Acta Geophysica vol. 55, no. 1, pp. 23-32 DOI 10.2478/s11600-006-0036-5 Measurements of armour layer roughness geometry function and porosity Jochen ABERLE Leichtweiss-Institute for Hydraulic Engineering, Technical University of Braunschweig Beethovenstr. 51, 38106 Braunschweig, Germany; e-mail: [email protected] Abstract The roughness geometry function of the interfacial sublayer of a gravel-bed armour layer was measured directly by filling water stepwise into a laboratory flume and indirectly from a digital elevation model (DEM) of the surface. The results of both methods are compared and show that the DEM can be used to reliably estimate the roughness density function for a wide range of the interfacial sublayer. The direct measurements revealed an absolute minimum of porosity at the level of the roughness trough which is significantly smaller than porosity in the undisturbed subsurface and porosity estimates obtained from relationships found in the literature. The significance of the results for hydraulic engineering and ecological applications is highlighted. Key words: Armour Layer, porosity, digital elevation model, flume study. 1. INTRODUCTION The roughness geometry of gravel-bed armour layers plays an important role in fluvial geomorphology and river hydraulics since it determines such flow properties as mean flow velocity, turbulence, and sediment transport. Its significance for the flow region between roughness top and trough (interfacial sublayer) becomes apparent from the double averaged (in temporal and spatial domains) momentum equations (e.g., Nikora et al. 2001, 2004, 2006a, b, Aberle and Koll 2004, Aberle 2006), which depend on the roughness geometry function φ. The roughness geometry function is defined as φ = Vf /V0 , where Vf = volume occupied by fluid within an averaging (total) volume V0. The definition of φ shows that it can also be interpreted as a statistical measure of ran© 2007 Institute of Geophysics, Polish Academy of Sciences Unauthenticated Download Date | 6/14/17 4:54 PM 24 J. ABERLE dom geometry of the bed surface (e.g., Nikora et al. 2001) as well as porosity. The latter is, similarly to φ, defined as the ratio of the space taken up by voids to the total volume (e.g., Bunte and Abt 2001). Porosity, on the other hand, is an important parameter for armour layer development, colmation processes, aquatic habitat studies and exchange processes between free surface and groundwater flows. For example, during armour layer development, the thickness of the active mixing layer is, for plane beds, controlled by the maximum grain size and porosity (Marion and Fraccarollo 1997). Hence, modelling bed development with the sediment continuity equation (e.g., Toro-Escobar et al. 1996, Bettes and Frangipane 2003) or analysing bed material samples extracted by, e.g., waxcoring methods (Marion and Fraccarollo 1997) requires the knowledge of porosity. Furthermore, porosity is also an important parameter for ecological applications. Fine sediment that passes the coarse armour layer may accumulate beneath the surface and during low discharge conditions a compact layer may develop that reduces porosity and hydraulic conductivity of the streambed. Such a compaction stabilizes the streambed against erosion (Veličković 2005) and is an important link between the sediment transport regime and the suitability of benthic habitat for stream organisms (Lisle 1989) as well as for exchange processes between free surface and groundwater flows (e.g., Veličković 2005, Vollmer 2005). Methods for the measurement of gravel-bed porosity are described in Milhous (2002) and Carling and Reader (1982). Milhous (2002) developed a simple method for the determination of porosity below the armoured surface using a wooden frame and plastic sheets. Carling and Reader (1982) analysed sediment cores sampled with a freezing technique to relate porosity to the mean grain diameter according to φ = 0.4665 dm–0.21 – 0.0333 (dm in mm) for poorly sorted sediments. A similar relationship was developed by Komura (1963), φ = 0.245+ 0.0864 dm–0.21 (dm in cm). Both relationships show that, below the surface, porosity is inversely related to mean grain diameter. Taking into account that the uppermost layer of an armour layer is coarser than the layers below (e.g., Sibanda et al. 2000), porosity is thus expected to increase with depth into the sediment. However, detailed measurements of roughness geometry function and