Wind Drag Coefficients for Highway Signs and Support Structures (2023)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

1   Wind Drag Coefficients for Highway Signs and Support Structures Overview The main objective of this research is to develop comprehensive methods for estimating wind loads and the associated drag coefficients for highway signs and overhead support structures for inclusion in the AASHTO Load and Resistance Factor Design Specifications for Structural Supports for Highway Signs, Luminaires, and Traffic Signals (LRFDLTS-1) (AASHTO 2015). This objective is motivated by the need to increase the design safety of highway sign support structures by using wind loads based on the best current under- standing of the phenomena. A theoretically sound numerically based approach is applied to estimate the wind loads acting on highway signs and on the monotubes and truss structure members supporting such signs. The numerical model is validated by using both existing wind-tunnel data for isolated signs and new wind-tunnel data obtained from experiments performed as part of the present research that address side-by-side signs and signs containing an add-on sign. Following validation, the numerical model is employed to generate the data needed to estimate drag coefficients for various highway signs. The proposed methods based on these data account, in a rational manner, for the shielding induced by highway signs and different members of the sign support structure and for the acceleration of airflow near the edges of highway signs and inside the gap region between neighboring signs. The method is developed to accurately determine drag coefficients—for isolated rectangular signs placed above the ground, including static signs; dynamic message signs (DMSs), also known as changeable message signs (CMSs); and variable message signs (VMSs)—for highway signs that include an add-on sign and for side-by-side signs. Drag coefficients can also be estimated for cases when these highway signs are placed on mono- tube, truss, and grade separation structures. Data from simulations conducted with highway signs attached to bridge-type, cantilever-type, and butterfly-type overhead structures are also used to estimate the variation of the drag coefficient along the chords of monotubes and trusses and the drag coefficients for secondary members and for gusset plates of truss structures. General methods are proposed to estimate (1) wind loads on highway signs and (2) wind loads on overhead monotube and truss structures that support one or multiple highway signs. The general approach is to calculate the drag coefficient for an element (e.g., chords, secondary members of trusses, highway signs), assuming that the element is iso- lated (e.g., free of interactions with other structural elements), and then to use modification factors to account for different effects (e.g., shielding, increased airflow velocities, suction) attributable to the presence of other highway signs, members of the support structure in the vicinity of that element, or both. The report also includes a comprehensive set of design examples that illustrate how to apply the proposed methods to estimate wind loads for highway signs and their support structures. S U M M A R Y

2 Wind Drag Coefficients for Highway Signs and Support Structures Research Approach The research approach is based on performing three-dimensional (3-D) numerical simu- lations of the airflow past highway signs and their support structures (e.g., trusses, mono- tubes) to first estimate the pressure and shear stress distributions on the individual signs and members and then to calculate the wind loads. The governing Navier-Stokes equations are solved on an unstructured grid where the level of grid refinement was increased in criti- cal regions needed to resolve the boundary layers on the traffic signs and the elements of the sign support structure. This approach is particularly well suited to obtaining the data necessary to develop a general method for estimating wind loads on structures that support single and multiple signs, including truss structures and monotube structures. The main data needed to develop such a method are the wind loads on each member and on each relevant part of the larger members (e.g., chords). Such data cannot be obtained experimen- tally but are critical to developing a method to accurately estimate the wind loads on sign support structures. These wind loads are then used to calculate the drag coefficients defined with the projected area and the incoming wind velocity away from the signs and the sup- port structures. Another advantage of the current numerical approach lies in allowing for performing of simulations in very wide and high domains, which ensures that flow blockage effects are negligible. Such effects are always a concern when estimating wind loads and drag coefficients for highway signs in the field based on data obtained from wind-tunnel studies. The accuracy of the numerical model is tested by comparing wind load predictions and corresponding drag coefficients with those measured experimentally in both wind tunnels and the field. Predicted drag coefficients for both thin and thick (DMS) rectangular signs are in good agreement with those resulting from the wind-tunnel experiments by Chowdhury et al. (2015) and from field measurements by Smith, Zuo, and Mehta (2014) and Quinn, Baker, and Wright (2001). Numerical predictions of the drag coefficient for a thin infinitely wide plate in free flow are also very close to values reported in the literature. Additional validation data for signs of a more complex shape are generated as part of the current project. Wind-tunnel experiments are conducted to determine mean wind forces on side-by-side signs with variable gap distances and on static rectangular signs that include an add-on sign. Very good agreement is observed between the estimates of drag coefficients based on experimental data and those predicted by the corresponding validation simulations. Drag Coefficients for Highway Signs A comprehensive parametric study is undertaken that produces detailed information on how the drag coefficient for rectangular signs placed above ground varies as a function of the sign aspect ratio; sign thickness; and h/(h + hg) where h is the sign height and hg is the ground clearance distance. Although the effect of the sign thickness is determined to be minor, a large increase in the drag coefficient is observed with increasing h/(h + hg). The value of the drag coefficient recommended in the current LRFDLTS-1 specifications for dynamic message signs is found to be much too conservative. Moreover, drag coefficient values for rectangular signs are a function of both the sign aspect ratio and h/(h + hg) but are not treated as such in the current specifications. In the case of a rectangular sign that includes an add-on sign, a slight increase in the drag coefficient for the full sign is observed. For side- by-side signs, the drag coefficient on the smaller sign can be subject to up to 30% amplifica- tion compared to drag coefficient values for a sign of the same size that is not situated in the vicinity of other signs. Drag coefficients for signs placed on a monotube or a truss structure are also larger than those calculated for the same signs without the support structure. Wind loads acting on a sign placed on a grade separation structure are strongly nonuniform over

