3-D CFD Prediction Playing A Key Role In The Aerodynamic .
3-D CFD predictionplaying a key rolein the aerodynamicdesign of highperformance fansDr. Wilfried Rick, Dr. Eribert Benz, Alain GodichonWhile experimental work still plays an important role in flow physics research, simulation isincreasingly becoming the method of choice for the detailed study of flow phenomena. At ABB, CFD(Computational Fluid Dynamics) simulations based on 3-D Navier-Stokes equations have beenincorporated in the design process for high-performance fans in order to find new, optimumcomponent configurations and reduce experimental investigation to a minimum. Overall fanperformance is also improved, while new products can be developed faster and with a smallerdegree of uncertainty. The simulations have further shown that CFD can be used routinely formatching machine components in many other applications.n addition to carrying out experimentalwork, most fan manufacturers todayrely heavily on empirical one- (1-D) ortwo-dimensional (2-D) flow equationswhen determining the fundamentalaerodynamic design of their machines [1].Since relatively little has beenpublished on the use of CFD for studyingthe internal flows of centrifugal fans, it canbe concluded that it is still commonpractice for the manufacturers of thesefans to apply the principle of similitudeand scale new machines directly frompreceding designs. This procedure makesuse of extensive databases and is known toIABB Review 2/2000be reliable, but it is also time-consumingand provides no insight into the machine’sinternal flow pattern. To find out moreabout the flow phenomena within theimpeller blade passages or inside the fancasing, tedious and costly experimentalinvestigations are necessary. ABB SolyventVentec has consequently started to useCFD simulations based on 3-D NavierStokes equations in fan design. CFD bringsa systematic approach to the design anddevelopment process, enabling new,optimum configurations to be found andexperimental investigation to be kept to aminimum. What is more, CFD allows amore comprehensive understanding ofthe flow phenomena, providing a basis forimproved fan performance and fasterdevelopment of even completely newproducts while reducing uncertainty andminimizing costs.Designing industrial fansfor new demandsIn the traditional market for centrifugalfans there has been a tendency in the pastto emphasize the first-time cost of themachines, and improving or controllingperformance usually has had to takesecond place to concerns about ease of77
Technology ReviewThe ABB Fan Group has several laboratories for tests according to international and national standards such asISO, AMCA, NF, BS.manufacture and installation. Now that thefan market has become global and highlycompetitive, operating costs are beinggiven a higher priority. The large forceddraft fans used in thermal power plants,for example, are expected today to exhibithigh aerodynamic efficiency over a muchwider range of operation than they did afew years ago.A case in point, and one whichillustrates how well CFD simulations aresuited for predicting performance andstudying fan aerodynamics in detail, is thedouble-inlet, high specific speedcentrifugal fan with adjustable inlet guidevane assembly (for flow regulation) andscroll casing shown in 2 .Seen in this diagram are the inlet boxeswith adjustable inlet guide vanes (IGV),the inlet mouth and the double inletimpeller, which discharges into a singletongue scroll housing of constant width.78The adjustable IGVs match the fan’s flowcharacteristic to the load by throttling theflow yet still ensuring a good inlet flow tothe impeller. The pre-swirl imparted to theinlet flow at constant rotational speed byvarying the stagger of the vanessimultaneously reduces the fan head, sothat the working point of the machine canbe controlled and shifted along constantlines of resistance.The IGV assembly, which is fitteddiagonally, is mounted in the conicaloutlet of the inlet box. Typically, the IGVsfor industrial fans consist of flat sheetmetal blades which are turned throughlarge angles of up to 90 degrees. Thissimple, low-cost design ensures ease ofmanufacture, low losses at zero staggerposition and almost zero flow. However,the meridional and simultaneouscircumferential turning of the flow withinthe IGV section at off-design load resultsin a complex flow exhibiting significantthree-dimensional effects that make onedimensional tools rather unsuitable forchecking for pressure losses or foranalyzing the IGV and impeller interaction.The shrouded radial double-inletimpeller shown is a high-flow design withhighly backward leaning, aerofoil-typeblades. Under design operating conditionsthe flow coefficient is 0.307 and the headcoefficient is 1, corresponding to a specificspeed of 1.65. The figures for the Machnumber and the Reynolds number are0.14 and 0.6 x 106, respectively, in eachcase referred to a stagnant inlet condition,the impeller tip speed and the tip width.Computational methodologybased on a commerciallyavailable CFD codeCFD is a very powerful tool for predictingcomplex flow structures and provides aABB Review 2/2000
