EXPERIMENTS OF INFLUENCE OF DISCHARGE PROCESS ON CHANNEL PATTERN

CHINESE GEOGRAPHICAL SCIENCEVolume 14. Number 4. PpP. 33336, 2004Science Press, Beijing, ChinaEXPERIMENTS OF INFLUENCE OF DISCHARGE PROCESSON CHANNEL PATTERNZHANG Jun-yong. CHEN Li(Stute Key Labordory 0/ W ater Resoure and Hyrpowver Engineering Science. w uhanUniversity, W uhan 430072. P. R. China)ABSTRACT: Based on experimental data and theory, by mcans of simplified discharge durations in a small flume,the infuence of discharge process on channel morphology and channel pattem was analyzed in this paper. It wasconeluded that on the same original channel, diferent discharge and channel conditions would end with diferent riv-er morphology, including thalwegs and radius of bends. Different discharge process resulted in two kinds of change:tiny change in the process of "igsmall-big" and distinct change in the process of malig-small". Flood dischargeduration was verified to be the main cause in the discharge process. Proper discharge process will change the mor-phologics of river, even can led to channel pattem transformatin. The influences based on the relationship betweenthe flow and the channel iself, inchuding slope and riverbed constitution. Although not be a main cause, originalchannel morphology may infuence its final channel pattem. Neglecting the influence of channel itself will hamper theunderstanding of channel ptterms.KEY WORDS: discharge process. channe patem, food durationCLC number: P931.1Document code: AAricle ID: 1002-0063(2004)04-0332-051 INTRODUCTIONthan long period of moderate flow; its remainder mayexist even after a long period of time. Different dis-Many researchers have worked on formation and trans-charge processes may induce different channel pat-formation of channel pattern by means of flume exper- terns, such as the rivers in north and south Blue Moun-iments since FRIEDKIN (1942), and a lot of achieve- tain, Jamaica (QIAN, 1987).ments have been gotten, but some problems still exist. It is well known that the building of reservoirs willTAN Ri-chang (1964) believed that the constitution of influence reaches downstream, even may cause chan-river channel was important to the formation of mean- nel pattem's transformation, such as the change fromdering rivers, especially the inconsistent constitution of braided river into meandering river in the lower reach-riverbank and riverbed. YIN Xue-liang (1965, 1995) es of the Han River after the building of Danjiangkousuggested that the flow and sediment inflow act prime Reservoir. That the reducing of sediment discharge is aparts in channel pattern' s formation. They both shaped main cause has been well acknowledged, but the influ-typical meandering rivers although together with some ence caused by change of discharge process is alwaysartificial measures. In the 1970s， SCHUUM and neglected and still kept unknown.KHAN (1972) successfully realized the formation ofIn the simplified flume experimental study on thedifferent channel patterms by changing the channel transformation of channel pattern downstream reser-slope and sediment discharge. By using of a mixture of voirs， some runs were designed to study the influencediatomaceous earth and kaolinite clay as grain materi- of the discharge process on channel patterms.al, SMITH (1998) succeeded in modeling high sinuosi-ty meanders in a relative small flume.2 EXPERIMENTSIn nature, it is common for the discharge process toreform rivers. One flood can influence the river more The experiments were performed in a glass flume withReceived date: 2004-08-27中国煤化工13 Program) (No.2003CB415205) and the National Natural Scienee Foundation of China (NBiography: ZHANG Jun yong (1977-), male, a native of Ji' an of JiangMHC N M H Gized in river engineer-ing. E-mail: zjy0796@sina.comExperiments of Infuence of Dischange Process on Channel Patterm333a changeable slope. The flume was 4m long, 1.2m wideAll channels were shaped by clear water freely. Aand 0.4m high (Fig. l). With that installation, dischar- straight channel with a triangle cross-section was ini-ge, slope and riverbed constitution all are changeable tialized before each run. The time of each dischargeaccording to the demand of experiments.process was long enough for the river to reach a stateof dynamic equilibrium. The range of Reynolds num-StrobeGlass flumeber was 1500- 14000 according to the average of allcross-sections.Reservoir3 RESULTS AND ANALYSISWater pump FlowmergLiftable brackt3.1 On Original ChannelAccording to the experiments, different discharge pro-Fig. I Isallations in simplified experimentcess may lead to the change of channel pattern，whichinitially began at duration 1- -on the same originalIn total 8 runs of experiments were performed. A- channel， different flows may cause variable channelmong them，three kinds of grains were adopted, in- morphologies.cluding sands and fine coal particles with the specific The rivers formed were firstly based on riverbed ma-gravity of 1.5 tm' . The median diameter of diferent terial. In runs 5, 6 for the coarsest sand, lite sinuousgrains was 0.42mm，0.095mm and 0.15mm in se- river was formed. With the increasing of median diam-quence. The discharge process was simplified as three eter of sand, rivers with more sinuous thalweg can bedurations. The durations in turm may be one flood dis- gotten in runs I, 2. Since the relative small density,charge duration