An OpenMETA model for the conceptual design of an autonomous transport aircraft
Table of Contents
- Summary
- A³ Vahana Configuration Trade Study
- OpenMETA Vahana Configuration Trade Study
- Getting Started with the OpenMETA Vahana Model
- Installing OpenMETA
- Downloading Xfoil
- Cloning the openmeta-vahana repository
- Opening the openmeta-vahana project
- Viewing a PET model
- Running a PET model
- Viewing PET model results in the Visualizer
- PETs included in openmeta-vahana
- Python Wrapper component scripts
- Documentation
- Viewing the Meta-Link CAD Model
- Future Plans
Inspired by Project Vahana from A³ by Airbus, we built OpenMETA models of the Vahana Tilt-Wing Multirotor aircraft using A³'s publicly released source code as a template.
A parametric CAD model of Vahana was also built within OpenMETA with the goal of using CAD analysis to improve the accuracy of future studies. The most recent draft of this CAD model is shown below with its rotors angled to transition between the hover and cruise modes.
Vahana CAD model:
The OpenMETA models were used to perform similar DOC analyses to those in the Vahana Configuration Trade Study as well as improve on the original study.
OpenMETA 'Optimizer' Results and Vahana Trade Study Results for Tilt-Wing Configuration at Range of 100 km:
DOC ($) | DOC (km/$) | rProp (m) | cruiseSpeed (m/s) | batteryMass (kg) | motorMass (kg) | mtom (kg) | |
---|---|---|---|---|---|---|---|
OpenMETA* | 92.78 | 0.93 | 0.98 | 48.6 | 304.2 | 55.6 | 799 |
Vahana Study | 92.78 | 0.93 | 0.98 | 48.6 | 304.2 | 55.6 | 799 |
*after correcting for double fuselage drag typo.
OpenMETA PET model results:
DOC vs. Range for Tilt Wing and Helicopter Configurations
DOC vs. Range for Tilt Wing Configuration with and without Fuel Weight Constraint
Vahana is an A³ campaign to create a low-cost, single-passenger, electric VTOL aircraft. As part of their design process, A³ conducted the Vahana Configuration Trade Study to better examine 2 different configurations (an electric helicopter and an electric eight-fan tilt-wing) using multidisciplinary design optimization (MDO).
A³ set up a MDO sizing problem in which they compared an electric helicopter model and an electric eight-fan tilt-wing model over a range of operating distances (10 km to 200 km in 10 km steps). At each distance, both vehicle models were optimized seperately for Direct Operating Cost (DOC) by varying 5 design variables: Cruise Speed, Rotor Radius, Battery Mass, and Takeoff Mass. The A³ team also provided the optimizer with 3 constraint equations (4 if the vehicle was a helicopter) that defined certain design requirements (e.g.: the vehicle's effective energy capacity had to be greater than the amount of energy required to execute a reserve ("worst case") mission).
As a result of their Sizing Trade Study, the Vahana team concluded that an eight-fan, tilt-wing configuration would best meet their broad design requirements for a low-cost, single-passenger, electric VTOL aircraft.
Here at MetaMorph, we set out to first replicate the Vahana Configuration Trade Study's results using the OpenMETA toolset. Since OpenMETA is designed for Multidisciplinary Design Analysis and Optimization, we thought that it would be interesting to see if we could reproduce the A³'s results using the OpenMETA toolset.
To read about the OpenMETA Vahana Configuration Trade Study, check out our blog post.
To start working with the OpenMETA Vahana Model, check out the sections below.
- Download the latest version of OpenMETA from https://openmeta.metamorphsoftware.com/
- Open the installer.
- Agree to the license terms and conditions.
- Left-click 'Install'.
Note: If this is your first time using OpenMETA, we recommend that you complete the PET Tutorial to develop a basic understanding of the tools!
For Windows:
- Download
XFOIL6.99.zip
(Xfoil 6.99 for Windows) from http://web.mit.edu/drela/Public/web/xfoil/. - Extract the contents of
XFOIL6.99.zip
toC:/OpenMETA/xfoil-and-nrel-codes/bin/
(you may need to create this directory).
