measure_lens_alignment

Measurement tools to interpret lens measurements from a Wenzel CMM.

Dependencies

• wenzel_database_parser

Setting up the configuration file

The configuration JSON file "config.json" contains, amongst other things, the necessary contextual information defining a set of measurements. There are four key parameters that are contained with the configuration file:

• PARAMS
• CONFIGURATIONS
• COORDINATE SYSTEMS, and
• DATA

PARAMS contains general information required for the program to run, e.g. where the database resides.

Each entry in CONFIGURATIONS defines a complete measurement set by specifying which datasets from DATA are to be used. Each has a unique identifier (id) and type (type). Each type will have a bespoke set of fields that will need to defined.

COORDINATE SYSTEM entries add extra information to a database coordinate system entry. For example, some may have an "angle" field which specifies the angle of the lens mount at which the measurement was done and a csMsRecNr to allow it to be mapped to entries in the database. For a coordinate system to be considered, it must be entered into this section of the configuration file.

Measurement datasets are defined in DATA. Each dataset must have fields containing a unique identifier (id), a measurement record range (elMsRecNr_range) and a list of elements within each measurement. Each of these elements should have a unique identifier (id) field, an element ID corresponding to that defined in the database (elId) and a description field (desc).

"rotational_analysis" type

The following fields are required:

• an identifier of the dataset containing rotational data (rotation_data)
• an identifier of the dataset containing error data (error_data)
• the element to be used for the front surface of the lens (lens_front_elId)
• the element to be used for the rear surface of the lens (lens_rear_elId)
• the element to be used for the mechanical front of the lens (mount_front_elId),
• the element to be used for the mechanical rear of the lens (mount_rear_elId)
• the z offset (pre-z direction reversal) to apply to the system, useful for defining the optical axis centre midway through the lens ring rather than at the front face (z_offset)
• the PCS id to be used for both rotational and error data (rotation_data_csId and error_data_csId)
• the two mount orientations to be used for hysteresis analysis (hys_idx_1 and hys_idx_2), with indexes defined from the array of angles
• whether the lens should be flipped in its holder, useful when the left and right lenses are measured flipped relative to the order of propagation in the optical system (flip_lens)
• whether the lens PCS x direction should be reversed (flip_PCS_x_direction)
• whether the lens PCS y direction should be reversed (flip_PCS_y_direction)
• whether the lens PCS z direction should be reversed, useful if the z direction of the PCS runs opposite to the z direction left-to-right convention (flip_PCS_z_direction)

The last of these four fields are particularly important for ensuring the consistency of the coordinate system between measurements. Elaborating on this further, consider this set of (real) examples:

Example 1:

Let us consider that a set of measurements are made with a part coordinate system (x, y, z) defined as being positive up, right, and towards the observer respectively. A lens is mounted such that the first surface measured corresponds to the front surface. Now let's consider that we have a model of the system in which the lens is to be placed. In this system, (x, y, z) is defined as being positive up, left, and away from the observer respectively. In order to match these coordinate systems, we'd need to set the flip_PCS_x_direction and flip_PCS_z_direction flags.

Example 2:

Now, as before, let us consider that another set of measurements are made with the same part coordinate system. However, this time a lens is mounted such that the first surface to be measured is the rear surface. In this situation, the optical axis is correctly aligned, but the axis points need to be reversed so that the program can correctly recognise which are the front and rear surfaces. Assuming the same model coordinate system as before, we'd need to set the flip_lens and flip_PCS_x_direction flags in order to match the coordinate systems.

rotation_analysis.py

The basic premise of this routine is to calculate the optical axes (OA, defined as a line connecting the two centres of lens curvature) and mechanical axes (MA, defined from the outer diameter of the lens itself) of lenses when mounted in their rings. To check for stability, these calculations are done for different rotations of the ring.

Obviously, the geometry of the measurements taken to ascertain the parameters that allow these two axes to be calculated will vary on the geometry of the setup. For ELT-PCS, we create a part coordinate system defined by the front surface of the lens ring, the outer diameter of the lens ring and an M6 thru' hole on the left side of the mount (looking face on). A line between the outer diameter of the ring and the thru' hole defines the x/y axes, with the front surface defining the z. Alongside this PCS, the program requires 4 elements to be measured:

1. An element of type SPHERE representing the front surface of the lens
2. An element of type SPHERE representing the rear surface of the lens
3. An element of type CIRCLE representing the mechanical front diameter of the lens
4. An element of type CIRCLE representing the mechanical rear diameter of the lens

With this, the program can generate plots and information in a variety of forms (pretty, json, raw) regarding the calcuation of the mechanical axis and optical axis respectively by invoking e.g.

python rotation_analysis.py -p2d -p -ma -oa -l lens_1

Caveats

• The z-origin of the PCS defined above is the "front" surface of the mount. For convenience, this is shifted so that z=0 corresponds to -mount_ring_thickness/2, i.e. z=0 roughly corresponds to the centre of the lens along the optical axis, but an error in this value will introduce an additional component to the decentre, the result of which means that even for a perfectly centred lens a constant decentre would be measured for all orientations.

• Tilts are defined after the coordinate system has been zeroed at the centre of the lens. If these tilts are used in other software (e.g. Zemax), the same definition of tilt will need to be applied (in Zemax, the tilt is done from the front surface, so you will need to move to a pivot at the centre of the lens prior to implementing the tilts defined here).