between two small discharge durations coal particle was too light to resist the shaping of flow;(small-big- small), or one small discharge duration be- rivers with coal particles had few steady bars, andtween two flood ones (big-small-big). Three discharges more characteristics of braided rivers were formed in(from 0.5-1.5m/h) and three slope (from 0.6%0- run 7 and 8, duration I.15%o; ) were adopted. The parameters of the experi-The amount of discharge will also influence the chan-ments are shown in Table I.nel morphologies. Three couple runs,1 and 2, 5 and 6,Table I Parameters of experimentsRun Graindo(mm) Slope (%0)Duration IDuration 2Duration 3Discharge (m/h) Time (h)__ Discharge (m/h)_ Time (h) Discharge (m/h) Time (h)Sand0.09503121.501060.7548391 .5014.0184659188.5199642sand0.4201980301130089Coal particle 0.1500.5098Coal paricle 0.150597 and 8, had the same experimental conditions except cess of "big-small-big"， the change of the river mor-discharge process. According to the appearance, chan-phology in the end of the first duration and the end ofnel made by flood was wider and larger in cross section the third duration was very small. It was observed thatarea, at the meantime the river's width-the bank-lines widened in a very limited extent and tinydepth-ratio decreased, the river tended to be shallow in change of morphologies in cross-sections was takenthe whole. Thalwegs and main flows were more distinct place.in rivers with lttle discharge than that with flood dis-The final bank lines in different durations of run 2charge. The radius of bends were also different, in were shown in Fig. 2. It was found that the later chan-rivers with flood discharge，the radius was 17.15%-nels coincided with the original channel in ,the final of31.2% bigger than that with lttle discharge.durat中国煤化工3.2 Influence of Small Discharge Duration3.3I1YHCNMHGrationIn the three runs (runs 2, 6, 8) with the discharge pro- In the tive experments (runs I, 5, 4, 5, 7) with the dis-334ZHANG Jun-yong, CHEN Licharge process of "sallig-small"， distinct change terminate after 48 hours of duration 3. Fig. 4 was the fi-appeared in the rivers between the first duration and the nal morphology after the "small-big-small" dischargethird duration. As shown in Fig. 2, the average width of process and a roughly meandering river was formed.channel may increase 41 .4% after the flood dischargeduration in run 1.3.4 Influence of Channel ConditionsAccording to the experiments, the influence of dis-charge process should also be based on the channel con-8*ditions especially the riverbed constitution and channel1400slope.120010003.4.1 Riverbed constitution800Because the channel was formed in the uniform experi-600mental sands, the riverbed conditions were mainly de-1 400cided by which sand we chose. Compared to the 8 ex-; 1000Run2perimental rivers in duration 3, with the different riverbed constitutes, the relationship between flow and chan-60nel may be shown as follows (Fig. 5):(1)Channels changed minutely in the small dischar-400sc100ge duration, and flow was restricted by channel (run 5).Longth (mm)(2) Channels evolved following the original channel,Fig. 2 Final bank line of dfferent durations (runs 1, 2)the channel restricted flow, at the mean time reshapedby the flow (run 1).The lateral morphology of rivers also changed during(3) Rivers evolved far beyond the original channelthe process of "smal-big small" . Fig. 3 shows the final and were fully reshaped (run 7).cross-section topography in different durations in runs I,It can be concluded that with the increase of mean3 and 4. With the procedure of different discharge pro- diameter of sands, which means a bigger stability, thecess, the channel changed both in width, width-depth- restrictive acting of channel to flow increases, at theratio and the position of thalwegs. In crossection 6#,mean time, the reshaping acting of flow to channel de-the position of the bottom transfered to and fro with ex-creases correspondingly. Run 5 has the thickest consti-tend of more than 0.2m. Changes happened both in du- tution of riverbed, the reshaping of flow to channel inration 2 and duration 3, compared to the topography induration 3 was the smallest. But in coal particle, thcrossection 8#, run 1, the channel was extended to channel cannot resist the reshaping of flows，and themore than 200mm in duration3; and a distinct ero-change was great.sion and deposition can be seen in cross-section 4#,3.4.2 Sloperun 4.As shown in Fig.2, three runs (runs 1, 3, 4), with thePhenomena in Run 1 were typical and end with trans-channel slopes of 1.0%0, 0.6%0 and 1.5%0 respectively,formation of channel patterns. Beginning with a rela- have different channel topographies after duration I,tively small discharge (0.75m/h) in duration I, no dis- and after the full discharge process, the discrepanciestinct mainflow was formed when the river had been sta-enhanced. The change in run 3 was relative small, but inbilized; the channel patterm was straight with some run I, the change can be distinct, even typical character-characteristics of debauchery. In duration 2, the channelistic of meandering river appeared in the lower reachexpanded but no distinct thalweg appeared. In the dura- (Fig. 4).tion 3，when the discharge reduced back to 0.75m/h,the river went through the thalweg formed in duration 2,3.5 Flood Duration and Stability of Barsshoals under water in