Hello Everyone,

We are looking into getting Numeca's tools to augment what we already have for turbomachinery. If any of you have experience pro/con with them I would appreciate any information that you would be willing to share.

Thank you

What do you currently use? CFX-Tascflow is also a pretty good option otherwise...

We chose FINE/Turbo 4.1 ~2.5 years ago over TaskFlow and Fluent primarily because of the relative ease (minimum number of key strokes per iteration) with which we could modify and optimize our radial-flow and mixed-flow turbines and compressors. We also found Numeca to be ~40% faster than CFX and three times faster than Fluent. We also did some experimental validation studies for micro nozzles and found better experimental agreement with Numeca, though the others were also quite close. After a year, we let our Numeca license lapse primarily because (1) we found the early releases of FINE/Turbo 5.2 to be rather buggy, (2) we needed combustion simulation capability, and (3) we needed better 3D blade geometry creation capability. We are now going back to Numeca FINE/Turbo 5.3, largely because of the advantages we mentioned above and because we are very optimistic about the usefulness of FINE/Design3D. Feel free to check back with us in a month, and we'll let you know how we feel after a month of working with FINE/Design3D.

I am a MIT graduate and I have been using Fine Turbo for a year now on compressor design. It is very user friendly and is pretty fast with the multigrid solver. The NUMECA team has also always been available to help me.

Jerome

We have used FINE/Turbo for about 4 years now, for use in a wide variety of single-phase fluid flows. We chose it based on the user-friendliness and the ability to interface the Concepts' 3-D blade design software (CCAD). In 1998 FINE was already available for a WindowsNT platform, which has proved to be a substantial benefit due to the low cost of hardware upgrades (which means more frequent upgrades). We have found very good agreement between the CFD results and test results. However, it is really Numeca's performance that allows us to get the most from the software. Over the last four years the grid generator, solver interface and the solver itself have seen very positive development, and Numeca has provided a high level of support. I would highly recommend Fine/Turbo.

I am working with Fine/turbo for three years now. After testing fluent I switched over to NUMECA's software which has been found to meet most of my requirements. The procedure of building up a CFD-model, solving it and postprocessing the results has been implemented in very comfortable and logical way. Concerning computation time, stability and accuracy I made good experiences. Especially, features which came up with fine 5.2 and 5.3 like the "full non matching boundary condition" makes NUMECA extreme valuable for the user. However, the grid generator IGG/Autogrid does not always build up good meshes. Especially if meshing small details IGG will run into accuray problems. Software Bugs have been reduced in version 5.3 in a high degree. Also I made good experiences with the support. They really take care of your problems.

Thanks for all of your comments. Wish there was a Numeca forum.

Are many others here also interested in a NUMECA forum? If you are please drop me a short email or respond to this message.

I haven't opened a NUMCEA forum because I did not think that there were enough users to support it. However, judging from the response to this post many visitors now seem to use NUMECA's software.

In my limited experience with Numeca software, I can provide some of my experiences. For myself, this was the first time that I had used any CFD software. Thus, initially the software took some getting use to. However, after some of the difficulties were worked out, I was quite impressed at the speed with which one could generate various geometries. I found, however, that at times, modifications were easier to implement in the IGG platform as opposed to the Autogrid window. However, for time and user friendliness, I am very impressed and would recommend the software to interested buyers. One other advantage that should be noted, is the level of detail that the user has over grid quality control. The fact that one could move one single node and regulate cell sizes in such a fine manner is quite impressive. I personally enjoy using the software.

L.A.

At our Institute of Jet Propulsion at the University of the Armed Forces Munich we have 4 or 5 years of experience with NUMECA and cooperated with NUMECA Germany in several projects. The CFD results were in a pretty good agreement with the measurements. I want to emphasise that the support of NUMECA Germany is of outstanding quality.

What type of calculations are you running? Steady state, transient, rotor/stator interaction ...?

What type of turbomachine can be simulated ?

As a key player in the field of high pressure turbocharging we have chosen the NUMECA Fine/Turbo software suite for its inherent capability of simulating transonic turbomachinery flows with excellently adapted numerical approaches. We highly appreciate the availability of both centered-in-space discretization schemes with sophisticated artificial dissipation as well as higher order upwind schemes involving Roe's flux difference splitting, where the latter is well-known for its best-in-class shock and shear layer resolution capabilities. As regards accuracy, robustness, and computational efficiency, Fine/Turbo exhibits excellent capabilities for our applications, such as full transonic centrifugal compressors and turbines including volutes as well as axial flow turbines.