Summary 3   the height of the sign, especially in cases when the top of the sign extends above the top of the barrier rail or separation rail. This outcome is observed both when the wind is oriented toward the front face of the sign and when the wind is oriented toward the back face of the sign. For typical distances between the sign and the grade separation structure, an overall amplification of the order of 30% is observed for the drag coefficient of a sign placed on a grade separation structure with the incoming wind direction toward the front face of the sign (when compared to the drag coefficient for the same sign with no bridge behind it). Wind Loads on Sign Support Structures A series of simulations is conducted to determine the drag coefficient for back-to-back circular cross-section and L-shaped members of infinite length and for isolated members of finite length. Findings indicate that the drag coefficient for an isolated member is a function of not only the Reynolds number but also the aspect ratio (i.e., the ratio between the length and the width or diameter of the member). The current research shows that the drag coefficient can decay significantly for relatively short members when compared to the values calculated for infinitely long members of the same cross-section and the same incoming wind velocity. Simulations are also performed with bridge-type, cantilever-type, and butterfly-type monotube and truss structures supporting one or two highway signs. Based on these simu- lations, estimations are made for (1) the lateral variation of the drag coefficients along the monotube and truss chords as a function of the distance from the lateral edges of the sign (or signs) mounted on them and (2) the wind loads acting on the gusset plates. These results demonstrate the amplification of the wind loads for the parts of the chords and secondary truss members situated in the flow-acceleration region that forms next to the lateral edges of the signs placed on the support structure and in the gap region between side-by-side signs. Estimates are made of the width of the flow-acceleration regions and the degree of amplification of the wind loads inside the flow-acceleration region compared to uniform- flow regions where the approaching wind velocity is equal to the wind velocity far from the sign support structure. For truss support structures, the simulations also indicate an important reduction of the wind loads acting on members shielded by other truss members compared to unshielded members directly exposed to the incoming wind flow. Based on the estimated wind loads, drag coefficients are defined by using the incoming wind velocity and the diameter or width of the member. These drag coefficients are normalized by the values expected for the same member, assuming no shielding or interactions with other members or highway signs. These simulation results provide the main information needed to develop methods to estimate wind loads on sign support structures. Proposed Methods for Estimating Wind Drag Coefficients and Wind Loads for Highway Signs and Support Structures A general method is proposed to estimate wind loads on highway signs and the associ- ated drag coefficients. Initially, the drag coefficient for an equivalent isolated rectangular sign, Cd0, is estimated as a function of the sign aspect ratio and h/(h + hg). For signs that include an add-on sign, the height of the full sign is corrected to account for the additional area of the add-on sign. Then, several modification factors are applied to calculate the drag coefficient for the sign Cd. These factors account for the effects of sign thickness (Kt), the presence of an add-on sign (Ka), the effects of proximity of another sign (Kp), and the presence of a sign support structure behind the sign (Ks). Rules on how to select values for

