Zur deutschen Seite

Mechanical and Thermal Design

By Dick Jansson, WD4FAB

Table of Contents


Space satellites performing missions varying from (these days) the mundane to the exotic, all have one thing in common - the spaceframe structure. Despite all of the sophisticated technologies that may be involved in the payload packages contained in the spacecraft, the spaceframe technology is a well founded Mechanical Engineering practice based upon aircraft and airframe experiences. This does not mean that innovative designs and fabrications are not used, quite the contrary, they are essential. The designer is forced to employ methods that are sound, but less than conservative, as might be used for some terrestrial object.

The paramount consideration is, however, that of smoothly blending the mechanical design technology needs of the spaceframe with all of the other technologies that comprise the complete object, a successful satellite fulfilling its mission. While this presentation will deal mostly with the Mechanical Engineering aspects of the Phase 3D spacecraft, the observer must keep in mind that none of this is done out of the context of all of the technologies involved in the whole mission.

Spaceframe and Launch Adaptor

The Phase 3D electronic equipment must be housed in something. That something is the spaceframe. In the case of Phase 3D, those designing and building the structure were continuously learning new techniques for such fabrications and getting quite an education in the process. Along the way a number of quite good lessons on light-weight aircraft structural construction methods were learned. While the end product is not as light as some would have liked (about 60kg), it has already demonstrated itself to be very strong. The spaceframes for OSCARs 10 and 13 weighed only 7kg.

The Phase 3D spaceframe is principally fabricated of thin-gauge sheet aluminum. Its formation, to rather unusually close tolerances for sheet metal structures, caused more than passing concerns by all who were involved in the effort. Typically these tolerances are in the range of 0.2mm (0.008in.). The secret to this type of construction is to place all of the load stresses into the shear plane of the sheet metal, where it is notably strong for its weight. An example of this are the six Divider Panels, one on each corner of the spaceframe. Three of these will be anchored to the launch vehicle which will get Phase 3D into space. Thus, during launch, these three points will be quite heavily loaded in all motions. The only machined parts in the spaceframe are the six Corner Posts at the outer ends of the Divider Panels. These must be robust enough to carry all of these launch thrust loads into the spaceframe, translating all of those forces into the plane of the 0.8mm thick sheet metal Divider Panels as sheer forces. It is difficult to completely convey these concepts in words and pictures, but those seeing the spaceframe in-person will be able to more readily grasp the concepts employed in the design.

Working with panels of this type is not all that easy when it comes to mounting massive single elements, such as an 11.5kg battery assembly. As such mountings are done to the flat surface of a Divider Panel, the panel will flex and bend at the slightest of influences, such as vibration. Such a mounting really needs to have its natural resonance frequency greater than 100 Hz, but without any modifications, such a panel resonance is less than 20 Hz. We found that by using light-weight aluminum stiffeners, formed into the shape of a hat, that we could bring the panel resonance up to 38 Hz. By placing a thin aluminum panel across the tops of the several stringers, we formed an aluminum sandwich, much like a honeycomb panel, and the resonance of the battery mass went up to 128 Hz. All of this added structure was done on the assembled spaceframe (not an easy way to so such a task) and joined with both rivets and epoxy bonding agents. Co-bonding structures like this is a very messy process.

Mention was made, above, of panel resonance frequencies. We did such tests in a manner that is one of the "flagstones" of AMSAT practice. The method, learned and 'borrowed' from our German compatriots, employs a high fidelity amplifier (a 45 year old Ultra Linear Williamson amplifier) and a speaker held inside of the spacecraft propellant bay. By driving the amplifier with a sine wave signal generator, the frequency was swept over the range of interest. On thin panels a light object, like a screw washer, will dance with panel resonance. Massive objects, such as the simulated battery, will not respond as well and required another method. This time we used an old ceramic microphone element attached to the 'battery' and to observe the output signal on an oscilloscope. Resonance frequencies are readily observable with such simple instrumentation. This is AMSAT.


