Elsevier

Aeolian Research

Volume 25, April 2017, Pages 11-21
Aeolian Research

An efficient, self-orienting, vertical-array, sand trap

https://doi.org/10.1016/j.aeolia.2017.01.003Get rights and content

Abstract

There remains a need for an efficient, low-cost, portable, passive sand trap, which can provide estimates of vertical sand flux over topography and within vegetation and which self-orients into the wind. We present a design for a stacked vertical trap that has been modelled (computational fluid dynamics, CFD) and evaluated in the field and in the wind tunnel. The ‘swinging’ trap orients to within 10° of the flow in the wind tunnel at 8 m s−1, and more rapidly in the field, where natural variability in wind direction accelerates orientation. The CFD analysis indicates flow is steered into the trap during incident wind flow. The trap has a low profile and there is only a small decrease in mass flow rate for multiple traps, poles and rows of poles. The efficiency of the trap was evaluated against an isokinetic sampler and found to be greater than 95%. The centre pole is a key element of the design, minimally decreasing trap efficiency. Finally, field comparisons with the trap of Sherman et al. (2014) yielded comparable estimates of vertical sand flux. The trap described in this paper provides accurate estimates of sand transport in a wide range of field conditions.

Introduction

There have been many attempts to design a passive sand trap for use in aeolian geomorphology. These include traps that integrate the vertical sand flux (e.g. Bagnold, 1938, Leatherman, 1978, Fryberger et al., 1984, Greeley et al., 1996) and segmented traps that provide estimates of sand flux as a function of height above the surface (e.g. Wilson and Cooke, 1980, Borówka, 1990, Dong et al., 2004, Basaran et al., 2011, Rotnicka, 2013, Sherman et al., 2014). It seems that few aeolian geomorphologists involved in empirical work have avoided the temptation to design a better trap, with the primary focus on trap efficiency. There has been less emphasis paid to cost, durability, portability and practicality in a range of field conditions.

All traps are solid structures and inevitably affect the air stream in front of the trap, which may in turn affect the trajectory of saltating grains and result in underestimates of sand flux (Pye and Tsoar, 1990, Rasmussen and Mikkelsen, 1998, Li and Ni, 2003). The efficiency of many traps has been assessed in wind tunnel or field conditions, with reported efficiencies ranging from 30% to over 90%. There are inevitable compromises, however. For example, Nickling and McKenna Neuman (1997) report a passive wedge-shaped sand trap that is efficient (>90%) over a range of wind velocities, provided incident winds angles remain <5°, a condition that is rare in the field. Other traps are efficient, the VTRAP (Namikas, 2002), for example, but require significant excavation of the bed to install the subterranean chamber.

Sand transport is highly variable over small temporal and spatial scales, particularly during near-threshold conditions and low transport rates (Ellis et al., 2012). It is difficult to imagine the larger and more complex and more expensive passive sand traps, the continuously weighing traps of Jackson (1996), or Bauer and Namikas (1998), for example, being deployed in spatial arrays adequate to account for this variation. The large footprints of some traps may render them unsuitable for use within or near vegetation in some situations (e.g. measuring sand flux on the stoss face of a foredune within Ammophila arenaria). Moreover, most of the traps described in the literature have been intended for use on relatively flat surfaces. Obtaining reliable estimates of sand flux between and within coastal dunes and dune systems requires traps that can be mass-produced and that are capable of providing estimates of sand transport in conditions of variable wind direction and speed.

Most passive sand traps are secured to the ground (e.g. the ‘streamer-type’ trap of Sherman et al., 2014), dug into the bed (e.g. the VTRAP and HTRAP of Namikas, 2002) or pressed into the bed (e.g. the segmented, unventilated trap used by Borówka (1990)) to attain stability. They are oriented into the wind with the assumption the wind field is unidirectional, or are deployed in sets intended (typically) to quarter the wind, such as the deployment of cylindrical traps (e.g. the ‘vertical rod sand trap’ of Leatherman, 1978). In fact, wind is not unidirectional at any time scale, and only meets assumptions of statistical (e.g., non-trending) unidirectionality over relatively short time scales. To accommodate changes in wind direction without direct intervention to reorient a trap, it must be able to orientate to variations in wind direction over seconds or minutes. A few traps of this type have been constructed (e.g. the Fryberger orientating trap, Fryberger et al., 1984) or can be partially rotated so they align with the ‘mean wind’ (e.g. the continuously-weighing, tipping-bucket T-BASS trap of Bauer and Namikas, 1998), but these are large traps that have a substantial foot-print and typically produce scour at their base (e.g. Greeley et al., 1996). This problem was largely overcome by the Wilson and Cook trap (WAC), developed by S.J. Wilson and R.U. Cooke in 1980, which has remained popular because of its simplicity, its efficiency (Goosens et al., 2000) and because it can be deployed at height on a mast (Petersen et al., 2011). The cyclone-style traps are efficient at a range of wind speeds and grain sizes (Basaran et al., 2011). Finally, omni-directional traps (e.g. Arens and van der Lee, 1995) appear advantageous, but have not yielded good results in coastal field conditions at very low or high speeds.

