www.seipub.

org/scea

Study of Civil Engineering and Architecture (SCEA) Volume 2 Issue 1, March 2013

ParameterSelectionsinSimulatingthe
PhysicalDiffusionPhenomenaofSuspended
LoadbyLowOrderDifferentialScheme
NumericalDispersion
ShuaiChen*1,JunxingWang2
StateKeyLaboratoryofWaterResourcesandHydropowerEngineeringScience,WuhanUniversity
Wuhan,Hubei,China
*1

liveangela@whu.edu.cn;2jxwang@whu.edu.cn

A
AS QS
S

ADx
(S r ' ) S
t
x x
x
t

Abstract
In this paper, the process of the introduction of numerical
diffusionisinvestigatedwhenloworderdifferentialscheme
inconvectiondiffusionequationfordiscretesuspendedload
is in use. According to the results of instances calculation,
theoptimalchoiceforparameterswhenphysicaldiffusionis
simulated by numerical diffusion is obtained. Finally,
comparison between several commonly used loworder
schemesisconductedandthemostviableloworderscheme
in simulating physical diffusion by numerical diffusion is
proposedindealingwithsuspendedload.

Introducingtheflowcontinuityequation,theequation
abovecanberewrittenas:

Keywords

Generally,fortransportationissueofsuspendedload,
the longitudinal diffusion of suspended load is much
smaller than the longitudinal conviction of sediment,
which is often overlooked. If the carrier velocity and
diffusion coefficient Dx are constants without

S Q S 1
S ( S r ' ) AS

ADx

t A x A x
x
t
A
where Dx the longitudinal diffusion coefficient, for
sedimentdiffusion,itcanbeapproximatedas

Dx 0.25u*h

DifferentialScheme;ConvectionDiffusionEquationofSuspended
Load;NumericalDiffusion

Introduction

considering the source item, the convectiondiffusion

equationwillbe

In general, using the loworder scheme for numerical

calculations will more or less introduce numerical
diffusion (Xie 1990). Different schemes have varying
degrees of numerical diffusion. In theory this is not
conducive to the exact solution of numerical
calculation,bututilizingthenumericaldiffusion,such
as to simulate the physical diffusion phenomenon, is
also an effective method (Yang 1993). As for how to
use the numerical diffusion to simulate the physical
diffusion,and how to select parameters toachievean
acceptablelevelofthesimulationresults,theyarethe
issuestobediscussedinthispaper.

S
S
2S
u
Dx 2
t
x
x
For this equation, the main numerical difficulty is to
calculate the convection term because it strictly
demands the conservation of matter, while the
differential method for numerical solution often
cannot achieve this. If the solution to convection
diffusion is considered under the premise of a good
solutiontoconvectiondiffusionterm,theprobability
of successful numerical solution would be higher. If
ignoring the diffusion term, we can get the pure
convectionequation

Convection-Diffusion Equation for OneDimensional Suspended Load

S
S
u
0
t
x

Equationforsuspendedloadmovementis(Zhengand
Zhao2001):

12

Study of Civil Engineering and Architecture (SCEA) Volume 2 Issue 1, March 2013

Establishment of Differential Scheme

Fromtheaboveproofitisclearthat,Upwindscheme
dispersesapureconvectionequationandanequation
similar to convective diffusion can be obtained.

Select Upwind scheme discrete pure convection

equation, when u 0 , the differential equation is
(Courant1928)

maketheequationcompatible,calculationshowsthat
the item also makes the solution of differential
equations no longer converge to the solution of the
original differential equation. Instead it converges to
the solution of the convectiondiffusion equation,
which means the implicit numerical diffusion,
resulting from thelimitation of x , t , which cannot
beinfinitelyclosetozero.However,sinceitisknown
that difference equations converge to the convection
diffusion equation, then it is used to identify the
required parameter values, so that the differential
equations can converge precisely to differential
equations, namely numerical diffusion can simulate
thephysicaldiffusionwell.

