🔭 astrophysics

Design and Testing of the Motorized 2-DoF Folding Mirror 1 for the VLT BlueMUSE Instrument

This paper presents the design, implementation, and testing of a motorized two-degree-of-freedom tip-tilt mount for the Folding Mirror 1 of the VLT BlueMUSE instrument, demonstrating that it successfully achieves the required precision and stability to correct environmental misalignments.

Original authors: Gloria Mellinand, Diane Chapuis Kerouanton, Malak Galal, Axel Nicolier, Aurélien Genin, Zeno Amann, Sébastien Pernecker, Rémi Giroud, Alexandre Jeanneau, Florence Laurent, Johan Richard, Jean-Paul Kne

Published 2026-06-18

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Gloria Mellinand, Diane Chapuis Kerouanton, Malak Galal, Axel Nicolier, Aurélien Genin, Zeno Amann, Sébastien Pernecker, Rémi Giroud, Alexandre Jeanneau, Florence Laurent, Johan Richard, Jean-Paul Kneib

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine the Very Large Telescope (VLT) as a giant, incredibly precise eye looking out into the universe. To see the faint, blue light of distant stars and galaxies, this eye needs a special new camera called BlueMUSE. Think of BlueMUSE as a high-tech prism that splits light into a rainbow to analyze what stars are made of.

However, light doesn't just travel in a straight line through this camera; it has to bounce off a series of 32 tiny mirrors (called Folding Mirrors) to fit into the instrument's compact shape. One specific mirror, Folding Mirror 1 (FM1), is the star of this story.

The Problem: The "Shaky Hand"

Even though the telescope sits in a desert with a stable climate, the temperature still changes between day and night. Just like a metal ruler expands when it gets hot and shrinks when it cools, the metal parts holding these mirrors warp slightly.

In the past, this warping was like a shaky hand holding a camera. The light would drift off-target by a few "arcseconds" (a tiny unit of angle, like the width of a hair seen from a football field away). To fix this, human operators had to wake up in the middle of the night, manually twist knobs, and realign the mirrors. This was slow, delicate, and took time away from actually taking pictures of the stars.

The Solution: The "Robot Finger"

The team at EPFL (a Swiss university) designed a motorized mount for this mirror. Instead of a human turning a knob, this mount uses tiny electric motors to act like a "robot finger" that can nudge the mirror with extreme precision.

Here is how the machine works, using a simple analogy:

The Gearbox: The motor spins, but if it spun the mirror directly, it would be too jerky. So, the team built a "gearbox" using a rack-and-pinion system (like the steering rack in a car).
The Ramp: The car's steering rack pushes a small cart up a very gentle, inclined ramp (a ceramic slope).
The Lever: A tiny ball bearing rolls up this ramp. Because the ramp is so steep and the lever arm is long, a tiny movement of the ball results in a very tiny, very precise tilt of the mirror.
The Result: This setup acts like a massive lever, turning a fast, rough motor spin into a slow, microscopic adjustment of the mirror.

The "Homework" (Testing)

Before building the final version, the team built a prototype out of stainless steel (like a kitchen sink) to test if their idea worked. They set up a lab experiment that looked like this:

They shone a laser beam at the mirror.
The beam bounced off the mirror, hit a fixed mirror, and bounced back to a high-tech sensor (an autocollimator) that acts like a super-accurate ruler for angles.
They put the whole setup in a plastic box and heated it up and cooled it down to simulate day and night temperature changes.

What They Found

Precision (Repeatability): The team wanted the mirror to land in the exact same spot every time they told it to move. The prototype was a huge success. It could hit the target with a precision of 0.25 to 0.5 arcseconds, which is far better than the required 2.5 arcseconds. It's like hitting a bullseye on a dartboard from across the room, every single time.
- How they did it: They used a "homing" trick. Before every move, the motor would touch a stop switch and back off slightly. This cleared out any "slop" or slack in the gears, ensuring the mirror always started from the exact same zero point.
Stability (The Heat Test): This was the tricky part. When they heated the prototype, the mirror moved more than they wanted (about 9.5 arcseconds instead of the allowed 0.5).
- The Twist: However, when they tested the empty setup without the mirror, it moved even more! This told them that the problem wasn't just the mirror mount; the test table itself was warping from the heat.
- The Fix: Computer simulations showed that if they switch the metal parts from stainless steel to Invar 36 (a special metal alloy that barely expands when heated, like a "temperature-proof" metal), the mirror will stay perfectly still.

The Bottom Line

The paper concludes that the mechanical design works perfectly. The "robot finger" can position the mirror with the incredible precision needed for the telescope. While the first prototype moved a bit too much when heated, the team knows exactly why (the materials and the test setup) and has a clear plan to fix it by using the special Invar metal and a better insulated test box.

In short, they have built the blueprint for a self-correcting mirror mount that will allow the VLT to take sharper, clearer pictures of the universe without needing a human to wake up and twist a knob in the middle of the night.