porosity within the interfacial sublayer and the subsurface layer of armoured gravel-beds are time consuming and therefore rare. Nikora et al. (1998, 2001, 2006b), Aberle and Koll (2004), and Aberle (2006) used topographic data of gravel-bed armour layers obtained with point gauges and laser scans to estimate φ for the application in the doubleaveraged momentum equations. These studies explicitly concluded that a correction function for φ is needed for such data, since φ = 0 is obtained at the lowest measured bed elevation due to the fact that a laser scan of the surface does not include measurements of the pore volume. Thus, in reality φ tends to a non-zero value of the bed porosity below roughness troughs and not to zero. In this paper, the requirement of a correction function for the roughness geometry function is investigated by comparing φ-estimates obtained from direct measurements in a big laboratory flume to φ-estimates derived from laser displacement data. Unauthenticated Download Date | 6/14/17 4:54 PM ARMOUR LAYER ROUGHNESS GEOMETRY FUNCTION AND POROSITY 25 2. EXPERIMENTAL SETUP The data analysed in this paper were obtained from laboratory experiments in a 20 m long, 0.90 m wide and 0.60 m high tilting flume in the laboratory of the LeichtweissInstitute for Hydraulic Engineering, Technical University of Braunschweig, Germany. These experiments were specifically designed to study the interaction between flow and roughness for flows with low and medium relative submergence based on the double-averaging approach. The detailed description of the hydraulic measurements is beyond the scope of this paper as it focuses on the analysis of the topography data. Preliminary results of the hydraulic measurements can be found in Aberle and Koll (2004), Aberle (2006) and Nikora et al. (2006b). In brief, the experimental methodology was based on a subsequent development of armour layers. At the first step of an experiment, the well mixed sediment (0.63 mm < d < 64 mm; see Fig. 1 for grain-size distribution) was placed in the flume over a length Ls = 13.44 m and flattened to a thickness of 0.20 m to ensure that the bed slope was parallel to the flume slope (S = 0.0027). The distance from the flow straightener at the flume inlet to the sediment bed was 1.00 m and the gap between was bridged with bricks to avoid extensive scouring at the inlet. The sediment body was separated from the bricks by a perforated metal plate. At the downstream end the sediment body was stabilized by a 20 cm high sill. The sill was also constructed of perforated metal to allow flow in the subsurface layer. The bed investigated in this study corresponded to the third subsequently developed armour layer which was armoured with a discharge of Q = 220 l/s (with preceding armouring discharges of Q = 120 l/s and 180 l/s, respectively). Each armouring discharge was run until bed-material transport rate became negligible, indicating a stable armoured bed surface. Photographs of the armour layers were used to obtain the grain-size distribution of the surface material applying a line by number method (Fehr 1987). The method was calibrated by manually sampling the surface layer for a selected preceding experiment and adjusting parameters to minimize differences between a sieve-based grain-size distribution and a distribution from photographs 100 Subsurface Material 90 Armour layer Percent finer [%] 80 70 60 50 40 30 20 10 0 1 10 d [mm] 100 Fig. 1. View of the flume during laser-scan (left) and grain-size distributions of subsurface material and armour layer (right). Unauthenticated Download Date | 6/14/17 4:54 PM J. ABERLE 26 (Oppermann 2005). Figure 1 shows the grain-size distribution of the subsurface material and of the armour layer under investigation. The bed topography was measured along the 2.40 m long and 0.716 m wide test section with an Optimess laser-displacement meter (see Fig. 1). The test section was located 8 m from the transition bricks-sediment so that neither the scour at the inlet section nor the downstream sill influenced the development of water worked roughness structure. Longitudinal bed profiles were recorded with a sampling interval ∆x = 1.0 mm, a lateral sampling interval ∆y = 4 mm, and with a vertical resolution of 0.1 mm. The different sampling intervals in x and y directions were due to the footprint of the sampling volume of the laser system (0.2 mm in flow direction and 4 mm in transverse direction). Figure 2 shows the measured topography of the armour layer under investigation. 