BUBBLE

BUBBLE is a simple python library to explore the laminar separation bubble losses in xfoil performance polar.

Installation

Typical python installation usually comes with all the packages required for this library. Otherwise, Use the package manager pip to install dependencies.

pip install pandas numpy matplotlib
sudo apt install xfoil

Usage

Bubble has two main scripts to analyze a given airfoil which helps to answer questions such as "Is this possibly be the best airfoil to use in this RE number and this cl range?"

1. aseq_analysis.py

This will sequence through a set of alphas or cls and dump boundary layer profiles into separate files. Then the result will be segregated into two charts which at glance shows the severity of the laminar separation bubble losses. At the end of the session, each data frame will be dumped into an excel. If further analysis is required all BL data can be accessed via the variable g_vplo_data_frames.

python -i aseq_analysis.py

The Function signature is as follows

xfoil_seq(airfoil, oper_type, reynolds_number, command, invoke_xfoil=True, smoothing=False, remove_bubble=True)

• airfiol is the airfoil name without .dat
• oper_type is the xfoil polar type
• RE: self-explanatory
• command : command to send to xfoil - ex; aseq 0 10 .5
• invoke_xfoil: if the BL dump files are already available no need to run xfoil
• smoothing: recommended for Wortmann airfoils for example
• remove_bubble: a separate target polar will be generated indicating what will happen if the bubble is removed by the forced transition.

In the first graph

• x-tr is the transition location.
• x_sep is where the laminar separation bubble starts.
• x_ret is where the laminar separation bubble ends and the flow gets re-attached.

In the second graph

• cd is the total drag.
• cdp is the profile drag.
• possible_cd: This is at best a crude value. generated by forcing the flow to transition at the beginning at the separation.
• possible_cdp: just as above but taking only the profile drag into account
• H_max: maximum H_12 value usually found inside the LSB for low Reynolds number flows.
• P2/P1: Momentum defect across the laminar separation bubble. [Ref_1]

2. vplo_plot.py

This is ideally an extension to the above if one finds a severe bubble loss at a design point. And it's best to limit its feed by probably copying the files of interest into a separate folder.

python vplo_plot.py

Reference

[1] XFOIL Yahoo group Message number 150

[email protected]
Feb 28, 2001

You can look at either the Cf(x) plot or the H(x) plot, both in the VPLO sub-menu in OPER. The bubble has Cf < 0, and approximately H > 4 over its extent.

I should point out that the length of the bubble does not directly influence how draggy it is. By far the best "badness" indicator of a bubble is the maximum value of H it has, which typically occurs just before reattachment at the end of the bubble. If H_max < 4, there is no laminar separation and no bubble. For H_max > 4, the bubble loss (additional CD suffered by the airfoil) can be approximated by

CD_bubble = A *(H_max - 4)^2

with the constant of proportionality A being somewhat dependent on the particular airfoil. Things get increasingly bad increasingly rapidly as H_max increases. Some rules of thumb I use:

H_max = 4 Ideal minimum drag situation.
H_max = 5 CD_bubble almost negligible.
H_max = 6 CD_bubble noticable, but may be tolerable.
H_max = 7 CD_bubble is a sizable fraction of total CD. Not good.
H_max > 8 CD_bubble dominates total CD. Really awful.

H_max will normally change with angle of attack in non-simple manner depending on the airfoil Cp distribution. H_max will always increase with decreasing Reynolds number and increasing Ncrit. It is very desirable for H_max to hover right around 4 over most of its alpha range. Michael Selig's more recent design approach seeks to control the H values directly, and thus allows control of the bubble drag among other things.

• Mark

Contributing

Pull requests or any suggestions for improvements are welcome.

License

MIT