2014 仿真 backward-facing step flows

2014 仿真 backward-facing step flows
2014 仿真 backward-facing step flows

RESEARCH PAPER

Numerical predictions of backward-facing step ?ows

in microchannels using extended Navier–Stokes equations

R.Sambasivam ?Suman Chakraborty ?F.Durst

Received:29January 2013/Accepted:13August 2013/Published online:5September 2013óSpringer-Verlag Berlin Heidelberg 2013

Abstract Flows in microchannels were successfully predicted,in the past,both analytically and numerically,employing the extended Navier–Stokes equations (ENSE).In ENSE,the self-diffusion transport of mass,together with the resulting momentum and heat transport,is taken into account properly and the same is omitted in the classical Navier–Stokes equations.The ENSE have been employed here to numerically predict backward-facing step ?ows in microchannels,and the predictions are summarized in this paper.The results obtained by employing ENSE are compared with the available literature data computed by both direct simulation Monte Carlo and slip-velocity-based simulations.The good agreement of the present results with those given in the literature evidently points out that the ENSE can be applied to gas ?ows through complex mi-crochannel geometries.

1Introduction

The ?ow conduits in Micro-Electro-Mechanical-Systems (MEMS)may not be simple straight channels and capil-laries.They often involve complex geometries including steps and abrupt turns and bends.Therefore,analyzing gas

?ows through complex microgeometries is of immense importance,and for this reason,this particular topic has continued to attract the interest of researchers over the years.Agrawal et al.(2009)simulated gas ?ows through 90o bends employing the slip ?ow theory and the lattice Boltzmann method (LBM).Simulation of gas ?ows over backward-facing steps in microchannels is considered to be one of the benchmark problems to demonstrate the effec-tiveness of theoretical models and computational tech-niques in microscale ?ows.A number of studies have applied the LBM or the direct simulation Monte Carlo (DSMC)method along with the ‘slip ?ow’theory employed in microchannels.Agrawal et al.(2005)simu-lated gas ?ows in microchannels with a sudden expansion or contraction in order to obtain insight into ?ows in complicated microdevices.They employed the LBM,and the computations were performed for several area and pressure ratios over a range of Knudsen numbers in order to assess the effects of compressibility and rarefaction.Wu and Lee (2001)demonstrated the usefulness of the DSMC method in simulating gas ?ows through backward-facing microstep geometries.Kursun and Kapat (2007)also employed the DSMC method to simulate ?ows through microchannels with a backward-facing step.The simula-tions in this case were conducted with Reynolds numbers of 0.03–0.64,Mach numbers of 0.013–0.083,and Knudsen numbers of 0.24–4.81.In these parameter ranges,they did not observe any separation region in the ?ow geometry.They employed an information preservation (IP)method to separate the macroscopic velocity from the molecular velocity,and hence,the typical statistical noise generated in DSMC simulations was minimized in this way.

Beskok (2001)also employed the DSMC method to simulate high-velocity gas ?ows through microchannels with a backward-facing step.Celik and Edis (2007)also

R.Sambasivam (&)

Technology Group (Global Wires and Longs),Tata Steel Ltd.,Jamshedpur 831001,India e-mail:rsambasivam@https://www.360docs.net/doc/134331974.html,

S.Chakraborty

Mechanical Engineering Department,Indian Institute of Technology Kharagpur,Kharagpur 721302,India F.Durst

FMP Technology GmbH,Am Weichselgarten 34,91058Erlangen,Germany

Micro?uid Nano?uid (2014)16:757–772DOI 10.1007/s10404-013-1254-1

performed backward-facing step?ow predictions using the same microchannel geometry as Beskok(2001),but employing a characteristic-based split Navier–Stokes?nite element solver,with second-order slip velocity and tem-perature-jump boundary conditions at the solid walls.The gas?ow characteristics as obtained by Celik and Edis (2007)are shown in Fig.1.It can be observed that the sudden expansion due to the change in the cross-sectional area results in a temperature drop,and hence,the high-velocity gas?ows through such microchannels cannot be assumed to be isothermal.Furthermore,a separation bub-ble near the bottom wall downstream of the step could also be observed,as shown in Fig.1.When the Reynolds number is very low,the?ow is not expected to separate near the bottom wall in microchannel geometries with backward-facing steps because of the introduced‘slip velocity’.However,the?ow tends to separate at higher Reynolds number,as is evident in Fig.1.

The above discussion indicates that,unlike the solutions obtained in straight microchannels,it is not possible to use a simple‘slip velocity’model in order to obtain reliable numerical solutions for gas?ows through complex mi-crogeometries,employing the conventional or classical Navier–Stokes equations(CNSE).One must resort to special techniques such as DSMC or LBM in order to obtain reasonable?ow predictions.In the published liter-ature,extensions to the CNSE are available through the work of Brenner(Brenner2005a,b,c)and Durst et al.(2006),Chakraborty and Durst2007,and Sambasivam and Durst2011).In this paper,by employing the extensions suggested by Durst and co-workers to the CNSE,referred to as the extended Navier–Stokes equations(ENSE),it is shown that accurate?ow predictions of ideal gas?ows in complex microgeometries are possible even using com-mercially available computational?uid dynamics software. Earlier,the ENSE have been employed to predict gas?ows in straight microchannels and one-dimensional shock waves,and excellent agreements were obtained with cor-responding experimental measurements(Sambasivam and Durst2011;Sambasivam2012;Dongari et al.2009).

In this paper,the results of numerical simulations of gas ?ows through microchannels with a backward-facing step, employing the ENSE for the predictions,are compared with the corresponding data obtained by the DSMC and the LBM simulations available in the literature.This method was chosen by the authors since the slip?ow theory, commonly employed to predict gas?ows through micro-channels,lacks a sound physical basis unless the wall is molecularly smooth,see Sambasivam(2012)and Sam-basivam and Durst(2011).Furthermore,the analysis of straight microchannels(Sambasivam and Durst2011) clearly indicated that the self-diffusion transport of mass caused the additional mass?ow rate observed in the experiments which is generally interpreted as the‘slip ?ow’.In several publications,the authors and their co-workers have shown that the ENSE allow ideal gas

?ows Fig.1Flow characteristics of Nitrogen gas?ow through microchannel with a backward-facing step;outlet Knudsen number,Kn o=0.0018,and pressure ratio,P R=2.32:as shown in Celik and Edis(2007)

through microchannels to be treated accurately without having to make any assumption regarding wall slip velocities(Sambasivam2012;Dongari et al.2009;Sam-basivam and Durst2011).The results of?ow through microchannels with a backward-facing step,presented in this paper,substantiate this claim further.The backward-facing step geometry employed by Beskok(2001)and Celik and Edis(2007)was chosen in the simulations pre-sented in this paper,and detailed analysis is provided in the subsequent sections.

2Governing equations and numerical simulations

2.1Governing equations

The extended Navier–Stokes equations(ENSE),in the total velocity form for the steady compressible?ows of ideal gases employed in all the simulations presented in this paper,are shown below.The detailed derivations of these equations are presented in Sambasivam(2012),and an introduction to the self-diffusion transport of mass and associated transport of momentum and heat is presented in Appendix1.

Continuity equation:

o q U T i

àá

o x i

?0e1TMomentum equation:

o o x i

q U T i U T j

o P

o x j

à

o

o x i

s C ijà

2

3

d ij_m D k U C k

e2T

Total energy equation:

o q U T i E T àá

o x i ?à

o_q i

o x i

à

o PU T i

àá

o x i

à

o s C ij U T j

o x i

e3T

Equation of state:

q?

P

RT

e4T

s C ij ?àl

o U C i

o x j

t

o U C j

o x i

!

t

2

3

d ij

o U C k

o x k

e5T

_q i?àk o T

o x i

e6T

In the above set of equations,U C i and U T i are the convec-tive and total velocity vectors,and they are related by the following relationship U T i?U C itU D i,where U D i is the diffusion velocity,as shown in Appendix1.q,P,and T represent the local density,static pressure,and temperature, respectively,and_m D k is the self-diffusion transport of mass in the k direction.Further,l and k are the local dynamic viscosity and thermal conductivity,respectively.In Eq.(3), the total energy based on the total kinetic energy E T is given by E T?et1

2

"U T

j

2

,where e is the local internal energy of the?uid and"U T j is magnitude of the local total velocity.In the considered?ow case,the in?uence of gravitational acceleration on the?uid?ow is neglected.

2.2Geometry and boundary conditions

Beskok(2001)and Celik and Edis(2007)employed identical backward-facing step geometries in their predic-tions using the DSMC method and the characteristic-based-split(CBS)algorithm-based simulations,respectively. Beskok(2001)obtained numerical discretizations by a spectral element-based algorithm for solving the com-pressible Navier–Stokes equations employing the?rst-order and higher-order slip boundary conditions at the solid walls.Furthermore,the DSMC method employed the var-iable hard sphere(VHS)collision model.Beskok(2001) also reported that while applying the DSMC algorithm to gas?ows through microchannels,slow convergence,large statistical noise,extensive number of simulated molecules, and lack of deterministic surface effects were encountered. These disadvantages were overcome by using relatively high-speed gas?ows in small-aspect-ratio geometries with a suf?ciently large number of simulated molecules.