0 200 400 600 0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200 2400 Fig. 2. Armour layer topography of the test section (flow direction from left to right; axes are scaled in mm). The laser data were used to estimate φ by assuming that the surface is composed of pillars with ground area identical to the sampling resolution (1×4 mm) and height of the corresponding reading. In a next step, the digital elevation model (DEM) was sliced into 100 vertical segments (with ground area equal to the overall size of the DEM) and φ was estimated for each segment from the ratio of the volume fraction occupied by the pillars and the total segment-volume. Further results related to the analysis of roughness structure of armour layers from these experiments are found in Aberle and Nikora (2006). In order to investigate the accuracy of the φ-functions obtained from the DEMdata, additional measurements were carried out for this study in which φ was measured directly. For this purpose, the flume was tilted into a horizontal position and sealed at the upstream and downstream end. The upstream section was sealed in a distance of 0.155 m to the sediment body by removing a row of bricks and fitting a plate into the flume cross-section. Another seal-plate was mounted at the downstream end of the flume in a distance of 0.545 m to the sill. Thus, two small basins with a total length LB = 0.70 m were created and the total length of the measuring section comprising the sediment body and the two basins was L = LS + LB = 13.44 m + 0.70 m = 14.14 m. Unauthenticated Download Date | 6/14/17 4:54 PM ARMOUR LAYER ROUGHNESS GEOMETRY FUNCTION AND POROSITY 27 This flume setup was used to measure φ directly by filling stepwise a known amount of water (measured by a water-meter) into the flume and monitoring corresponding water levels. The latter were measured in the downstream basin and over the roughness trough (of the total sediment body) using point gauges. Readings were taken only when the water level was stable. The knowledge of the water volume added in each step VSt and the corresponding water level increment ∆h enables the calculation of φ = Vf /V0. The volume fraction of water added to the sediment body Vf can be estimated by subtracting the volume of water added to the basins VB = wLB ∆h (with w = flume width) from the volume of water added to the flume VSt in the step. The total volume of the averaging region of the sediment bed is known from V0 = wLs ∆h. Hence, porosity can be calculated by the following equation: φ= Vf V0 = VSt − VB V − wLB ∆h V L = St = St − B . V0 wLS ∆h V0 LS (1) In total, four additional experiments were carried out to determine φ. Before the measurements, the sediment body was drained and air dried for a period of 3-4 days. The experimental runs 1-3 were designed to determine φ solely above the roughness trough and the additional fourth experiment was carried out to measure porosity also in the subsurface layer. In runs 1-3, the sealed flume section was slowly filled with water until water became visible at the roughness trough. From this point onwards, water was successively filled into the flume in steps of 10-50 litres; 10 l steps were applied in the lower part of the armour layer (i.e., close to roughness trough) and 50 l steps in the upper part. For the fourth experiment, water was filled into the flume in steps of 50 l starting from the flume plastic bottom. Close to the roughness trough, the step width was reduced to 10 l and increased again to 50 l at some distance to the roughness trough. 3. RESULTS AND DISCUSSION Flume measurements Figure 3 shows the vertical distributions of φ obtained from the flume measurements. The results in Fig. 3 show that the measurements are characterised by a high degree of reproducibility, although the beds were only air dried. All curves decrease monotonically from the roughness tops towards the roughness trough. Slightly above the roughness trough, φ is characterised by an absolute minimum value of 6%, indicating at poor sorting and a densely packed layer. Below the roughness trough, φ increases for approximately 0.04 m before reaching a constant value (φ ≈ 18-22%) within an approximately 0.06 m thick subsurface layer. The observed successive increase of φ below the roughness trough is associated with a change in the sediment matrix and, thus, with the finer composition of