Since Phase 3D will be a secondary payload on an Ariane launch vehicle, it must conform to whatever space the European Space Agency (ESA) can make available. ESA has the conical 1194V Adaptor which interfaces between the (2624mm ((103.3inches) diameter bolt circle on the Ariane upper stage to a (1194mm clamp-band used for payloads. Although hollow, the conical adaptor does not provide sufficient space to house Phase 3D, or any other reasonably sized payload. Accordingly, ESA offered the amateur satellite community the opportunity to launch aboard the new Ariane 5 vehicle if we would provide a cylindrical "spacer" that could be mounted between the (2624mm diameter bolt circles on the bottom and the conical section on the top. Phase 3D could then ride to orbit inside this cylinder while a prime payload satellite would attach to the conical adaptor with the (1194mm clamp-band. Thus, they require that our cylindrical section must be able to support the launch loads of this prime payload fellow passenger. This means that this AMSAT provided (2624mm diameter Specific Bearing Structure (SBS) must be able to withstand the load forces imposed by a 4.7 T (metric ton, 10,350 lb.) satellite load. In order to assure ourselves that our design is capable of handling such a load, extensive structural Finite Element Analysis (FEA) computer work has been performed (on the same home computer used to accomplish the thermal analysis).

Construction of two of these SBS units has been completed in Utah, mainly at Weber State University. Each of the $10k machined flange rings, or Frames, on each end of the SBS cylinder were done in Florida, obviously on a very large lathe. These were made from the highest quality of ring-forged aircraft aluminum. The Frames were mounted to a purpose-built precision steel table, holding flatness and roundness values to less than 0.05mm (0.002in.), far better than ESA requires. The 4mm thick sheet aluminum rolled shell sections are attached to the Frames with a total of 976 special fasteners. A total of 2600 additional rivet fasteners are also used in each assembly.

Tests have been conducted at Weber State University to verify the operation of the mechanism to separate the spacecraft from the SBS, and the launch vehicle. For this purpose, a 500kg Mass Mockup Unit (MMU) was constructed to take the place of the P3D satellite. The Separation Nut units were operated using Nitrogen gas, rather than the pyrotechnically generated gas sources that will be used in flight. In this manner, we could reuse the Nuts over a number of tests. After a little travail at the start of the tests, all went very well over quite a number of repeated tests, with concurrence of the three Nut separations being within 3ms (0.003 second) of each other. Witnessing the exiting of a 500kg object from the SBS was impressive, to say the least.

Thermal Control Subsystem

Three axis stabilization satellite operation planned for Phase 3D drives the thermal design of the spacecraft. Most of the commercial communication satellites spin in order to keep any side from getting too hot and the opposite side too cold. Once in its desired orbit and orientation, Phase 3D will not spin, but will be oriented in three-dimensional space, with the antennas continually facing Earth. This continual attitude adjustment, with one side facing the sun, causes some interesting thermal design problems. The Phase 3D solution to overcome most of these problems is through the use of the four heat pipes. A heat pipe is a thermal linkage of very high conductivity consisting of a closed, evacuated tubular chamber with walls lined with a wick and partially filled with a pure fluid. The fluid used in Phase 3D is anhydrous ammonia. The fluid is vaporized at the hot end. The vapors then move through the hollow core of the tube, and condense at the cold end; from which the resulting liquid is returned through the wick to the hot end by capillary action. By this process, heat is transported from the hot to the cold end. Heat pipes typically offer heat transport characteristics that are many times greater than the heat transfer capability of the best heat conducting materials, while maintaining an essentially uniform temperature. The process requires no power and operates to its maximum performance in a zero-gravity environment.

In the case of the Phase 3D spacecraft, the internal ring-shaped heat pipes remove heat from one part of the spacecraft and re-distribute the energy to other parts where it is ultimately transported through the sides of the spacecraft and radiated to space - the ultimate "heat sink". What is felt to be a unique feature of this heat pipe system, as employed on Phase 3D, is that none of the pipes come in direct contact with space-facing panels. Instead they depend upon indirect re-radiation of the heat from internal equipment mounting panels to side panels that are deliberately allowed to become cold. All along, however, the electronic equipment modules maintain their desired temperatures because of the thermal influence of the heat pipe system, regardless of whether those modules are mounted on the solar heated side, or on the space-cold backside of the spacecraft. These design concepts put to practice the concepts of energy conservation, eliminating any need for active thermal control or supplemental heating.

