Sie sind auf Seite 1von 19

Release Note

FINETM/Turbo Package v6.2-9


including: IGGTM v4.9-2 AutoGrid v4.9-2 FINETM GUI v6.2-9 Euranus v5.1-9 CFViewTM v4.2-7

-May 2005-

NUMERICAL MECHANICS APPLICATIONS

Release Note
FINETM/Turbo Package v6.2-9
documentation version 6.2-d

NUMECA International 5, Avenue Franklin Roosevelt 1050 Brussels Belgium Tel: +32 2 647.83.11 Fax: +32 2 647.93.98 Web : http://www.numeca.com

NUMERICAL MECHANICS APPLICATIONS

NUMERICAL MECHANICS APPLICATIONS

Contents

IGG (v. 4.9-2) 1 Introduction 2 Corrected Problems AutoGrid (v. 4.9-2) 1 Introduction 2 Corrected Problems FINE GUI (v. 6.2-9) & EURANUS (v. 5.1-9) 1 Introduction 2 Improvements Thermodynamic Tables for Water (Steam) 4 Corrected Problems CFView (v. 4.2-7) 1 Introduction 2 Corrected Problems

1 1 1 3 3 3 6 6 7 7 7 12 12 12

Contents

ii

IGG (v. 4.9-2)

IGG (v. 4.9-2)

Introduction

No new feature or improvement has been integrated in IGG v4.9-2, but fix #4259 has been introduced. Details are listed in the next section.

Corrected Problems

IGG v4.9 includes a number of fixes from IGG v4.7-5. The update is both based on customer feedback and continued NUMECA internal testing. The information is summarized in the following table. The first column indicates the ID associated to the bug/suggestion (NUMECA internal reference). The second column indicates the version number in which the bug is solved. The third column provides a description of the bug/ suggestion. ID 1662 2064 2300 2716 2893 3865 4259 Solved in 4.9-1 4.9-1 4.9-1 4.9-1 4.9-1 4.9-1 4.9-2 Description error when printing as postscript or bitmap postscript search connection not scripted in 2D projects deletion of butterfly mesh can lead IGG to crash connections may not be found in some circumstances in 2D projects NMB connections modified while using /split block option Problem to map edge on a curve by using additional vertices Python command added for /grid/create block/smoothing capability

Few additional bugs and suggestions have been outlined and partly fixed but, in accordance with NUMECA quality procedures for releases, have not been integrated in IGG v4.9-2 to prevent any instability. They include:

Rotate all blocks with duplicate geometry option does not enable any further deletion of the curves (#598) Management of extra-characters (accents,...) in naming of blocks is not supported nor warned. As a result,
it can possibly lead to a crash of EURANUS solver (#857)

Release Note

IGG (v. 4.9-2)

Loading a 2D project in IGG from FINE GUI does not display curves. Workaround is to use Show
All option (#1668)

Use of Transform Block option from a group of blocks can lead to inadequate rotation of few edges.
Workaround is to make the rotation using the Topology from Grid option (#1875)

Shortcut <Alt-S> and menu /Geometry/View/Surfaces act differently while they should not (#2290) Name of geometrical object is not properly stored in file when /save is selected (#3278) Task Manager hangs in when running IGG (#3444) Batch problems because of display (#3642)
Degenerated surface found in search for NMB connection (#3968)

Face grid projection on a surface that has been deleted (#4072)

Release Note

AutoGrid (v. 4.9-2)

AutoGrid (v. 4.9-2)

Introduction

No new feature, improvement or fix has been integrated in AutoGrid v4.9-2. Details of fixes included from AutoGrid v4.7-5 are listed in the next section.