For Mac:
- Download
xfoil.dmg
(Independent 3rd-party build of Xfoil for Mac) from http://xfoil4mac.altervista.org/. - Drag-and-drop
Xfoil.app
into your Applications folder.
- Copy the following key onto your clipboard:
git@github.com:metamorph-inc/openmeta-vahana.git
- Open Git Bash in your desired project directory.
- Run the following command in Git Bash:
git clone git@github.com:metamorph-inc/openmeta-vahana.git
- Open the openmeta-vahana folder.
- Double-click on the openmeta-vahana.xme file.
- GME will open and display a pop-up 'Import to new project' window. Select 'Create project file' and click 'Next >'.
- The 'Save As' window will open. Click 'Save' to save 'openmeta-vahana.mga' inside the openmeta-vahana folder
- Within GME, to your right, there should be a 'GME Browser' window with a single 'RootFolder' object inside. Click on the '+' to expand the root folder.
- Left-click on the '+' next to 'Testing' to expand the testing folder.
- Left-click on the '+' next to 'ParametricExploration' to expand the parametric exploration folder.
- You should now see a number of PETs
. - Within the GME Browser window, double-click on 'TiltWingPET' to open it.
- Left-click the CyPhy Master Interpreter button located on the top toolbar.
- The 'CyPhy Master Interpreter' window will open. Make sure 'Post to META Job Manager' is checked and left-click 'OK'.
- The 'Results Browser' window will open. The running PET will be listed under the 'Active Jobs' tab. Blue means the Master Interpreter is still running, red means the Master Interpreter failed, and green means that the Master Interpreter succeeded.
- Once the DemoVahanaTiltWingPET finishes running, left-click the PET tab of the Results Browser.
- Information from the PET run will be displayed to your right within the Results Browser window.
- Left-click the 'Launch in Visualizer' button in the bottom-right corner (of the Results Browser window) to view the results in the PET Visualizer.
- The Visualizer will open in a browser window. Left-click the 'Explore>Single Plot' tab.
- Under the 'Variables' section, set the X-Axis to 'Range' and the 'Y-Axis' to 'DOCPerKm'.
The following PETs are included in the openmeta-vahana project in the RootFolder>Testing>ParametricExploration
directory.
Parametric Study + Optimizer | These PETs optimize the Tilt-Wing model for DOC over a series of ranges |
---|---|
TiltWingPET | Base Model, re-creates Vahana Trade Study tilt-wing configuration results |
TiltWingPETOrganized | Model organized by subsystem using sub-PET blocks |
TiltWingPETExposedParameters | Model with exposed top-level design parameters for easy experiments |
TiltWingPETFuelWeightCon | Model with additional constraint limiting fuel weight to 1/3 of MTOW |
TiltWingPETHigherCruiseSpeed | Model with increased minimum cruise speed of 60 m/s |
Optimizer | These PETs optimize the Tilt-Wing model for DOC at a single range |
---|---|
TiltWingOptimizerPET | Base model |
TiltWingOptimizerPETCostBreakdown | Model with cost breakdown metrics recorded by driver for analysis |
TiltWingOptimizerPETMassBreakdown | Model with mass breakdown metrics recorded by driver for analysis |
Parametric Study | This PET explores the Tilt-Wing model design space via a parametric study |
---|---|
TiltWingParametricStudyPET | Base model |
Parametric Study + Optimizer | These PETs optimize the Helicopter model for DOC over a series of ranges |
---|---|
HelicopterPET | Base model, re-creates Vahana Trade Study helicopter configuration results |
HelicopterPETFuelWeightCon | Model with additional constraint limiting fuel weight to 1/3 of MTOW |
Optimizer | These PETs optimize the Helicopter model for DOC at a single range |
---|---|
HelicopterOptimizerPET | Base model |
HelicopterOptimizerPETCostBreakdown | Model with cost breakdown metrics recorded by driver for analysis |
HelicopterOptimizerPETMassBreakdown | Model with mass breakdown metrics recorded by driver for analysis |
All the Python Wrapper component scripts used to build the openmeta-vahana PETs are located in the openmeta-vahana/scripts
folder. We encourage you to compare them with the A³ MATLAB scripts. To learn more, check out OpenMETA's Python Wrapper documentation as well as OpenMDAO's documentation.