Celik and Edis(2007)employed a modi?ed CBS-algo-rithm to take into account the slip velocity and temperature-jump boundary conditions encountered in compressible ?ows through microsized geometries.The CBS is a general algorithm that can be employed to solve both compressible and incompressible?ow problems.In the CBS solution procedure,the numerical instabilities present in the standard Galerkin formulation of the discretized governing?uid?ow equations in space were stabilized by discretizing the equa-tions along the characteristic of the total derivative. Although the solution procedure employed by Celik and Edis (2007)was different from other conventional algorithms,the boundary conditions,arising out of the slip?ow theory,were employed at the solid boundaries.

The microchannel geometry of the backward-facing step,employed by Beskok(2001)and Celik and Edis (2007),is shown in Fig.2,and the salient geometric fea-tures of the channel are summarized in Table1.As can be seen in Fig.2,all the linear dimensions of the channel were non-dimensionalized with the height of the channel at the outlet h,and it was chosen to be 1.25l m.When the Knudsen number at the outlet is speci?ed,its actual value can be obtained by changing either the channel dimensions or the static pressure.It is generally assumed in the mi-crochannel literature that when comparing different

simulations from geometrically similar channels and if the outlet Knudsen number and pressure ratio do not change, the?ow characteristics obtained with various outlet pres-sure and channel height combinations are identical.How-ever,since the diffusion velocity is a function of the local pressure and its gradient(Fig.2and also explained in Sambasivam2012and Sambasivam and Durst2011),it is possible to obtain different solutions for the same outlet Knudsen number and pressure ratio combinations by changing the channel dimensions.Since the exact channel height employed in the studies mentioned in the literature was not given,the channel dimension at the outlet was chosen to be1.25l m based on Celik and Edis(2007).The dimensionless total length L of the channel was 5.6h. Furthermore,the dimensionless step height S was taken to be0.467.A non-dimensional approach length l’of0.86 was chosen at the beginning of the channel.The step was positioned at a dimensionless distance of1.81from the inlet of the channel.The grid divisions at different regions of the geometry are also given in Table1.

The following boundary conditions were employed in the simulations as shown in Fig.2:

1.The outlet pressure of the channel was calculated

based on the outlet Knudsen number Kn o=0.018.

Since during the iterative solution procedure the outlet temperature varied signi?cantly,the average outlet temperature was calculated frequently during the solution procedure to correct the outlet pressure in order to obtain the prescribed outlet Knudsen number.

2.The inlet static pressure was calculated based on the

speci?ed pressure ratio P R across the channel and the calculated outlet pressure.The dynamic pressure

calculated based on the speci?ed inlet Mach number was added to the static pressure in order to obtain the total pressure at the inlet.

3.Similarly,the dynamic temperature was calculated

based on the speci?ed inlet Mach number and was added to the inlet static temperature of330K to obtain the correct thermal boundary conditions at the inlet.

4.The symmetry boundary condition was employed

along the approach length at both the top and bottom boundaries.

5.Since ENSE were solved in the total velocity form as

given in Eqs.(1–3),the total velocity at the solid boundaries needed to be speci?ed.Since the convec-tive velocity components satisfy the no-slip boundary condition at the solid wall,it was suf?cient to specify the diffusion velocity as the total velocity boundary condition at the solid wall.At all the horizontal walls, the streamwise component of the diffusion velocity

was speci?ed,given by U D1?àl

q

1

P

o P

o x1

,and the normal component of the diffusion velocity was zero due to the impermeability condition at the walls.

Similarly,at the vertical wall,the cross-stream com-ponent of the diffusion velocity was speci?ed,given by

U D

2

?àl

q

1

P

o P

o x2

,and the streamwise component was zero.The basis,for arriving at the above-mentioned wall boundary conditions,is explained in Appendix2.

Furthermore,a constant static temperature of300K was speci?ed at the solid walls.

Second-order upwind discretization scheme was employed for all?ow variables,and SIMPLE algorithm was used for pressure–velocity coupling.The

simulations

were continued till the residuals of all?ow variables reached less than10-10.Simulations were carried out with the commercially available CFD code FLUENT6.3,and the special boundary conditions,needed for the ENSE computations,were implemented in the software through user-de?ned functions(UDF).Similarly,the additional terms in the governing equations were also incorporated as volumetric source terms.

3Results and discussions

3.1Comparisons with Beskok(2001)data

In Beskok(2001),the?ow was maintained subsonic,and the outlet Mach number was evaluated as an average value, i.e.,Ma o\1.Further,the average outlet Knudsen number of the channel Kn was calculated based on the following equation.

Kn?

cp

2

1=2Ma

Re_m T

e7T

where c is the speci?c heat capacity ratio of the gas,Ma o the average Mach number at the outlet of the channel,and Re_m T the Reynolds number of the?ow in the channel based on the total mass?ow rate.By employing Eq.(7),it is possible to calculate the maximum possible(or limiting) Reynolds number,for a given outlet Knudsen number and the outlet Mach number of unity,in a given geometry.For an outlet Knudsen number of0.04,as shown in Fig.8of Beskok(2001),the maximum possible Reynolds number of the channel can be only37,whereas the Reynolds number was stated to be80in Beskok(2001).For a Reynolds number of80,the maximum possible outlet Knudsen number is only0.0185and not0.04as mentioned in Bes-kok(2001).

Since Celik and Edis(2007)also compared their results with those of Beskok(2001),it was assumed that the?ow conditions given by the former author were identical with the simulation conditions of the later authors.Hence,the simulations with the ENSE presented in this paper were carried out with the?ow conditions given by Celik and Edis(2007).In Figs.3,4,and5,the contours of pressure, density,Mach number,static temperature,and Knudsen number,obtained in the numerical simulations of a nitro-gen gas?ow through a microchannel with a backward-facing step,employing the ENSE,are shown,respectively. On comparing them with Fig.1and other pro?les given in Beskok(2001)and Celik and Edis(2008),it can be observed that the pro?les obtained by solving the ENSE are almost similar to those presented in the literature.In Fig.4, the strong expansion of the gas can also be observed,near the step wall,due to the sudden increase in the cross-sec-tional area.

The local static pressure decreases signi?cantly at the sharp corner,and one would expect that this would result in enhanced diffusion effects.As observed in Fig.4,the?ow is accelerated rapidly near the sudden expansion of area of the?ow geometry and the?ow becomes locally super-sonic.However,the?ow becomes gradually subsonic again further downstream of the channel.Toward the exit of the channel,the?ow again accelerates and the velocity increases due to the gradual decrease in pressure.Fur-thermore,the recirculation region close to the bottom wall of the channel can be observed behind the step.In addition, the drop in the local temperature near the step,caused by the sudden expansion,is evident from the temperature contours in Fig.4.After a gradual recovery of the tem-perature downstream of the step,it decreases again toward the exit of the channel due to the continued reduction in local pressure values.However,the temperature near the top and bottom walls does not decrease very much because of lower velocities and the presence of a re-circulatory region,respectively.

The local Knudsen number pro?les are shown in Fig.5, and as expected,the Knudsen number increases toward the exit of the channel.However,the maximum Knudsen number happens to be at the step because of the strong pressure gradients near the step,caused by the sudden expansion of the?ow.In general,the pro?les of the?ow variables,computed with the ENSE,followed the expected trend and are in good qualitative agreement with the DSMC simulations.

In Fig.6,the streamwise velocity pro?les are plotted as a function of the cross-stream coordinate at two dimen-sionless streamwise locations,one in the upstream direc-tion and the other in the downstream direction of the

Table1Summary of simulation conditions

Data Value No.of grid

divisions

Channel dimensions

Height at outlet,h l m 1.25110 Dimensionless height at inlet,1-S0.53350

Step height,S0.46760

Total length of channel,L 5.6260 Length of approach section,l00.8640

Inlet Mach number0.45

Inlet static temperature,T S K330

Pressure ratio,P R 2.32

Gas used Nitrogen

Molecular diameter,m 3.77910-10

Outlet Knudsen number,Kn o0.018

Wall temperature,T w K300

backward-facing step.The velocity pro?les obtained in the DSMC simulations of Beskok(2001)are also shown for comparison.It can be observed in Fig.6a,b that the gen-eral agreement between the two pro?les obtained by the two computational methods is very good.However,the streamwise velocity at the top wall is higher in the DSMC simulations than in the present solutions,i.e.,the calculated slip velocity is larger.Since the outlet Knudsen number is 0.018,the local Knudsen number in the middle of the channel is expected to be even lower.One can observe in Fig.6that the local Knudsen number before the expansion is only about0.008.Hence,it is possible to conclude that

Pressure in Pa Density in Kg/m3

Static Pressure

Density

Fig.3Contours of static pressure(top)and density(bottom)computed by employing the ENSE;Kn o=0.018;P R=2.32

Temperature in K

Mach number

Static temperature

Mach number

Fig.4Contours of Mach number,Ma?U T

1 ?????????

c

p

(top)and static temperature(bottom)computed by employing the ENSE;Kn o=0.018;

P R=2.32

the slip velocity must be very small at this location.Through the ENSE-computations,very low velocity mag-nitudes at the top and bottom walls were obtained,unlike the DSMC solutions of Beskok (2001).A detailed analysis of this feature of the ?ow is given below.

In order to further understand the above-mentioned discrepancy in the predicted slip velocity,the calculated wall velocity pro?les were analysed as a function of the streamwise coordinate and are shown in Fig.7.The con-tinuous line represents the velocity at the top wall,and the dotted line denotes the velocity at the bottom wall.The slip velocity values calculated by Beskok (2001),using dif-ferent wall slip models,are also shown for comparison.The solid symbols represent the velocity values at the top wall,and the open symbols denote the values at the bottom wall.The O(Kn )pro?les represent the ?rst-order Maxwell slip model.The l and (3/2)l pro?les represent higher-order slip models where the slip information was estimated based on the velocity at a distance of l and (3/2)l from the wall,and l represents the molecular mean free path.It is inter-esting to note in Fig.7that the wall velocity calculated by the ENSE is signi?cantly lower than the values predicted by the various slip velocity models.Since the Knudsen number simulated in this ?ow case is small,it is not pos-sible for strong self-diffusion effects to be present and hence the wall velocity must be small.At the ?rst glance,it was puzzling to note the huge discrepancy shown in the computed data.

