[Jonathan]

Hi,

I have an issue I was wondering whether somebody may have encountered previously, or may have a short answer to.

I am modelling a multistage gas turbine using the MRF and mixing plane models. At this point, i seem to get relatively good results for the upstream and downstream nozzles around the blade row, however, get poor efficiencies compared to the experimental data for the rotor row, stage, and overall turbine efficiency.

Obviously it seems that the poor global and stage efficiencies are as a result therefore of the poor rotor efficiency prediction.

In setting up the models, i used area-averaging for the mixing plane which i have subsequently changed to mass averaging to get better mixing plane profiles, but am not using swirl or enthalpy conservation across the mixing plane, basically as I am not sure how great an effect they will have on the rotor blade row efficiency. I was wondering what your thoughts regarding trying to get a better prediction for the blade row might be (i.e. use swirl conservation??) but then in the long run, try to get the global efficiencies sorted out.

I gather i should probably be conserving the enthalpy across the mp's, but then would i be correct in assuming this would require me to increase the no of stators / nozzles to match the rotor domain (i.e in reality there are 30 stators -20 blades -30 stators) - at the moment i am only modelling a single stator-rotor-stator to keep model same and CPU time down.

So basically, i was wondering whether anyone had any thoughts on this in general or whether perhaps there are a set of 'standard' modeling assumptions/options which are used by some of the industry guys in modeling gas turbines especially when using the MRF and mixing plane models in Fluent?

any thoughts much appreciated,
cheers
jon