CN109460605B - A method for predicting the flow rate of a large-scale low-lift pump - Google Patents

Abstract

The invention discloses a kind of methods for predicting large-scale low-lift pump flow, comprising steps of S1 calculates large-scale low-lift pump pressure difference by test model；S2 passes through cloud computing computation model large size low-lift pump pressure difference；S3 calculates prototype large size low-lift pump pressure difference by cloud computing；S4 predicts the discharge coefficient and truncated error of large-scale low-lift pump；S5 predicts large-scale low-lift pump field measurement flow.Cloud computing is introduced in the method for predicting large-scale low-lift pump flow, solving by the numerical computation method of cloud computing leads to the imponderable problem of work station due to generating flood tide grid using model meshes size in the calculating of prototype numerical value, avoids since hardware problem will calculate the problem that grid interval increases and causes analog result inaccurate.The present invention can accurately predict large-scale low-lift pump field measurement flow, so that it is determined that pump assembly efficiency and energy unit consumption in pump station engineering, scientifically and accurately complete pump installation performance test.

Description

A method for predicting the flow rate of a large-scale low-lift pump

technical field

The invention belongs to the field of hydraulic engineering, in particular to a method for predicting the flow rate of a large-scale low-lift water pump.

Background technique

The flow rate of large-scale low-lift pumps is a key parameter to determine the efficiency and unit energy consumption of the pumping station. It is related to whether the performance test of the prototype pump can be completed scientifically and accurately, so as to test the research, design and manufacturing level of the large-scale pump. Because the water inlet channel of the large-scale low-lift pump is short, the section is irregular, and it is made of reinforced concrete cast-in-place, the measurement and monitoring of its flow has always been an unsolved problem in the field of pump station engineering.

At present, the flow data of large-scale low-head pumps are basically obtained from the measured water levels in the inlet and outlet pools, and the performance curve of the prototype pump device after the device head is similar to the performance curve of the model pump device. Due to limitations, the original model cannot unify the similarity criteria, resulting in the existence of scale effects, data fitting and interpolation errors in the original model similarity conversion, etc., it is difficult for this flow estimation method to achieve high measurement accuracy; the field flow test method mainly has five holes The probe flow measurement method, the flow measurement method around the flow tube, the inner-mounted ultrasonic flowmeter, the flow meter method, the brine concentration method, etc., these methods are affected by the measurement accuracy or the high cost, which limits their use in large-scale low-head pumps. It is widely used in water flow channels; scholars have calculated the relationship between flow rate and pressure difference through theoretical analysis, and summed up the flow measurement method of differential pressure in the water inlet flow channel, which is a relatively simple flow measurement method and is more suitable for large-scale low lift. On-site flow monitoring of the pump, but it needs to determine the relationship between the flow and the pressure difference in advance. Usually, the flow meter method or the brine concentration method used in the field test of the pump station is used to determine the relationship between the flow and the differential pressure. This calibration method has a large workload and is limited. There are many conditions and the error is difficult to meet the accuracy requirements, which makes it difficult to popularize and apply the differential pressure flow measurement method of the inlet flow channel. In addition, the selection of pressure measurement points also limits the simplicity of the flow measurement method. In recent years, CFD numerical calculation has been applied in fluid dynamics. There are many applications. Some scholars try to obtain the coefficient of field measured flow and differential pressure through CFD calculation to predict the measured flow of the pump inlet channel. The method first obtains the relationship coefficient k _mt between the flow Q _mt and the pressure difference ΔP _mt through a physical model test. , and then obtain the relationship coefficient k _mc between the flow rate Q _mc and the pressure difference ΔP _mc of the model in the numerical calculation through numerical calculation, and then obtain the relationship coefficient k _pc between the flow rate Q _pc and the pressure difference ΔP _pc of the prototype in the numerical calculation through numerical calculation, and finally K _pc is determined by the relationship between _kmt and _kmc to obtain k _pt , and finally the field measured flow rate is obtained through the differential pressure formula. The above method of obtaining field measured flow rate through CFD calculation is unscientific in that the truncation error of the fitting equation cannot be determined according to the proportional ratio, and secondly, on the premise of ensuring the correct turbulence model, grid quality and boundary layer processing method, the computer cannot carry large-scale low-voltage For the calculation of the massive grid of the head pump prototype, the computer hardware requirements can only be met by enlarging the grid size, which greatly reduces the accuracy of the CFD numerical calculation.