t / x
where Cr isCroutnumberand Cr C
Thedifferentialequationcanbereducedto

S nj S nj 1
x

The first item in the above equation is the forward

difference quotient of S / t , and the second term is

u
backward difference quotient of S / t , then C
can be drawn. We will exam the compatibility,
stability and convergence of the scheme as follows.
n 1

Expand S j

and S j 1 onpoint(j,n)

n 1
j

S
1 2S
S ( )tn t ( 2 ) nj t 2 ...
t
2 t

n
j 1

S
1 2S
S ( ) nj x ( 2 ) nj x 2 ...
x
2 x

The Instance Analysis of Parameters Effect

n
j

As differential equation is similar in form with

convectiondiffusion equation, Dn might be called
numerical diffusion coefficient. It is clear that the

n
j

closeness between numerical diffusion coefficient Dn

Take them into the differential equation, and use the

pureconvectionequationtoget

and physical diffusion coefficient Dx can determine

the accuracy of the numerical diffusion in simulating
physicaldiffusion,thustheproblemistofindtheright

S
S x 2Cr (1 Cr ) 2 S

t
x
2t
x 2
Assume Dn

parameters,sothat Dn canbecloseto Dx (Papadakis

and Metaxas 2011). On the other hand, three

x 2Cr (1 Cr )

2t

parameters can determine Dn which are velocity u ,

space interval x and time step t , where flow rate
should be in accordance with the actual engineering
value, and x is subject to the restrictions on the
boundarysimulationaccuracy(itwilltakeondifferent
values in accordance with the different size of the
project and the complexity of the border. Usually,for
example, the smaller the size of x is, the more
precise the boundarysimulation results will be. But it
willincreasetheamountofcalculation),onlythetime
step t has wider choice. By selecting different t
we can get the corresponding numerical diffusion

S
S
2S
Then
u
Dn 2
t
x
x
Comparing to the pure convection equation, only the
right term of the equation tends to be zero when the
space and time step are small enough, which means
theyarecompatible.Taking t Cr x / C into Dn ,we
canget D xC (1 Cr )
n
2

where C u , 0 Cr 1 , when x 0 , t 0 , and

Dn 0 . So the difference equations and differential

coefficient Dn . Then comparing Dn with different

equationsarecompatible,asthestabilityconditionfor
Upwind is u t / x 1 . Using Lax equivalence

longitudinal diffusion coefficient Dx , the impact of

t on calculated
(Gasiorowski2013).

theorem again we can get that the solution of

differential equation converges to the solution of
differentialequations.

Although the differentiated items Dn S / x can

S nj 1 S nj Cr ( S nj S nj 1 )

S nj 1 S nj

www.seipub.org/scea

results

can

be

examined

Buildthefollowingmodel:

13

www.seipub.org/scea

Study of Civil Engineering and Architecture (SCEA) Volume 2 Issue 1, March 2013

In the results (Table 2) when t is greater than x ,

numerical diffusion coefficient is negative. It will
repeatthelastcalculationlawwhentheyareequal.In
factiftheformulaischangedfor Dn ,wecanfindthe

Model1
Length is 10km, width is 2m, water depth is 4m,
Chezy coefficient C is 25, and flow rate is 0.5m/s,
x =200m. Take respectively t =400s, 399.5s, 399s,
360s,300s,200s

contradictions.

Thecalculationresultsareshownasfollows(Table1).

Dn

TABLE1CALCULATIONRESULTS(U=0.5M/S)

x 2Cr (1 Cr ) ux u 2 t

2t
2

When u =1m/s, Dn ( x t ) / 2 , obviously x needs

t

Parameters

(s)

to be larger than t to make Dn positive, in fact, this

Cr

Dn

Dx

400

0.063246

399.5

0.99875

0.0625

0.063246

399

0.9975

0.125

0.063246

360

0.9

0.063246

300

0.75

12.5

0.063246

iscausedbythefailuretosatisfythestabilitycondition
u t / x 1 . And, from another perspective, the
accuracy of Dn shows direct linear relationship with
the precision of x as well as t , which demands
highly for x and t , in the actual calculation, it is
oftenimpossibletomeetsucharequest,sootherflow
ratesshouldbetakentocalculateagain.
TABLE2CALCULATIONRESULTS(U=1M/S)

