Cp on a duct with VSPAERO

[loc...@gmail.com]

While making aerodynamic analysis on a duct with panel method,i find some panels with extremely small Cp on the trailing edge on the right side. Since  Alpha and Beta are all zero, i think all the wakes should be the same. Can anyone help me with this  phenomenon? Thank you.

[Rob McDonald]

You'll also notice that there is a 'seam' in front of that point leading forward.

The duct component as implemented as a scripted user-defined component isn't great.

It isn't a perfect body of revolution -- the circles it makes aren't quite circular.  The error is greatest where they start/end.

You'll also notice that you don't have tremendous control of the overall shape of the duct -- or on the paneling resolution.

I've been meaning to make a new body of revolution component for some time, but haven't had a chance.  I don't know when I'll be able to do that.

In the meantime, depending on your needs, I suggest you try using a stack or fuselage component with the cross-section policy set to 'Loop'.  This will give you _much_ more control of your duct.

Rob

On Fri, Dec 9, 2016 at 6:32 AM, <loc...@gmail.com> wrote:

While making aerodynamic analysis on a duct with panel method,i find some panels with extremely small Cp on the trailing edge on the right side. Since  Alpha and Beta are all zero, i think all the wakes should be the same. Can anyone help me with this  phenomenon? Thank you.

--
You received this message because you are subscribed to the Google Groups "OpenVSP" group.
To unsubscribe from this group and stop receiving emails from it, send an email to openvsp+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[loc...@gmail.com]

Rob,thanks for your reply

I tried to use stack or fuselage to replace the duct.But it seems that the VSPAERO cannot draw grids for the model when the section is not a point at the first and last section.

What's more,I do doubt whether it can identify the Kutta nodes even if grids are successfully drawn. Because I think VSPAERO relies on the definition of airfoils 

to find Kutta nodes(I'm not sure). For now, I may ignore the error caused by the 'seam' if the result for CL and CDi doesn't differ too much from that generated by Panair on the whole aircraft.

Hoping for your updates.

Lock 

[Rob McDonald]

I think you're mixing apples and oranges.

VSPAERO can run in a thin-surface VLM mode or a thick-surface panel mode.

The thin-surface geometry generation will have trouble with nacelles.
However, the panel method (your earlier screenshot was like this)
should have no problems.

I've attached a nacelle like I suggest that was extracted from a
Transport aircraft model on the hangar.

As it so happens, it fixes the 'seam' I described, but it still
exhibits the larger problem you noticed. I'm not sure what is
happening here, but I'll try to pass it on to see if we can figure
something out.

Rob

> --
> You received this message because you are subscribed to the Google Groups
> "OpenVSP" group.
> To unsubscribe from this group and stop receiving emails from it, send an

> email to openvsp+u...@googlegroups.com.

[loc...@gmail.com]

Thank you for pointing out my mistake and supplying the nacelle model. 

I noticed the disordered wakes you mentioned  and I will make some  comparison for investigation.

I'm very grateful for your quick response and  I will keep following your progress.

               Lock