In the solutions obtained by employing the ENSE,the wall velocity values were computed ‘exactly at the wall’.However,the computed slip velocity values appear to be estimated a small normal distances away from the wall.Therefore,it is possible that the wall velocity pro?les obtained by the ENSE and the slip velocity models shown in Beskok (2001),shown in Fig.7,may not have been calculated at the same dimensionless cross-stream location.This is also understood by the dimensionless cross-stream coordinates of the top and bottom walls as 0.9875

and

Fig.5Contours of Knudsen number computed by employing the ENSE;Kn o =0.018;P R =

2.32

Fig.6Comparison of

streamwise velocity pro?les obtained with the ENSE

(present work)and the DSMC simulations by Beskok (2001);Kn o =0.018;P R =2.32;dimensionless axial distance,x /h :a 1.7and b 2.1

0.01675instead of1and0,respectively,given in Table2, Beskok(2001).In order to verify this,the wall velocity pro?les shown in Fig.7were plotted at the dimensionless cross-stream locations of0.01675and0.9875to represent the bottom and top walls,respectively,as done by Beskok (2001).The new pro?les were compared with the pro?les predicted with the slip velocity models given by Beskok (2001),as shown in Fig.8.The comparison of the two models is surprisingly good,and the wall velocity values, predicted by the two methods,are almost identical.

Hence,it can be concluded that the‘real’wall velocity is not very high at the given Knudsen number,but one obtains larger velocity values,as‘apparent’slip velocity,at small distances away from the walls.This is the reason for the discrepancy in the results presented in Fig.7.It can be observed that the discrepancy of the streamwise velocity at the wall,between the ENSE and the DSMC results of Beskok(2001),as shown in Fig.6,also occurs for this reason only.Further,the dimensionless reattachment location(x/h)was found to be2.8in the DSMC simulations by Beskok(2001)and the same value was observed in the simulations employing the ENSE,as seen in Fig.8.

The cross-stream velocity pro?les obtained by solving ENSE at two different dimensionless streamwise locations are shown in Fig.9.The pro?les predicted by the DSMC simulations of Beskok(2001)are also given in this?gure. The velocity pro?les,obtained by the two very much dif-ferent methods,agree very well as shown in Fig.9.Since

Top wall, ENSE

Bottom wall, ENSE

3/2 l, Higher order slip

l, Higher order slip

O(Kn), 1st order slip

velocity pro?les along

streamwise coordinate obtained

with the ENSE(present work)

and as shown in Beskok(2001);

closed symbols represent values

at the top wall,whereas open

symbols denote values at the

bottom wall;Kn o=0.018;

P R=2.32

Top wall, ENSE

Bottom wall, ENSE

3/2 l, Higher order slip

l, Higher order slip

O(Kn), 1st order slip

the impermeability condition at the solid wall needs to be satis?ed,the cross-stream velocity at the solid wall is zero and hence the no-slip boundary condition is satis?ed for the wall-normal velocity.The local Mach number pro?les,the streamwise velocity normalized by the local velocity of sound,as a function of the non-dimensional

streamwise

stream velocity pro?les obtained with the ENSE

(present work)and the DSMC simulations by Beskok (2001);Kn o =0.018;P R =2.32;x /h a 1.7and b

2.1

Fig.10Comparison of local Mach number,U T 1 ?????????c

ENSE

Fig.11Comparison of dimensionless static pressure,P .0:5q in U T in

àá2h i

along the streamwise distance at different

dimensionless cross-stream locations;Ma in =0.45;Kn o =0.018the symbols represent the DSMC data of Beskok (2001),and the lines show the results obtained with the ENSE

coordinate at different dimensionless cross-stream loca-tions are plotted in Fig.10.The Mach number pro?les obtained by the DSMC solutions by Beskok (2001)are also plotted for comparison.The two sets of pro?les agree remarkably well in most regions of the channel.The wall velocity at the top wall was found to be underpredicted by ENSE near the step in comparison to the DSMC simulations.

3.1.1Pressure pro?les

The dimensionless pressure pro?les at different cross-stream locations are plotted as a function of the dimen-sionless streamwise coordinate in Fig.11.The local static pressure was non-dimensionalized using the dynamic

pressure 12q in U T in

àá2

at the inlet.The pressure pro?les obtained from the DSMC simulations by Beskok (2001)are also shown for comparison in this ?gure,and it can be

observed that the agreement between pro?les obtained,employing the two very different computational methods,is very good.

3.1.2Temperature pro?les

The static temperature pro?les obtained with the ENSE are shown at two dimensionless streamwise locations in Fig.12.The temperature pro?les obtained by the DSMC simulations and the slip velocity models are also shown in this ?gure for comparison.It can be observed in Fig.12a that the temperature values predicted by the ENSE are much lower than those given by the DSMC simulations.Due to the expansion caused by the sudden increase in the area,the thermal energy is converted into kinetic energy,and hence,the temperature was signi?cantly reduced near the step.In the absence of local measurements,it will be dif?cult to conclude which method predicted the temper-ature pro?les

accurately.

temperature pro?les obtained with the ENSE and the DSMC simulations by Beskok (2001);Kn o =0.018;p =2.32;x /h a 1.7,b 2.1

C&E_0.01456 C&E_0.2476C&E_0.4978 C&E_0.755 C&E_0.9901

Fig.13Comparison of local Mach number,U T 1 ?????????c

represent the simulation data of Celik and Edis (2007),and the lines depict the results obtained with the ENSE

3.2Comparisons with Celik and Edis (2007)

Celik and Edis (2007)employed the slip velocity model in the CBS scheme in order to provide solutions to gas ?ows through microchannels with a backward-facing step.In Fig.13,the local Mach number values obtained with the ENSE are compared with those calculated based on the CBS algorithm by Celik and Edis (2007).The agreement between the two pro?les is good.There is a discrepancy in the pro?les close to the top wall in the neighbourhood of the step because of the differences in the calculated wall velocity,as described in Figs.7and 8.The agreement between the pro?les obtained by the two methods toward the outlet of the channel is also very good.

In Fig.14,the dimensionless pressure pro?le obtained with the ENSE along the streamwise direction at the middle section of the channel is shown.The pro?les

obtained by the simulations employing the slip velocity theory carried out by Celik and Edis (2007),Beskok (2001),and Baysal and Aslan (2002)are also depicted in this ?gure for comparison.The pro?les agree with one another exceedingly well throughout the entire length of the channel.The small discrepancy observed near the inlet could be attributed to the minor variations in the inlet conditions since the exact boundary conditions used in the simulations were not mentioned in the literature.

Further,the local Knudsen number pro?le obtained with the ENSE along the streamwise direction at the middle section of the channel is compared with the pro?les obtained by Celik and Edis (2007)and Baysal and Aslan (2002)in Fig.15.The agreement between the pro?les is excellent throughout the length of the channel except in the region close to the step.As observed in Figs.12and 14,the static pressure and temperature pro?les obtained by

the

Fig.14Comparison of dimensionless pressure,P .0:5q in U T in

àá2h i

along the streamwise distance at the

middle section of the channel;Ma in =0.45;Kn o =0.018;the symbols represent the

simulation data of Celik and Edis (2007),Beskok (2001)and Baysal and Aslan (2002),and the lines depict the results obtained with the

ENSE

Fig.15Comparison of local Knudsen number,Kn =(l /h )along the streamwise direction at the middle of the channel;

Ma in =0.45;Kn o =0.018;the symbols represent the simulation data of Celik and Edis (2007)and Baysal and Aslan (2002),and the lines depict those obtained with the ENSE

ENSE and the slip-velocity-based theory did not match in the neighbourhood of the step.It is felt that in the absence of accurate experimental measurements,it is not possible to ascertain the validity of the pro?les obtained by either the ENSE or the DSMC simulation or any other method.It may be argued that unlike other methods employed in the simulations of micro-gas-?ows,the ENSE do not have empirical assumptions regarding the slip velocity,hence the possibility of the pro?les obtained with these equations being correct is high.It is suggested that this aspect of the gas?ow through backward-facing step geometries needs to be investigated comprehensively in future studies of this kind of?ows.

Based on the various discussions mentioned in this section,it can be concluded that the ENSE were success-fully employed to accurately predict all the characteristics of gas?ows through microchannels with a backward-fac-ing step.Based on the comparisons with the DSMC and slip-velocity-theory-based simulations,it was evident that the ENSE can be employed satisfactorily,in the Knudsen number range studied,to predict microchannel?ows with complex geometries.