200
100
50
0.5

0.5
0.25
0.125
0.00125

25
37.5
43.75
49.9375

0.063246

(s)

Cr

Dn

Dx

400

100

0.126491

300

1.5

50

0.126491

200

0.126491

199.5

0.9975

0.25

0.126491

199

0.995

0.5

0.126491

198

0.99

0.126491

100

0.5

50

0.126491

50

0.25

75

0.126491

0.5

0.0025

99.75

0.126491

0.063246
0.063246
0.063246

In the above table, only when t =399.5s, the

differencebetweennumericaldiffusioncoefficient Dn
and longitudinal diffusion coefficient Dx is the
smallestat0.000746.Atthesametime,thedifference
between Cr and 1 is 0.00125, when t decreases by
0.5sto399s.Thetwodiffusioncoefficientsdiffersinan
order of magnitude, indicating that only when Cr is
verycloseto1thephysicaldiffusioncanbesimulated
more accurately. On the other hand, the numerical
diffusion coefficient Dn will increase with the

Model3

decrease of Cr . When Cr tends 0, a solution with

Flow velocity u =0.8m/s, taking t =250s, 249.5s, 249s,

200s, 100s, 50s The remaining parameters are
unchanged.

largererrorwillbegot.
Here the time step is accurate to the extent of 0.5,
which is not conducive to calculation. Since the time
step may beinfluenced by the flow velocity, the flow
rateischangedandthemodelismodifiedasfollows:

Model4
Flow rate u =0.2m/s, taking t =1000s, 999.5s, 999s,
998s, 900s, 800s The remaining parameters are
unchanged.

Model2
Flow velocity u =1m/s, taking t =400s, 300 s 200s,
199.5s, 199s, 198s other parameters are unchanged.

14

Parameters

CalculationresultsareshowninTable3andTable4
respectively.

Study of Civil Engineering and Architecture (SCEA) Volume 2 Issue 1, March 2013

u =0.5m/s;

TABLE3CALCULATIONRESULTS(U=0.8M/S)

Parameters

(s)

Cr

Dn

Dx

250

0.101193

249.5

0.998

0.16

0.101193

249

0.996

0.32

0.101193

200

0.8

16

0.101193

100

0.4

48

0.101193

50

0.2

64

0.101193

0.5

0.002

79.84

0.101193

(s)

When u is in the interval (0,1], although the

closer Cr is to 1, the better the simulation is,
however within a small range close to 1, it is
not in line with this rule. As shown in the
followingtable.
TABLE5VARIATIONS IN PARAMETERS WITH DIFFERENT U

(m/s)

(s)

0.2

Parameters

When u is in the interval (0,1], the simulation

is better when the time step value is 0.5s
smaller than the t that makes Cr =1, but
thedifferenceof0.5sisnotideal;

TABLE4CALCULATIONRESULTS(U=0.2M/S)

www.seipub.org/scea

Parameters

Cr

Dn

Dx

D x Dn

999.5

0.9995

0.01

0.025298

0.0153

60.48

0.5

399.5

0.99875

0.0625

0.063246

0.0007

1.11

Cr

Dn

Dx

0.8

249.5

0.998

0.16

0.101193

0.0588

58.11

1000

0.025298

1.0

199.5

0.9975

0.25

0.126491

0.1235

97.64

999.5

0.9995

0.01

0.025298

2.5

999

0.999

0.02

0.025298

998

0.998

0.04

0.025298

900

0.9

0.025298

u=0.2m/s
u=0.5m/s
u=0.8m/s
u=1.0m/s

Cr

1.5
1
0.5

800

0.8

0.025298

500

0.5

10

0.025298

100

0.1

18

0.025298

0.5

0.0005

19.99

0.025298

0
10

100
lgt(s)