A few remarks to pre and postprocessing: 1. Turbomachinery cascade gridding is highly convenient by virtue of the IGG/Autogrid module 2. Complex geometries, e.g. volutes or ancillaries are gridded by the IGG module, which exhibits numerous convenient features to set up complex multi block grids involving full non matching boundaries 3. Post processing is greatly enhanced by an easy to learn macro language, which includes all required options

Finally, we would like to emphasize excellent and fast support by NUMECA representatives in Belgium as well as in Germany.

Hannes Benetschik MAN B&W Diesel AG

I am an aerodynamics design engineer in the Turbine Aerodynamics group at Honeywell ES&S. Here at Honeywell, we have access to many CFD codes, both commercial and government owned, and I have had had the opportunity to work with most of them. Based on my past experiences with all these CFD tools, I feel fortunate that we have actually chosen to adopt Fine/Turbo as our standard aero design and analysis tool in the turbine aero group.

We decided to adopt Fine/Turbo as our standard CFD tool after a systematic evaluation of various commercial and non-commercial CFD tools. We rated several codes based on the ease of model generation, mesh quality, robustness and speed of the solver, and postprocessing capabilities. After also comparing the results from various codes to our rig test data, we have decided to adopt Fine/Turbo over the others.

Although very diverse and complex, most of our work is in steady, multiblade row analysis of axial and radial turbines. Numeca's automated mesher, AutoGrid is an excellent tool for generating 3D mesh around axial/radial airfoils and splitter/tandem vanes. One can generate smoothed H-type and O type grids very quickly using templates. The multigrid accelerator allows for rapid convergence. For big tasks, I found the parallel capability to be also very useful. The postprocessing capability is very user friendly, visual and also quantitative. One can easily select surfaces or create cutting planes and plot velocity vectors or contours of various quantities on these planes. Various macros, allow us to generate mass averaged quantities, both 1D and radial.

But in my mind the biggest winner is Numeca's general purpose mesher, IGG. Thanks to that, I am actually using Fine/Turbo not only for multi-stage turbine analysis but for all my CFD needs. The beauty of IGG is in its ability to generate mesh blocks that attract themselves to the geometric surfaces. Usually, the geometry is developed by a designer and passed on to us as an IGES file and it is not always the best quality. I know that with many CFD tools one has to spend too much time trying to clean up this geometry, because their meshers actually mesh the geometry using the surfaces, curves, and points from the IGES model. Since with IGG you are actually meshing the blocks that are attracted/wrapped around/and projected onto the geometry, you don't need to spend time with geometry cleanup. I find this extremely useful. For example if there is a gap in the geometry, to use geometry based meshers you would need to close that gap. With IGG, you generate a block right across the gap, and the mesh is still generated. A lot of times, these geometric details do not even need to be resolved. Why waste time? Just cross over them with the edge of a block.

My first experience with IGG was to analyze a 3D turbine scroll with a radial nozzle assembly just downstream of it. I had had no formal training with IGG before (nor later). Despite that, I was able to take an IGES file of the scroll, generate blocks of orthogonal butterfly mesh inside the scroll with IGG, mesh the radial nozzle using AutoGrid quickly, and merge the two mesh blocks of the scroll and the nozzle with Fully-Non-Matching grids. The convergence rate was great. I actually evaluated three different scroll designs and finished the design well ahead of time. With the butterfly mesh, the mesh quality was so orthogonal and refined near the solid surfaces that I actually dared to calculate the heat transfer coefficients to support the cooling design.

The support engineers are also very responsive and technically strong. They are easily accessible and always return my calls to help with any issues/questions I may have.

Happy Holidays and a New Year to you all. It is too bad they shut down the machines around this time of the year for maintenance. We have to stop running those iterations!

Dear Jonas,

On behalf of NUMECA I would like to thank you for this proposal. We also think that such a dedicated forum would be useful for our current users worldwide and for potential future users. We would highly appreciate if you could put in place such a forum. Let me know if there is anything we should do from our side to start this forum.

I wish you and all readers a very happy and succesful year 2003 !

Bruno