FLOOD07353 - Laserfiche WebLink

Home Browse Search

.' . Effects of particle Shape on Bedload Transport M. Moorel and P. Diplas2, Member, ASCE Abstract The effects of particle shape on bedload transport in gravel-bed streams are examined ".sing a similarity al?proach and fractional transport analysl.s for data from P~ceance Creek Colorado, which contains flat, low density, shale parti~les. The reference transport critical shear stress for the median surface grain size, "tr5o.*' for the flat shale particles in piceance Creek is approximately 2.5 times higher than those for more spherical particles. This indicates a lower susceptibility of disc-like particles to initial entrainment and lower transport rates for given flow conditions than more rounded particles. Introduction The experiments on which the widely used Shields' curve for incipient motion is based are for nearly spherical, uniform grains on a flat bed. Since most bedload transport relations do not specifically account for the effects of particle shape on grain motion these effects must be studied for streams containing either a variety of particle shapes or an abundan,ce. of a par~icular sh~pe. Particle shape may be quant~f1ed by uS1ng the Z1n/~g classification or the Corey Shape Factor, CSF = c/(ab)' , where a, h, and c are the longest, intermediate, and shortest grain axes respectively. IGraduate student, Department of Civil Engineering, Virginia polytechnic Institute and State University, B1acksburg, VA 24061 2Associate Professor, Department of Civil Engineer~nq, Virginia polytechnic Institute and State Univers1ty, B1acksburg, VA 24061 800 -.- . . EFfECTS OF PARTICLE SHAPE SOl Research on the relative mobility of discs versus sub- or well-rounded particles has produced conflicting results. However, the majority of the investigations suggest that discs are less mobile. Flat, disc-like particles have been observed to be imbricated, whereby one grain rests atop another with onB end tilting up in the direction of flow. Mantz (1980) suggests that flat, imbricated grains have increased bed stability and lower transport rates relative to non- imbricated grains for the same flow conditions. Komar and Li (1986) attribute their increased stability near threshold conditions to higher pivot angles exhibited by imbricated and flat grains. Similarly, Lane and Carlson (1954) found that disc shaped grains are less susceptible to motion than spherical particles of equal weight. Ashworth and Ferguson (1989), noticed that spherical particles moved farther than flatter particles, which indicates lower transport rates for discs. carling, et al (1992) found lower grain velocities for discs than for spheres for low flow velocities, however, at higher flow velocities the trend was reversed. However, Magalhaes and Chau (1983) concluded that flat, low density shale particles have lower resistance to initial motion. The shale sediments had critical shear stresses 15% lower than those of Shields' diagram and 40- SO, lower than those reconunended by the U. S. Bureau of Reclamation for the design of channels. Current study on the Effects of Particle S~ Bedload transport data from Piceance Creek, Colorado, a qr~vel-bed stream containing mostly flat shale particles, prov1de an opportunity to further investigate the relative m~bility of disc-like versus more spherical grains. Plc~ance Creek has an average CSF = 0.3 for the largest graIns, an average grain specific gravity of 2.1, a bulk mixture median grain size, 050 = 5.05 mm, and a surface layer median grain size, 0508 = 14 mm. The Piceance Creek data are used to calculate fractional bedload transport rates with a similarity al?proach analogous to that of Parker, et al (1982) dnd Dll?l~s (1987) for Oak Creek. Reference transport rate crlt1cal shear stresses are determined for comparison with other data to examine the relative mobility of the flat particles of Piceance Creek versus more rounded particles found in other streams. Dimensionless transport rates for each size range, Wi.' are plotted against dimensionless shear stress for each size range, '(i*' in Fig. 1. Wi* and '(1* are given by Eqs. 1 & 2, where the subscript i refers to the ith grain size r~nge; qe1 is the volumetric bedload transport rate per unit w1dth: q8i. is the Einstein bedload parameter; Oi and l " ..