1000

FIG.1 DISTRIBUTION OF Cr WITH GROWTH OF THE TIME STEP t (IN LOG

FORM)
120
u=0.2m/s
u=0.5m/s
u=0.8m/s
u=1.0m/s

100

Combining the above four models; we can draw the

followingrules:

Dn

80

When u isintheinterval(0,1],thelongitudinal

40

diffusion coefficient Dx increases as u

20

increase, and the maximum value is close to

0.126 (this is the maximum value when the
waterdepthis4mandChezycoefficientis25,
dependingonthecircumstances,thevaluewill
bedifferent);

60

0
10

100
lgt(s)

1000

FIG.2 DISTRIBUTION OF Dn WITH GROWTH OF THE TIME STEP t (IN LOG

FORM)

Obviously, in the descending process of u , Cr is

When u is in the interval (0,1], if u is small,

the numerical diffusion simulation of physical
diffusion is better, which is the closest when

gettingcloserto1,but Dx Dn isthenminimumwhen

u =0.5m/s.Theexistenceofasmallerdifferencecanbe

15

www.seipub.org/scea

Study of Civil Engineering and Architecture (SCEA) Volume 2 Issue 1, March 2013

tested by changing the value of u , however it is no

longerstudiedhere.

the current model parameter conditions to examine

whether a scheme can achieve a relatively satisfying
stepvalue.

In fact, in order to get the parameters conditions of

numerical diffusion that accurately simulate physical
diffusion, formulas are used to reason reversely,
assume Dx Dn

Comparison among Various Difference

Schemes
Although various differential schemes have different
compatibility and stability conditions, as long as
extracting numerical diffusion coefficient after
dispersing pure convection equations respectively,
then equivalent unify with the physical diffusion
coefficient, we can obtain the calculation formula for
stepsize(Table6).

Weget:
ux u 2 t
u
Dn
Dx 0.25u*h 0.25 g h
2
C
Simplifyittobe:
x u t

0.5 g h
C

Obviously, these three schemes of Dobbins, Upwind

and implicit have the same method to select the step
size, and they also have the same characteristics
represented by the Upwind scheme. The loworder
Preissmann scheme is more suitable for the large
stride length, and the possibility of an integral step
length has been improved in this way, where =0.6,

Byassuming k 0.5 g h ,wecanobtain:

C

x k

Obviously, t isalinearfunctionof x ,where k isa

constant related to the parameters of the physical
model. Take u =0.5, h =4, C =25, x =200m into the
formula then the calculation shows t =399.499 which
is very close to the above 399.5, so the difference
betweenthetwodiffusioncoefficientsisinsignificant.
Butthisstepsizeisveryunfavorabletocalculation; as
it is tendency to take an integer as much as possible.
Although in different engineering cases, different
parameters of the physical model can improve the
value of t , its value depends completely on the
differential scheme itself. In order to keep the
generality, the following will horizontally contrast a
variety of commonly used differential schemes under

=0.4, t =19900s, and Dn =0.05, which shows a

difference of 20.94% with Dx . When t =19870s and

Dn =0.065, Dn show a difference of 2.77% with Dx

andprecisioncanmeettherequirements.Instead,Lac
Friedrichsschemeismoresuitableforsmalltimestep.
Thoughthestepissmaller,thesimulationaccuracyis
not improved, such as when t =247s and Dn =0.097,

t is only0.1ssmaller than247.1s, while the relative

errorofdiffusioncoefficientreachesanunsatisfactory
level,52.83%.

TABLE6RESULTS OF DIFFERENT CALCULATION FORMULATIONS

Schemename

Dn expression

t expression

Dobbins

x 2Cr (1 Cr )
2t

x k
u

399.5

Upwind

x 2Cr (1 Cr )
2t

x k
u

399.5

k x
u

399.5

Implicit

Preissmann

LacFriedrichs

x 2Cr (1 Cr )
2t
2
x Cr (2 1)
Cr (2 1)
2 t
x 2 (1 Cr Cr2 )
2t

k x (1 2 )x 2
2(2 1)

Calculationresults

5x 2 2k x k 2 k x
2u

19874.8

20125.2
247.1

Note:
1.