4Conclusions,?nal remarks and outlook

The results of?ow computations,presented in this paper, clearly show that the extended transport equations suggested by Durst et al.(2006)pose a sound physical basis.The good agreement obtained between the ENSE computations and the DSMC simulations clearly shows that the continuum treat-ment of microchannel?ows is possible beyond the Knudsen numbers of0.1.However,to make microchannel?ows treatable by the continuum-based basic equations of?uid mechanics,it is necessary to introduce the self-diffusion of mass into the continuity equation.Its introduction also affects the diffusion transport of heat and momentum,and the extended equations for these quantities also need to be taken into account.In this way,the ENSE have been derived and these equations have been successfully employed in numerically computing microchannel?ows with a back-ward-facing step.The obtained results compare very well with the corresponding DSMC results available in the liter-ature.This?nding readily suggests employing the ENSE for all computations of microchannel?ows with complex?ow geometries.The application of the ENSE does not require any ad hoc assumption regarding wall slip velocities to obtain the?ow characteristics.As a matter of fact,‘apparent’slip velocities result from the ENSE computations,arising due to pressure-gradient-driven diffusive?ows.It is sug-gested that solving microchannel?ows with the ENSE can reveal interesting features of?ow characteristics,as dem-onstrated in this paper.Appendix1:Derivations of diffusive transport equations of mass,heat,and momentum

Usually,the diffusion transport of mass in gases is considered in the classical Navier–Stokes equations only when multiple species are under consideration(multi-component diffusion). It is well known that diffusion transport of mass also occurs in a single-component gas,when density and/or temperature gradients are present,known as self-diffusion.Self-diffusion is always neglected in gases with the presumption that its in?uence is insigni?cant in comparison with the convective transport of mass.However,in the presence of strong density and temperature gradients,one can easily demonstrate that the self-diffusion of mass cannot be neglected and at times can attain magnitudes of the order of convective mass transport. Therefore,the authors argue that?ows through microchannels and capillaries are?ows where the self-diffusion of mass plays a signi?cant role.The authors introduced the self-diffusion transport of mass and the associated momentum and heat transports in the classical Navier–Stokes equations and obtained the extended Navier–Stokes equations.In this appendix,the derivations of expressions for self-diffusion transport of mass,momentum,and heat are shown based on Boltzmann’s statistical considerations.

Diffusion transport of mass

To analyse the molecular transport processes,one can consider Fig.16.The self-diffusion transport of mass can be estimated from two planes separated by a distance l from the plane under consideration,x i.Applying the kinetic theory of gases,the molecular transport of mass can be written as_m D x

i

tl

?1

6

q x itl

eT"u M x itl

eTand_m D x

i

àl

?1

6

q x iàl

eT"u M x iàl

eTfrom the planes located at(x i?l)and (x i-l),respectively.Here,q is the local density of the gas and"u M is the local molecular mean velocity,de?ned as "u M?

??????????

8j T

p m M

r

e8T

where T is the absolute temperature,j the Boltzmann constant,m M the molecular mass,and l the local molecular mean free path of the considered ideal gas.The net diffusive mass?ux can be expressed as

_m D

i

?

1

6

q x iàl

eT"u M x iàl

eTàq x itl

eT"u M x itl

eT

??

e9TIt can be clearly observed from Eqs.(8)and(9)that when there is no spatial density and temperature gradients in the ?ow?eld,the net diffusive mass?ux_m D i?0,i.e.,there is no self-diffusion transport of mass.However,when density and temperature gradients are present,the net diffusive mass?ux_m D i?0is present,and it can be evaluated as shown below.

Expanding Eq.(9)using Taylor series,the following expression can be obtained after truncating the higher-order terms:

_m D i ?16"q x i eTto q o x i àl eT "u M x i eTto "u

M

o x i

àl eT àq x i eTto q o x i l eT "u M x i eTto "u

M

o x i l eT

#e10TTaking into consideration the isotropy of the molecular

motion and by including only the product terms that contain ?rst-order derivatives,the diffusive mass transport in the x i direction can be written as

_m D i ?à13l "u M o q o x i tq o "u

M

o x i !e11TDe?ning a diffusion coef?cient D ?1=3l "u

M eT,Eq.(4)can be rewritten as

_m D i ?àD o q o x i tq "u o "u

M

o x i !e12TFurther,substituting the dynamic viscosity,l =D q for ideal gases and employing Eq.(8),the following form of Eq.(12)can be obtained:

_m D i ?àD o q o x i tq 2T o T o x i

!?àD q eT|?{z?}l

1q o q o x i t12T o T

o x i !?àl 1P o P o x i à

12T o T o x i

!e13T

The results of the above derivations are well known,and they yield the following terms:Fick’s mass diffusion term:_m F i ?àD

o q

o x i

e14T

Soret’s mass diffusion term :_m S i ?àD q 2T o T

o x i e15T

where D S =(l /2T )is the Soret diffusion coef?cient.Diffusion transport of heat

It can be observed that the presence of self-diffusion

transport of mass results in a corresponding diffusive transport of heat associated with it.Therefore,the diffusion transport of heat from the planes located at (x i ?l )and (x i -l )can be written as

_q D x i tl ?1

6q x i tl eT"u M x i tl eTe x i tl eTand _q D x i àl

?1

6

q x i àl eT"u

M x i àl eTe x i àl eTe16T

where e is the local internal energy of the gas.

Further,for an ideal gas,the internal energy e can be written as e =C V T ,where C V is the speci?c heat at con-stant volume.Therefore,the net diffusive heat ?ux can be expressed as

_q D

i ?C V 6h q x i àl eT"u

M x i àl eTT x i àl eTàq x i tl eT"u M

x i tl eTT x i tl eTi

e17TFurther,after Taylor series expansion of the terms and

simpli?cation,the following expression can be obtained:

_q D i ?àC V

3l "u M q àáT q o q o x i tT "u M o "u M o x i to T o x i !e18TFrom the de?nition of "u M given by Eq.(8),one can derive o "u M =o x i

as o "u M o x i

???????????8j p m M r 12????T p o T

o x i e19TSubstituting Eq.(19)in Eq.(18),the following expression for _q D i can be obtained._q D i ?àC V D q |??{z??}k

0@1

A o T o x i àC V T D q eT1q o q o x i t12T o T o x i

!

|?????????????????{z?????????????????}à_m D i

e20THence,the total diffusive heat transport in an ideal gas can

be expressed as _q D i ?àk o T

o x i |???{z???}_q F i

t_m D i C V T |???

?{z????}_q

SD i

e21T

In Eq.(21),

_q F i

?àk

o T

o x i

represents the Fourier’s heat

diffusion term and _q SD i ?_m D i C V T is due to the effect of the

self-diffusion mass transport.k is the local

thermal

Fig.16Self-diffusion mass transport due to density and temperature gradients

conductivity of the?uid.Interestingly,Eq.(20)can also be written as

_q D i ?à

3

2

k

o T

o x i

|?????{z?????}

Temperatureàdriven

heat transport term

àC V T

eTD

eT

o q

o x i

!

|???????????{z???????????}

Densityàdriven heat

transport term?Dufouràterm

?àk eff

o T

o x i

àD f

o q

o x i

e22T

The Dufour term shown in Eq.(22)is often considered in treatments of binary diffusion problems,see Coelho and Silva Telles(2002),but not for‘‘self-diffusion’’problems.

Diffusion transport of momentum

Similarly,the net diffusive transport of j momentum in the i direction can be written as

s ij?1

6

q x iàl

eT"u M x iàl

eTU C j x iàl

eT

h

àq x itl

eT"u M x itl

eTU C j x itl

eT

ie23T

After expanding and simplifying,the diffusive momentum transport can be expressed as

s ij?à1

3

l"u M q

àáU C j

q

o q

o x i

t

U C

j

"u M

o"u M

o x i

t

o U C j

o x i

"#

e24T

Equation(24)can be simpli?ed further as given below:

s ij?àU C j

1

l"u M q

|???{z???}

l

B B

@

1

C C

A

1o q

i

t

1o T

i

t

1

U C j

o U C j

i

"#

e25T

After rearranging,Eq.(25)takes the following form.

s ij?àl o U C j

o x i

àl U C j

1

q

o q

o x i

t

1

2T

o T

o x i

!

?àl

o U C j

o x i

t_m D i U C j

e26T

Similarly,the net diffusive transport of i momentum in the j direction can be written as

s ji?àl o U C i

o x j

t_m D j U C ie27T

Similarly,it is also possible to derive extended volume dilation terms,and incorporating all these terms into the diffusive momentum transport tensor,the following equation can be obtained.

s T ij ?àl

o U C j

o x i

t

o U C i

o x j

!

t

2

3

d ij l

o U C k

o x k

t_m D i U C jt_m D j U C ià

2

3

d ij_m D

k

U C

k

e28T

The extended Navier–Stokes equations

The self-diffusion transport of mass can be introduced into

the classical continuity equation as given below.

o q

o t

t

o q U C i

àá

o x i

t

o_m D i

o x i

?0e29T

From Eq.(29),it can be observed that when the molecular

mass diffusion_m D i is negligible in comparison with the

convective mass transport(when the density and/or

temperature gradients are small),then the classical

continuity equation is obtained.Further,introducing

diffusion velocity U D i,the self-diffusive mass transport

_m D

i

can be written as_m D i?q U D i.Now,let us introduce a

‘total velocity’,de?ned as the vector sum of convective

and diffusion velocities such as U T i?U C itU D i.By

employing this de?nition of the total velocity,the

extended continuity Eq.(29)can be rewritten as

o q

o t

t

o q U T i

àá

o x i

?0e30T

By following the above-mentioned procedure,the extended

momentum and energy equations can be obtained as

o q U T j

t

o q U T i U T j

i

o P

j

à

o

i

s C ijà

2

d ij_m D k U C k

tq g j

e31T

o q E T

eT

o t

t

o q U T i E T

àá

o x i

o_q i

o x i

à

o PU T i

àá

o x i

à

o s C ij U T j

o x i

e32T

where s C ij?àl o U C j

o x i

to U C i

o x j

t2

3

d ij l o U C k

o x k

and_q i?àk o T

o x i

.

