peterdsharpe
commented
4 years ago Ah! This is a great question and definitely possible in the latest version, at least with CasLL1 (might be possible with CasVLM1 too, but not tested)!
You'll basically want to implement alpha as a free variable and set a constraint such that ap.CL == 0.7 or whatever your desired value is. Then, it'll solve the entire system of nonlinear equations together.
Here's a sample runfile! I've highlighted some key lines below it:
import copy
from aerosandbox import *
from aerosandbox.library.airfoils import generic_airfoil, generic_cambered_airfoil
opti = cas.Opti() # Initialize an analysis/optimization environment
# Define the 3D geometry you want to analyze/optimize.
# Here, all distances are in meters and all angles are in degrees.
airplane = Airplane(
name="Peter's Glider",
x_ref=0, # CG location
y_ref=0, # CG location
z_ref=0, # CG location
wings=[
Wing(
name="Main Wing",
x_le=0, # Coordinates of the wing's leading edge
y_le=0, # Coordinates of the wing's leading edge
z_le=0, # Coordinates of the wing's leading edge
symmetric=True,
xsecs=[ # The wing's cross ("X") sections
WingXSec( # Root
x_le=0, # Coordinates of the XSec's leading edge, relative to the wing's leading edge.
y_le=0, # Coordinates of the XSec's leading edge, relative to the wing's leading edge.
z_le=0, # Coordinates of the XSec's leading edge, relative to the wing's leading edge.
chord=0.18,
twist=2, # degrees
airfoil=generic_cambered_airfoil, # Airfoils are blended between a given XSec and the next one.
control_surface_type='symmetric',
# Flap # Control surfaces are applied between a given XSec and the next one.
control_surface_deflection=0, # degrees
),
WingXSec( # Mid
x_le=0.01,
y_le=0.5,
z_le=0,
chord=0.16,
twist=0,
airfoil=generic_cambered_airfoil,
control_surface_type='asymmetric', # Aileron
control_surface_deflection=0,
),
WingXSec( # Tip
x_le=0.08,
y_le=1,
z_le=0.1,
chord=0.08,
twist=-2,
airfoil=generic_cambered_airfoil,
),
]
),
Wing(
name="Horizontal Stabilizer",
x_le=0.6,
y_le=0,
z_le=0.1,
symmetric=True,
xsecs=[
WingXSec( # root
x_le=0,
y_le=0,
z_le=0,
chord=0.1,
twist=-10,
airfoil=generic_airfoil,
control_surface_type='symmetric', # Elevator
control_surface_deflection=0,
),
WingXSec( # tip
x_le=0.02,
y_le=0.17,
z_le=0,
chord=0.08,
twist=-10,
airfoil=generic_airfoil
)
]
),
Wing(
name="Vertical Stabilizer",
x_le=0.6,
y_le=0,
z_le=0.15,
symmetric=False,
xsecs=[
WingXSec(
x_le=0,
y_le=0,
z_le=0,
chord=0.1,
twist=0,
airfoil=generic_airfoil,
control_surface_type='symmetric', # Rudder
control_surface_deflection=0,
),
WingXSec(
x_le=0.04,
y_le=0,
z_le=0.15,
chord=0.06,
twist=0,
airfoil=generic_airfoil
)
]
)
]
)
airplane.set_spanwise_paneling_everywhere(30) # Set the resolution of your analysis
alpha = opti.variable()
opti.set_initial(alpha, 0)
ap = Casll1( # Set up the AeroProblem
airplane=airplane,
op_point=OperatingPoint(
density=1.225, # kg/m^3
viscosity=1.81e-5, # kg/m-s
velocity=10, # m/s
mach=0, # Freestream mach number
alpha=alpha, # In degrees
beta=0, # In degrees
p=0, # About the body x-axis, in rad/sec
q=0, # About the body y-axis, in rad/sec
r=0, # About the body z-axis, in rad/sec
),
opti=opti # Pass it an optimization environment to work in
)
opti.subject_to([
ap.CL == 0.7
])
# Solver options
p_opts = {}
s_opts = {}
# s_opts["mu_strategy"] = "adaptive"
opti.solver('ipopt', p_opts, s_opts)
# Solve
try:
sol = opti.solve()
except RuntimeError:
sol = opti.debug
raise Exception("An error occurred!")
# Postprocess
# Create solved object
ap_sol = copy.deepcopy(ap)
ap_sol.substitute_solution(sol)
# ap_sol.draw(show=True) # Generates a pretty picture!
print("CL:", ap_sol.CL)
print("CD:", ap_sol.CD)
print("CY:", ap_sol.CY)
print("Cl:", ap_sol.Cl)
print("Cm:", ap_sol.Cm)
print("Cn:", ap_sol.Cn)Snippet of output from the file above:
total | 1.18 s ( 1.18 s) 1.18 s ( 1.18 s) 1
CL: 0.699999999999972
CD: 0.027870764828169817
CY: 0.0
Cl: 0.0
Cm: 0.018719090893452202
Cn: 0.0Key lines from the file above:
alpha = opti.variable()
opti.set_initial(alpha, 0)
ap = Casll1( # Set up the AeroProblem
airplane=airplane,
op_point=OperatingPoint(
density=1.225, # kg/m^3
viscosity=1.81e-5, # kg/m-s
velocity=10, # m/s
mach=0, # Freestream mach number
alpha=alpha, # In degrees
beta=0, # In degrees
p=0, # About the body x-axis, in rad/sec
q=0, # About the body y-axis, in rad/sec
r=0, # About the body z-axis, in rad/sec
),
opti=opti # Pass it an optimization environment to work in
)
opti.subject_to([
ap.CL == 0.7
])
Describe the solution you'd like Set a lifting coefficient instead of an angle of attack to perform the calculations.
Describe alternatives you've considered For the moment I am iteratively changing the angle of attack until I converge to the desired CL.