the subsurface bed material (see Fig. 1). The minor differences in the φ-values within the constant φ-layer are most likely related to the change in the volume of added water from 50 to 10 l within this particular (fourth) ex- Unauthenticated Download Date | 6/14/17 4:54 PM J. ABERLE 28 periment. Reasonably assuming that the distance from the roughness tops to the constant φ-layer corresponds to the active layer during the armouring process, the active layer thickness is approximately 2dmax for the bed under investigation. Height over plastic bottom [m] 0.25 0.2 Roughness trough 0.15 1st Run 2nd Run 0.1 3rd Run 4th Run 0.05 0 0 0.2 0.4 0.6 0.8 1 1.2 φ = Vf /V0 Fig. 3. Estimates of roughness density and porosity for the experimental runs 1-4. Below the constant φ-layer and in a distance of 0.06 m to the flume plastic bottom, φ-values increase again with depth into the sediment. Bearing in mind that the maximum grain size corresponds to dmax = 0.064 m, it can be concluded that the matrix in this region is influenced by the plastic bottom and dmax. The plastic bottom prevents thorough mixing of the particles and, thus, larger pore volumes may develop in this region. Such an increase of φ can also be observed for densely packed, single layered gravel particles of the same diameter (Stephan 2001). For this case φ reduces from the gravel tops to approximately half the particle diameter where the single-layered bed is characterised by the maximum volume of solids. In the region below, the volume of solids reduces again and hence φ increases towards the impermeable flume bottom (φ → 1), where finally φ = 0 due to the impermeable surface. Thus, the observed increase of φ towards the plastic flume bottom in Fig. 3 may be associated with a similar mechanism. The sediment matrix may act as a filter for larger particles close to the flume bottom, leading to increased pore volumes. However, smaller particles such as sand and fine gravel particles may still pass this filter region and settle at the bottom. Such a behaviour would explain the observed decrease of φ for the measurement point closest to the plastic bottom in Fig. 3, although further data is needed to explore this behaviour in more detail. A closer examination of the results shown in Fig. 3 reveals that the φ-curves are characterised by smaller scatter (with values φ > 1) in the region of the absolute roughness top. This scatter is associated with measurement accuracy. In this region the volume of solids is small compared to the volume of added water since it is composed Unauthenticated Download Date | 6/14/17 4:54 PM ARMOUR LAYER ROUGHNESS GEOMETRY FUNCTION AND POROSITY 29 only by the tops of a few protruding particles. Hence the corresponding φ-estimates are prone to measurement errors. In fact, an inaccuracy in the point gauge reading of 0.0001 m results in an error for the φ-estimate of approx. 3% in this region. However, it is worth mentioning that the error is reduced in regions further below the roughness tops; the same measurement error corresponds to an error in φ-estimates of approximately 1% in the mid-region and 0.2% close to the roughness trough. The measurements can also be used to analyse bulk porosity values. For runs 1-4, the bulk porosity, estimated from the plastic bottom to roughness trough, corresponds to 22.2, 24.1, 23.2, and 23.3%, respectively. An additional measurement, where the bed material was placed in a 10 l bucket, resulted in an estimate of bulk porosity φ = 28%. This difference can be attributed to sample preparation in the bucket, in which the sediment was not compacted in the same way as in the flume nor characterised by an armour layer. Using the relationships of Carling and Reader (1982) and Komura (1963), bulk values for φ of 26.4% and 33.4%, respectively, are obtained for the subsurface material. These values are reasonably close to the bulk and bucket estimates of φ as well as to estimates found in the literature (φ = 30-40% for static armour layers, e.g., Marion and Fraccarollo 1997, Bettes and Frangipane 2003). However, using the grain-size distribution of the surface layer results in estimates of φ = 20.9 and 31.8%, respectively, being significantly larger than the observed minimum value of φ = 6% at the roughness trough. Thus, this result shows that the minimum value of porosity of the surface layer may be significantly overestimated using bulk relationships. Laser displacement data Figure 4 shows