Technical Summary: Design and Testing of the Motorized 2-DoF Folding Mirror 1 for the VLT BlueMUSE Instrument

Problem Statement
The BlueMUSE instrument, a blue-optimized integral field spectrograph for the Very Large Telescope (VLT), requires high thermal stability to maintain optical performance. The instrument utilizes 32 folding mirrors (FMs) distributed across its relay optics. Specifically, Folding Mirror 1 (FM1) units provide tip and tilt adjustments (two rotational degrees of freedom) to correct beam misalignments. While the VLT site at Cerro Paranal has a stable climate, diurnal temperature variations induce thermo-mechanical deformations in precision mounts, leading to beam deviations on the order of arcseconds. The current MUSE instrument relies on manually adjusted FM mounts, requiring time-consuming and delicate recalibration by operators between exposures. The project requires a compact, automated, and thermally stable motorized mount for FM1 capable of arcsecond-level repeatability and stability (specifically < 0.5 arcsec stability under ±2°C variations) to eliminate manual intervention.

Methodology
The authors developed a custom motorized mount architecture and a dedicated experimental testbench to validate the design.

Mechanical Architecture: The design employs a high-reduction motion system to convert motor rotation into fine angular displacement. The mechanism utilizes a rack-and-pinion drive to generate linear motion, which actuates a high-stiffness ceramic slope (EKasic silicon carbide). A ball transfer unit contacts this slope, transmitting vertical displacement to a lever arm connected to the mirror support. This lever system converts linear motion into the required tip and tilt rotations ( $\Theta_{XF M}$ and $\Theta_{YF M}$ ). Finite Element Analysis (FEA) was conducted to compare stainless steel and Invar 36 for structural components, with Invar 36 selected for its low coefficient of thermal expansion to minimize thermally induced drift.
Control System: The system is controlled via a Beckhoff EtherCAT environment using DC motors, encoders, and limit switches. The control strategy operates in an open-loop manner but relies on a homing sequence (reversing past a limit switch) before every movement to eliminate backlash and mechanical flexure errors. A calibration look-up table is planned to correct for non-linearities in the mechanical transmission.
Experimental Setup: A stainless-steel prototype was manufactured and tested using an autocollimation-based optical testbench. The setup measures angular displacement by reflecting a beam twice off the FM1 prototype, multiplying the sensitivity (measured angle = $4\theta$ $4 θ$ ). Two distinct test modes were employed:
1. Repeatability: Measured by commanding the motor to a specific position after homing and comparing the result to a reference beam.
2. Stability: Measured by monitoring angular drift over time under ambient conditions and controlled thermal cycling (using heaters inside an acrylic enclosure) to simulate environmental variations.

Key Contributions

Custom Mechanical Design: The paper presents a novel, compact 2-DoF actuation mechanism using a rack-and-pinion, inclined ceramic slope, and lever arm system, achieving a theoretical transmission ratio of 0.153 arcsec/rad.
Material Selection Analysis: The study provides comparative FEA data demonstrating that replacing stainless steel with Invar 36 reduces predicted thermal angular variation from 4.8 arcsec to 0.17 arcsec, satisfying the project's strict thermal stability requirements.
Prototype Validation: The authors successfully manufactured and tested a first-generation prototype, characterizing its mechanical performance and identifying critical factors such as backlash, friction, and slope flexing.
Testbench Development: A specialized optical testbench was developed to measure arcsecond-level stability, including a differential beam measurement technique to decouple mirror motion from testbench instability.

Results

Repeatability: The prototype demonstrated excellent repeatability. The $\Theta_{XF M}$ axis achieved 0.5 arcsec repeatability, and the $\Theta_{YF M}$ axis achieved 0.25 arcsec. Both values are well within the 2.5 arcsec requirement, confirming that the homing strategy effectively mitigates backlash and elastic effects.
Ambient Stability: Under ambient laboratory conditions (approx. 1°C variation), the $\Theta_{XF M}$ axis remained stable within ~~1.5 arcsec over several days. The $\Theta_{YF M}$ axis showed higher sensitivity (~~6 arcsec), which the authors attribute largely to the testbench setup rather than the mechanism itself.
Thermal Stability: Initial thermal cycling tests (3.5°C variation) resulted in angular drifts exceeding the 0.5 arcsec requirement (9.5 arcsec for $\Theta_{XF M}$ ). However, the authors note that these results are preliminary. Control tests without the FM1 mechanism revealed that the optical testbench itself exhibited significant thermal sensitivity (drifts an order of magnitude higher than requirements). Consequently, the authors cannot yet definitively attribute the observed drift solely to the FM1 mechanism.

Significance and Claims
The paper claims that the proposed mechanical architecture is capable of meeting the BlueMUSE project's angular precision requirements, as evidenced by the successful repeatability tests. The results confirm that the design can achieve arcsecond-level positioning when operated with a consistent homing-based strategy.

Regarding thermal stability, the authors maintain a modest stance. While the stainless-steel prototype exhibited thermal drift exceeding specifications, FEA results suggest that the final design using Invar 36 will meet the 0.5 arcsec stability requirement. The authors emphasize that the initial thermal test results are inconclusive due to limitations in the testbench (thermal gradients, acrylic enclosure, and optical table sensitivity). Therefore, the paper concludes that while the repeatability performance is validated, the intrinsic thermal stability of the final design requires further characterization using an improved test setup (featuring a PEEK support plate, better insulation, and distributed temperature sensors) and the implementation of final materials (Invar 36 and Zerodur).