Inthetable k 0.5 ( gh ) / C ,where h =4m,Chezycoefficient C =25,flowrate u =0.5m/s, x =200m;

2.

Preissmann scheme only consider the loworder condition, i.e. , is not equal to 0.5, , , are respectively corresponding to the
differentvaluesof , ,whereiscorrespondingto =0.6,and =0.4,iscorrespondingto =0.4,and =0.6.

16

Study of Civil Engineering and Architecture (SCEA) Volume 2 Issue 1, March 2013

increasing the calculation accuracy, which is a

recommendedmethod.

Conclusions
The paper starts from Upwind scheme, stating the
process that generates numerical diffusion when
dispersing suspended load convectiondiffusion
equation with a lowlevel difference scheme, and it
tries to simulate the physical diffusion phenomenon
using numerical diffusion, simulation accuracy under
different parameters, such as t , x and u , are
calculated. And by horizontally comparing different
differential schemes we can get the following
conclusions:
1)

2)

www.seipub.org/scea

3)

Numericaldiffusionofdifferentialschemecanbe
used to simulate the actual physical diffusion
phenomenon, but it has certain requirements on
the parameter selection. For suspended load
convectiondiffusion equation, the simulation
method is to disperse the pure convection
equation first, and then propose diffusion term.
Since the coefficient in front of diffusion term is
the numerical diffusion coefficient, by assuming
whichequalstothephysicaldiffusioncoefficient,
wecangettherelationshipequationcomposedof
some parameters that meet the simulation
accuracy.

Thoughthisarticleonlyanalyzedsuspendedload
convection diffusion equation, it can be
speculated that bedload convectiondiffusion
equation, as well as the convectiondiffusion
phenomena of pollutants in the water, can be
analyzed using the same method. Similarly, the
method can be employed to solve the general
problems by means of its unique superiority of
differential equations clear mathematical
foundation, simple calculation and easy
programming,whichisstillfavourablycompared
toamoreprecisenumericalmethod.

REFERENCES

Gasiorowski D. Balance Errors Generated by Numerical

Diffusion in the Solution of Nonlinear Open Channel
Flow Equations. Journal of Hydrology 476 (2013): 384
394.
Papadakis Antonis P., and Metaxas Andrew C. Optimum
Mesh Dependent Diffusion Coefficient Proof from High
to Low Order Upwind Schemes Utilized in Plasma
Discharges. 10th WSEAS International Conference on

Ingeneral,inthevariousparameters,onlyspatial
interval x and time step t which are related
to the calculation can be chosen randomly, and
the selection of x needs to be compatible with
the size of the project and the complexity of the
boundary, so the key is the time step t . By
calculation,thispaperhasfoundtherelationship
of t that meets the accurate simulation under
various differential schemes, but usually the
calculated step sizes are with decimals, which is
not conducive to calculation, so an instance is
used to calculate the relative error between the
numerical diffusion coefficient and the physical
diffusion coefficient when t takes a similar
integer value. It is found that as the step of low
levelPreissmannschemeislarger,thepossibility
of an integral step is greatly increased, thereby

EHAC11 and ISPRA11, 3rd WSEAS Int. Conf. on

Nanotechnology, Nanotechnology11, 6th WSEAS Int.
Conf. on ICOAA11, 2nd WSEAS Int.Conf. on
IPLAFUN11,2011.
RichardCourant,KurtOttoFriedrichs,andHansLewy.On
the Partial Difference Equations of Mathematical
Physics.Math.Ann.100(1928):32.
XieJianheng.RiversSimulation.Beijing:WaterPowerPress,
1990.
Yang Guolu. Rivers Mathematical Model. Beijing: Ocean
Press,1993.
Zheng Bangmin, and Zhao Xin. Calculation of Water
Dynamics.Wuhan:WuhanUniversityPress,2001.

17