The Giant Magellan Telescope (GMT) primary mirror subsystem (M1S) consists of seven 8.4m diameter borosilicate primary mirrors that must be maintained at the ambient nighttime air temperature as it changes throughout the observing night to prevent seeing effects at the mirror surface. Additionally, thermal gradients internal to the mirrors must be minimized to prevent figure errors caused by distortions of the mirror due to the non-zero thermal expansion coefficient of the glass. To address these requirements, the GMT M1S team is fabricating a prototype thermal control system design that consists of a sub-critical refrigeration system utilizing high pressure (~30 to ~60 bar) CO2 (R744) refrigerant. This paper describes the design and status of the M1 Subsystem Thermal Control (M1STC).
GMTO has developed a full-scale prototype of the cell that can house an 8.4-meter borosilicate mirror. This test cell is populated with all the active support control hardware and a mass simulator that simulates the mirror weight and the moment of inertia. GMTO has implemented the control software with all the core features needed to operate the active support system. A series of tests have been carried out to verify the functions, performance, and safety of the active support control system. The tests were carried out at several different orientations of the cell to demonstrate that the active support system works with the changing zenith angle and location of the mirror on the telescope mount. This paper describes the results of important safety and dynamic response tests of the active support system.
KEYWORDS: Software development, Control systems, Design, Software engineering, Observatories, Telescopes, Industry, Project management, Control systems design, Control software
The Giant Magellan Telescope (GMT) Software and Controls (SWC) team is responsible for designing, implementing, and maintaining the GMT Observatory Control System (OCS). GMT software modules are developed either in-house, or in collaboration with GMT partner institutions, following an Agile software development process. However, these software industry best-practices require significant tailoring to integrate well with other Engineering disciplines on a large, complex project such as GMT. In this paper we explore the various challenges in managing software development and how we are tackling them at GMT. Key areas include building the right team, handling programmatic challenges, streamlining development processes and engaging with customers and stakeholders. We’ve learned that people are at the heart of what we do, and the health of the team directly affects our ability to deliver high quality software on time and within budget. Also, managing limited resources is a common theme, requiring many different solutions in different domains. We have found the most effective to be a combination of process-optimization, resource-loaded scheduling, agile development, drastic overhead reduction and regular review of top priorities to help the team focus on what is important. Lastly, active engagement and efficient communication with customers and other stakeholders from the very beginning, help to set clear expectations and sets the team up for success. The team has made tremendous progress in the last few years in these areas and will continue to do so in the future due to a commitment to continuous improvement.
The Giant Magellan Telescope (GMT) is a next-generation ground-based segmented telescope. In the last few years, significant progress has been made by the GMT team and partners to design a natural guide-star wavefront control strategy that can reliably correct wavefront error, including the discrete piston aberration between segment gaps. After an extensive set of simulations and external reviews, the team proposed a design of a Pyramidal Wavefront Sensor (PWFS) combined with a Holographic Dispersed Fringe Sensor (HDFS) and started building a prototype for integrating a GMT simulator (High Contrast AO Testbed) with a PWFS and an HDFS. The prototype was developed in collaboration with the University of Arizona, INAF-Arcetri, and the GMT observatory. The software development of the adaptive optics controllers and the interfaces between all testbed components were done using the GMT software frameworks, as they will be implemented for the final observatory software. The GMT framework is model-based, and the software component interfaces are defined using a domain-specific language (DSL). In this paper, we show how the design of the testbed software fits within GMT's component-based architecture and what each partner was responsible for delivering. We discuss the challenge of a multidisciplinary team from multiple institutions in different time zones working together on the same software, describe how the software architecture and development process helped to ensure seamless integration and highlight other accomplishments and lessons learned.
The Giant Magellan Telescope (GMT) is a complex observatory with thirty major subsystems, many low-level subsystems, components, external contracts, and interfaces. Almost all subsystems require software and controls to operate. An important goal for GMT is to have software and control subsystems that are easy to develop, test, integrate, operate, and maintain. To provide consistency across all controlled subsystems, a set of standards and a reference architecture are provided. Software components are specified using a Domain Specific Language (DSL), which enables code-generation in several languages and automatic validation of architectural conformance and interfaces. Some of the main observatory control subsystems have already been modeled using this approach, and initial implementations are currently being tested. The most advanced control subsystem is the primary mirror Device Control System (M1 DCS), which is currently under testing before the integration of the optical mirror in the test cell. This paper describes the status of the GMT control system, the main lessons learned, and the future steps in the development of the GMT control system.