4 Wind Drag Coefficients for Highway Signs and Support Structures these factors are given as a function of the main geometrical variables. A method is also proposed to estimate wind loads on the relevant subzones of highway signs placed on a grade separation structure when the incoming wind direction is oriented toward the front face or the back face of the sign. In addition, a general method is proposed to estimate wind loads on monotube and truss structures supporting one or multiple highway signs. For each chord and secondary member of the support structure, a coefficient Cd0 is calculated, assuming that the member is isolated (i.e., it has no interactions with other members or highway signs). This coefficient is given as a function of the Reynolds number and the aspect ratio of the member. The chords are split into several subregions. Using simple geometrical rules, each subregion of the chords and each secondary member are assigned to a uniform-flow region, flow-acceleration region, behind-the-sign region, or gap region. Then, the drag coefficients for each chord subregion and the secondary truss members are calculated by applying one or two modification factors to Cd0 based on the shielding effects of the signs and on the shielding effects and interac- tions with the other members of the support structure. Rules on how to select values for the modification factors are also specified. A similar procedure is used to estimate the drag coefficient for the gusset plates subject to nonnegligible wind loads. Recommendations are also made on how to estimate the wind loads on vertical columns supporting monotubes and trusses and the wind loads applied transverse to the plane of the sign support structure. Design Examples The report includes a set of design examples illustrating how to apply the proposed methods and recommendations to estimate wind drag coefficients for highway signs and their support structures. The main applications covered by the design examples address estimat- ing the following: • Normal wind loads for an overhead bridge-type monotube structure supporting one static sign and one dynamic message sign. • Normal and transverse wind loads for an overhead bridge-type 3-chord truss structure supporting one dynamic message sign, with the posts and columns supporting the truss included in the calculation. • Normal wind loads for a cantilever-type 4-chord truss structure supporting two traffic signs, including estimation of the wind loads on the gusset plates. • Normal wind loads for an overhead bridge-type monotube structure supporting two static signs, with one of the signs containing an accompanying add-on sign. • Normal wind loads for an overhead bridge-type monotube structure supporting two static signs and a dynamic message sign. • Normal wind loads for a traffic sign attached to a grade separation structure that includes a barrier rail or a separation rail, with the normal wind loads evaluated for cases when the wind is directed toward the front face of the sign or the back face of the sign. Overall Conclusions and Recommendations General methods to estimate mean (time-averaged) wind loads on highway signs and their support structures are developed. These methods are targeted for strength design, and they can be applied to a wide range of support structures for single or multiple signs and can account for the interactions and shielding effects between the signs and support structures. This study generates more accurate and detailed information on wind drag coefficients for highway signs, facilitating an understanding of how the presence of other signs or support

Summary 5   structure members influences the drag coefficient for each sign. Such information is not available from previous studies. The new methods to estimate wind loads on highway signs and their support structures fill an important research gap because the current LRFDLTS-1 specifications address only a subset of the relevant geometrical and flow parameters that affect the drag coefficient for commonly used highway signs. Furthermore, the current AASHTO specifications contain only very simplified rules to estimate wind loads on sign support structures. Accounting for the complex interactions (e.g., shielding, flow acceleration, suction) between the signs and the different members of the support structure can lead to significant changes in the calcu- lated values for total wind loads acting on the signs and their support structure compared to the values derived from the simpler procedures in the current LRFDLTS-1 specifications. The design examples in this report indicate that, in some cases, the wind loads calculated by using the new methods are much higher than those calculated based on the procedures in the existing LRFDLTS-1 specifications. The proposed numerically based methods for determining drag coefficients and wind loads on highway signs and sign support structures are based on high-resolution compu- tational fluid dynamics (CFD) simulations. These methods are very general and can be extended to other types of signs, luminaries, and support structures (e.g., vertical columns of different cross-sections, trussed towers). More research is needed to refine the proposed method for structures supporting multiple traffic signs and signals on both their upwind and downwind sides.