Corrected Problems

AutoGrid v4.9 includes a collection of fixes and suggestions from AutoGrid v4.7-5. The updates are both based on customer feedback and continued NUMECA internal testing. The information is summarized in the following table. The first column indicates the ID associated to the bug/suggestion (NUMECA internal reference). The second column indicates the version number in which the bug is solved. The third column provides a description of the bug/suggestion. ID 1357 1478 1937 3233 3661 Solved in 4.9-1 4.9-1 4.9-1 4.9-1 4.9-1 Description number of blades is not adequately stored in the ".trb" file hardly difficult to move leading edge point in meridional and blade-to-blade views. Problem appears when the user loads a new geometry in a new template. impossible to treat blade without blunt as if a blunt was present in order to enhance mesh quality save template including a leakage can lead AutoGrid to crash AutoGrid crash on a customer geometry when asking for a new template without splitter

Few additional bugs and suggestions have been outlined and partly fixed but, in accordance with NUMECA quality procedures for releases, have not been integrated in AutoGrid v4.9-2 to prevent any instability. They include:

Hardly difficult to "define a corner" on hub or shroud (PC-dependent). Workaround is to set the meridional view in full screen mode (#821)

Tip gap definition files cannot extend from inlet to outlet but well from the leading to trailing edge (#972)

Release Note

AutoGrid (v. 4.9-2)

AutoGrid may not split the patches adequately when hub leakage is defined. Workaround is to reserve
geometry orientation, mesh and subsequently reverse inlet / outlet patches at the end of the meshing process (#1100)

HOH topology may be inadequate in combination with high pitch and stagger angles. Workaround is to
use H/I topology (#1129)

GUI can be frozen when spanwise direction is defined from shroud to hub (PC-dependent) (#1304) Impossible to select a curve from the 3D view. Workaround is to select curve in the meridional view and
then define from selection (#1324)

Seal Leakage - Change in number of grid points (streamwise direction) in set up box cannot be seen in the
GUI. Workaround is to deactivate and further reactivate mesh creation (#1336)

Shortcut <Ctrl-a> is not saving the grid (as expected) but well removing the Quick Access Pad (#1345) Blade-to-Blade Mesh Control/Hub-to-Shroud parameter (default value = 0.5 at inlet/outlet) has no impact
on the final mesh (#1378)

Running HOH mesh can create discontinuities in the mesh at inlet. Workaround is to move the leading
edge (#1451)

In some circumstances, AutoGrid can define PERNM boundary condition as PER boundary condition
(#1543)

Blade to blade mesh is created straightforwardly once the streamline control is ensured. This prevents the
user to act on the extension shape controls, if needed. Workaround is to act after blade-to-blade mesh generation. No workaround is known is system crash during the blade-to-blade mesh generation (#1665)

Cell length is not updated in the M-Theta grid control box. Workaround is to close the M-Theta grid control box, modify the cell length and click on the blade-to-blade mesh icon (#1672)

In some circumstances, PER connections may not be well defined. Workaround is to search the periodic
connection in IGGTM (#1762)

Curve names stored in the mesh can be different between WIN32 and LINUX platforms (#1972) In some circumstances, the clustering may not be properly applied when the template is re-run. Workaround is to go back to the streamlines control and press apply (#2029)

Some patches may be undefined due to the location of the leading edge (#2040) First cell size selected by user may not be conserved when multiblock smoothing is applied (#2121) Hub leakage may lead AutoGrid to crash (#2311) Cannot move RZ corner at inlet. Workaround is to select default orientation first (#2313) Reset distribution at inlet/outlet may fail with large hub clearance. Workaround is to create an external hub clearance curve (#2436) dependent) (#2522)

In some circumstances, use of meridional control with H/I topology can lead AutoGrid to crash (PC
AutoGrid does not take blanks into account in pathname (#2831) AutoGrid crash when generating blade-to-blade mesh (#2954) When "Tip Definition" checked in ".trb" file, not tip curve is defined in ".geomTurbo" file (#3472) AutoGrid crash when hub section of the blade is defined right on the hub surface. Workaround is to extend blade in the hub (#3565)

AutoGrid crash when geometry is presenting a splitter blade and a bulb at outlet. Workaround is to change
blade, hub and shroud streamwise orientation to have bulb at inlet (#3663) Coordinate problem between ".dat" imposed and resulting 3D mesh (#3769)

Negative cells within the LINUX mesh while not present within Windows (#3851) Display problem when loading a second ".geomTurbo" file (#3876)

Release Note

AutoGrid (v. 4.9-2)

Mesh discontinuities when using leading and trailing edge fitting (#4022)

Release Note

FINE GUI (v. 6.2-9) & EURANUS (v. 5.1-9)

FINE GUI (v. 6.2-9) & EURANUS (v. 5.1-9)

Introduction

This note summarizes the main improvements introduced in FINE GUI v6.2-9 and EURANUS v5.1-9. Improvement include:

Update of thermodynamic tables for water (steam), with enhanced robustness and accuracy, especially at
low pressures (locally < 1000 Pa) (also referred to as #3899).