Industrial fans geta boost from R&DFans come in an enormous variety oftypes and sizes, with designs for everyconceivable use. In thermal processes,they may be used to transport gases,vapors and gas/solids mixtures for thepurpose of ventilation, cooling, drying,air-conditioning or combustion.Industrial sectors that depend heavilyon them include power plants, highwayconstruction (in tunnels), cementmaking, chemical/petrochemical plantsand mining. Smaller fans are used to2 Cut-away drawing of a forced draft fan, showing the inlet boxes with thecool electronic components and motors.adjustable guide vanes, the double inlet impeller and scroll housingThe competitive forces driving today’sglobal economy have resulted in manybetter understanding of the flowcharacteristics in turbomachinerycomponents. ABB Solyvent-Ventec andABB Corporate Research1 have made useof this fact by incorporating CFD in thedesign process for centrifugal fans. Theresults have shown that CFD is accurateenough to be used to calculate theperformance characteristics of these fans.A commercially available CFD code wasused to study the aerodynamics of thedescribed fan. Incompressible fluidproperties were assumed. The code,which solves the Reynolds averagedNavier-Stokes equations on structuredgrids, was applied to a coupled singleblade passage of the IGVs and the impelleras well as to the coupled system of the fullrotor and the scroll housing. The surfacegrid of the computational domain for thescroll housing is given in 3 , which showsone half of the impeller and scrollABB Review 2/2000housing. In order to show the bladedesign better, part of the impellercenterplate has been cut away.All steady-state flow calculations carriedout are based on turbulence modelingusing the standard ‘high-Reynoldsnumber’ K-ε model in conjunction with awall function approach for the closure ofthe Reynolds-averaged Navier-Stokesequations.In all cases the CFD models usedcaptured the effects of the recirculatingflow along the impeller shroud caused bythe clearance flow through the gapbetween the stationary inlet mouth andthe rotating impeller shroud.Generally, a mixing plane approachwas used at the interfaces between theof the newest technologies beingintroduced progressively, starting withthe aerospace industry and leading tostationary gas turbines for power plants,then to the industrial compressors andfans. As a leading fan supplier, ABB isduty-bound to keep abreast of all thelatest developments.Recent R&D work has yielded importantnew knowledge, which ABB uses,as this article shows, to improve theaerodynamic design of its fans. Anotherinvestment, in ‘environmentally safeengineering’, has produced fans that runon less power. These include axial fansfrom ABB Fläkt with power inputs of upto 2.5 MW and radial fans built by ABBSolyvent-Ventec, with wheels as large1 ABB Corporate Research activities in this fieldhave since been transferred to ABB ALSTOMas 5 meters in diameter and powerinputs of up to 10 MW.POWER TECHNOLOGY Ltd.79
Technology Reviewmethods gives the designer a reliable toolwith which to optimize turbomachinerycomponents.As is well known, the incidence withrespect to the impeller blading is animportant parameter in fan aerodynamics.For instance, if the blade loading or thelevel of positive incidence (stagnationpoint shifted towards the blade pressureside) is high, adverse pressure gradientsacting on the blade suction surfaceincrease the boundary layer growth, and3 Computational grid for the scroll housing and impeller. Such CFD modelsare used to capture the flow effects in the fan components.stationary and the rotating grids. At theinterface to the stationary grid of thescroll housing, however, a frozen rotorapproach was used as the circumferentialvariation of the flow could be largecompared with that across an impellerblade passage at the tip. (The mixingplane approach is based on