Figures 1, 2 and 3. Basics of photogrammetry
The study of shape deformations in large spacecraft structures such as the antennas of communication satellites and large space reflectors, is essential in the design of any space system. Thus, there is a great interest in accurately characterising the performance of the structures under operational conditions. A photogrammetry system can accurately monitor the structural deformation of these large objects under sun- induced thermal stress in space conditions. To do this, a large number of targets are affixed to the object under study and their motion is followed with a number of very high-precision optical photogrammetry cameras. An important feature of photogrammetry is that the measurements can be made remotely, i.e. without any physical contact with the specimen. The design and performance characteristics of a photogrammetry camera system, accommodated in canisters, to be used under space environmental conditions, are described here.
The ESA Test Centre, located at ESTEC in The Netherlands, is equipped with a series of environmental test facilities that allow a complete test campaign to be executed on the ESTEC site. Mechanical, electrical and thermal tests can be performed in sequence. A key element of the ESA Test Centre is the Large Space Simulator (LSS) which enables the performance of a variety of tests under vacuum, such as solar simulation, temperature cycling tests and deformation measurements.
The main features of the LSS test facility are:
Close-range photogrammetry is being used more and more in various applications such as aircraft manufacturing and surveying of telescopes, but it is being done in the open air. ESA has carried out a feasibility study to evaluate the use of a photogrammetry system in a vacuum, in the LSS. Such a system can provide a powerful tool for monitoring the structural deformations of large objects under sun-induced thermal stress and simulated space environmental conditions.
Photogrammetry can also be used to measure dimensions to verify compliance with design specifications (e.g. for deployable or inflatable objects). The selection of an adequate camera system is determined by the high accuracy requirement of one part in 10 to 5 of the diametre of the object to be tested. This demanding accuracy stems from the surface accuracy requirements for antenna reflectors..
The conclusions of the feasibility study are:
The essential advantages of a photogrammetric measurement, compared for instance with a theodolite measurement system, are:
Installing and operating cameras inside the LSS vacuum chamber, however, leads to basic requirements such as:
Based on the results of the feasibility study, ESTEC decided to procure and evaluate a photogrammetry camera and to develop a vacuum-proof canister to enclose the camera as a protection against the harsh environmental conditions present in the LSS during a testing operation. The camera model chosen is the Rollei R-Metrika.
The photogrammetric measuring process involves three essential phases:
Photographs of a targeted object are taken with one or more cameras from several directions. The photographic images of the object target array represent perspective projections of the three-dimensional scene onto a two-dimensional plane. The image coordinates (xy) of the targets in turn are used to reconstruct the coordinates of the marked points in object space (XYZ). The photogrammetric bundle triangulation is illustrated in Figure 1.
As shown in Figure 2, object point P is projected via the perspective centre O (camera lens) onto the image plane (P'). The position of the perspective centre O is defined in image space by the perpendicular distance ck between O and the image plane, the principal distance, and by the principal point H, where the camera axis meets the image plane. The position of the perspective centre O in image space, identified by ck and H, is called the interior orientation of the camera. The location of the perspective centre in object space and the orientation of the axes of image space with respect to those of object space define the exterior orientation of the photogrammetric camera.
The unknown object coordinates as well as the orientation parametres of the images are determined by least square adjustment, so that the bundles built by the object points and the perspective centres intersect with the measured image points in the best possible way. By treating the image space parametres as unknown, the recording cameras can be calibrated simultaneously. The accuracy and reliability of the triangulation process is greatly influenced by the number of photographs on which a target point is imaged. The configuration of camera stations, i.e. the photogrammetric network, is of even greater importance.
After the object has been photographed, the coordinates of the image points on the film are measured by means of an automatic film-reseau scanning process. A CCD-array sensor is used for digital automatic measurement of the image coordinates. The reseau grid is also measured to serve as a reference for the image coordinates (Fig. 3).
The digital image is stored in a real-time image processor board in the microcomputer. Using modern image-processing methods, the accuracy of each measuring point and reseau cross is better than 1 micro metre.
Figures 1, 2 and 3. Basics of photogrammetry
The design requirements for the photogrammetry system can be summarised as follows:
The overall concept of the camera system installed in the LSS is shown in Figure 4.
Figure 4. Camera system in the Large Space Simulator (LSS)
An overview of the photogrammetry system is given in Figure 5.
The camera-canister system consists of the following subsystems:
Camera system
The camera system is the combination of the camera body, its lens, a CCD viewfinder and the electronic
control unit. The camera features a vacuum pull-down system for flattening the film on a high-density cross
reseau grid. The reseau grid is imaged on the film and permits the correction of any film unflatness and
possible shrinkage (during development), a crucial step for precise measurements.
An annular flash is placed around the free end of the lens. The flash is powered from the outside. Its ignition circuit receives the command to fire via a dedicated line. The viewfinder, a small CCD television camera, is directly above the camera.
The electronic camera control box is attached to the canister's inner wall. Accessed from the outside via an RS232 link, this box controls the camera, for example, it takes a picture, advances the film, and sets the reseau exposure level. When interrogated, it relays back information on the camera's general status, such as the current shutter speed and the reseau exposure level.