Re-introduction of MPICH library 1.2-4 for smoother management of semaphores running multi-processors calculations (restricted to LINUX platforms) (also referred to as #4002).

FINE GUI v6.2-9 and EURANUS v5.1-9 also fixes some known reported problems that were present in previous versions. Details are listed in the last section.

Release Note

FINE GUI (v. 6.2-9) & EURANUS (v. 5.1-9)

2
2-1

Improvements
Thermodynamic Tables for Water (Steam)

This section briefly describes the new features that have been introduced in FINE GUI v6.2-9.

The thermodynamic module aims at the modelling of the real thermodynamic properties of a given fluid by means of interpolation of the variables from dedicated tables. The approach that has been adopted in EURANUS consists of using a series of thermodynamic tables, one table being required each time a thermodynamic variable must be deduced from two other ones. This implies the creation of many tables as input, but presents the advantage that no iterative inversion of the tables is done in the solver, with as a consequence a very small additional CPU time. An update of thermodynamic tables for water (steam) has been included in FINE GUI v6.2-9, with the objective to enhance robustness and accuracy of corresponding calculations, especially at low pressures (< 1000 Pa locally). The range of variation of admissible values for thermodynamic variables has been extended and both the discretization of the tables and the nature of the interpolation algorithms have been revisited. Relative mean and maximum errors are now attached to the set of water (steam) thermodynamic tables. More details can be found in the Appendix D of FINE v6.2-9 User Manual.

Corrected Problems

FINE v6.2-9 and EURANUS v5.1-9 also include a collection of fixes and suggestions from FINE v6.1 and EURANUS v4.8. The updates are both based on customer feedback and continued NUMECA internal testing. The information is summarized in the following table. The first column indicates the ID associated to the bug/suggestion (NUMECA internal reference). The second column indicates the version number in which the bug is solved. The third column provides a description of the bug/suggestion. For FINE v6.2: ID 852 1983 1984 2057 2288 2312 2507 2721 3448 3509 3566 3580 1763 Solved in 6.2-1 6.2-1 6.2-1 6.2-1 6.2-1 6.2-1 6.2-1 6.2-1 6.2-2 6.2-2 6.2-2 6.2-2 6.2-3 Description British units are not updated in the steering cannot select initial solution file in /design2D page cannot kill computation in /design2D mode location of slots in /cooling page depends on grid level coordinates in /cooling page and mesh do not have same units merging not correct when periodicity is different in FINETM GUI and IGGTM in some circumstances, IRSCHM parameter can be set to 3 while should be 2 remote computation using rcp drives to /tmp, invalid on WIN32 platforms (PCdependent) add expert parameter NSTAGE prescribed total temperature is defined in relative mode for Cooling Flow wrong rotation speed saved in specific cases automatic modification of the sweeps not working properly appearance of FNMB patches in boundary conditions page

Release Note

FINE GUI (v. 6.2-9) & EURANUS (v. 5.1-9)

ID 3596 2627 3682 3704 3729 3733 3745 3758 3765 3853 3861 3875 4162 3946 4021 4112

Solved in 6.2-3 6.2-7 6.2-7 6.2-7 6.2-7 6.2-7 6.2-7 6.2-7 6.2-7 6.2-7 6.2-7 6.2-7 6.2-7 6.2-8 6.2-8 6.2-8

Description adapt Reynolds definition based on Sutherland Law no warning when no file specified as initial solution from file import cooling-holes file in FINE GUI is not working no blade parameters in transition mode when merging mesh impossible to define two identical injectors on two sides of a blade problem when renaming injectors slot definition problem when defined on several patches cooling/bleed display problem when merging a new mesh within the project in some circumstances, removing computation does not remove related directory in some circumstances, rotor/stator ID is not appearing in initial solution page problem with IDOMSC value when using domain scaling approach computation steering reset automatically when restarting computation Task Manager in parallel crashes when suspending the task Initial solution block selection (right-click in block list) not working properly renaming injector may modify patch selection (cooling/bleed module) /save as command does not update patch and name in *.mf and *.me.cfv files

