Constraint analysis scripts for initial sizing #398

stankarpuktubs · 2020-12-27T11:11:12Z

Pull Request Checklist

Merge in the latest Develop branch changes to your branch
Remove .pyc files from your repository:
- Linux/Mac
  - find . -name *.pyc -delete
- Windows
  - del /S *.pyc
Run automatic regression and makes sure everything is passing
Did you make a new regression test that covers your new code?
Did you update your docstrings?
Did you update your headers to include your name and date?
Do a final compare with the Develop branch to be sure what you're changing

coveralls · 2020-12-27T12:17:22Z

Coverage increased (+0.01%) to 89.998% when pulling 693adfe on stankarpuktubs:develop into 4f83c46 on suavecode:develop.

A deep methods update and a regression script

Updated the propeller aircraft constraint analysis regression

…o develop

planes

The two big overall comments:
Split things out into methods and make consistent with SUAVE vehicle data structures.

Overall good work. I think this is a good contribution.

Tutorials-master/tut_payload_range.py

Tutorials-master/tut_solar_uav.py

regression/automatic_regression.py

regression/scripts/constraint_analysis/constraint_analysis.py

trunk/SUAVE/Analyses/Constraint_Analysis/Constraint_Analysis.py

planes · 2021-11-15T18:19:38Z

trunk/SUAVE/Analyses/Constraint_Analysis/Constraint_Analysis.py

+        return e
+
+
+    def estimate_max_lift(self,highlift_type):


I think this could be useful for many other calculations. Can you make this more general and move to methods?

I see that SUAVE already has the maximum lift estimation function. Is it better to have this function as a stand-alone or to integrate it into the SUAVE default maximum lift function?

If you're able to integrate it, that would be great. You can make a setting that will switch between types.

planes · 2021-11-15T18:20:11Z

trunk/SUAVE/Analyses/Constraint_Analysis/Constraint_Analysis.py

+            raise ValueError(highliftmsg)
+
+
+    def compressibility_drag(self,mach,cl):


call the existing scripts since this is much the same

Here, I am struggling with the fundamental difference between functions of SUAVE and requirements for the constraint diagram. All default functions require definitions of settings, states, etc, since thez are created for the mission segment analysis. Here, I just need 2-3 inputs for the execution with minimum input, so the constraint diagram can be initialized and executed as simple as possible. For the latest update, I still have not managed to integrate these functions into the default SUAVE

I understand you're trying to do a simple analysis here. I'm not trying to change that. I bet you think this is frustrating because you coded this with a certain intention of this being simple.

For this function, you can easily replace it with the original. Settings are not used, hence they become None. Geometry only requires wing that has a thickness to chord and sweep. State is just a dictionary that saves the outputs and provides a CL. It's not tricky to use the existing script.

We've found that having reproduced code eventually leads to bugs and just bloat. It's more confusing and it all becomes a mess eventually.

planes · 2021-11-18T05:42:57Z

If you haven't seen already, there's an issue with versions of a Python package unrelated to SUAVE. This is giving the red X. When the issue is fixed, I'll rebuild and we can finish off the pull request.

planes

Just adding a few new comments.

planes · 2021-11-20T20:22:39Z

regression/scripts/constraint_analysis/constraint_analysis.py

-    ca.geometry.high_lift_type_clean   = None
-    ca.geometry.high_lift_type_takeoff = 'double-slotted Fowler'
-    ca.geometry.high_lift_type_landing = 'double-slotted Fowler'
+    ca.geometry.aspect_ratio                 = vehicle.wings['main_wing'].aspect_ratio


I was hoping it would look something like this:

ca.geometry = B737_vehicle()

That way a user doesn't need to do all of this manually.

I think a really neat workflow for a new user who would be interested in your contributions here could be to draw a vehicle in OpenVSP and very quickly produce the constraint diagram. That would enable this.

I am actually working on the neat integration now, so the constraint diagram script will accept the entire vehicle with some pre-defined constraint diagram settings. I can also do it the way you suggested. Hope, I will commit the deep update by the end of next week

Great! Do what you think makes most sense here.