Figure 5. Functional system diagram of the photogrametry system
Canister
The canister is a vacuum-proof aluminium cylinder protecting the camera against the harsh environmental conditions present in the
LSS during a test (Fig. 6). The main window assembly consists of two different quartz sections joined together using an optically opaque
glue, to avoid any light reflections from reaching the lens when the flash is operated. The main window is held in place by two O-ring
seals. The seals mechanically detach the window from the canister's front flange, accommodating any differential expansion or
contraction between the quartz and the aluminium. The view-finder window is built using a single quartz element.
All arriving electrical wires and air tubes enter the canister via the rotary feed-through located on the rear
flange.
Figure 6a. Camera canister assembly
Figure 6b. Camera canister assembly in HBF3
Drive assembly
The canister's roll-drive is governed by a stepper motor, located in its vacuum-proof housing, with warm
air arriving through an umbilical tube.
The motor assembly consists of a gear reduction box, a temperature sensor, a brake, two limit switches, two emergency limit switches and two arresting bolts. A drive chain connects mechanically the roll motor to the canister. When the motor is powered up, its brake is disengaged via a solenoid. Once the motor has reached the desired position, its power is cut off with the break re-engaging automatically. Emergency switches prevent the motor from moving any further when, due to a software failure, either of the final positions is overridden. Arresting bolts provide additional redundancy in case the emergency limit switches fail.
Thermal control
The need for temperature stability of the camera and its film drives the design of the air-conditioning unit:
the temperature at the camera body must remain constant, at close to 30 °C. Furthermore, the air-conditioning
system should ensure that no water condenses on either the lens or the window. An air-conditioning unit,
located outside the LSS, monitors the air temperature at the canister's inlet and adjusts accordingly the flux,
humidity and temperature of the supplied input airstream. An enclosure surrounds the camera; its purpose is
to improve the heat removal properties of the airflow and to ensure the temperature stability of the camera.
In order to avoid further condensation problems, both windows can be warmed using independent annular window heater sets. A dedicated control unit, located in the central control console, can command the window heaters to turn on. Temperature sensors are affixed to key items in the canister. Two are attached to each of the window's borders and one is located at the air inlet to monitor the temperature of the incoming airstream.
Control console
In addition to the air-conditioning unit, the following equipment is located outside the LSS:
All these elements are integrated in the control console. Commands can be sent from the computer to the roll-drive motor via the motor controller. The computer calculates the angular positions of the canister by counting the number of steps the motor has moved away from the last end position. The computer has access also to all the temperature sensors through the temperature monitor.
Through the RS232 multiplexer, the computer is able to address the camera s control unit to give commands or request information about the camera. It is also able to assure a correct synchronisation between the exposure and the flash.
In view of the relative complexity of the system and the interfaces with the LSS, it was decided to design a prototype of the canister and use it as a development model. The preliminary design requirements were drafted in close liaison with ESTEC's Test Operations Section, which is responsible for operating the LSS and the camera.
The camera was developed by the Engineering Section of ESTEC's Testing Division, in close liaison with Rollei and ESTEC experts in various specialised fields, such as optics, thermal control, safety and reliability. During the design process, various aspects had to be analysed. In the following, some significant topics are highlighted.
Camera window
One of the most critical elements of the canister is the camera window. In order to minimise the distortion,
quartz glass of the Homosil type was chosen. An overall analysis of the window has shown that the thermal
gradients in the window and the pressure gradient of 1 bar has no notable degrading effects on the image
quality.
Thermal control system
Because of the stringent requirements with respect to temperature stability and humidity inside the
canister, a detailed design analysis of the Thermal Control System (TCS) has been carried out. It has shown
that the TCS with an open air-loop will guarantee the required camera environment.
The main features of the TCS are:
In this concept, the requirement of the temperature stability will be fulfilled in the camera enclosure only.
Safety and reliability aspects
Failure Mode Effect and Criticality Analysis (FMECA) and hazard analyses are currently being carried out.
Some remarks, however, can already be made.
Safety aspects
From a safety point of view, the quartz window is one of the most critical elements. A pressure gradient
of 1 bar across the window in the high vacuum conditions of the LSS constitutes a hazardous situation. Failure
of the window under these circumstances could have catastrophic results. Therefore, preventive measures have
been taken to minimise the hazard:
Secondly, in case of window failure, the consequential damage is minimised by means of an automatic closure of the air supply and outlet. Furthermore, prior to any operation of the canisters in vacuum, a leak test on the canister will be a basic checkpoint in the operation procedures.
Reliability aspects
Wherever possible, off the shelf equipment of proven reliability has been chosen in order to achieve an
optimum reliability in a cost effective way. The air-conditioning unit will supply all four canisters and ensure the
thermal control of the cameras. From a system point of view, this sub-system constitutes a single point failure,
which may result in a failure of the test session. Therefore redundancies in the air-conditioning system will be
implemented.