For a more detailed description of the derivations of the

extended Navier–Stokes equations,one may refer to Durst

et al.(2006)and Sambasivam(2012).

Appendix2:Wall boundary conditions employed

in solving ENSE

It is a well-known practice to employ the no-slip boundary

condition at the solid wall when solving the CNSE.How-

ever,the deviations in the measured mass?ow rates in

experiments led to the introduction of‘slip velocity’at the

solid wall in gas?ows through microconduits in order to

account for the additional mass?ow rates in the simula-

tions.The‘slip coef?cient’has to be experimentally

determined and is generally considered to be a function of

the local Knudsen number.

Before discussing the wall boundary conditions employed in the simulations using the ENSE,it is neces-sary to understand the molecule–wall interactions.As shown schematically in Fig.17a,unlike the molecules moving in far away distances from the solid wall,the molecules near the roughness elements in the wall are expected to be in?uenced by the multiple bombardments within the roughness.The molecules in this region tend to lose their momentum (in excess of the internal energy characterized by the local static temperature)completely in this process,and hence,the molecules are likely to attain the velocity of the wall,i.e.,the no-slip velocity condition is satis?ed.It is shown clearly by Mo and Rosenberger (1990)that unless the wall is atomically smooth,the no-slip boundary condition for the velocity is valid at all the engineering surfaces.The roughness elements are a mini-mum of two orders of magnitude higher than the diameter of molecules in a typical microchannel,see Arkilic et al.(2001),and the no-slip boundary condition is expected to be valid even in typical microchannel gas ?ows.

Based on the above discussions,the convection,diffu-sion,and total velocity pro?les can be examined,and it is stressed here that all the three velocity pro?les have to satisfy the no-slip boundary condition at the wall,i.e.,at the proximity to the root of the roughness elements.As shown schematically in Fig.17b,the convection velocity pro?le is expected to increase gradually from a value of zero at the mean height of roughness elements at the wall to the maximum at the mid-section of the channel under fully developed ?ow conditions.However,the diffusion velocity is expected to reach its maximum value with a sharp jump immediately beyond the roughness elements of the wall,as shown in Fig.17c.Since the static pressure remains constant along the cross-stream direction gas ?ows through straight microchannels,the diffusion velocity pro?le is practically constant along the cross-stream direction.

Now it is easy to observe in Fig.17d that the total velocity pro?le,the sum of convection,and diffusion velocity pro?les display a slip-like behaviour at the solid wall.It is important to note that the physical reason for this ‘slip-like’non-zero velocity at the wall,i.e.,at the mean height of roughness elements,is the self-diffusion mass transport driven by pressure/density gradients and not the

(b)

(a)(d)

(c)roughness roughness l

l a W l

l a W l

l a W l

l a W representation of ?ow characteristics near solid wall in gas ?ows through microchannels velocity pro?le;c diffusion velocity pro?le;and velocity pro?le

tangential momentum accommodation-based‘slip-concept’adopted by the‘slip?ow’models proposed in the micro-channel literature.Unlike the‘slip?ow’theory,the velocity at the solid wall is not determined by experimental measurements but from calculations based on the?rst principles.Based on the above arguments,the boundary conditions at the wall were introduced in the simulations as shown in Fig.2.

References

Agrawal A,Djenidi L,Antonia RA(2005)Simulation of gas?ow in microchannels with a sudden expansion or contraction.J Fluid Mech530:135–144

Agrawal A,Djenidi L,Agrawal A(2009)Simulation of gas?ow in microchannels with a single90°https://www.360docs.net/doc/134331974.html,put Fluids.doi:10.

1016/https://www.360docs.net/doc/134331974.html,p?uid.2009.01.004

Arkilic EB,Breuer KS,Schmidt MA(2001)Mass?ow and tangential momentum accommodation in silicon micro-machined channels.

J Fluid Mech437:29–43

Baysal O,Aslan AR(2002)Computing separated?ows in MEMS devices.In:Proceedings of ASME?uids engineering division summer meeting,Montreal,Quebec,Canada,Paper No.

FEDSM2002-31157,July14–18

Beskok A(2001)Validation of a new velocity-slip model for separated gas micro-?ows.Numer Heat Transf Part B40: 451–471

Brenner H(2005a)Navier–Stokes revisited.Phys A349:60–132 Brenner H(2005b)Kinematics of volume transport.Phys A 349:11–59

Brenner H(2005c)Nonisothermal Brownian motion:thermophoresis as the macroscopic manifestation of thermally biased molecular motion.Phys Rev E72:061201Celik B,Edis FO(2007)Computational investigation of micro backward facing step duct?ow in slip regime.Nanoscale Microscale Thermophysical Eng11:319–331

Chakraborty S,Durst F(2007)Derivations of extended Navier–Stokes equations from molecular transport considerations for compressible ideal gas?ows:towards extended constitutive forms.Phys Fluids19(1–4):088104

Coelho RML,Silva Telles A(2002)Extended Graetz problem accompanied by Dufour and Soret effects.Int J Heat Mass Transf45(15):3101–3110

Dongari N,Sambasivam R,Durst F(2009)Extended Navier–Stokes equations and treatments of microchannel gas?ows.JSME J Fluid Sci Technol4(2):1–14

Durst F,Gomes J,Sambasivam R(2006)Thermo-?uid-dynamics:do we solve the right kind of equations?In:Hanjalic K,Nagamo Y, Jakirlic S(ed)5th symposium on turbulence,heat and mass transfer,vol1.Dubrovnik,p3,September2006

Kursun U,Kapat JS(2007)Modelling of microscale gas?ows in transition regime part I:?ow over backward facing steps.

Nanoscale Microscale Thermophys Eng11:15–30

Mo G,Rosenberger F(1990)Molecular dynamics simulation of?ow in a two dimensional channel with atomically rough walls.Phys Rev A42:4688–4692

Sambasivam R(2012)Extended Navier–Stokes equations:deriva-tions and applications to?uid?ow problems.PhD thesis, submitted to Technische Fakultaet,University of Erlangen-Nuremberg,Germany

Sambasivam R,Durst F(2011)Ideal gas?ows through microchan-nels—revisited.In:SK Mitra,S Chakraborty(eds)Micro?uidics and nano?uidics handbook:chemistry,physics and life sciences principles

Wu JS,Lee F(2001)Pressure boundary treatment in micromechan-ical devices using the direct simulation Monte Carlo method.

JSME Int J Series B44(3):439–450

2014届毕业设计任务书(道路)

2014届毕业设计任务书(道路)

本科生毕业设计任务书 (工科及部分理科专业使用) 题目: 题目来源:□省部级以上□校级□横向□自选 题目性质:□理论研究□应用与理论结合研究 □应用基础研究□实际应用研究学院:建筑工程系:土木 专业:土木工程 班级: 学号: 学生姓名: 起讫日期:2014.2.7 ~ 2014.6.6 指导教师:职称: 系分管主任: 审核日期:

说明 1.毕业设计任务书由指导教师填写,并经专业学科组审定,下达到 学生。 2.进度表由学生填写,至少每两周交指导教师签署审查意见,并作 为毕业设计工作检查的主要依据。进度表中的周次是指实际的毕业设计进程中的周次。 3.学生根据指导教师下达的任务书独立完成开题报告,于3周内提 交给指导教师批阅。 4.本任务书在毕业设计完成后,与论文一起交指导教师,作为论文 评阅和毕业设计答辩的主要档案资料,是学士学位论文成册的主要内容之一。

路面结构设计参数表 层位结构层名称厚度 (cm) 高温抗压模量 (Mpa) 极限强度 (Mpa) 1 沥青混凝土7 1200 1.30 2 沥青碎石10 700 0.62 3 石灰土设计层500 0.25 4 级配砂砾1 5 160 土路基35(干燥)、30(中湿) 注:厚度可在许可范围内进行调整,不要所有人都一样。 二)桥梁、涵洞设计 1、根据资料:对路线上的中小桥梁进行孔径、桥确定(水文计算从略)。 2、根据资料:自选一座主线桥梁进行桥梁设计,要求对墩台下部结构进行计 算。绘制“桥位平面图”、“上部一般构造图”、“下部一般构造图”、“主梁 钢筋(或钢束)配置图(要求至少绘一张钢筋图)”;本图的工程数量表应 绘制在设计图上,每座桥汇总的数量表应单独绘制,设于该桥设计图前。 3、根据资料提供的涵洞孔径和就近桩号的横断面确定涵洞底部标高和纵坡, 计算涵长和工程数量。选择一道圆管涵,一道钢筋混凝土盖板涵绘制一般 构造图。 4、编制“桥梁表”、“涵洞表”。 三)路线交叉设计 1、根据平面图上与主线相交的公路、人行道路,确定被交道路在主线上应设 置的跨线桥或下穿通道的桩号、与路线交角、跨线桥的跨度、通道孔径、采用的结构形式等。 2、对交叉结构物进行设计,自选一座分离式立交桥进行桥梁设计,不进行结 构计算,仅绘制“桥型布置图”、“上部一般构造图”、“下部一般构造图”。或选择一座通道绘制立面图和断面图。 四)施工组织设计 1、编制“人工、主要材料及机具设备安排表” 2、编制“工程进度横道图” 3、编制“临时工程一览表” 五)编写设计说明书:按设计顺序编写(注意书写格式),并打印。后附下列表格。 道路部分设计所出的表格主要有以下三种: 1、路基设计表 2、土石方计算表 3、曲线要素表 桥梁部分表格有:桥梁表、涵洞表