the φ-function estimated from the DEM together with the φ-function obtained from the 2nd experimental run of the flume measurements. The latter has been included to directly compare the φ-estimates of both methods. The laser-data φ-curve decays similarly to the flume-curves monotonically from φ = 1 at the roughness tops towards the roughness trough. However, in contrast to the flume measurements, φ = 0 is obtained at the roughness trough although φ = 6% would be expected. This difference is due to the fact that the laser scan of the surface does not include measurements of the pore volume. Nonetheless, both curves are almost identical for φ > 12% (z > 0.018 m above roughness trough). For distances to the roughness trough larger than half the roughness range (φ > 60%), the laser data result in slightly underestimated values of φ compared to the values obtained from the flume measurements. These differences are associated with (i) the measurement accuracy of the flume experiments and (ii) the different ground areas of both methods. The deviation of the curves becomes apparent at the level above the roughness trough where the volume of added water was increased from 10 to 50 l. Thus, the aforementioned inaccuracies in point gauge readings may become more significant. The second explanation is related to the different measurement scale. Unauthenticated Download Date | 6/14/17 4:54 PM J. ABERLE 30 0.25 Height over plastic bottom [m] 2nd Run Laser Displacement Meter 0.2 0.15 0.1 0 0.2 0.4 0.6 0.8 1 1.2 φ = Vf /V0 Fig. 4. Comparison of roughness density function obtained from laser data and obtained from direct measurements of run 2. For the flume measurements, φ was estimated for the total sediment body while, for the indirect estimates using the laser data, φ-estimates are restricted to the ground area of the DEM. The absolute maximum and minimum elevation of the sediment body was not located within the DEM-section. This becomes apparent from the different maximum and minimum bed elevations for both methods in Fig. 4. Furthermore, differences in bed geometry such as the scour at the flume inlet or the accumulation of bed material at the downstream sill may slightly bias the results from the flume measurements. Nonetheless, the above results show that the laser scan provides an accurate estimate of the roughness geometry function for almost the total range of the interfacial sublayer. Thus, it can be concluded that a correction function for φ for gravel-bed armour layers is required only for the region close to the roughness trough. However, the derivation of an appropriate correction function requires more data sampled over different armour layers. 4. CONCLUSIONS In this paper, a direct and an indirect method for estimation of the roughness geometry function φ within the interfacial sublayer of a gravel-bed armour layer were compared. For the direct method, water was filled stepwise into a big laboratory flume and φ was calculated from the volume of added water and the associated increment of water levels. The indirect method was based on the analysis of an armour layer DEM. The comparison of both methods showed that DEM data can be used to accurately estimate φ for a wide range of the interfacial sublayer (φ >12%), although the pore volume is Unauthenticated Download Date | 6/14/17 4:54 PM ARMOUR LAYER ROUGHNESS GEOMETRY FUNCTION AND POROSITY 31 not reflected in DEM data. It was concluded that a correction function for φ estimated from DEM’s is required only for the region close to the roughness trough. The direct measurements of the roughness geometry function and subsurface porosity revealed an absolute minimum value of φ at the level of the roughness trough of φ = 6% for the bed under investigation. This low value of porosity was associated with the dense packing of the surface particles due to static armour. Bearing in mind that this study was carried out with particles larger than 0.63 mm, it can be concluded that armour layers in the field may be characterised by even lower values of φ. Furthermore, the observed minimum value of φ is significantly smaller than φ-values estimated from relationships found in the literature. Thus, the results of this study are important for applications concerning armour layer development, exchange processes between subsurface and surface flows as well as benthic habitat studies, where porosity is an important parameter. More data on this topic would be