SUMMARY OF THE INVENTION

Purpose of the invention: In view of the problems existing in the prior art, the purpose of the present invention is to provide a method for predicting the flow rate of a large-scale low-lift pump, so as to accurately and conveniently predict the measured flow rate of a large-scale low-lift pump station, thereby determining the efficiency of the pump station device With the unit energy consumption, the performance test of the prototype pump device is completed scientifically and accurately.

In order to achieve the above purpose of the invention, the technical solution specifically adopted in the present invention is: a method for predicting the flow rate of a large-scale low-lift water pump, comprising the following steps:

S1. Calculate the pressure difference of the large-scale low-lift water pump through the test model;

S11. Establish a physical test model of the water inlet channel of a large-scale low-lift pump;

Further, as a preferred solution of the present invention: the large-scale low-lift water pump inlet channel test model described in the step S11 adopts the overall normal hydraulic model, is designed according to the gravity similarity criterion, and comprehensively considers the resistance of the model water flow. The selection of the square area and the model pump is to choose the model linear scale λ _l and the pump model scale λ _n , the flow scale Velocity scale λ _v =λ _l λ _n , the slope and roughness of the model are obtained by scaling the linear scale;

S12. Select two suitable pressure measuring points R and W in the water inlet channel of the large-scale low-lift pump;

Further, as a preferred solution of the present invention: the position of the two pressure measuring points in the step S12 satisfies that the pressure difference between the two points in the model is large enough to facilitate the calibration of the relationship between the pressure difference and the flow rate of the water inlet channel, and at the same time. Satisfy the rationality and feasibility of the pressure measurement technology at the engineering site;

S13. Install high-precision sensors in the two pressure measuring points R and W in the water inlet channel of the large-scale low-lift pump;

S14, carry out the physical model test, the flow rate is Q _j , where Q is the flow rate, the unit is m ³ /s, and j is the number of operating conditions;

S15. Collect the pressure of the two pressure measuring points in the physical model test unit Pa;

S16. Obtain the pressure difference between the pressure measuring points R and W under the physical model test conditions

S2. The pressure difference of the large-scale low-lift water pump is calculated through the cloud computing model;

S21, constructing a three-dimensional model of a large-scale low-lift water pump inlet channel through the three-dimensional modeling software SolidWorks;

S22. Use the cloud computing software _ICEM to divide the grid, the grid scheme is Ni, and the corresponding grid overall layout ratio is And the corresponding maximum grid size is ω _i , the corresponding total amount of generated grids is M _i , respectively, and ω decrease with the increment of i, and M increases with the increment of i, where i is the ordinal number of the grid scheme;

Further, as a preferred solution of the present invention: the grid type of the grid described in the step S22 is a tetrahedral unstructured grid with strong adaptability, and the area near the measuring point is locally encrypted;

S23, calculate the calculation files of the grid scheme N _i and N _i+1 output in step S22 by cloud computing Fluent;

Further, as a preferred solution of the present invention: the turbulent flow model is Realizable k-ε, the inlet adopts the velocity inlet condition v _mj , the outlet boundary adopts the free outflow condition outflow, the free water surface adopts rigid cover assumption symmetry, and the side wall adopts wall, The steel roughness is 0.012, the discrete method of the control equation is the finite volume method, the diffusion term adopts the second-order central difference format, the convection term adopts the Quick format, the coupling of pressure and velocity adopts the SIMPLEC algorithm, and the calculation method adopts the parallel calculation;

S24. Collect the pressures of the pressure measuring points of the grid schemes N _i and N _i+1 of the model large-scale low-head water pump in the cloud computing through cloud computing CFD-Post post-processing unit Pa;

S25. Obtain the pressure difference between the pressure measuring points _R and W of the grid scheme Ni and Ni ₊₁ of the inlet flow channel of the large-scale low-lift pump of the cloud computing model

S26, S26 when If satisfied, output the result, otherwise return to S22;

S27 outputs the pressure difference ΔP _mcj of the pressure measuring points R and W under the condition of the model workstation, and the optimal grid scheme N _i ;