2014年土木工程专业地铁车站毕业设计任务书

土木工程专业 城市地下空间工程方向毕业设计任务书 中南林业科技大学土木工程与力学学院 二0 一四年三月

XX地铁车站初步设计 一、毕业设计目的 毕业设计是按教学计划完成理论教学和相关实践教学之后的综合性教学,是对专业方向教学的继续深化和拓宽,是培养学生工程实践能力的重要教学阶段,其目的在于全面培养、训练学生运用已学的专业基本理论、基本知识、基本技能,进行本专业工程设计或科学研究的综合素质。 二、毕业设计基本要求 1、按设计课题的要求,独立完成设计任务,做出不同的设计方案,交出各自的成果。 2、认真设计、准确计算、细致绘图、文字表达确切流畅。 3、树立科学态度,注重钻研精神、独立工作能力的培养。 4、严格按照有关文件要求进行毕业设计管理,努力提高毕业设计质量。 5、图纸绘制要求:全部采用A3 图纸(可加长);计算机出图必须有3 张;图 纸布局要协调,要紧凑而不拥挤;线条粗细要正确,位置要准确; 6、注重资料的收集、分析和整理工作,设计完成后,设计成果应按如下要求装订成册:(1)《毕业设计计算书》A4 一份;(2)《毕业设计图纸》A4 一份。 7、图纸装订顺序:封面,目录,设计总说明,设计图纸、表格。 8、设计计算书装订顺序:封面、目录、中英文摘要、设计总说明、设计计算的全部内容、致谢(300 字左右)。 三、设计任务与要求 (一)、设计资料 1、车站地质勘察报告 2、预测客流(见附表) 3、车辆外形尺寸:A 型车或B 型车。 4、车辆编组:设计时采用远期列车6 辆编组。 5、防水等级:一级;二次衬砌混凝土抗渗等级不小于S6。 6主要技术标准:执行《地铁设计规范》(GB50157-2003)的有关技术标 准。

毕业设计任务书

任务书 1、本毕业设计课题应达到的目的: 本课题是基于资源丰富(environmentally abundant)的碳基氧化镍超级电容电极优化设计研究。要求学生了解超级电容器的发展和研究背景,重点掌握氧化镍电极材料材料、碳基氧化镍复合电极材料的制备、改性处理及分析、测试方法,对新能源材料的研究进行初步了解,提高学生的实验操作能力,发现和解决问题的能力,培养学生的团队协作能力,激发学生对新能源材料的研究兴趣,为科研工作做好充分准备。 2、本毕业设计(论文)课题任务的内容和要求(包括原始数据、技术要求、工作要求等): 内容: ⑴氧化镍制备及其性能研究 ①通过水热法合成氧化镍:通过大量实验,采用控制变量法,设定不同水热反 应参数(pH、表面剂、不同水热反应温度和时间、热处理温度和时间),制备的样品不少于12组; ②研究氧化镍微观形貌的控制手段,并结合至少4组SEM图像进行分析; ③对氧化镍进行物相、形貌分析和相关电化学性能测试,比较不同的水热反应 条件对氧化镍形貌和性能的影响,确定最佳的水热反应参数。 ⑵复合电极制备及其性能研究 ①选择电化学性能较好的氧化镍与碳源复合(石墨烯、碳纳米管、生物碳荷叶、 椰壳),研究复合方法,制备的样品不少于4组; ②对复合材料进行形貌分析和电化学性能测试,比较不同碳基材料对于氧化镍 的形貌改变,及电化学性能的改善作用。 技术和工作要求; ⑴原始数据要求:要求学生学会记录并分析在样品制备、样品改性、性能表征、 测试时的一系列实验变量和实验数据,尤其是在电化学性能测

试时对恒流充放电测试曲线以及阻抗曲线分析;学会使用正交 设计助手安排实验进展、寻求最佳实验参数;学会使用Origin 等软件分析大量实验数据和作图,及时总结规律; ⑵技术要求:要求学生掌握样品制备、样品改性、性能表征及测试过程中相关实 验方法、基本操作步骤和实验仪器的使用维护,并能指出并改进在 这过程中出现的技术问题,提高实验精准度,优化相关工艺参数; ⑶工作要求:要求学生严格按照任务进度表开展毕业设计工作,秉着严谨大胆、 虚心踏实的工作态度完成毕业设计的全部任务,并独立撰写毕业设 计说明书一份。 实验用品与仪器如下: 名称/化学式生产厂商纯度 Ni(NO3)2·6H2O 国药集团化学试剂有限公司分析纯 AR CO(NH2)2 国药集团化学试剂有限公司分析纯 AR C2H5OH 国药集团化学试剂有限公司分析纯 AR NMP 国药集团化学试剂有限公司分析纯 AR 十二烷基硫酸钠(SDS)国药集团化学试剂有限公司分析纯 AR 乙炔黑太原迎泽力之源电池销售部电池级聚偏氟乙烯(PVDF)- 电池级 泡沫镍梧州三和新材料有限公司- 名称生产商型号水热反应釜天合科研协作中心250 度 F4 内衬 250mL 电子天平赛多利斯科学仪器有限公司SQP 磁力搅拌器江苏省金坛荣华仪器制造有限公司HJ-6 型 真空干燥箱上海索谱仪器有限公司DZF6020 真空管式烧结炉合肥科晶材料技术有限公司OTF-1200X-S 真空石英管式炉合肥科晶材料技术有限公司SWGL-1400X 砂芯过滤装置建湖县近湖镇亚东玻璃仪器厂100ml 油压式压片机- - 电化学工作站上海辰华仪器有限公司CHI660C X 射线衍射仪德国 BRUKER-AXS 公司Bruker-AXS D8 Advance 扫描电子显微镜美国 FEI 公司FEI Quanta 250FEG

工业设计毕业设计任务书(模板)

湖南工学院2012届毕业设计(论文)课题任务书系:建筑工程与设计艺术学院专业:工业设计专业 指导教师学生姓名 课题名称 内容及任务一、设计目的 毕业设计是工业设计专业本科培养计划中最重要的综合性实践教学环节。旨在培养工业设计专业本科学生综合应用所学基础课、专业基础课及专业课知识和相应技能,解决具体工业设计问题所需的综合能力和创新能力。本设计方向主要培养学生工业设计中基于人文关怀的创新设计能力。 二、选题背景 该课题主要在2008下半年,对株洲市长江路公共设施现状进行研究,针对其表现出来的不足,加以思考、设计,以达到长江路公共设施摆放统一、美观、实用、能够体现株洲市文化特色等要求。在株洲市公共设施的整体建设不容乐观的情况下,普遍存在:①缺乏独特性,不能够反映地域特色;②公共设施本身缺乏系统性;③存在资源使用不足和浪费的情况;④设施使用过程中存在很多不人机的现象,例如株洲市属于南方城市,夏季炎热多雨、冬季阴冷,现有设施不能较好地解决这一问题;⑤一条道路上的公共设施原则上要统一,但也要根据周围环境的不同而有所变化,这也是目前株洲市公共设施存在的一大问题。 针对这些情况,要求对株洲市长江路这一主干道为重点调研对象,通过实地调研走访,分析该道路上公共设施的设计方向,提出自己的创新方案。 三、设计定位 1、设计风格:从城市地域特色中选取适当元素,在注重公共设施满足市民生活需求的前提下,更加注重公共设施本身所传达出的城市文化信息。不同地段有不同的城市特色,这些都可以融入到公共设施的设计当中去。 2、人文关怀:针对株洲市市民的生活以及出行习惯,尤其关注弱势群体,在公共设施的设计和摆放上加以参考以上数据,不仅能使公共设施达到一个比较高的利用率,更方便了不同人群的生活。 3、模块化设计:公共设施是一种户外家具,运用模块化的设计理念,通过模块的组合配置,就可以创建不同需求的设施,来满足城市人群不同的需求,更可以降低成本,缩短生产周期。 4、设计目标:主要通过对长江路现有公共设施调研找出不足,加以创新设计,尤其是在公交系统的设计上,通过这些设计来给市民的出行和生活带来更大的便利,也通过这些设计来改善市民的户外生活质量;同时通过这些公共设施的设计,使公共设施更加易于使用和维护,并且使公共设施在造型和色彩上和周围的环境更加协调,更富于株洲市的文化特色。

毕业设计(论文)任务书及指导书

武汉工业学院毕业设计(论文)任务书及指导书 二、指导书(撰写参考内容,字数不限,可自拟标题) (1)可行方案的筛选方法提要(设计类);研究方法的思路(论文类) 1)参数设定 种群中个体数目s=30,交叉概率pc=0.9。 2)对机器人路径空间建模 采用删格法对机器人工作空间进行划分,用序号标识栅格,机器人可在其中运动的栅格称为自由栅格,包含障碍物的栅格称为障碍物栅格。 3)生成初始种群 生成30个个体的初始种群 4)适应度函数设置 设定为f=19-n 5)选择 6)交叉 7)达到终止条件终止。终止条件为适应度值最 (2)已学过的相关知识提要以及与本课题有关的新知识 系统学习过智能控制这本教材,以及在网上浏览过很多关于机器人的论文 (3)毕业设计(论文)进度安排