The Vera C. Rubin Observatory is currently under construction on Cerro Pachón, in Chile. It was designed to conduct a 10-year multi-band survey of the southern sky with frequent re-visits (with both an intra- and extra-night cadence) to identify transient and moving objects. The mirror cell assembly was designed in Tucson, Arizona by the Rubin Observatory engineering department, and was tested twice in Tucson. The first testing campaign was performed at CAID industries, where the mirror cell was fabricated, using a steel mirror surrogate that has the same geometry and mass of the glass mirror2,4. The glass mirror is a single monolith that contains both the primary and tertiary mirrors on a single substrate. The testing results confirmed that the mirror support system was performing within the design specifications, and that it was safe to install the glass mirror on the cell. The second test campaign was performed at the Richard F. Caris Mirror Lab of the University of Arizona using the actual glass mirror16. This test campaign was performed under the test tower, which contains a vibration insensitive interferometer to measure mirror figure. This confirmed the mirror support system could achieve proper optical surface figure control for both primary and tertiary mirrors. After successful test campaigns at CAID, and the mirror Lab, the mirror cell assembly was disassembled, packed and shipped to the Rubin Observatory site at the Cerro Pachón summit in Chile. Upon arrival, the mirror cell has been integrated with the mirror surrogate once again to perform the third test campaign that confirmed the system has arrived safe and operational to the summit. This integrated system will be tested on the telescope mount assembly to verify that it still meets it requirements under the effects of variations in gravitational orientation, and dynamic (slewing) loads.
The Vera C. Rubin Observatory Primary Tertiary Mirror (M1M3), together with the fully-assembled mirror support system, underwent two optical testing campaigns at the University of Arizona Richard F. Caris Mirror Lab. The objectives of the testing campaigns were: (1) optimizing the M1M3 surfaces with support forces, and (2) characterizing how the surfaces respond to actuator forces, including measuring the bending modes and single actuator influence functions. Both objectives were successfully achieved. The differences between the measured bending modes and the Finite Element Analysis (FEA) predicted modes were shown to be less than a few percent. The surface optimizations routinely resulted in Root-Mean-Square (RMS) surface errors below 30 nm for M1 and M3, simultaneously. The entire system was shown to be robust and repeatable. In this paper, we present the results of the optical testing and the analyses performed using the data acquired.
The Large Synoptic Survey Telescope is an 8.4m telescope now in construction on Cerro Pachón, in Chile. This telescope is designed to conduct a 10-year survey of the southern sky in which it will map the entire night sky every few nights. In order to achieve this goal, the telescope mount has been designed to achieve high accelerations that will allow the system to change the observing field in just 2 seconds. These rapid slews will subject the M1M3 mirror to high inertial and changing gravitational forces that has to be actively compensated for in order to keep the mirror safe, aligned, and properly figured during operations. The LSST M1M3 active support system is composed of six “hard point” actuators and 156 pneumatic actuators. The hard points define the mirror position in the mirror cell (with little or no applied force) and hold that position while observing in order to maintain the alignment of the telescope optics. The pneumatic actuators provide the force-distributed mirror support plus a known (static) figure correction as well as dynamic optical figure optimizations coming from other components of the Active Optics System. Optimizing this mirror support system required the introduction of innovative control concepts in the control loops (Inner and Outer). The Inner Loop involves an extensive pressure control loop to ensure precise force feedback for each pneumatic actuator while the Outer Loop includes telescope motion sensors to provide the real-time feedback to compensate for the changing external inertial and gravitational forces. These optimizations allow the mirror support system to maximize the hard point force-offloading while keeping the glass safe when slewing and during seismic events.
The Large Synoptic Survey Telescope (LSST) primary/tertiary (M1M3) mirror cell is a 25-ton, 9-meter x 9-meter x 2- meter steel weldment that supports the 19-ton borosilicate M1M3 monolith mirror on the telescope and acts as the lower vessel of the coating chamber when optically coating the mirror surfaces. The M1M3 telescope mirror cell contract was awarded to CAID Industries, Inc., of Tucson, Arizona in October 2015. After the mirror cell final acceptance in October 2017, the integration of the mirror support system started. The M1M3 cell assembly with the surrogate mirror will take place in a dedicated controlled-environment area at CAID Industries. All components of the mirror support system that were developed and tested by the LSST Telescope and Site M1M3 team at the NOAO offices in Tucson have been moved to CAID premises and have been integrated into the cell by a team of LSST, CAID and Richard F. Caris Mirror lab personnel. After completion of the cell integration and its assembly with the surrogate, a test phase that includes zenith and offzenith testing for the mirror support system will be carried by the LSST team. These tests aim to verify that the active support system components, mirror control, and software are performing as expected and the mirror support system is safe for the next step, the M1M3 cell to borosilicate glass assembly and tests at the RFC Mirror Lab of the University of Arizona.
The Large Synoptic Survey Telescope (LSST) Telescope and Site software team has adapted a previously described component template (SysML/UML model and code) to accommodate the project’s selected middleware (Service Abstraction Layer using the Data Distribution Service), message classification scheme, top-level state machine, operating system, command response paradigm, and extended settings requirements. The extended implementation easily accommodates extension for the use of any publish-subscribe protocol and isolates this behavior to make it easier to use. The revised component template remains a complete working application that developers extend in a precise manner to add application-specific behaviors. We report on the progress made designing and developing system components using the template and its application in the project workflow.
Construction of the Large Synoptic Survey Telescope system involves several different organizations, a situation that poses many challenges at the time of the software integration of the components. To ensure commonality for the purposes of usability, maintainability, and robustness, the LSST software teams have agreed to the following for system software components: a summary state machine, a manner of managing settings, a flexible solution to specify controller/controllee relationships reliably as needed, and a paradigm for responding to and communicating alarms. This paper describes these agreed solutions and the factors that motivated these.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.