conservativecircumferential averaging at the gridinterfaces; the frozen rotor approach givesframe changes across the interfacewithout relative changes in the gridpositions over time.) Mass flow andturbulence levels were imposed at theinflow boundaries of the computationaldomain.Calculations were performed for four80different operating points at constantrotational speed over a range of 80% to140% design flow. The zero staggerposition of the IGVs was used toinvestigate the design and high-flowoperating conditions. For the purpose ofcomparison, stagger angles of 60 degreesand 45 degrees measured from themeridional plane were used with 80% and90% design flow, respectively.Detailed flow analysisSome examples of actual CFD resultsobtained with a radial fan are used in thefollowing to show how CFD can be usedto investigate complex flow physics andhow the accuracy of modern CFDthereby the risk of boundary layerseparation. This can lead to blade stall,which is associated with high losses. Onthe other hand, it is known that the profileloss is insensitive to the incidence at lowinlet Mach numbers, this being particularlytrue for profiles with blunt leading edges.A uniform incidence distribution alongthe span can nevertheless contribute to asignificantly increased efficiency andpressure rise [1]. To check the bladedesign for the radial impeller considered,with its typically strong meridional flowturning upstream of the blade's leadingedge and a clearance jet impacting ontothe leading edge at the impeller shroud,details of the flow on the blade's leadingedge are needed.Design flow conditionsThe flow pattern close to the impellershroud and impeller exit under designconditions is shown by the velocitydistribution in 4 . Clearly seen is thetypical formation of the jet/wake patternin the downstream parts of the impeller,which results in an accumulation of lowenergy fluid in the near shroud/suctionABB Review 2/2000
4 Velocity distribution at design load.5 Vector flowfield at design load (impeller shroud). The CFD analysis shows thatThis figure shows the flow pattern closeeven under design conditions the boundary layer flow along the pressure side isto the impeller shroud and the impellerclose to separating - a condition known as ‘corner stall’.discharge.side area and high-velocity fluid in thecenterplate pressure-side region. Thegeneration of streamwise vorticity alongthe impeller passage can be explained bythe centrifugal forces prevailing in theaxial-to-radial bend region (these movelow-momentum fluid towards the shroudfrom the blade pressure and suctionsurfaces) and the Coriolis forces acting inthe radial part of the passage (forcing lowenergy fluid to migrate from thecenterplate and shroud walls onto theblade suction surface). However, from 5 ,which shows the corresponding vectorflowfield, it is seen that there is nobackward flow involved in this flowprocess. Nevertheless, the CFD analysisindicates that even under designconditions the boundary layer flow alongthe pressure side is close to corner stallABB Review 2/2000(flow separation), being caused by thenegative incidence onto the blade'sleading edge and the imposed diffusionfurther downstream. Details of thismechanism are discussed in [2].contrast to the situation with design flow,the low-speed region at a position on theblade pressure side at half chord lengthand the vector flowfield at the impellershroud together reveal a radially inwardflow in this part of the impeller 7 . ThePart-load flow conditionsAt 80% design flow operation with preswirl imparted to the inlet flow by 60degrees staggering of the IGVs, bladeloading is reduced and the wake flowpattern becomes less pronounced 6 . In6 Velocity distribution at part load withpre-swirl imparted to the inlet flow by60-degrees staggering of the inlet guidevanes. It can be seen that the wake flowpattern is less pronounced.81