(You must have a CREO Parametric license to view the Meta-Link CAD Model)
- In the openmeta-vahana project, navigate to
RootFolder>ComponentAssembly>VahanaV2
. - Left-click on the Meta-Link button located on the top toolbar.
- The 'CyPhyMLSync' windows will pop up. Left-click 'OK'.
- CREO Parametric will start (this may take a few seconds).
- For best viewing in CREO, turn off all the axis displays and turn on 'Shading With Reflections'.
- Changing the 'Canard_Rotation', 'rProp', and 'Wing_Rotation' parameters in the OpenMETA VahanaV2 model will automatically update the CAD model in CREO Parametric.
For additional information regarding the OpenMETA toolset, please consult the documentation.
Quick links:
Introduction
PET Tutorial
PET Nesting
Results Browser
Visualizer
Since we were able to produce similar trends to those in the Vahana Configuration Trade Study using OpenMETA, we wanted to explore ways in which to improve on the Trade Study's results. Outside of exercising the OpenMETA toolset, there was little reason to recreate the Vahana Configuration Trade Study - it had already served its purpose and A³ had already progressed to the next stage in the design process. Therefore, our next goal was to see how we could improve the Trade Study's model and provide more detailed anaylsis for design purposes.
A CAD model can provide a more accurate representation of a vehicle's geometry, mass, and mass distribution. In creating a CREO model of the Tilt-Wing Configuration, we hoped to provide a more model that could eventually be used to calculate mass, center of gravity, and production costs.
The model (shown below) is based on the sketches of the Tilt-Wing configuration that A³ released in the Vahana Trade Study Report. This model is composed within GME and contains parametric features that align with the design requirements outlined within the Vahana Trade Study's MATLAB code. The rotational orientation of the wings and canards can be varied between the cruise and hover positions (or 0-90 degrees).
Vahana transitioning from hover mode to cruise mode:
The model assembly is located at RootFolder/ComponentAssemblies/Vahana2
within openmeta-vahana.xme. By changing the values of the Canard_Rotation
and Wing_Rotation
parameters, a user can be change the orientation for a specific analysis. The rProp
parameter allows the user to change vary the length of the propeller blades. Changing the propeller blade radius also changes the span of both flight surfaces as well as the positions of the rotors in order to maintain an appropriate spacing between the neighboring rotors.
When exploring the Vahana model's design space with the TiltWingParametricStudyPET, we observed that less than 0.04% of the design generated met our design requirements. This low success rate was primarily caused by designs failing the first two design constraints:
- The vehicle battery must contain more energy than the energy required to complete the reserve mission
- The vehicle motors must be capable of providing at least 1.7 times the maximum power required for hover mode
The Vahana Trade Study set battery mass and motor mass as design variables while using assumed values for battery specific energy and motor specific power. Since 95% of all 'Parameter Study' PET runs failed as a result of the first two constraints, we looked into making battery mass and motor mass dependent variables which could be solved using an iterative method and the other design variables.
Using Euler's linear method of numerical integration (method shown here) the required battery mass and motor mass can be solved for within 0.01% of the theoretical value in 10 interation for ranges up to ten times larger than the assumed payload of 150 kg. A future model may implement these mass calculations in order to allow a more efficient parametric exploration of the design space.
OpenMETA Vahana Case Study - Part 1
OpenMETA Installer
OpenMETA Documentation
A³ Vahana Configuration Trade Study Part - 1
A³ Vahana Configuration Trade Study Part - 2
A³ MATLAB Code