help and advice needed on using Gambit and Fluent in AC yacht analysis

[D_Mylonas]

Hello,

I am a 5th year Naval Architect student at the Universities of Glasgow & Strathclyde, currently doing my MEng Group project. For this project, we have to optimize the design of an America's Cup under the new version of the rules (that we designed ourselves) using the available tools.

My part,at the moment,is focused on the appendages, that is the keel, the bulb and the rudder, and I need to analyse the forces, the resistance and the lift/drag coefficients applied to the appendages through CFD analysis with Gambit and Fluent.

The problem: I can import the designs as IGES files in the Gambit (version 1.3 by the way with some computers on version 2.2.30) and I create a box around say the rudder to create a volume and I define the rudder as a volume, but whatever I try, I cannot mesh it and everytime it comes up with an error message.

The people supposed to help, well, are not really helpful and I'm starting to get desperate. I have access to the tutorials in Gambit from the official website, but I really need some help and thus thought of posting in this forum.

Anyone who's faced a similar case in modelling a yacht or else in Gambit and Fluent will be of great help.

Thank you

Dimitris

[burakreis]

hi dimitris,

there are many possible reasons for your problem. but this sounds to me like a geometry issue. first check your surface normals, all surface normals must be inward in your domain (volume) so check your rudder, keel surfaces. gaps between surfaces may also a reason (if your geometry contains trimmed surfaces specially check the iges export tolerances) so you have to check and repair your geometry first, i dont know if you can repair in gambit because i am not using it. if available in your university use gridgen which is far more powerful then gambit...

regards,

[Tim B]

You should be able to do this with GAMBIT with little hassle. Are you using structured grids? if so think about the geometry of the surface grid. If it is multi-block think about the blocking and how it is best fitted. Remember, that all your blocks must have 6 faces, and within blocks cells must have 6 faces and positive volumes.

For Unstructured grids I can't see a problem unless it is trying to grid the wrong side of the surface. Check everything carefully.

Bad Geometry is the first cause of bad results... it's a shame it's never addressed in any great depth.

Tim B.

[nico]

Hey Dimitris,

I have just started a PhD at Strathclyde in Cfd and i am from a yacht design background. I am upstairs in the research area. Or send me an email. nicolas.rousselon AT strath DOT ac DO uk