I'm looking forward to this. I think your work here fills an important need in SUAVE.

planes · 2021-11-20T20:24:22Z

trunk/SUAVE/Analyses/Constraint_Analysis/Constraint_Analysis.py

+from SUAVE.Methods.Constraint_Analysis.compute_cruise_constraint    import compute_cruise_constraint
+from SUAVE.Methods.Constraint_Analysis.compute_ceiling_constraint   import compute_ceiling_constraint
+from SUAVE.Methods.Constraint_Analysis.compute_landing_constraint   import compute_landing_constraint
+from SUAVE.Input_Output.Results.plot_constraint_diagram                     import plot_constraint_diagram


planes · 2021-11-20T20:25:54Z

trunk/SUAVE/Input_Output/Results/plot_constraint_diagram.py

+    f.close()    
+
+
+# ----------------------------------------------------------------------


Delete this lower bit. This is our fault for leaving this in our style.

1. Moved the geometry and energy network definition into the Vehicle (where was possible) 2. Defined the constraint analysis variable set within the vehicle 3. Allocated analyses scripts within the Aero folder 4. Allocated the post-processing code into the Input/Output folder 5. Fixed many parts to make the structure work

stankarpuktubs · 2022-01-10T10:15:39Z

Emilio,

I think, I have changed almost all parts where needed and embedded the constraint analyses definitions into the vehicle, so, as you described, the designer can start from minimum data and then build the vehicle step-by-step.

planes · 2022-01-11T06:28:34Z

trunk/SUAVE/Input_Output/Results/plot_constraint_diagram.py

@@ -0,0 +1,103 @@
+## @ingroup Input_Output-Results


this should be moved to this folder: https://github.com/suavecode/SUAVE/tree/develop/trunk/SUAVE/Plots/Performance

planes · 2022-01-11T06:29:02Z

trunk/SUAVE/Input_Output/Results/plot_constraint_diagram.py

+    f.close()    
+
+
+# ----------------------------------------------------------------------


planes · 2022-01-11T06:31:07Z

trunk/SUAVE/Methods/Aerodynamics/Common/Fidelity_Zero/Helper_Functions/oswald_efficiency.py

+    """  
+
+    # Unpack inputs
+    taper = vehicle.wings['main_wing'].taper


I would prefer if this parsed for the main wing rather than look for a name. The code is simple:

from SUAVE.Components.Wings.Main_Wing

main_wing = None
for wing in wings:
if isinstance(wing,main_wing):
main_wing = wing

planes · 2022-01-11T06:34:09Z

trunk/SUAVE/Methods/Constraint_Analysis/compressibility_drag_constraint.py

@@ -0,0 +1,78 @@
+## @ingroup Methods-Constraint_Analysis
+# Oswald_efficiency.py


This header is wrong

planes · 2022-01-11T06:34:18Z

trunk/SUAVE/Methods/Constraint_Analysis/compressibility_drag_constraint.py

+import numpy as np
+
+# ------------------------------------------------------------------------------------
+#  Compute maximum lift coefficient for the constraint analysis


This title is wrong

planes · 2022-01-11T06:38:06Z

trunk/SUAVE/Methods/Constraint_Analysis/compute_climb_constraint.py

+            T_W[i] = Vy/Vx + q/W_S[i]*cd0 + k/q*W_S[i]
+            P_W[i] = T_W[i]*Vx/etap
+
+            if eng_type == ('turboprop' or 'Turboprop'):


check for instances

planes · 2022-01-11T06:39:33Z

trunk/SUAVE/Methods/Constraint_Analysis/compute_cruise_constraint.py

+    eng_type  = vehicle.constraints.engine.type
+    cd_min    = vehicle.constraints.aerodynamics.cd_min_clean 
+    W_S       = vehicle.constraints.wing_loading
+    AR        = vehicle.wings['main_wing'].aspect_ratio


planes · 2022-01-11T06:42:52Z

trunk/SUAVE/Methods/Constraint_Analysis/compute_turn_constraint.py

+        if etap == 0:
+            raise ValueError('Warning: Set the propeller efficiency during turn')
+
+    for i in range(len(W_S)):


2 things I see.

