[Altocirrus]

Hi,

I’m trying to simulate a glider wing (15m span, 0.7m chord, Eppler 603 airfoil) flying at 25 m/s (Reynolds number = 1200000). I'm teaching myself Star-CCM+ for this purpose, but with very little information to go on and no support I'm having a hard time getting accurate results. I’m validating against XFLR5, as well as performance values for a glider using this wing. Induced drag is the primary interest of the project, so the bounding box extends beyond the wingtip to simulate tip effects. I’d really appreciate your thoughts on the following:

• The size of the bounding box: The Star-CCM+ documentation suggests that a chord length away from the wing surface is sufficient, but this seems close to me. My boundaries are currently 3x the chord length away from the wing.
• Mesh density/ cell size: I don’t think this is the main issue, but I was wondering whether there is a ‘rule of thumb’ for setting these values.
• Mesh type: I’ve tried polyhedral and trimmer meshes, and polyhedral has so far seemed to give more accurate results (my trimmer meshes seem of low quality for some reason). However they use much more memory and simulations take far longer, so there is more scope for increasing the trimmer mesh density.
• Prism layer mesh: I’m unsure on the number/ thickness I should use.
• Turbulence modelling: As a glider wing, I imagine the flow over the wing’s surface will be predominantly laminar, so do you believe the laminar flow model would be sufficient? If not I’ll use the Spalart-Allmaras model combined with turbulence suppression. Any tips or suggestions for values I should for these models?
• Any other things I should bear in mind for finite wing simulation?

Sorry for asking so much, but I’d be extremely grateful for any advice! Better yet, if anyone knows of any relevant tutorials or (better still) example .sim files to have a look it would be a huge help. I've spent weeks on end trying to improve accuracy, and it seems mad to start from scratch when this is a fairly standard aerospace problem that has probably been done throusands of times before.

Thanks

[Ash_D]

Hi,

If you can send me your email address I can get one of our technical team members to send you some suggestions for your modelling requirements.

Best Regards

Ash

ashkan.davoodi@uk.cd-adapco.com