This message is for users (aside from Numeca's personnels) of Numeca's Fine/Turbo package.

What is your overall impression of the solver in terms of speed, memory requirement, accuracy, and robustness? I am in the process of evaluating this package for turbomachinery applications. Any feedbacks from experienced users would be greatly appreciated.

Thanks in advance.

Hi, I am using Fine Turbo , modelling a low speed research compressor, 4 1/2 stage, Ma 0.2, Re 500000, massflow 25 kg/s, pressure ratio 1.1

Fine Turbo is quite fast, especially using the Multigrid features, in terms of memory requirement ist seems to be reasonable, I was not forced buy memory. 288 MB RAM, 324 MB Swap at a DEC Alpha 3000/600 let me do meshes up to 500000 nodes.

the interface is easy to use, includes a lot of features (and sources of user-errors) Baldwin-Lomax-Turbulence-model is very robust, others in my case not.

Great feature is Autogrid, I model a 500000 nodes mesh of a real Turbomachine in 1-2 hours.

Ask details, if you need more information.

Ralf

Hello Hoang Vinh,

we have been using FINETurbo for 2 years at our institute and we made very good experiences with this package. We are mainly working on 2D- and 3D-flow simulations in Turbine- and compressor-blade-cascades and succeeded in validating FINE/turbo by experimental data (by the way the only commercial code to fit the measurements). So we decided to make it the main commercial tool for this purpose. Numeca did a lot during this two years to increase speed tremendously and they enormously reduced memory requirements of the code. The GUI now is very stable and the integrated grid generator (igg) is one of the best you can get (automatic grid generation, local mesh-refinement, powerful smoothing algorithms). We also made very good experience referring the support (Mr. Hildebrandt in Germany; e-mail: Th.Hildebrandt@t-online.de) what I do think is one of the most important points in that matter.

Hope I could help you,

Markus

Hi,

first of all, I'm NOT a very experienced user of the FINE/Turbo package with only few weeks of using it. Although in that time we have calculated a whole characteristic (6 points) of the 5 1/2 transonic front stages of a 10 stage axial compressor. I think that is saying a lot about the level of automation of the IGG-Griding functions, especially for turbomachines. It took about 24h to 30h for a converged calculation of 1 point on the characteristic (SGI Origin with enough(?) memory, approx. 1.7 Mio grid points). A drawback of the solver is that it is not parallelized yet, although Numeca is working on it. The post processing side is, I think, the best part of the whole package. You will find there all kind of turbomachinery specific functions, mass averaging, performance calc. etc.

To sum it up I can state that if You are "tired of meshing around", need a robust and easy to use CFD-package and You are dealing with turbomachinery, Numeca's FINE/Turbo is, at present, the best choice.

Regards Marius

Dear Hoang Vinh,

besides other CFD packages we have been using FINE/TURBO for 1 year to calaculate different types of turbomachinery applications. Within this time I calculated axial stages, radial impellers, complete radial stages and also volute flows with that program. Meshing all these different types of geometries in reasonable times requires a highly automatized pre-processor. IGG/Autogrid is a very powerful pre-processor which allows automazid grid generation for a large variety of turbomachinery bladings. I do not know any other grid generator which is such powerful for automatic turbomachinery grid generation like IGG/Autogrid. The general grid generaor IGG is also very userfriendly. CFVIEW gives you the opportunity to create turbomachinery specific views very comfortable by just a view mouse click's. Beyond that the macro language enables you to make your individual post-processing in an automatized way.

The flow solver itself is quite fast (with an speed up factor of 3 from version 3.01 to 3.02) and the multigrid functionality provides the very useful feature to run problems on different grid levels without meshing around. Meshing, calculating (meshsize 50 000 points) and post-processing 3 different performance characteristics (each with 5 points) of different radial impeller versions can be done in one day (SGI octane).

Last but not least, the support from NUMECA by Mr. Hildebrandt who is our direct contact person is very good, and also the support of NUMECA international. I fully agree with the other author, that this point should not be disregarded.

Hope I could help you Matthias Grapow

[Deb]

I note that this thread is more than 10 years old. However, I am now trying to assess CFX versus Fine Turbo for automotive turbocharger applications. Any words of wisdom would be most appreciated.

In particular I am also interested in interfacing with one of the FEA codes for thermal-stress and fatigue calculations. Keeping this in view are there any comments?

Fine Turbo has non linear harmonics, useful for studying unsteady interactions between guide vances and the blade row. CFX has an approach that is phase lagged. Any comments regarding this difference between the two codes.

Many thanks in advance