duration 2 came out of the water. The influence of discharge process on channel pattern,Thalweg and lateral current appeared. With time goingwhich mainly means the influence of flood duration inon, the phenomenon of deposition in inner bank and our experiments, may be caused by two main factors:erosion in outer bank also became distinct. Sinuosity of the change in flood duration and the stability of bars.the river increased to more than 1.5 especially in thercess of "small-big-lower river. The curve ratio kept increasing and the riv- sm |中国煤化工duration, all riverser had expanded to the edge of the flume near cross-sec-:ha:YHC N M H G of flood discharge,tion 8# in the lower river, experiment one was forced to wheuer cnange une caui s imophology directly，orExperiments of Influence of Dischange Process on Channel Pattern3354#(Run 1)6#(Run 1)8#(Run 1)27090 110315070 90 110 1350050Widtb(cm)6# (Run3)8# (Run3)09040 6030 100 12060 80 100 124# (Run 4)6# (Run4)8# (Run4)trt406080100120406080100120- + Duration 1-Duration 2- + Duration 3Fig. 3 Final crosection topography in dfferent durations口3.0cm2.5cm2.0cm1.5cm1000-1 .0cm0.5cm0.0cm .-0.5cm-1.0cm500-2.0em .s00100015002000 .3000Length (mm)Fig. 4 Final morphology in run 1, duration 3Original morphologyFinal morphologyduring this process.The key of the influence may lie in the stability ofRunI.bars formed in the flood duration. According to the ex-Run5periments, most rivers in flood duration had sinuousthelwags, distinct or not. When the discharge decreasedback to be small, flows fall into the thelwags, bars e-merged from underwater, and a transitory meandering6____ 7Bars under water， Bars beyond waterriver formed. The bars were surely to be reshaped bythe new small flow, some were distinguished, but be-Fig. 5 Sketches of channel change in duration 3cause of the different capacities of shaping of floodand small discharge, most of them existed in some ex-change the physical characteristic or stress structure of tend. When the transitory meandering river can keepchannels by means of erosion, moisten, so as to break meandering， channel patterm transformation emerged,the equilibrium formed in the small discharge duration.such as the river in run 1When the discharge decreased back to small dis- Acq中国煤化工ments, the intrinsiccharge in duration 3，flow and channel had to begina resulme as a reshapednew process of adjusting from un-equilibrium to equi- chant:YHC N M H Gns the influence oflibrium, and channel patterm transformation may appear original channel morphology on channel patterns. On336ZHANG Jun-yong, CHEN Limeandering river, PRUS-CHCINSKI (1966) believed tions and the channel slope. For dferent river condi-that the existence of a bend is also infuenced by the tions, this infuence may be dfferent.bends in sequence, and YANG (1971) believed that to(4) The key of the influence may lie in the stability ofmaintain a bend needs less energy than to form one.bars formed by the flood duration, which means the in-In recent, on the basis of bankfull specific stream, fuence of original channel morphology on channel pat-the proposed distinction between meandering and braid- terns. Neglecting the influence of channel itself willed river channel pattemrs was studied, LEWIN (2001) hamper the understanding ofchannel pttrms.analyzed and rejected this point, and concluded that theREERENCESuse of a single-stage stream power measure and bed ma-terial size alone is unlikely to achieve. We believe thisFRIEDKIN J F, 1942. A laboratory study of the meandering ofmay be the result of neglecting the influence of channelAlluviol River [R]. U.s. Water, Exp. Sta.itself.LEWIN J, BREWER P A, 2001. Prediting channel pttrms [凹Geomorphology, 40: 329 -39.4 CONCLUSIONSPRUS-CHCINSKI T M, 1966. Discussion on "Critica analysisof open-channel resistance" by H.Rouse []. ASCE Jourmal(1) On the same original channel, different dischargesof Hydraulics Diversion, (3) 389- -393.and dfferent rivertbed materials will end up with dffer- QIAN Ning, ZHAN Ren, zHOU Zhide, 1987. Eolvememt ofent river morphologies, including thalwegs and radius。Rives [M]. Beiing; Seience Pess. (in Chinese)SCHUMM S A, KHAN H R, 1972. Experimental study of chan-of bends.nel patterns []. Bulletin of the Geological Society of Ame-(2) Different discharge processes may reshape the riv-rica, 83: 1755-1770.er and end up with different river morphologies. The in-SMITH C E, 1998. Modeling high sinuosity meanders in a smallfluence of discharge process was mainly laid on the flume [J]. Geomorphology, 25: 19- 30.flood discharge. In the discharge process of "big-small- TAN Ri-chang, 1964. Elemental analysis and experimentalbig", the river morphology kept nearly unchanged. Af-study on the formation of meandering rivers []， Yangteter the flood discharge duration in process of "small-big-River, (2): 13-21. (in Chinese)small"，channel shaped by small discharge durationYANG Chin Ted, 1971. On river meanders [] Journal of Hy~-drology, 13: 231-253.may be fully reformed; even channel pattern can beYIN Xue-liang , 1965. Experimental research on channel pattermtransformed when the conditions were appropriate.[J]. Journal of Geography, 31(4): 287-303. (in Chinese).(3) The extension of this infuence is also decided by YIN Xue-liang, 1959. Characer of Lower Yellw River [M].the river conditions itself including the channel condi-Bejing: China WaterPower Press. (in Chinese)中国煤化工MYHCNMHG