First, this section of code is used repeatedly in several scripts. This indicates this could be functionalized.

Second, I believe you could vectorize this for loop.

planes · 2022-01-11T06:44:40Z

trunk/SUAVE/Vehicle.py

+        self.plot_tag           = False
+
+
+        # Defines default constraint analyses


vehicle.analyses breaks a fundamental assumption of SUAVE having data/analysis orthogonality

planes · 2022-01-11T06:45:09Z

trunk/SUAVE/Vehicle.py

+        self.propeller.OEI_climb_efficiency = 0.0
+
+        # Define aerodynamics
+        self.aerodynamics = Data()


This is basically an analysis

Updated the data constraint analysis structure Fixed all errors, relocated useful files Added a dummy turboprop energy network (similar to ICE, but with a different power lapse function)

stankarpuktubs · 2022-01-15T11:39:57Z

@planes Have updated the analyses, used existing scripts as much as possible, and corrected many small things in the code

planes · 2022-01-31T23:29:30Z

regression/scripts/Turboprop/Turboprop_CS_Test.py

@@ -0,0 +1,228 @@
+# ICE_Test.py


update header

planes · 2022-01-31T23:32:30Z

regression/scripts/Turboprop/Turboprop_Test.py

+# 
+# Created: Jan 2022, S. Karpuk 
+
+""" setup file for a mission with a Beech 99 with a turboprop energy network


planes · 2022-01-31T23:35:27Z

regression/scripts/Turboprop/Turboprop_Test.py

+sys.path.append('../Vehicles')
+# the analysis functions 
+
+#from Beech_99  import vehicle_setup  


remove commented out code

planes · 2022-01-31T23:59:32Z

regression/scripts/Vehicles/Cessna_172_turboprop.py

+    #   Fuel
+    # ------------------------------------------------------------------    
+    # define fuel weight needed to size fuel system
+    fuel                                        = SUAVE.Attributes.Propellants.Aviation_Gasoline()


planes · 2022-02-01T00:03:22Z

regression/scripts/constraint_analysis/constraint_analysis.py

+    # Sample jet airplane
+    # ------------------------------------------------------------------
+
+    # Define the vehicle and required constrain analysis parameters


planes · 2022-02-01T00:10:12Z

regression/scripts/constraint_analysis/constraint_analysis.py

+    ca.propeller.ceiling_efficiency   = 0.85
+    ca.propeller.OEI_climb_efficiency = 0.5
+
+    # Define aerodynamics (an example case with max lift calculation for differnet flap settions)


different flap settings

planes · 2022-02-01T00:58:06Z

trunk/SUAVE/Components/Energy/Networks/Turboprop_Constant_Speed.py

+#  Network
+# ----------------------------------------------------------------------
+## @ingroup Components-Energy-Networks
+class Turboprop_Constant_Speed(Network):


This should inherit from this network: https://github.com/suavecode/SUAVE/blob/develop/trunk/SUAVE/Components/Energy/Networks/Internal_Combustion_Propeller_Constant_Speed.py

@planes I see. Indeed this script is identical to the ICE. The only difference is inside the turboprop converter (https://github.com/stankarpuktubs/SUAVE/blob/develop/trunk/SUAVE/Components/Energy/Converters/Turboprop.py) where I am using the power lapse formulation for a turboprop. Shall I then modify the ICE network to work with the turboprop converter to avoid unnecessary copies of the code?

planes · 2022-02-01T00:58:49Z

trunk/SUAVE/Components/Energy/Networks/Turboprop.py

+#  Network
+# ----------------------------------------------------------------------
+## @ingroup Components-Energy-Networks
+class Turboprop(Network):


this should inherit from here: https://github.com/suavecode/SUAVE/blob/develop/trunk/SUAVE/Components/Energy/Networks/Internal_Combustion_Propeller.py

planes · 2022-02-01T01:14:14Z

trunk/SUAVE/Methods/Constraint_Analysis/compute_ceiling_constraint.py

+            isinstance(prop, Nets.Internal_Combustion_Propeller_Constant_Speed) or isinstance(prop, Nets.Turboprop):
+
+            P_W  = np.zeros(len(W_S))
+            etap = ca.propeller.cruise_efficiency


ceiling efficiency

planes · 2022-02-01T01:18:42Z

trunk/SUAVE/Methods/Constraint_Analysis/compute_climb_constraint.py

+
+            for i in range(len(W_S)):
+                error      = 1
+                tollerance = 1e-6


tolerance is spelled wrong

Merged a sample turboprop and an ICE into one energy network. modified the constraint analysis script to include changes of the energy network