(4) (5)本题目的重点和难点 本题目重点是选择好参数进行选择交叉操作,难点是调试仿真程序 (6)若有同组其它学生参加同一课题应指明所做题目之间的关系 没有同学和我做的课题相同 (7)列出主要参考文献和研究与设计内容的检索关键词(中英文) 参考文献如下: [1]孙立宁. 机器人技术发展状况.机电论文.2008年6月 [2]张寒松.郑桐等.现代机器人技术与中国国民经济发展.天津职业技术师范学院学报.2004年4月 [3]张琳.移动机器人局部路径规划方法的研究.济南大学硕士论文.2008年6月 [4]唐潮.彭范.聂江天. 关于机器人的论文.硕士论文.2007年5月 [5]陈志兴. 用于多机器人路径规划的多目标遗传算法的研究. 中南大学硕士学 位论文.2009年5月 [6]李贻斌.周凤余.李彩虹等.移动机器人在线路径规划算法研究.系统工程 与电子技术。2000. 2),79~81 [7]廖平.傅杰.基于遗传算法的水下机器人的一种路径规划方法.中国工网.2005年9月 [8]孙树栋.曲彦宾.遗传算法在机器人路径规划中的应用研究[J].西北工业 大学学报,1998,16(1):79-83. [9]赵峰.动态环境下移动机器人的路径规划.北京工业大学硕士学位论文, 2003年5月

本科毕业设计任务书(范本)

(说明:请把红色字体部分根据个人题目的不同进行更改) 广州大学华软软件学院 本科毕业设计任务书 设计题目浅析计算机病 毒的免杀技术 系别网络技术系 专业网络工程 班级10网络设计与管理(1)班 学号1040217901 学生姓名郑天骄 指导教师田宏政 下发时间:2014年10月28日

毕业设计须知 1、认真学习和执行广州大学华软软件学院学生毕业论文(设计)工作管理规程; 2、努力学习、勤于实践、勇于创新,保质保量地完成任务书规定的任务; 3、遵守纪律,保证出勤,因事、因病离岗,应事先向指导教师请假,否则作为缺席处理。凡随机抽查三次不到,总分降低10分。累计缺席时间达到全过程l/4者,取消答辩资格,成绩按不及格处理; 4、独立完成规定的工作任务,不弄虚作假,不抄袭和拷贝别人的工作内容。否则毕业设计成绩按不及格处理; 5、毕业设计必须符合《广州大学华软软件学院普通本科生毕业论文(设计)规范化要求》,否则不能取得参加答辩的资格; 6、实验时,爱护仪器设备,节约材料,严格遵守操作规程及实验室有关制度。 7、妥善保存《广州大学华软软件学院本科毕业设计任务书》。 8、定期打扫卫生,保持良好的学习和工作环境。 9、毕业设计成果、资料按规定要求装订好后交指导教师。凡涉及到国家机密、知识产权、技术专利、商业利益的成果,学生不得擅自带离学校。如需发表,必须在保守国家秘密的前提下,经指导教师推荐和院领导批准。

课题名称浅析计算机病毒的免杀技术 完成日期:2015年4月30日 一、题目来源及原始数据资料: 随着计算机技术的飞速发展,信息网络已经成为社会发展的重要保证。有很多是敏感信息,甚至是国家机密。所以难免会吸引来自世界各地的各种人为攻击,窃取、篡改、删添等。随着时代的发展,网络已经成为了一个我们生活的必需品。而Web站点已经随处可见,其应用也是遍及各个领域,并已和我们日常生活息息相关。但是针对站点的渗透攻击也是缕缕出现,给我们带来了很大的危胁。因此我们必须展开对Web站点渗透技术的研究。 教师根据学生对站点的内部结构研究结果,分析可能成功的渗透技术,通过模拟攻击过程展示渗透成功之后的效果并寻求解决办法,进而提出一套行之有效的防护措施,顺利完成本次毕业设计任务。 二、毕业设计要求: 要求:详细的Web站点渗透技术的研究。大致可分为以下七部分: 1、网络安全现状的分析; 2、常见的站点结构组成; 3、常见的渗透技术分析; 4、模拟主要的攻击技术; 5、提出防范思路并设计解决方案; 6、必要的实现过程展示; 7、总结与未来工作的展望; 具体要求如下: 1、分析国内、国外的网络安全现状,了解网络安全方面主要存在的问题。 2、了解常见的Web站点结构、机制和原理。 3、了解针对站点的渗透技术。 4、分析主流的渗透攻击技术; a、文件与内存特征码定位; b、压缩整容,加壳免杀;

毕业设计(论文)任务书范文

西南科技大学城市学院 本科毕业设计(论文)任务书 题目名称 当代汽车服务营销策略的探究----以北京现代汽车4S店为例学生姓名专业班级学号 题目来源□教师科研□社会实践 □实验室建设■其它 题目 类型 □理论研究■应用研究 □设计开发□其它 选题背景及目的 到2010 年,我国汽车服务市场总额可以达到3000 亿元,汽车服务产业已经进入中国国民经济主流,成为一个战略性支柱行业。能否建立起具有国际竞争力的汽车业服务体系,关系到我国汽车行业的发展和未来。特别是当国际销售潮流已经出现的历史性变化都与越来越细化的服务有关时,一个全新的服务的世纪就到来了。有观点认为,21 世纪的世界经济是出售服务的经济,在这个“出售服务”的年代,谁能取得最后的胜利,就看谁的服务更到位、更有特色、更受消费者欢迎。服务营销工作本身可以给公司带来丰厚的经济效益,更为重要的是,有正确战略指导的、全方位的服务营销工作,可以确保客户的满意度和忠诚度。因此,服务的竞争必将成为未来汽车行业竞争的主战场。以服务为导向,服务营销为途径,创建4S店自身的品牌是汽车4S店发展之战略措施。 多年以来,对汽车的热门话题几乎都倾注在制造业和产品上,对汽车销售及其服务业的关注度相对较小。然而随着汽车市场的发展,单凭产品的独立优势赢得竞争己非常困难,汽车售后服务建设已成为当今所关注的重心。本文运用市场营销、服务营销等理论,通过对现代汽车4S店服务营销策略的研究分析,构建汽车4S店服务营销策略的理论框架,为4S店服务营销实践提供理论资源和指导。 工作任务及要求 (一)总体要求 1.选题符合专业培养目标,难易度适当,具有理论意义或实际价值。 2.论文必须文题相符,概念清楚,思路清晰,层次分明,论据充分、可靠,引用正确,论证有力。 3.论文符合写作规范。 4.论文的字数要求在0.9万字以上。 5.论文必须清楚反映自己的学术观点和学术水平,严禁抄袭。 (二)进度要求 2011年10月中旬——2011年11月中旬:选题,写论文提纲和开题报告; 2011年11月下旬——2012年4月初:写作论文初稿,论文修改及定稿; 2012年4月中旬——4月下旬:毕业论文答辩。

毕业设计任务书范本

武汉大学本科毕业论文(设计)任务书 毕业论文(设计)题目学科本体构建方法研究——语文学科本体的构建 姓名:李鑫学号: 200431500120 学院:武汉大学计算机学院 一、毕业论文(设计)题目的来源 本次毕业设计题目来自教师科研课题中的部分研究内容。 二、毕业论文(设计)的主要内容 随着教育信息化和计算机技术的发展,人们开发出了大量教育资源并构建了许多教育资源库,为教师自主制作个性化的教学课件提供了丰富的素材。但是,不同的人对同一事物的描述存在一定的差异,在资源标注中也存在同样的问题,这严重影响了教育资源的使用效率。相同问题也出现在网络信息资源的管理中。为了解决这一问题,人们引入本体的思想,以提高资源的管理效率。 学科本体的构建,试图在学科范围内建立起一种领域的规范,为资源制作者描述资源,为资源使用者查询资源提供统一的、规范的、明确的描述机制。共同的描述机制将大大提高资源查询时的查全率和查准率。本次毕业设计拟以语言类学科中的语文学科为例,在分析语文学科知识的基础上,抽取语文学科的概念术语,建立起术语间的关系,在此基础上,构建语文学科本体,并采用一定的描述方式予以表达,为语言类学科资源的构建与管理提供一定的借鉴。 三、毕业论文(设计)的基本要求及应完成的成果形式 通过本次毕业设计,要求完成并提交包含如下内容的光盘: ⑴关于本体相关内容学习的综述报告,包括以下内容: ·本体的概念、特性、构成、描述方法及本体的应用背景简介; ·现有本体构建方法论及其对比分析; ⑵完成语文学科本体的构建,并采用一定的描述方式予以表达: ·在学习和构建本体过程中各阶段的规范文档; ·语文学科本体。

毕业设计任务书模板

毕业设计(论文)任务书 学院专业班学生学号 一、毕业设计(论文)题目: 二、毕业设计(论文)工作规定进行日期:2014 年 3 月17日起至2014年5月30日止 三、毕业设计(论文)进行地点: 四、任务书的内容: (Ⅰ)任务背景和意义: Web 应用程序正变得越来越复杂,越来越重要。为了帮助管理这种复杂性,需要为 Web 应用程序建模。近年来,作为概念建模语言的UML为语义Web建模,已经成为一个重要的研究主题。在尝试用 UML 为 Web 应用程序建模时,很明显它的一些构件不能与标准的 UML 建模元素一一对应。为了让整个系统(Web 构件,以及传统的中间层构件)使用同一种建模表示法,必须扩展 UML。因此对扩展UML的研究成为了人们注意的焦点。本课题涉及Web程序构建的过程以及UML的具体应用,包括架构定义、需求分析、系统设计、实施等。