Technology Reviewthe boundary layer and wraps itself aroundthe leading edge. The course of the stagnation streamline and the saddle point,which divides the separation streamlineon the pressure side and suction side leg7 Vector flowfield at part load (impeller shroud). A radially inward flow can be seenin this part of the impeller.reason for this was found to be atip/leading edge corner stall vortexdeveloping on the blade pressure surfaceand originating from leading edgeseparation at the blade tip due to a largenegative incidence. This vortex interactswith the shroud end wall flow and leads tothe mentioned backward flow along theshroud at the blade pressure surface. Adetailed study of the clearance flow, whichis injected from the fan-side cavity throughthe clearance between the stationary inletmouth and the rotating shroud into theinlet flow in this region, indicates that itcontributes to the negative incidence ontothe blade's leading edge at the impellershroud. Also, the clearance flow, whichleads to a circulating flow along theimpeller shroud, is another cause ofvolumetric loss.The flow at the leading edge close tothe impeller centerplate has also beenanalyzed in detail at this off-designworking point. In contrast to thepredicted strong negative incidence at theblade tip, the vector flowfield in 8 reveals82a large positive incidence onto the leadingedge, resulting in the formation of a leading edge vortex. The underlying reason isthat the incoming boundary layer flow onthe centerplate cannot withstand thepressure gradient in front of the bluntleading edge and starts to separateupstream of it. This process is associatedwith the formation of a vortex that rolls upof this leading edge vortex, is seen in 8 .The boundary layer, which re-developsbeyond the separation line, experiences astrong expansion around the blade'sleading edge. This results in separationfurther downstream in thecenterplate/suction-side corner, beingcaused by an adverse pressure gradient.Matching the impellerto the scroll casingTo predict the overall fan efficiency andreveal where the highest losses are locatedin the fan, it is also necessary to investigatethe flow in the scroll casing.In accordance with the impellerdischarge swirl that the scroll is designedfor, there is just one working point per8 Vector flowfield at part load (impeller centerplate). The large positiveincidence onto the leading edge results in the formation of a leading edgevortex.ABB Review 2/2000
9 Vector flowfield of the impeller and scroll casing at design10 Vector flowfield of the impeller and scroll casing at 140%load. The stagnation point is located on the leading edge of thedesign flow. Fluid is pushed back into the diffusing exit duct oftongue of the scroll casing.the casing, so that the stagnation point is located inside thescroll passage.constant-speed line (normally the designpoint) for which the scroll imposes analmost circumferentially uniform pressuredistribution at the impeller tip. During anexcursion from this operating point, aperipheral asymmetric pressure fieldoccurs in the scroll housing, causing a circumferential variation in the blade loadingand mass flow distributed to the differentblade passages and leading to strong nonsteady interaction between the impellerand scroll flow. The quasi-steady frozenrotor approach with the rotor positionfixed in relation to the scroll tongue isused for simplification and to reduce turnaround times for the computations.At the design flow rate the stagnationpoint of the streamline, which divides theflow entering the scroll casing from theflow being discharged, is located on theleading edge tip of the scroll tongue. Athigher flow rates the casing is too small toaccommodate the flow, which thenaccelerates circumferentially. The staticpressure at the first cross-section of theABB Review 2/2000scroll passage is consequently higher thanoutside, ie at the inlet to the diffuser. Thestagnation point is shifted inside the scrollcasing and the corresponding staticpressure gradient pushes fluid back intothe diffuser. Conversely, at low flow ratesthe flow is decelerated up to the diffuserexit, so that fluid is pushed inside thescroll casing under the tongue and thestagnation point moves along the casingwall towards the diffuser exit.Details of the flowfield at the tongue ofthe scroll casing at design load are given in9 . As stated above, the stagnation point islocated on the leading edge of the tonguedue to the circumferentially uniformpressure at the scroll inlet. The bladepassages approaching the scroll tongueare affected locally by a higher backpressure, resulting in a reduced throughflow that is associated with higher bladeloading and large positive incidenceangles. As already mentioned, at a highmass flow rate (140% design flow), fluid ispushed back into the diffusing exit duct ofthe casing. Thus, the stagnation point islocated inside the scroll passage. Theassociated backflow experiences strongexpansion around the leading edge of thetongue, followed by a massive flowseparation due to the adverse pressuregradient prevailing in this exit section ofthe casing 10 .Performance characteristic –measurements versus CFDpredictionABB Solyvent-Ventec conductedperformance measurements on a scaledtest rig and used the data to compare theCFD predictions with the globalmeasurements of total fan pressure riseand total efficiency. From 11 and 12it can be seen there is very goodagreement between the measured andpredicted values. The uncertainties inoverall total efficiency (neglecting thebearing losses) predicted for flows rangingfrom 80% to 140% design flow were foundto be smaller than 1.5%. The results show83