…o develop

stankarpuk93 added 2 commits December 27, 2020 12:07

Constraint analysis scripts for initial sizing

89ae097

changed the regression directory

9ef71de

stankarpuk93 and others added 13 commits June 3, 2021 20:58

Updated the constraint diagram

8e383df

A deep methods update and a regression script

regression update

8524c5a

Updated constraint analysis regression

18b80d9

Updated the propeller aircraft constraint analysis regression

Update automatic_regression.py

9c315e4

Merge branch 'develop' into develop

3915cf9

Removed ADRpy

6763a63

Merge branch 'develop' of https://github.com/stankarpuktubs/SUAVE int…

c1e05fa

…o develop

Updated regression

b00ef8c

Corrected the regression

46c71a8

Update constraint_analysis.py

4f69130

Checked the code structure and extended regression cases

99662fb

Update Constraint_Analysis.py

e5a58f4

Update constraint_analysis.py

2bb4b16

planes reviewed Nov 15, 2021

View reviewed changes

Modified the code structure. Removed unnecessary stuff.

2e4335b

planes reviewed Nov 20, 2021

View reviewed changes

stankarpuktubs and others added 6 commits January 2, 2022 14:48

Corrected the analysis initialization script

58cb34a

Corrected Methods init

f3ca997

Added cases to improve coverage

b95734b

Update constraint_analysis.py

69e1aa3

Merge branch 'develop' into develop

8e33952

stankarpuktubs requested a review from planes January 7, 2022 17:36

planes reviewed Jan 11, 2022

View reviewed changes

Deep update

3a239aa

Updated the data constraint analysis structure Fixed all errors, relocated useful files Added a dummy turboprop energy network (similar to ICE, but with a different power lapse function)

stankarpuktubs added 4 commits January 14, 2022 21:12

Dates update

d459ff0

Changed the compressibility drag calculation reference

5c2771f

Removed cache import

955ac8f

Extra formating

693adfe

stankarpuktubs requested a review from planes January 15, 2022 11:38

stankarpuktubs closed this Jan 22, 2022

stankarpuktubs reopened this Jan 22, 2022

planes reviewed Feb 14, 2022

View reviewed changes

stankarpuktubs and others added 4 commits February 24, 2022 23:50

Corrected grammar and minor errors

36971a3

Merge branch 'develop' into develop

3f5ae8d

Updates of the Internal Combustion Energy Network

cee054b

Merged a sample turboprop and an ICE into one energy network. modified the constraint analysis script to include changes of the energy network

Merge branch 'develop' of https://github.com/stankarpuktubs/SUAVE int…

c7d0e0f

…o develop

		raise ValueError(highliftmsg)


		def compressibility_drag(self,mach,cl):

		f.close()


		# ----------------------------------------------------------------------

		@@ -0,0 +1,78 @@
		## @ingroup Methods-Constraint_Analysis
		# Oswald_efficiency.py

Constraint analysis scripts for initial sizing #398

Are you sure you want to change the base?

Constraint analysis scripts for initial sizing #398

Conversation

stankarpuktubs commented Dec 27, 2020

Pull Request Checklist

coveralls commented Dec 27, 2020 • edited Loading

planes left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

planes commented Nov 18, 2021

planes left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

stankarpuktubs commented Jan 10, 2022

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

stankarpuktubs commented Jan 15, 2022 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

stankarpuktubs Feb 14, 2022 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

coveralls commented Dec 27, 2020 •

edited

Loading

stankarpuktubs commented Jan 15, 2022 •

edited

Loading

stankarpuktubs Feb 14, 2022 •

edited

Loading