In order to ensure the proper functioning of the camera/canister system, a test programme has been implemented, comprising a calibration in ambient, a functional test in thermal vacuum and a test of the overall performance in the LSS.
Calibration
The calibration of the camera/canister in ambient was successfully implemented by means of a calibration
reference object, provided by the German National Standards Laboratory (PTB).
Functional test under thermal vacuum conditions
The purpose of this test was to verify the proper functioning of the complete photogrammetry system under
thermal vacuum conditions. In particular, the thermal control system had to be tested. The main objective of
the thermal control system is to maintain an environment in the canister such that the temperature level of the
canister ambient lies between 0 °C and 40 °C and that the temperature shift is less than 1 °C per hour. Moreover,
there should be no condensation on any surface inside the canister.
The test has shown that the photogrammetry system meets all its design specifications and is able to work properly under thermal vacuum conditions. Temperature and humidity were monitored in some 17 locations. As an example, Figure 7 shows the temperature fluctuations of four relevant locations on the camera body. In particular, sensors T3 and T10 are located in the area of the reseau plate and T2 on the film container. All temperatures stay within the range of 29 33 °C. The temperature variations in time comply with the requirement of (Delta T) < 1 °C per hour.
Figure 7. Temperature fluctuations at four locations on the camera body (top) and of the thermal vacuum environment (bottom)
Verification of measuring performance under thermal vacuum in LSS
The purpose of this test was to verify the measuring performance of the photogrammetry system under
thermal vacuum conditions. In particular, the overall measuring accuracy had to be verified.
Since only one camera/canister is available, a photogrammetric network with several cameras had to be simulated by rotation of ±30° of the test structure around its vertical axis. The test structure is made of CFRP rods and is a flat, quadratic grid truss measuring 3 m x 3 m. Thestructure, which has a high specific stiffness and a very small thermal expansion coefficient (a<0.2 x 10 to -6 mm °C), was mounted on a turntable inside the LSS (Fig. 8).
The test has shown that no significant distortions occurred in the quartz window, due to the pressure difference of 1 bar, as the effect of the distortions is below the measurement accuracy of the system. The RMS standard deviation of the object coordinates of all measurements is better than 50 µm.
Figure 8. Test set-up to verify the photogrammetry systems measuring performance in a thermal vacuum
Deformation measurement of Artemis' antenna reflector
After the successful verification measurements in the LSS, a thermal distortion measurementunder vacuum
was carried out for the EM-model of the Artemis Common Large Reflector Dish, designed and manufactured
by CASA. Some 300 retro-reflective targets were fixed on the Dish, which has a size of 3.3 x 2.8 m (Fig. 9).
During the test, the dish was rotated around its vertical axis ±30° on the rotary table in the LSS. All planned
photos were taken with the dish in three different rotated positions. The measuring accuracies obtained (RMS
standard deviation) were 20 µm in the plane of the reflector and 45 µm perpendicular to it.
The performance tests in the LSS have shown that the photogrammetry system meets its design specifications well in thermal vacuum. With RMS standard deviations on average better than 40 µm in the direction of the optical axis and better than 30 µm perpendicular to it, high-precision deformation measurements in a thermal vacuum are possible. By carrying out photogrammetric network simulation for the actual object, even better accuracies can be achieved, which was proven using the Artemis reflector.
Figure 9. Artemis antenna reflector (left) during testing in the LSS
Single-camera system At present, with a single-camera configuration, measurements can be performed by rotating the test object around its vertical axis to obtain the different recording directions. The performance tests in the LSS have shown that such a single-camera system achieves sufficiently accurate results, in the order of one part in 10 to 5 , to meet the requirements for the deformation measurements of communication spacecraft antennae.
The single-camera system can be used in the LSS for large objects (Fig. 8) as well as in the smaller vacuum facility, HBF3 (Fig. 6b) to test objects measuring up to 3.5 metres.
The prototype camera/canister is being upgraded to a fully operational system, which will be ready for use in both facilities by the spring of 1995.
Multi-camera system In order to achieve even greater precision and to be able to perform deformation measurements during solar simulations, it is planned to install a multi-camera system in the LSS. Once such a system is in place, it will also be possible to monitor structural deformations of large spacecraft structures under thermal stress in space conditions.
A prototype canister system has been developed and built for a Rollei photogrammetry camera to be used in thermal vacuum. Functional and performance tests of the system have shown that the camera/canister system meets all of its requirements. The tests have shown that accuracies of on average better than 30 m can be achieved for an object with a diametre of 3 m. By carrying out a simulation of the actual object prior to the test, the photogrammetric network can be optimised and even better accuracies can be achieved, as was proven with the Artemis reflector.
At present, with a single-camera configuration, high-precision distortion measurements in a thermal vacuum can be performed by rotating the object around its vertical axis. The camera system can be used in both ESTEC s Large Space Simulator and in the medium-size HBF3 vacuum facility. Once a multi-camera system has been installed in the LSS, it will be possible to monitor dynamic structural deformations of large spacecraft structures under sun-induced thermal stresses in space conditions.
ESA Bulletin Nr. 81.