Technology Review2000100N 950 rpmN 950 rpm801500pF [Pa]η [%]F6040ExperimentCFD; IGV 0 degCFD; IGV 45 degCFD; IGV 60 deg201000ExperimentCFD; IGV 0 degCFD; IGV 45 degCFD; IGV 60 deg02000051015202505Q[m3/s]10152011 Performance map showing the good agreement between12 Fan pressure versus flow rate. Comparison ofmeasurements carried out on a scaled test rig and CFD predictionsmeasurements and CFD predictionsηF Fan efficiencypF Fan pressureQQFlow ratethat CFD is able to correctly predict thecomplex flow structure. In addition, thedetailed analysis of the local flow structuresprovides a better understanding of theperformance characteristic of a specific fan.At the same time, in-depth knowledge ofthe flow gives the designer a betteroverview of the performance characteristicof radial fans over their full operating range.Benefits for evaluatingdesign candidatesAs the results of the described investigationshow, CFD flowfield predictions permit avery detailed study of internal flowphenomena. This would otherwise requireexpensive and time-consumingexperimental work.25Q[m3/s]Flow rateWhat is certain is that such flow effectscannot be captured using empirical 1-D or2-D flow equations, although the use ofthese methods can still be justified for preoptimization in the early design phases.The CFD code should be looked attogether with other preliminary designtools as part of an integrated system foreva
CFD simulations based on 3-D Navier-Stokes equations in fan design. CFD brings a systematic approach to the design and development process, enabling new, optimum configurations to be found and experimental investigation to be kept to a minimum. What is more, CFD allows a more comprehensive u
refrigerator & freezer . service manual (cfd units) model: cfd-1rr . cfd-2rr . cfd-3rr . cfd-1ff . cfd-2ff . cfd-3ff . 1 table of contents
430 allocation to elianto cfd o&m 20,577.32 440 allocation to trillium west cfd o&m 27,267.00 450 allocation to west park cfd o&m 70,008.22 460 allocation to festival ranch cfd o&m 177,790.54 480 allocation to tartesso west cfd o&m 27,809.17 481 allocation to anthem sun valley cfd o&
A.2 Initial Interactive CFD Analysis Figure 2: Initial CFD. Our forward trained network provides a spatial CFD analysis prediction within a few seconds and is visualised in our CAD software. A.3 Thresholded and Modified CFD Analysis Figure 3: Threshold. The CFD is thresholded to localise on
performing CFD for the past 16 years and is familiar with most commercial CFD packages. Sean is the lead author for the tutorial and is responsible for the following sections: General Procedures for CFD Analyses Modeling Turbulence Example 3 - CFD Analysis
CFD Analysis Process 1. Formulate the Flow Problem 2. Model the Geometry 3. Model the Flow (Computational) Domain 4. Generate the Grid 5. Specify the Boundary Conditions 6. Specify the Initial Conditions 7. Set up the CFD Simulation 8. Conduct the CFD Simulation 9. Examine and Process the CFD Results 10. F
The CFD software used i s Fluent 5.5. Comparison between the predicted and simulated airflow rate is suggested as a validation method of the implemented CFD code, while the common practice is to compare CFD outputs to wind tunnel or full-scale . Both implemented CFD and Network models are briefly explained below. This followed by the .
Emphasis is on comparing CFD results, not comparison to experiment CFD Solvers: BCFD, CFD , GGNS Grids: JAXA (D), ANSA (E), VGRID (C) Turbulence Models: Spalart-Allmaras (SA), SA-QCR, SA-RC-QCR Principal results: Different CFD codes on same/similar meshes with same turbulence model generate similar results
Unit 39: Adventure Tourism 378 Unit 40: Special Interest Tourism 386 Unit 41: Tourist Resort Management 393 Unit 42: Cruise Management 401 Unit 43: International Tourism Planning and Policy 408 Unit 44: Organisational Behaviour 415 Unit 45: Sales Management 421 Unit 46: Pitching and Negotiation Skills 427 Unit 47: Strategic Human Resource Management 433 Unit 48: Launching a New Venture 440 .
