Innovation & Quality


We're pleased to present FINE™/Turbo and FINE™/Design3D 12.1.

This release comes with a documentation included in the software package, available in our customer area on www.numeca.com. FINE/Agile™, AutoBlade™, AutoGrid5™ and CFView™ are also available as stand-­alone.

 

Highlights

FINE™/Admin: New admin and licensing wizard

The new administration wizard allows our users to:

Configure the license server and/or client and activate licenses

  • Set User Configuration
  • Manage Network Configuration
  • Diagnose and fix problems

AutoGrid5™: Parallel mesh generation

AutoGrid5™ high-quality meshes can be now generated in parallel with shared memory (multithreading). Most CPU-expensive processes have been parallelized including block interpolation, 3D mesh generation, grid quality checks and block negative cells computation.

From now, our users can generate a 100M cells mesh in about 2 minutes on a standard workstation.

This new feature requires an add-on license. Contact your local sales representative for more information.

 

AutoGrid5™: Butterfly topology for B2B offset fillet creation method

The fillet geometry can be created using the B2B section offset method when the fillet topology is set to butterfly.

This method is a more robust fillet geometry creation when the specified fillet radius is equal or bigger than the leading/trailing edge radius.

 

FINE™/Turbo: Nonlinear Harmonic method for Conjugate Heat Transfer

The challenging task when modelling conjugate heat transfer for turbomachinery application is that it requires accurate simulation of unsteady flows in conjunction with heat transfer simulation of the airfoil solid structure, leading to large time-scale mismatch between fluid and solid domains which consequently requires large computational resources.
To overcome this problem, the nonlinear harmonic (NLH) method has been extended to conjugate heat transfer (CHT) simulation to provide an effective tool for turbomachinery thermal design and analysis.

 

 

FINE™/Turbo: GPU Acceleration (BETA)

With the new GPU acceleration option you obtain the same high-quality results, yet even faster!
FINE™/Turbo takes advantage of your Linux workstation or cluster node Graphical Processing Unit (GPU) to speed-up your simulations.

FINE™/Design3D: New optimization kernel and data-mining tools

Multidisciplinary Design Optimization has a considerable impact on the design by increasing performance, lowering life-cycle cost and shortening design time for complex products. With the integration of Minamo, FINE™/Design3D extends its capabilities with the most advanced genetic algorithms, powerful design of experiments techniques, efficient nonlinear surrogate models and comprehensive monitoring and data analysis tools.

Minamo module requires an additional license, provided upon request by your local sales representative.

FINE™/Design3D: Robust Design Optimization

Real applications are always subject to uncertainties, either being:

  • Uncertain operating conditions (inflow conditions, pressures, fuel mass flow…)
  • Geometrical shape variability (erosion, damage, fouling, thermal expansion…)
  • Manufacturing tolerances (milling, forging, assembly tolerances…)

When performing a deterministic optimization, the input uncertainties are not considered. The resulting optimum may be highly sensitive to the input uncertainties. As a result, this optimum performances present a wide Probability Density Function. This means that the operational performances may be much lower than expected.

On the contrary, the robust design optimization formulation does not only seek to minimize the mean value of the quantity but also to minimize its variability with respect to the input uncertainties. As a consequence, it allows a better control on:

  • Component failure probability
  • Reliability of performance
  • Performance over product lifetime

Example: Robust Design Optimization of the Rotor 37:

Optimum design performances are compared to the deterministic optimum and the initial design. Two of the objectives relates to maximizing mean efficiency while minimizing its standard deviation (eg. reach to the bottom right of the figure).

The blue line indicates the efficiency of the deterministic computation whereas the yellow star abscissa indicates the mean efficiency taking into account the input uncertainties, obtained with a UQ evaluation of the initial design. One can see that a deterministic simulation can overestimate the design performance by neglecting uncertainties.

On the left: original design, on the right: robust optimum


 

Fixes and improvements

AutoGrid5™

  • AutoGrid4 is removed from package
  • Distortion of the mesh on the nozzle
  • Far field configuration wizards merging
  • Batch information about 3D generation status
  • Saving standard upgraded to CGNS v3.3 ADF & HFD5

FINE™/Turbo

  • Usability improvements of expert parameters settings
  • Convergence Steering: group name selection via drop-down list
  • Residual output to Monitor per row
  • Nonreflecting Inlet/Outlet BC for RANS Turbo applications
  • Python: Task manager & Convergence steering
  • NLH: usability improvements
  • FSI: Fluid-Structure Interactions compatible with HPC partitioning
  • FSI: local outputs from modal simulation
  • FSI: modal approach with complex modes

 

Get ready to upgrade

Before you upgrade, check out the Release Notes for important info about this release and see the full list of issues resolved.

See also

NUMECA customer area: http://portal.numeca.be/products

NUMECA online documentation: http://portal.numeca.be/docs/Default.htm

For more information, please contact sales@numeca.be

 

Share this article

Author

Yannick Baux

Yannick graduated as Energy and Environmental engineer from the French engineer school INSA Lyon. After working for six years as a CFD consultant and part-time lecturer, he joined NUMECA in 2016 as turbomachinery products and applications manager.

Comments

Leave first comment here...

LEAVE A COMMENT