The current paper describes a new style of passive sand trap designed to overcome the limitations described above. It is simple and inexpensive to construct, easy to deploy, extremely durable, is highly efficient and self-orientating, and it has a small footprint. It can be used within or over a vegetation canopy and is suitable for use over complex dune morphologies and during light rain.

Section snippets

Design & materials

The traps are manufactured from a PVC non-pressure water pipe (Fig. 1) (42 mm external diameter, 2 mm wall thickness), cut into 210 mm lengths. Thinner pipe could be used. One end of the pipe is warmed using a 2000 W hot air gun (or similar), then moulded by inserting a wooden die (turned on a lathe from a 38 × 38 mm square-section) to achieve a round to square section transition. A 10 mm hole is drilled through the top and bottom of the moulded section, 20 mm (centre) from the square opening. This hole

Orientation and catch-bag resilience

The trap assembly was tested in the University of Canterbury’s open circuit wind tunnel facility to determine (i) the tendency for the trap to orient to the flow and (ii) the resilience of the catch-bag assembly. Traps were placed at 45° to the air flow and flow speed through the tunnel increased in 2 m s−1 increments (Fig. 2). Each experiment was videoed (from above), with a range of angles marked on the floor of the wind tunnel, to allow the alignment of the trap to be associated with wind

The tendency of the trap to orient into the wind

The swinging trap orients into the wind because pressure on the upstream side of the trap and trap bag is greater than the downstream side, creating a turning force and rigid body motion. The results of one of the wind tunnel tests, with the initial angle of the trap plus catch bag (empty) at 45° to the flow, are typical. The trap commenced moving at 1 m s−1, and oriented to 70° (where 90° is flow parallel) at 4 m s−1, 80° at 7 m s−1, and 88° at 10 m s−1. The trap is, therefore, close to parallel

Discussion

The current paper describes a swinging trap system that is suitable for deriving potentially omni-directional estimates of the vertical distribution of sand flux. The trap has a number of notable advantages and limitations. It is efficient, inexpensive and easy to construct, easy to deploy and suitable for making sand flux and relative transport rate estimates where topography might limit the use of passive box-type traps. The low cost of manufacture and ease of deployment allows for many more

Conclusions

We have designed and evaluated a passive sand trap that provides reliable estimates of vertical sand flux profiles, time-integrated transport rates and vertical distributions of grain size. It is a small trap, designed to be deployed over periods of seconds to few minutes. The attributes of this trap are similar to the mesh-style traps of Sherman et al. (2014), and others. However, it is significantly cheaper than the Sherman design and able to be deployed more rapidly. More importantly, it can

Acknowledgements

The authors would like to acknowledge the design acumen of Mr Bill Ingram, Workshop Technician, Department of Applied Sciences, University of Otago, who helped developed multiple trap prototypes, including the trap described in the current paper. Sherman’s participation in the project was supported by the William Evans Visiting Fellowship at the University of Otago – New Zealand. Thanks also to Mr Brent Beaven of the Department of Conservation, Rakiura National Park, for supporting fieldwork at

Cited by (19)

  • Measuring Sediment Transport by Wind

    2022, Treatise on Geomorphology
  • Field test of the accuracy of laser particle counters to measure aeolian sediment flux

    2021, Aeolian Research
    Citation Excerpt :

    During the last 80 years several instruments have been developed, trying to improve the efficiency of the measurements. The instruments or devices to measure aeolian sand transport can be described in three main groups, according to their measurement principles: Sand traps (Leatherman, 1978; Arens and van der Lee, 1995; Sherman et al., 2014; Hilton et al., 2017), Impact sensors (Baas, 2004; de Winter et al., 2018; Ellis et al., 2009; Rezaei et al., 2020) and Optical sensors (Duarte-Campos et al., 2017; Etyemezian et al., 2017; Hugenholtz and Barchyn, 2011). Several studies have compared the different kinds of instruments of a same group (van Pelt et al., 2009; Poortinga et al., 2015).

View all citing articles on Scopus
View full text