S3. Calculate the pressure difference of the prototype large-scale low-lift water pump through cloud computing

S31, constructing a three-dimensional model of a prototype large-scale low-lift water pump inlet channel through the three-dimensional modeling software SolidWorks;

S32. Use the cloud computing meshing software ICEM to mesh the three-dimensional model of the prototype large-scale low-head pumping station, and the mesh size is the same as the optimal mesh scheme of the model;

S33, calculate the output file in step S32 by cloud computing Fluent, and the Fluent setting is the same as in step S23;

Further, as a preferred solution of the present invention: prototype flow The import area is Then v _pj = v _mj , the turbulence model is Realizable k-ε, the inlet adopts the velocity inlet condition v _pj , the outlet boundary adopts the free outflow condition outflow, the free water surface adopts the rigid cover assumption symmetry, the side wall adopts the wall, and the reinforced concrete roughness is 0.013, the discrete method of the governing equation is the finite volume method, the diffusion term adopts the second-order central difference format, the convection term adopts the Quick format, the coupling of pressure and velocity adopts the SIMPLEC algorithm, and the calculation method adopts parallel computing;

S34. Collect the pressure at the pressure measuring points R' and W' of the prototype large-scale low-head water pump in cloud computing through cloud computing CFD-Post post-processing unit Pa;

S35. Collect the pressure of the cloud computing prototype large-scale low-lift water pump inlet channel at the pressure measuring points R' and W', and process the data;

S36. Output the pressure difference between the pressure measuring points R' and W' under the cloud computing conditions of the prototype

S4. Predict the flow coefficient and truncation error of the large-scale low-lift water pump;

S41. The error of the pressure difference between the model test and cloud computing is the same as the error between the field measurement and the cloud computing pressure difference, then: Get:

S42. According to Bernoulli equation:

where z _W′ and z _R′ are the positional potential energy at the pressure measuring points R′ and W′, m;

S43. Fit the regression equation of the flow rate Q _pj and Δh _ptj of the large-scale low-head water pump of the cloud computing prototype: b is the truncation error of the fitting equation;

S5. Predict the on-site measured flow of the large-scale low-lift pump: the coefficients of the on-site measured flow and pressure difference k = k _p , b = b _p , then the on-site measured flow is complete.

Beneficial effects: The present invention can accurately and conveniently predict the field measured flow of a large-scale low-lift pump station, thereby determining the efficiency and unit energy consumption of the pump station in the project, and scientifically and accurately complete the performance test of the pump device in the project.

Description of drawings

FIG. 1 is a flow chart of the method for predicting the flow rate of a large-scale low-head water pump according to the present invention.

Detailed ways

The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

Figure 1 shows the flow chart of the method for predicting the flow rate of a large-scale low-head water pump of the present invention. In the method for predicting the flow rate of a large-scale low-lift water pump in this embodiment, a physical test model of the inlet water channel of the large-scale low-lift water pump is first established, two suitable pressure measuring points are selected in the water inlet channel, and high-precision sensors are installed at the two pressure measuring points. , Collect the pressures of the two pressure measuring points in the physical model test under different flow conditions, and obtain the pressure difference of each flow rate at the two pressure measuring points under the physical model test conditions; then build a model through the three-dimensional modeling software SolidWorks. Using cloud computing ICEM to divide the model grid, perform numerical analysis through cloud computing Fluent and post-processing through CFD-Post to obtain the pressure difference of the corresponding flow model in cloud computing and the corresponding optimal grid scheme; then construct The three-dimensional model of the inlet flow channel of the prototype large-scale low-head water pump, using cloud computing to obtain the pressure difference of each prototype flow after similar conversion at the pressure measuring point, the grid size of each prototype flow and the corresponding model The optimal grid scheme is the same; because the error of the pressure difference between the model test and the cloud computing model is the same as the error between the field prototype measurement and the cloud computing prototype pressure difference, the pressure difference of the field measured prototype is obtained, and then calculated according to the Bernoulli equation. The pressure head difference corresponding to the actual measured prototype is fitted to the regression equation of the flow rate and pressure difference of the cloud computing prototype large-scale low-head pump to obtain the relationship coefficient and truncation error of the field-measured prototype flow rate and pressure difference, and finally the large-scale low-head pump site is predicted. measured flow. The specific steps are:

S1. Calculate the pressure difference of the large-scale low-lift water pump through the test model;

S11. Establish a physical test model of the water inlet channel of a large-scale low-lift pump;

Further, as a preferred solution of the present invention: the large-scale low-lift water pump inlet channel test model described in the step S11 adopts the overall normal hydraulic model, is designed according to the gravity similarity criterion, and comprehensively considers the resistance of the model water flow. The selection of the square area and the model pump is to choose the model linear scale λ _l and the pump model scale λ _n , the flow scale Velocity scale λ _v =λ _l λ _n , the slope and roughness of the model are obtained by scaling the linear scale;

S12. Select two suitable pressure measuring points R and W in the water inlet channel of the large-scale low-lift pump;

Further, as a preferred solution of the present invention: the position of the two pressure measuring points in the step S12 satisfies that the pressure difference between the two points in the model is large enough to facilitate the calibration of the relationship between the pressure difference and the flow rate of the water inlet channel, and at the same time. Satisfy the rationality and feasibility of the pressure measurement technology at the engineering site;

S13. Install high-precision sensors in the two pressure measuring points R and W in the water inlet channel of the large-scale low-lift pump;

S14, carry out the physical model test, the flow rate is Q _j , where Q is the flow rate, the unit is m ³ /s, and j is the number of operating conditions;

S15. Collect the pressure of the two pressure measuring points in the physical model test unit Pa;

S16. Obtain the pressure difference between the pressure measuring points R and W under the physical model test conditions

S2. The pressure difference of the large-scale low-lift water pump is calculated through the cloud computing model;

S21, constructing a three-dimensional model of a large-scale low-lift water pump inlet channel through the three-dimensional modeling software SolidWorks;

S22. Use the cloud computing software _ICEM to divide the grid, the grid scheme is Ni, and the corresponding grid overall layout ratio is And the corresponding maximum grid size is ω _i , the corresponding total amount of grids generated is M _i , and M increases with the decrease of i, where i is the ordinal number of the grid scheme;

Further, as a preferred solution of the present invention: the grid type of the grid described in the step S22 is a tetrahedral unstructured grid with strong adaptability, and the area near the measuring point is locally encrypted;

S23, calculate the calculation files of the grid scheme N _i and N _i+1 output in step S22 by cloud computing Fluent;

Further, as a preferred solution of the present invention: the turbulent flow model is Realizable k-ε, the inlet adopts the velocity inlet condition v _mj , the outlet boundary adopts the free outflow condition outflow, the free water surface adopts rigid cover assumption symmetry, and the side wall adopts wall, The steel roughness is 0.012, the discrete method of the control equation is the finite volume method, the diffusion term adopts the second-order central difference format, the convection term adopts the Quick format, the coupling of pressure and velocity adopts the SIMPLEC algorithm, and the calculation method adopts the parallel calculation;

S24. Collect the pressures of the pressure measuring points of the grid schemes N _i and N _i+1 of the model large-scale low-head water pump in the cloud computing through cloud computing CFD-Post post-processing unit Pa;

S25. Obtain the pressure difference between the pressure measuring points _R and W of the grid scheme Ni and Ni ₊₁ of the inlet flow channel of the large-scale low-lift pump of the cloud computing model

S26, S26 when If satisfied, output the result, otherwise return to S22;

S27 outputs the pressure difference ΔP _mcj of the pressure measuring points R and W under the condition of the model workstation, and the optimal grid scheme N _i ;

S3. Calculate the pressure difference of the prototype large-scale low-lift water pump through cloud computing

S31, constructing a three-dimensional model of a prototype large-scale low-lift water pump inlet channel through the three-dimensional modeling software SolidWorks;

S32. Use the cloud computing meshing software ICEM to mesh the three-dimensional model of the prototype large-scale low-head pumping station, and the mesh size is the same as the optimal mesh scheme of the model;

S33, calculate the output file in step S32 by cloud computing Fluent, and the Fluent setting is the same as in step S23;

