openvogel / tucan Goto Github PK

The connection to the fuselage is sometimes wrong. This might have something to do with the reduced number of nodes in the cross sections.

The current results panel in this Beta version contains valuable data, but it could be improved.
Here some of the tasks for the static part flap:

• Add more information to the air-loads distribution graph: local chord, local Reynolds, local polar, ect.
• Allow the user to switch between dimensionless coefficients and real magnitudes for the three components of the force and the moment.
• Allow the user to switch the units (this could also be done globally in the whole solution).

In the aeroelastic flap we could also improve the graph and provide some more information about the modes.

The fuselage structured meshing algorithm exits with exception after attaching a wing anchor

NOTE: this issue is affecting the calculation of rotor-blades, propellers and wind turbines. It is not critical for airplanes models.

When the free air stream is only generated by the angular velocity vector (null constant velocity vector), there is an exception while computing the air loads, because the reference dynamic pressure is zero.
To solve this there must come a different (more general) way of compute the reference dynamic pressure.
A solution could be setting and imposed value.

While trying to model the blades of a wind turbine, I realized we sometimes need a different sequence of rotation. The current sequence is ZYX, but it can be that for many purposes an XYZ sequence or any other transformation is easier to handle.
In the EulerAngles class, we need to add the rotation sequence, and take this into account when generating the rotation matrix.
Background: recall that a different rotation sequence will in general result in a different orientation, since finite rotations are not commutative.

Use only the library from the dependencies

Observed in VS 2013, Resource file "VisualModel\GUI\UserControls\TotalForcePanel.resx" cannot be found.

We are currently using the Newmark method for numerical integration of the modal ODEs. However, the implementation of this algorithm has not been exhaustively tested yet.
We need to run test cases to verify the associated classes are working appropriately.

There is an issue with the new code in VortexRing3. The resulting velocity and Cp values are wrong. Probably a mistake in the source or potential functions.
To be further investigated. The workaround is to avoid them by using small segments at the two extremes of the fuselage.

While playing with the E-Fan model, I realized that the mesh of the fuselage is sometimes not generated due to an issue in the point mapping. If the Z coordinate is slightly outside the range covered by the cross sections, the sub exits with an exception because it cannot map the point.

• There are some duplicated methods in the basic surface (like for generating the lattices)
• Rename "VortexSegment" to "Segment"
• Check other issues that might be source of confusion

• We need to provide a way to import custom profiles.
• If there is a standard format, we should implement it.

We have to check the composition of the stiffness and mass matrices at both, element and structural levels.
This is critical to leave the Beta-phase.

• When adding a new section, it should be located at positive Z.

• When adding a new section between two sections of one node, a circular reference section should be inserted.

To allow new elements in the future, we need a more flexible structure.

• Create a global interface for all kind of finite elements
• Create a namespace for the type "elements" and add classes for each one of them

A new feature has to be introduced to analyze the load over a particular set of panels.
In postprocessing mode, the user would select a group of panels and create a group. Then, the total force an moment around a given axis for that group of panels would be displayed in a specialized form.
This would allow users to study the air loads over ailerons, flaps and elevators.

The results for the sphere_24 problem in Apame are different than the ones in Tucan. This must be further investigated.

The form representing the aeroelastic histogram should be improved. Up to now, we only have the basic data but the representation in the plotting area is quite poor.
The form is located in the following path:
OpenVOGEL/AeroTools/VisualModel/GUI/Forms/FormHistogram.vb
This form is based in the user controls located at:
OpenVOGEL/AeroTools/VisualModel/GUI/User controls/Histograms/

To do:

• Replace S/N of surfaces by ID of type Guid
• Fix binary IO regarding polar curves IDs

NOTE: this is not a critical issue, but it would improve the representation of the results.

• The pressure coefficient color scales should be different for Dirichlet and Neumann panels, since they mean different things (absolute pressure .vs. pressure jump across the surface).

• The pressure vectors on Dirichlet panels should always point outwards.

We are currently working with the inverse iteration method with Gram-smith orthogonalization.
To improve performance we should implement the subspace iteration method.
There are already some subs to do this, but they have to be checked and tested.

My most recent contribution brings a new module where global units can be selected for each Implemented magnitude. There is however still no GUI for this, so although several magnitudes and units are available in the code, users cannot access them.

• We need now to create a form where users can select their favorite units for each magnitude, and we need to save this setting in the XML. Would anyone like to build this important feature?

• We also need to implement the global units in the design forms, such as for instance, in the lifting surface editor.