(Ⅱ)内容与要求 (1)广泛收集与阅读有关该课题的最新国内 外文献资料(导师给定的阅读资料或通过自己收集 的资料),了解基于UML软件度量技术的最新进展; (2)分析Web应用的特点和目前建模所面临 的一些问题,提出基于Web应用的开发模型, (3)结合扩展UML建模方法,开发一个基于WEB的软件应用系统,达到理论研究与实际应用问 题相结合的目的。 (Ⅲ)进度安排 第4周:查阅与题目有关的资料,完成开题报告。 第5~9周:深入学习并掌握UML统一建模语言,掌握基于UML的建模方法,能够在分析建模开 发过程中熟练应用;分析Web应用的特点和目前Web应用建模所面临的一些问题,提出基于Web应 用的UML扩展,建立开发模型;确定要做的Web 应用项目,进行Web应用项目的基本建模。 使用上所做的UML扩展。撰写论文正文。 并准备答辩。

毕业设计任务书Word

XXXX大学 毕业设计(论文)任务书 设计题目_____________________ 或论文题目 ________________________ 学院 ________________ 专业名称__________________ 班级 ________________ 姓名 ________________ 指导教师__________________ 内封面 日期填写本学期开学时下达任务时的日期:应与内容要求的日期一致 年月日

内容和要求 计算机打印。但是教师必须亲笔签字 第一页 指导教师签字:本人签字 年月日填写日期为下达任务时 的日期应与上一页任务书的日期 一致 毕业论文(设计)评价表(一) 指导教师对毕业论文(设计)的评价:

指导教师手工填写评语,不能打印。注意填写成 绩,满分:25分 指导教师签字:本人签字年月日填写日期为写 评语时日期 评阅人对毕业论文(设计)的评价: 评阅人手工填写评语,不能打印。注意填写成绩, 满分:25分。 第二页 评阅人签字:本人签字(注:不能是指导教师中的任何一人) 年月日填写日期为写评语的日期,应与上栏同期或在之后。 答辩小组成员(由答辩秘书统一填写本表)

毕业论文(设计)评价表(二) 答辩小组评语及成绩: 经答辩小组讨论决定评语和成绩。由秘书填写,组长签字满分35分。

答辩小组组长签字: 年月日填写日期为答辩当日 答辩委员会给定成绩: 由院答辩委员会审议评定的成绩为最终成绩。 学校要求: 优秀20%,良好40%,中等30%,及格和不及格10% 第四页 答辩委员会主任签字: 年月日填写日期为当日毕业论文(设计)起止日期:年月日至年月日记住填写日期 毕业论文(设计)答辩日期:年月日填写答辩当日 毕业论文(设计)评价表(三) 专家对毕业论文(设计)的评语: 专家签字: 年月日

2014本科毕业设计任务书(修改)

2014本科毕业设计任务书(修改)

毕业设计(论文)任务书 学院电气与电 子工程学 院 指导 教师 陈俊 职 称 讲师 学生姓名专业 班级 学 号 设计 题目 张紧器测试平台伺服电机控制设计

设计内容目标和要求一.设计内容目标: 近年来,汽车发动机的链式正时传 动越来越广泛地采用链传动系统, 其尺寸紧凑,高稳定性、高耐磨性 的显著特征是齿轮传动和带传动所 不具备的,显示了其广阔的应用前 景。张紧器测试平台模拟张紧器运 行环境,设计两个静态试验,检验 汽车张紧器的形变情况。 二. 设计内容要求: 1. 查阅国标或企业标准确定合理方案; 2. 设计系统的硬件电路,CAD绘制 图纸; 3. 熟悉CX-P软件,编写PLC控制 程序; 4. 编写上位机界面。 三.设计进度: 1. 2月~3月:查阅相关资料,阅读 文献,完成开题报告; 2. 4月:系统方案确定,硬件电路设 计,软件编写与调试;

3. 5 月~5月中:完成毕业论文初稿; 4. 5月中~6月1日:修改毕业论文, 并最终定搞,完成毕业答辩; 5. 完成3000字专业英文翻译。 指导教师签名: 月日 基层 教学单位审核 学院 审核 毕业设计(论文)任务书

学院电气与电 子工程学 院 指导 教师 陈俊 职 称 讲师 学生姓名专业 班级 学 号 设计 题目 张紧器测试平台异步电机控制设计 设计内容目标和要求一. 设计内容目标: 近年来,汽车发动机的链式正时传动越来越广泛地采用链传动系统,其尺寸紧凑,高稳定性、高耐磨性的显著特征是齿轮传动和带传动所不具备的,显示了其广阔的应用前景。张紧器测试平台用异步电机模拟张紧器运行环境,使其在怠速、加速、减速等交变速度过程中的受到高冲击,检验汽车张紧器的耐久性能。 二. 设计内容要求: 1. 查阅国标或企业标准确定合理方案; 2. 设计系统的硬件电路,CAD绘制 图纸;

2014毕业设计任务书

青岛农业大学毕业论文(设计)任务书 论文(设计)题目智能门人员计数器设计 要求完成时间 2014.06.16 论文(设计)内容(需明确列出研究的问题):查阅相关资料(10 篇中文,两篇外文),了解和掌握人员计数器的工作特点与原理,完成 开题报告;选择合适的控制方法,完成智能门人员计数器的软件和硬 件设计。 系统采用单片机作为核心控制芯片,选择合适的器件,搭建外围 电路:包括电源电路、单片机系统电路、人员感应装置、人机交互接 口电路、显示电路等。所设计硬件电路采用PROTEL绘图软件完成原理 图和PCB图绘制,所设计的主要控制时序经过实验系统验证。最终完 成理论依据充分、设计正确、叙述清晰、格式规范的毕业论文撰写。 资料、数据、技术水平等方面的要求 1)设计能够供系统正常工作的不同电压等级的电源,包括5V、12V 及供A/D使用的参考电源。 2)设计模式正确的系统电路,能够实现有线或者无线数据收集。 3)能够与电脑上位机通信,报告人员数量及工作状态 4)具有按键操作和基本信息提示,便于操作;显示器件自行选择。 5)能够区分多人进入的异常状态,采用不同检测与计算策略,实 现计数器的正常工作。通过响应上位机发送指令,进行相应操作。结 合实际考虑软件设计结构。 指导教师签名: 2014 年 2 月 24 日

青岛农业大学毕业论文(设计)任务书 论文(设计)题目基于CAN的数据采集系统设计 要求完成时间 2014.06.16 论文(设计)内容(需明确列出研究的问题):查阅相关资料(10 篇中文,两篇外文),了解和掌握CAN总线的通信协议、工作原理,完 成开题报告;选择合适的设计方法,完成数据采集系统的软件和硬件 设计。 系统采用单片机作为核心控制芯片,选择合适的CAN总线通信器 件,与节点传感器,搭建外围电路:包括电源电路、单片机系统电路、 无线通信电路、人机交互接口电路等。所设计硬件电路采用PROTEL绘 图软件完成原理图和PCB图绘制,所设计的软件程序经过实验系统验 证。最终完成理论依据充分、设计正确、叙述清晰、格式规范的毕业 论文撰写。 资料、数据、技术水平等方面的要求 1)采用DC12V供电,设计能够供系统正常工作的不同电压等级的 电源,包括5V及供AD使用的参考电源。设计电路要满足设备低功耗 的要求。 2)设计2个设备节点,代表不同信号采集点,并进行互相通信。 3)一个采集点作为集控点,具有按键操作和基本信息提示,便于 操作;显示器件自行选择。该节点具有频率采集功能和开关量采集功 能。 4)一个子节点具有温度、湿度等监测功能,并通过总线发送给另 一节点,并在另一节点上显示温湿度信息。 5)可根据实际应用需求考虑系统应用软件的设计。 指导教师签名: 2014 年 2 月 24 日

毕业设计任务书

北京联合大学毕业设计(论文)任务书 题目:基于Qt的成品粮仓储管理信息系统设计 专业:物流工程指导教师:刘景云 学院:自动化学院学号:2011100358220 班级:物流1102 B 姓名:董豪 一、主要内容和基本要求 仓储是现代物流和供应链系统的关键环节,提高其信息化程度与数据采集速度是提升物流管理水平与效率的重要保障。本课题将条码技术应用到仓储系统,构建基于条码技术的仓储管理信息系统,实现货物信息的自动采集,提高作业效率、减小出错率、改善仓储作业流程。 本课题设计一种基于条码技术的仓储管理信息系统系统,完成以下设计任务:探讨条码技术在仓储管理中的应用模式,进行信息系统进行需求分析,采用Qt与SQL Server 2005作为系统和数据库开发工具,设计用户管理、入库、出库、在库管理、统计管理等功能模块。 本课题需要学生综合运用软件开发及数据库技术等理论和知识来完成本课题内容,能够围绕课题结合实际的业务与技术条件进行分析,完成课题需求及设计过程,编写和调试程序,实现系统功能要求,进行总结、归纳,最终完成毕业论文。 开发环境: 硬件:PC机,Honeywell 1900条码扫描枪 软件:Qt SDK for 4.7.0,SQL Server 2005 基本要求: 1、系统功能完整; 2、各功能达到指标要求; 3、系统具有较好的设备适应能力; 4、具有一定的扩展性。 二、调研资料情况 1、研究背景 成品粮,即经过加工后的符合一定标准的成品粮食。我国作为一个世界粮食大国,做好对成品粮食的储存管理工作有着极其重大的意义,不仅关系到国家安全,也关系到社会稳定。随着时代的进步,运用传统的工作方式对粮食进行管理,效率低,而且

相关文档
最新文档