For EURANUS v5.1: ID 1896 1939 1973 2028 2146 2211 2285 2315 2319 2618 2868 2989 3008 3030 3095 3242 3591 3646 706 1421 Solved in 5.1-1 5.1-1 5.1-1 5.1-1 5.1-1 5.1-1 5.1-1 5.1-1 5.1-1 5.1-1 5.1-1 5.1-1 5.1-1 5.1-1 5.1-1 5.1-2 5.1-2 5.1-3 5.1-6 5.1-6 Description calculation with low Reynolds number turbulence model may lead to crash when finest mesh level is reached in some circumstances, EURANUS may crash in running multi-processor calculations, while serial runs are satisfactory cannot allocate more than 1 Gb on IBM (IBM-dependent) inlet mass flow boundary condition with absolute velocity directions assumes surface of revolution at inlet heat transfer condition on a solid wall is not respected difference in numbering the rotor-stator interfaces between FINE GUI and EURANUS total temperature condition at inlet may not be respected in /condensable gas module introduction of slots (/cooling page) at shroud may lead EURANUS to crash cannot read thermodynamic tables for steam (PC-dependent) throughflow initial solution may fail on SGI (SGI-dependent) large increase in turbulent viscosity over a rotor-stator interface with Spalart-Allmaras turbulence model wall distance may be different when running serial/multi-processors EURANUS may hang while writing the solution in meridional view wall distance may be incorrect nearby hybrid FNMB (solid/connection) incorrect units for length in /blade-to-blade module EURANUS crashes when COOLFL=2 with validation cases vtf123 and vtf124 azimuthal averaging algorithm failure on a confidential customer case k-epsilon restart problem (BUG#3647) non-zero heat fluxes on adiabatic walls heat flux deviation on a solid wall near NMB connection

Release Note

FINE GUI (v. 6.2-9) & EURANUS (v. 5.1-9)

ID 3092 3573 3616 3687 3688 3689 3718 3731 3732 3734 3857 3887 3905 3928 4003 4002 4071 4086 4097 3899 4216 4230 4136 4261

Solved in 5.1-6 5.1-6 5.1-6 5.1-6 5.1-6 5.1-6 5.1-6 5.1-6 5.1-6 5.1-6 5.1-6 5.1-6 5.1-6 5.1-6 5.1-7 5.1-8 5.1-8 5.1-8 5.1-8 5.1-9 5.1-9 5.1-9 5.1-9 5.1-9

Description in some circumstances, restart problem on LINUXALPHA and ALPHA platform in some circumstances, gas condensable restart leads to negative density values set to zero in ".mf" when restarting with "0" iterations no solid data "heat flux" on solid blocks when using CHT module crash of transition module under some circumstances computation starts but gets stuck in initialization without obvious reason lagrangian not working in multiblock axisymmetric project Excessive integer memory requirement for R/S case with high number of blocks in parallel impossible to save outputs when computing a phase-lagged multiple rows project ICYOUT not working properly with Phase-Lagged method impossible to read relative total enthalpy when using gas condensable computation with more than "500" blocks starts but gets stuck in initialization warning message in ".std" file when saving meridional average different in parallel compare to sequential empty field in meridional output of a multistage applications run in parallel (restricted to Itanium processors) use of pressure adaptation boundary condition may not reach the desired mass flow in some circumstances semaphores are not clean while multi-processor run is stopped properly mismatch between number of passages and periodicity in domain scaling cannot switch from unsteady phase-lagged to steady calculation using /duplicate project error proportional to fluid density in running unsteady calculations using the k- turbulence model. limitation in water (steam) thermodynamic tables leads solver to blow up y+ distribution at walls can be discontinuous in some circumstances. Related to incorrect treatment of artificial dissipation fluxes along I=1 and I=Imax faces backflow treatment (BCKLFO) not compatible with cylindrical mode cannot start calculation of transitional distances in multi-processors runs wake induced transition model not compatible with multi-processors runs

Few additional bugs and suggestions have been outlined and partly fixed but, in accordance with NUMECA quality procedures for releases, have not been integrated in FINE v6.2-7 to prevent any instability. They include:

Host definition may be modified by EURANUS (#603) Keeping the GUI opened under WIN 2000 is CPU consuming. Work around is to increase priority of
EURANUS task in Windows Task Manager (#873)

Acces to online help from the GUI is prohibited (#880) Cannot create run files with extra characters ("()", "+"," ",...) in project name (#1039, #1080) Host connection from PC to SOLARIS is not working properly (#1212) User is not warned when inputs in the steering are not adequately bounded (#1344) Quantity static pressure at constant IJK surface is not properly reported in the steering (#1372)

Release Note

FINE GUI (v. 6.2-9) & EURANUS (v. 5.1-9)

Boundary condition - when used with mass flow at inlets, velocity direction is automatically set as a modulus, although user is free to force a positive or negative direction (#1379)

Inlet boundary conditions for turbulent quantities may be shown while not appropriate (#1419) Convergence history curve may not be continuous after restart (#1789, #3535) Adapt the way to impose thermodynamic properties of real gas (#3625) Random behaviour (spurious jumps) of convergence history plotted in the Steering (#3748) Automatic definition of meriodional angles in boundary conditions (#3783) Wrong block repetition vector in ".run" file after merging new mesh within project (#3849) Not possible to define a slot on multiple patches with grid coordinates (#3910) GUI can crash when creating ANSYS oyutput files for centrifugal compressor applications (#4070)

Few additional bugs and suggestions have been outlined and partly fixed but, in accordance with NUMECA quality procedures for releases, have not been integrated in EURANUS v5.1-9 to prevent any instability. They include:

Not possible to allocate more than 1 Gb on PC WinNT (#515) Not possible to allocate more than 1.4 Gb on HP-UX11 64 bits (HP L3000, 4 Gb memory) (#721) User is not warned when hard disk is full (#813)
Peak memory can be high when pre-processing of FNMB data on large scale grids (several 106) (#1032) FNMB connections cannot be defined when crossing an AutoGrid O-block (#1117) A patch cannot be referenced in more than one FNMB connection (#1117) Performances of parallel solver are largely degraded on RedHat 8. Workaround is to go back to RedHat 7.2 (#1200) Plot3D format is not compatible with TECPLOT v8 (#1330) Automated RAM estimation for EURANUS calculations is only valid in sequential mode RAM is indicated in Mb in GUI while it is stored in words in .log file (#1491) FNMB may be successful in IGG, but not in EURANUS (#2106) In some circumstances, free stream quantities may impact on the flow solution (#3048) Automated RAM estimation for EURANUS calculations may fail when meridional outputs are included (#3368) Intuitivity in use of phase-lagged module must be improved (#3435) NPLOUT parameter does not work with rotor-stator interfaces (#3499) Transition model takes unacceptable time when computing wall distance (#3610) Runtime error when continuing a Windows 2000 computation under LINUX (#3639) Problem of writing ".p4pg" files on Windows (#3645) Several contributions to one cell in the Cooling Flow module may lead to lack of robustness (#3680) Cavitation cannot start when ICAV=0 (#3696) Mixing Plane or Frozen Rotor Approach not working in some circumstances (#3883) Use of radial equilibrium with a mesh generated within AutoGrid 5 is erroneous (#3948) Impossible to add machine with XP service pack 2 (#3987) - workaround: uninstall service pack 2 ZR effects within AutoGrid 5 generate invalid FNMB connections (#4012) Increasing memory consumption of nitaskmgr.exe (#4052) Backward compatibility criteria not ensured for LH2 with seal calculations (#4104)

10

Release Note

FINE GUI (v. 6.2-9) & EURANUS (v. 5.1-9)

Backflow treatment (parameter BCKFLO) not compatible with use of cavitation module (#4277) Cannot restart in the same computation when running multi-processors calculations (#4284)

11

Release Note

CFView (v. 4.2-7)

CFView (v. 4.2-7)

Introduction

No new feature, improvement or fix have been introduced in CFViewTM. Details of fixes included from CFViewTM v4.2 are listed in the next section.