Further, as a preferred solution of the present invention: prototype flow The import area is Then v _pj = v _mj , the turbulence model is Realizable k-ε, the inlet adopts the velocity inlet condition v _pj , the outlet boundary adopts the free outflow condition outflow, the free water surface adopts the rigid cover assumption symmetry, the side wall adopts the wall, and the reinforced concrete roughness is 0.013, the discrete method of the governing equation is the finite volume method, the diffusion term adopts the second-order central difference format, the convection term adopts the Quick format, the coupling of pressure and velocity adopts the SIMPLEC algorithm, and the calculation method adopts parallel computing;

S34. Collect the pressure at the pressure measuring points R' and W' of the prototype large-scale low-head water pump in cloud computing through cloud computing CFD-Post post-processing unit Pa;

S35. Collect the pressure of the cloud computing prototype large-scale low-lift water pump inlet channel at the pressure measuring points R' and W', and process the data;

S36. Output the pressure difference between the pressure measuring points R' and W' under the cloud computing conditions of the prototype

S4. Predict the flow coefficient and truncation error of the large-scale low-lift water pump;

S41. The error of the pressure difference between the model test and cloud computing is the same as the error between the field measurement and the cloud computing pressure difference, then: Get:

S42. According to Bernoulli equation:

where z _W′ and z _R′ are the positional potential energy at the pressure measuring points R′ and W′, m;

S43. Fit the regression equation of the flow rate Q _pj and Δh _ptj of the large-scale low-head water pump of the cloud computing prototype: b is the truncation error of the fitting equation;

S5. Predict the on-site measured flow of the large-scale low-lift pump: the coefficients of the on-site measured flow and pressure difference k = k _p , b = b _p , then the on-site measured flow is complete.

The above descriptions are only preferred embodiments of the present invention, and are not intended to limit the present invention in any form. Any person skilled in the art, without departing from the scope of the technical solution of the present invention, can Any non-essential modifications or equivalent changes made by the above embodiments still fall within the protection scope of the claims of the present invention.

Claims (5)

1. a kind of method for predicting large-scale low-lift pump flow, it is characterised in that: step includes that S1 is calculated by test model Large-scale low-lift pump pressure difference；S2 passes through cloud computing computation model large size low-lift pump pressure difference；S3 passes through cloud computing meter Calculate prototype large size low-lift pump pressure difference；S4 predicts the discharge coefficient and truncated error of large-scale low-lift pump；S5 prediction is big Type low-lift pump field measurement flow；

The step S1 includes specific steps, and: S11 establishes large-scale low-lift pump water inlet flow channel physical test model, S12 into Two suitable pressure tap R, W are chosen in water flow passage, S13 installs high-precision sensor at pressure tap, and S14 carries out physical model Test, S15 acquires pressure at pressure tap R, W, and handles data, and S16 exports the pressure difference Δ P of pressure tap R, W_mtj；

The method of determination of the two pressure tap positions step S12 determines multiple groups pressure tap in a model first, then passes through work Make station Fluent and carry out tentative calculation, then is post-processed by work station CFD-Post, several groups of sufficiently large survey pressures of output pressure difference Point, finally selection meets pressure tap R, W of engineering site measuring point pressure measurement technique reasonability and feasibility；

The step S2 includes specific steps: S21 establishes Three-dimension Numerical Model, S22 benefit to test model by SolidWorks With cloud computing ICEM subdivision grid, trellis schemes are respectively N_i, S23 cloud computing Fluent numerical analysis, S24 cloud computing CFD- Post post-processing, S25 acquire pressure at pressure tap R, W, output pressure difference Δ P_mcji、ΔP_mcji+1, S26 works asIt is full Otherwise sufficient then output is as a result, return to S22, the pressure difference Δ P of pressure tap R, W under the conditions of S27 output model work station_mcj, optimal net Lattice scheme N_i；

The step 22 trellis schemes N_i, corresponding grid total arrangement ratio isAnd corresponding maximum mesh having a size of ω_i, the grid total amount correspondingly generated is respectively Reduce with ω with being incremented by for i, M is incremented by with i's And increase, wherein i is the ordinal number of trellis schemes, and trellis-type is tetrahedron unstructured grid, is carried out to measuring point near zone Local cypher；

