pole wall perpendicular to the wall surface. Finite element studies were used to model the connection, resulting in stress analysis and deformation. The figure below shows the deformation due to horizontal movement of the mast arm (out of plane). Examination of vertical mast-arm (in-plane) movement indicates smaller oil-can deformation and associated stresses.
Previous research has focused on the cause of in-plane movements and the associated fatigue damage. Field studies of the crack patterns indicate that cracks can occur in any corner of the connection with no statistical preference. This behavior indicates that both in-plane and out-of-plane motions are important in the development of fatigue stresses. Moreover, visual observations of the pole tip motion easily confirm that the tip moves in an elliptical manner (or figure eight). Field data taken from an instrumented pole located in Laramie, WY illustrate that the out-of-plane (horizontal) motion can create significantly higher bending moments (forces) than the in-plane moment. A typical in-plane and out-of-plane moment record is illustrated in the figure. This behavior presents an interesting situation where both the load effect and the associated stresses may be significantly different than previously thought. The moments in the out-of-plane direction may have been underestimated and their effect on the fatigue damage may be significantly more important than the same moment applied in the in-plane direction.
To establish the load effects (moments), poles are being monitored under service conditions. Wind speed and direction are recorded along with the in-plane and out-of-plane moments. The field testing is intended to provide a better understanding of the loads and associated number of cycles imposed on structures typical of the Wyoming installations.
Research is also being conducted to establish the fatigue characteristics of the in-service connections. Physical testing is being conducted at the University of Wyoming on poles that have been in service and those with new, more fatigue resistance details. See figure. Finite element modeling parallels the testing. It is hoped that S-N curves can be developed for the existing and improved connection details. WYDOT uses a new connection design for new poles that is significantly stiffened by extending the top and bottom plates from the box connection around the pole. (See figure).
Non Destructive Testing
In conjunction with the physical testing, non-destructive tests are being conducted with acoustic emission (AE), ultra sonic, and magnetic particle. The Colorado DOT is helping with the latter two NDTs. With the AE, sensors are placed on the pole around the connection that digitally record the emissions from crack propagation. When a static load is applied to a connection that has fatigue cracks, the cracks will grow slightly, causing the emission of stress waves. These stress waves can be detected using sensitive sensors mounted on the steel surface (see figure). The authors are hopeful that AE can be used as an inspection method where the pole tip can be quickly pulled and AE used to detect cracking, mitigating the need for time-consuming visual inspections, dye penetrate, and/or ultrasonic tests. Perhaps the inspection with AE can be performed during routine maintenance cycles for light replacements, thereby simplifying traffic control. The illustration below shows the sensors located on a test pole. Such inspections are routinely used for other systems such as pressure vessels and utility truck booms.
Collaboration with the various NDTs, fatigue crack growth observations, field performance, and modeling all play an important and integrated role in the assessment of the behavior, design of improved systems, and the development of test methods for improve inspection.
Eliminating the Problem - Retrofit
Retrofitting poles that may, or may not, have exhibited cracking is a possible solution. The pole could be either strengthened and/or modified to help eliminate the number of fatigue cycles due to wind. The authors suggest that strengthening retrofits would be ill advised for system where cracks have initiated. It is unlikely that modifications can be substantial enough to redirect stresses away from the cracks - thereby slowing crack growth. As an alternative to strengthening, extensive research is currently in progress to increase the damping of the system. Numerous retrofits have been investigated; several showed promise. Two of the more notable were:
Free Vibration Testing
A typical free-vibration test is illustrated. The solid line illustrates the response of the non-retrofitted pole. The elastomeric pad worked best for out-of-plane motion and the strut worked well for in-plan motion. The inherent damping is approximately 0.15% and the retrofitted solutions increase the damping significantly Â– to over 6% for the strut system. More field testing and monitoring of in-service retrofitted poles is planned for the coming year. Numerous other retrofit options were investigated as well.
In collaboration with WYDOT and CDOT engineers, University Wyoming engineers are attempting to better understand a critical infrastructure problem, which confront their state. Several approaches are being taken to address this difficult problem. Hopefully, this work will be helpful to other transportation agencies and will become more important to them as their inventories age. Additional research information may be obtained from the authors.