Laserfiche WebLink
<br />- <br />. . <br /> <br />.. <br /> <br />supercritical Flows Versus Subcritical Flows <br />in Natural Channels <br /> <br />By Douglas J. Trieste, P.H.' <br /> <br />Abstract <br /> <br />Many analysts model large discharges and/or <br />high-qradient reaches as supercritical flow. While <br />this assumption may be valid for man-made channels of <br />smooth, non erosive materials, and some smooth <br />uniform, bedrock natural channels, it is questionabl~ <br />for most natural channels typically modeled. More <br />research is needed in this area to assess flow regime <br />in natural channels, and to properly evaluate flow <br />resistance especially for large flO107s, and/or <br />high-gradient channels. <br /> <br />Introduction <br /> <br />There are two schools of thought on flow regimes in <br />natural channels. One is that super critical flows do <br />not occur except for short channel reaches. The other <br />is that both supercritical and subcritical flO107e do <br />occur in long channel reaches depending on flow <br />resistance and slope. <br />Occurrence of supercritical flow was not questioned <br />until relatively recent times (around the last <br />10 years). But, more analysts are performing open- <br />channel hydraUlic studiee under the premise that <br />supercritical flow does not occur in natural channels <br />except for short reach lengthe. <br />Also, because most open-channel flow computations <br />use some form of Manning'S equation for computation of <br />velocity, traditional Manning'S n values (n-values) are <br />being questioned. That. is, for certain sit.uat.ions, <br />n-values may be underestimated resulting in <br />supercritical flow assumptions, whereas, flows are <br />truly in the suhcritical range. <br /> <br />Hydraulic Engineer, Bureau of Reclamat.ion, <br />P.O. Box 25007, Denver CO 80225 <br /> <br />132 <br /> <br />I <br /> <br />. <br /> <br />. <br /> <br />SUPERCRITICAL v, SUBCRITICAL FLOWS <br /> <br />133 <br /> <br />This paper discusees both sChools of thought. on <br />flow regime. It pertains to flow regime assumed, or <br />calculated, in reach lengths used for open-channel <br />flow model st.udies (e.g., inundation studies, flow <br />routings) . <br /> <br />subcritical Flow <br /> <br />within high-gradient channel reaches, supercrit.ical <br />flow may occur for very short distances, say, 1 to <br />25 ft (0.3 to 7.6 m), but it normally changes back to <br />subcrit.ical flow because of extreme energy dissipation <br />such as hydraulic jumps, turbulence, obstructions, et.c. <br />Thus, flows (both large and small) in steep natural <br />channel reaches usually alternate between supercritical <br />and subcritical. Jarrett and costa (1986) noted the <br />occurrence of predominantly subcritical flow for a <br />dambreak flood in a channel reach wit.h slopes t.hat. <br />average O. 10. In a discussion of an analysis of <br />hydraulic data on high-gradient streams, Jarret.t (1984) <br />states that. streams thought. to be in the supercritical <br />flow range were actually in the subcritical flo" range. <br />An approach for estimat.ing Manning n-values (n-values) <br />for large floods (Trieste and Jarrett, 1987) is the <br />hypothesis that. for large floods supercritical flow <br />generally does not occur in natural channels except for <br />short distances. <br />A review of the literature reveals no papers that <br />verify the occurrence of supercritical flow channels <br />over a significant reach length. Also, years of <br />exp~,ience in open-channel flow modeling and personal <br />communication with other modelers has revealed few <br />situations where supercritical flow ie justified along <br />a channel reach longer than around 25 ft (7.6 m). This <br />can be seen by observing flows along most any high- <br />gradient stream and noting t.he length of reach where <br />supercritical flow is occurring. <br />Theoretically, the larger the discharge the more <br />likely flows are to be in the supercritical range for <br />a constant slope and n-value (Fread, 1988). However, <br />this does not appear to be the case in the natural <br />environment. (Trieste, 1992). <br />The main premise for non-occurrence of super- <br />critical, flows in natural channels is that. flow <br />resistance increases to the level needed for <br />predominat.ely subcritical flow to occur. As channel <br />gradient increases, and thus energy of flow, so do the <br />effects of the factors contributing to flow resistance <br />thereby checking velocity and maintaining flows in the <br />subcritical range. <br />Many factors, in addition to streambed and bank <br />characteristics (Le.. boundary frict.ion) come into <br />play when discharge and/or slope increases (Trieste and <br /> <br />-.. <br /> <br />" <br /> <br />. <br />