|
[Sponsors] |
July 8, 2007, 05:50 |
Aerofoil analysis
|
#1 |
Guest
Posts: n/a
|
Hi All!
I am trying to optimise one axial flow pump with very low head. Its a 360 deg steady state case. I have extended the size of rot domain 3 mm from leading edge and trailing edge sides. Rot domain meshed with tet and st with hex. Defined frozen rotor ggi between rot and stationary domain without any pitch change. So total geom looks like a pump in a pipe. Defined static pressure inlet with flow direction as zero gradient and turbulence also to zero gr in rot domain. Used SST in both domains with no HT. Initialisation is automatic. Getting convergence But still no validation done coz it will require a prototype. but i dont know whether this rot domain size,Turbulence model is right to predict the phemomenon or should i try Transition model in rot dom and k-E in stationary dom? any thing missing ? waiting for responses. thanks in advance. |
|
July 8, 2007, 19:26 |
Re: Aerofoil analysis
|
#2 |
Guest
Posts: n/a
|
Hi,
What is the blade Reynolds number? Glenn Horrocks |
|
July 9, 2007, 03:24 |
Re: Aerofoil analysis
|
#3 |
Guest
Posts: n/a
|
Hi Glenn!
Reynold No. is 2.2304E+06. I tried transition model with default coefficients but with coarse mesh i.e around 100000 nodes in the rot domain.Results shows no remarkable differences. And results with fine mesh around 600000 nodes shows remarkable differences. Whether prisms could make it different? Thanks. Kumar. |
|
July 9, 2007, 19:23 |
Re: Aerofoil analysis
|
#4 |
Guest
Posts: n/a
|
Hi,
Unless your blade is designed for extensive laminar flow it is unlikely there is much laminar flow at that Reynolds number. This means I would not bother with the turbulence transition model as the flow will be almost entirely turbulent. Also - there is no point running the transition model with a coarse mesh. It needs a fine and high quality mesh to work. Regards, Glenn |
|
July 10, 2007, 02:05 |
Re: Aerofoil analysis
|
#5 |
Guest
Posts: n/a
|
Hi Glenn!
Agreed with your suggestions. Now i m using only sst. Here mesh densities make significant differences. Tet with 700000 nodes and tet+prism with 300000 nodes(3prism layers 0f size 0.5) predict diff results. One more thing I would like to ask abt rot domain size is, how to choose its size? When i use Turbogrid it creates its own fluid zone,but i wants to model a 360 deg case.inlets and outlets from that fluid zone are not in a plane(slightly curved shape). Whether we can believe on that rot domain size? Waiting for Reply. Thanks in advance. Kumar. |
|
July 11, 2007, 19:29 |
Re: Aerofoil analysis
|
#6 |
Guest
Posts: n/a
|
Hi,
3 prisms is not generally enough prism layers to get good boundary layer resolution. Have a look in the documentation about recommendations for wall meshing. Glenn Horrocks |
|
Thread Tools | Search this Thread |
Display Modes | |
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
Analysis of separation bubble over an low Re flow 2D Aerofoil | Ashish_Mittal | FLUENT | 0 | August 9, 2011 10:14 |
effective size of flow boundary for aerofoil analysis | nangless | CFX | 3 | January 16, 2010 17:12 |
3D analysis of Ahmed body | Irshad22 | FLUENT | 0 | December 17, 2009 05:33 |
axial compressor -aerofoil analysis | richard | Main CFD Forum | 1 | November 1, 2002 09:15 |
Short Course: Computational Thermal Analysis | Dean S. Schrage | Main CFD Forum | 11 | September 27, 2000 18:46 |