The method of determination of the step S27 Bestgrid scheme, the trellis schemes that step S22 is exported by cloud computing Fluent N_iAnd N_i+1Calculation document calculated, then post-processed by cloud computing CFD-Post, acquire that model in cloud computing is large-scale low to be raised Journey water pump trellis schemes N_iAnd N_i+1Pressure tap pressureUnit Pa obtains cloud computing mould Type large size low-lift pump water inlet flow channel trellis schemes N_iAnd N_i+1In the pressure difference of pressure tap R, WWhenMeet then output as a result, otherwise return step S22, it is final to determine Bestgrid scheme N_i；

The step S3 includes specific steps: S31 establishes Three-dimension Numerical Model, S32 cloud computing to prototype by SolidWorks ICEM subdivision grid, S33 cloud computing Fluent numerical analysis, S34 cloud computing CFD-Post post-processing, S35 acquisition pressure tap R ', Pressure at W ', and data are handled, the pressure difference Δ P of pressure tap R ', W ' under the conditions of S36 output prototype cloud computing_pcj；

The step S4 includes specific steps: S41 according to etc. relative errors find out Δ P_ptj, S42 finds out Δ according to Bernoulli equation h_ptj, S43 fitting Q_ptjWith Δ h_ptjEquation find out k_p, b_p；Relative errors, that is, the model test such as described and cloud computing pressure difference Error and field measurement are identical as the error of cloud computing pressure difference；The k_pFor pump capacity Q_pjWith Δ h_ptjCoefficient, b_pTo cut Disconnected error；

The step S5 includes specific steps: predicting large-scale low-lift pump field measurement flow；

In step S5, coefficient k=k of field measurement flow and pressure difference_p, b=b_p, predict large-scale low-lift pump field measurement Flow is

2. a kind of method for predicting large-scale low-lift pump flow according to claim 1, it is characterised in that: the step S11 establishes large-scale low-lift pump water inlet flow channel physical test model, whole normal state hydraulic model is used, by the similar standard of gravity It then designs, Selection Model linear ratio ruler λ is intended in the selection for comprehensively considering water flow in region of quadratic resistance law and model pump_lAnd pump mould Type is than ruler λ_n, flow-rate ratio rulerSpeed is than ruler λ_v=λ_lλ_n, scale to obtain model gradient and roughness according to linear ratio ruler.

3. a kind of method for predicting large-scale low-lift pump flow according to claim 1, it is characterised in that: the step ICEM, Fluent, CFD-Post method are all made of cloud computing technology in S2~step S3, and wherein Fluent turbulence model is Realizablek- ε, import use speed condition for import, and outlet border uses free discharge condition, and table is using just lid It is assumed that symmetry, side wall uses wall, and the discrete way of governing equation is finite volume method FVM, and diffusion term is using in second order Heart difference scheme, convective term use Quick format, and the coupling of pressure and speed uses SIMPLEC algorithm, and calculation is using simultaneously Row calculates.

4. the method according to claim 1 for predicting large-scale low-lift pump flow, it is characterised in that: the step S32 Described in mesh generation mode using trellis-type corresponding with Bestgrid scheme described in step S27, size of mesh opening and Mesh refinement mode is all the same.

5. the method according to claim 1 for predicting large-scale low-lift pump flow, it is characterised in that: the step S43 Middle k_p, b_pMethod of determination, first pass through cloud computing Fluent and output file in step S32 calculated, Fluent setting and step It is identical in rapid S23, it is then post-processed by cloud computing CFD-Post, acquires cloud computing mesarcs large size low-lift pump pressure tap Pressure at R ', W 'Unit Pa, obtain cloud computing prototype large size low-lift pump water inlet flow channel pressure tap R ', The pressure difference of W 'According to the error and field measurement and cloud computing pressure of model test and cloud computing pressure difference Strong poor error is identical, then:It acquires:According to Bernoulli Jacob side Journey:Acquire the head difference of R ', W 'Z in formula_W′、z_R′It is former for position potential energy m at pressure tap R ', W ' Type flow isWherein Q is model flow, unit m³/ s, j are operating condition ordinal number, and final fitting cloud computing prototype is big Type low-lift pump flow Q_pjWith Δ h_ptjRegression equationK has been determined_p, b_pValue.