Corrected Problems

CFView v4.2 also includes a collection of fixes and suggestions from CFView v4.2. The updates are both based on customer feedback and continued NUMECA internal testing. The information is summarized in the following table. The first column indicates the ID associated to the bug/suggestion (NUMECA internal reference). The second column indicates the version number in which the bug is solved. The third column provides a description of the bug/suggestion. ID 1778 2616 2854 3079 3292 3376 1753 2932 2888 3426 3581 3670 Solved in 4.2-1 4.2-1 4.2-1 4.2-1 4.2-1 4.2-1 4.2-2 4.2-2 4.2-7 4.2-7 4.2-7 4.2-7 Description inconsistency between Vxyz_Z and Vz solution fields starting streamlines animation without streamlines leads CFView to crash batch mode cannot manage to post-process time-accurate projects (pop up) cannot import PLOT3D files with IBLANKING viewPlaneX() and plotPlaneX() macro commands are inverted vector on grid nodes button in QAP disabled but corresponding menu enabled CFView GUI frozen when performing specific steps in Animation/Cutting Plane Scrolling menu problem to compute the force on a solid object impossible to edit hub/shroud to recompute blade-to-blade view cannot edit title text of colormap (#3787) crash when changing the X-coordinate of the Cartesian plot resulting from a "plot data" (surface averaged output from FINE GUI) cutting plane polygon remaining after closing selection window

12

Release Note

CFView (v. 4.2-7)

ID 3674 3739 3742 3916

Solved in 4.2-7 4.2-7 4.2-7 4.2-7

Description "OpenCGNSProject" command not working properly solid data not available on the connections with solid blocks new quantity not correctly recorded in template Mesh within AutoGrid5 including domains meshed on 360 degrees leads to a failure of the pitch average process

Few additional bugs and suggestions have been outlined and partly fixed but, in accordance with NUMECA quality procedures for releases, have not been integrated in CFView v4.2-7 to prevent any instability. They include:

Polygon mode cut planes does not extend over the whole geometry (#903) Title may be cut in the banner in some circumstances (#952) Not possible to consider part of FNMB connections as solid patches. Work around is to define a new
quantity factor as exp ( -wall distance / K) with K being a big value. Since wall distances are set to zero on FNMBs, the factor will go to zero while it will tend to unity on solid surfaces (#1107)

Short cuts may not be consistent between FINE GUI and CFView (#1136) Animations cannot be stored in MPEG files (#1141) Can be difficult to see negative sign when using local values (#1445) No text appears in the banner and NUMECA logo is removed when printing a postscript picture with banner (PC dependent) (#1943) Text type editor does not account for the selected preferences (#1945) In some circumstances, relative velocity vector field in blade-to-blade view may not be correct (#2261) Cannot create cartesian plot /along solid boundaries from blade-to-blade cut (#2720) Cannot create postscript file with a banner (2825) Macro /load validation data does not run due to pop-up (PC-dependent) (#2828) Menu /Template and Menu/Macro do not apply to the turbomachinery module (#2897 - 3819 - 3820) Cannot convert a vector line to a curve (#3071) Flat color contours may lead to black screen (#3209) Macro /updatecurvetype does not bring the expected results (PC-dependent) (#3294) Error when running macro involving several views (#3297) Cannot open pitchwise averaged view when blank space is inserted (#3432) Batch problem due to the display (#3643) Surface orientation problem in project with AutoGrid5 mesh (#3736) Print Output is different on LINUX and Windows (#3759) Cross stream surface is not computed correct in a radial machine (#3821) Vectors are not shown correctly in blade to blade view (#3822) In meridional view, vector lines are stopped on connection (#3949) crash when opening customer project on SOLARIS (#3959) Particles do not cross a periodic boundary (#3964) "tmpfile-###" not removed in .numeca directory (#4069) No display in CFview when launched from FINE GUI on IBM (#4094) In some circumstances, CFview may required user intervention in running templates (#4190)

13

Release Note

CFView (v. 4.2-7)

CFview may not start running OS SuSe 9.1. Work-around is to modify the environment variable
LD_ASSUME_KERNEL = 2.4.20 (or 2.4.21) (#4247) Streamlines may have discontinuous behaviour (polylines) in tip gaps (#4267)

Running B2B view, hub projection may hang is some circumstances (#4298)

14

Release Note

Das könnte Ihnen auch gefallen