desirable to generalise the findings of this study. However, such data is not available because simultaneous direct measurements of porosity and bed geometry are time consuming and therefore rare. A c k n o w l e d g m e n t s . The research was conducted under contract DI-651/4-3 from DFG (Deutsche Forschungsgemeinschaft). U. Ecklebe and M. Oppermann provided technical support. The author is grateful for useful discussions and suggestions to V. Nikora, D. Pokrajac, K. Koll, A. Dittrich and to all participants of the “doubleaveraging” workshops. References Aberle, J., 2006, Spatially averaged near-bed flow field over rough armour layers. In: River Flow 2006, Proc. of 3rd Intern. Conf. on Fluvial Hydraulics, 6-8 Sept. 2006, Lisbon, Portugal. Aberle, J., and K. Koll, 2004, Double-averaged flow field over static armor layer. In: M. Greco, A. Carravetta, R. Della Morte (eds.), River Flow 2004, Proc. of 2nd Intern. Conf. on Fluvial Hydraulics, June 2004, Napoli, Italy, 225-233. Aberle, J., and V. Nikora, 2006, Statistical properties of armored gravel-bed surfaces, Water Resources Research (in press). Bettes, R., and A. Frangipane, 2003, A one-layer model to predict the time development of static armour, J. Hydraul. Res. 41, 2, 179-194. Bunte, K., and S.R. Abt, 2001, Sampling surface and subsurface particle-size distributions in wadable gravel- and cobble-bed streams for analyses in sediment transport, hydraulics and streambed monitoring, Gen. Tech. Rep. RMRS-GTR-74, Fort Collins, CO, U.S. Department of Agriculture, Forest Service, Rocky Mountain Research Station. Carling, P.A., and N.A. Reader, 1982, Structure, composition and bulk properties of upland stream gravels, Earth Surface Processes and Landforms 7, 349-365. Fehr, R., 1987, Geschiebeanalysen in Gebirgsflüssen, Mitt. der Versuchsanstalt für Wasserbau, Hydrologie und Glaziologie, ETH Zürich 92. Unauthenticated Download Date | 6/14/17 4:54 PM 32 J. ABERLE Komura, S., 1963, Sediment transport mechanics: Introduction and properties of sediment, J. Hydraul. Div. 89, 1, 263-266. Lisle, T.E., 1989, Sediment transport and resulting deposition in spawning gravels, North Coastal California, Water Resour. Res. 25, 6, 1303-1319. Marion, A., and L. Fraccarollo, 1997, New conversion model for areal sampling of fluvial sediments, J. Hydraul. Eng. 123, 12, 1148-1151. Milhous, R.T., 2002, Measurement of the bed material of gravel-bed rivers, Hydraulic Measurements & Experimental Methods 2002, Estes Park, Colorado, USA (papers on CDrom). Nikora, V.I., D.G. Goring and B.J.F. Biggs, 1998, On gravel-bed roughness characterisation, Water Resour. Res. 34, 3, 517–527. Nikora, V.I., D.G. Goring, I. McEwan and G.A. Griffiths, 2001, Spatially averaged open channel-flow over rough bed, J. Hydraul. Eng. 127, 2, 123-133. Nikora, V.I., K. Koll, I.K. McEwan, S.R. McLean and A. Dittrich, 2004, Velocity distribution in the roughness layer of rough-bed flows, J. Hydraul. Eng. 130, 7, 1036-1042. Nikora, V., I. McEwan, S. McLean, S. Coleman, D. Pokrajac and R. Walters, 2006a, Double averaging concept for rough-bed open-channel and overland flows: Theoretical background, J. Hydraul. Eng. (accepted). Nikora, V., S. McLean, S. Coleman, D. Pokrajac, I. McEwan, L. Campbell, J. Aberle, D. Clunie and K. Koll, 2006b, Double averaging concept for rough-bed open-channel and overland flows: Applications, J. Hydraul. Eng. (accepted). Oppermann, M., 2005, Bestimmung von Sohlrauheiten bei Deckschichtbildung, Thesis, Leichtweiß-Institut für Wasserbau, TU Braunschweig. Sibanda, E., I. McEwan and A. Marion, 2000, Measuring the structure of mixed-grain-size sediment beds, J. Hydraul. Eng. 126, 5, 347-353. Stephan, U., 2001, Zum Fliesswiderstandsverhalten flexibler Vegetation, PhD thesis, Technische Universität Wien, Vienna. Toro-Escobar, C., C. Paola and G. Parker, 1996, Transfer function for the deposition of poorly sorted gravel in response to streambed aggradation, J. Hydraul. Res. 34, 1, 35-53. Veličković, B., 2005, Colmation as one of the processes in interaction between the groundwater and surface water, Facta Universitatis, Series: Architecture and Civil Engineering 3, 2, 165-172. Vollmer, S., 2005, Einfluß der Oberflächenströmung auf die permeable Gewässersohle, Mitt. des Instituts für Wasser und Gewässerentwicklung, Heft 231, Universität Karlsruhe (TH), Karlsruhe. Received 19 September 2006 Accepted 17 October 2006 Unauthenticated Download Date | 6/14/17 4:54 PM
© Copyright 2026 Paperzz