The earlier Phase 3A, B and C satellites employed several multi-layer thermal insulation blankets to assist these spacecraft through the thermal rigors of spaceflight. Quite simply, such blankets are a first-class nuisance to fabricate, as the required assembly technology is very exacting. In the case of Phase 3D, the side panels of the spacecraft will be painted to provide the necessary radiative heat rejection. The top and bottom panels will be mostly solar energy absorbing metallic finishes of several different types, depending upon the location and desired temperatures of that section of the spacecraft. In general, the thermal design calls for the mean spacecraft temperatures to be between -5° and +20°C for the expected range of sun angles (() from -80° to +80°.

Extensive computer thermal analyses of the Phase 3D spacecraft have given us a very comfortable confidence that this design will provide the desired results, without the use of the kind of thermal blankets used on most other satellites. These thermal analytic computations were accomplished on a home computer to produce a series of temperature performance curves as functions of ( angle.

More Power Needed

One of the design features intended to make Phase 3D more accessible to smaller ground stations will be the use of higher power transmitters. This increased power carries with it another set of problems. First, the power must be generated. This means more solar array area. To achieve this, Phase 3D will employ four deployable solar panels in addition to the two mounted on the spaceframe. This will be the first use of deployable panels on an amateur satellite. Of course, deployable panels means mechanisms to initiate the unfolding plus appropriate hinges and latches to achieve the desired final configuration. This type of hinge is able to swing both ways but always return to the desired center position. One of the German members of the Phase 3D Design Team first suggested the use of this type of hinge, and actually obtained one at a hardware store to demonstrate the utility of the principle on the model of the former Phase 3D "Falcon" spacecraft design. As there is not the luxury of a lot of excess space around the current Phase 3D spacecraft when installed in the SBS, this hinge design had to go through several gestations in order to achieve the desired device in a compact manner. This effort included finding a spring wire able to withstand the metallurgical and thermal rigors of anticipated operation at temperatures as low as -100°C.

Power System

Since satellites must get their primary power from a the sun, the only practical means of obtaining power is the use of semiconductor solar panels. Generating the power needed to support the transmitters aboard Phase 3D requires large solar panels. The design which evolved calls for a total solar panel area of 4.46 m2 (48 ft2) and BSFR Silicon solar cells of 14.3% efficiency. This array will produce about 620 Watts of power at the beginning-of-life (BOL) and at optimum sun angle ((=0°). After 10 years in orbit, this power number will still be about 350 Watts at a (=45°). This amount of power is still sufficient to operate at least two transmitters and the other necessary spacecraft systems. Like almost anything else, solar arrays deteriorate with age. This is why their performance after a specified number of years is an important design consideration. The cells for Phase 3D are being obtained through a very attractive agreement with DASA, the German Space Agency. These cells are of US manufacture and were surplus inventory from a prior satellite program. While other sources and configurations of solar cells were considered, it was concluded that this one represents the best trade-off between performance and cost. Solar cells, and their assembled panels, represent one of the single highest cost items which go into building a spacecraft.

While solar panels are satisfactory as a sole source of power, some form of energy storage must also be provided. This is accomplished with a battery. Energy storage is necessary, not only to power the spacecraft during times that the sun is eclipsed by the earth, but also to operate the arc-jet thruster. It's power requirements exceed the capability of the solar arrays, even under the best of conditions. Actually the Phase 3D satellite will carry two batteries, a "main" and an "auxiliary". This is to provide redundancy in case of failure of the main battery. The Phase 3D design team evaluated several sources and types of batteries. A final decision was made to select a more or less conventional nickel-cadmium battery, albeit with a new plate design, as proposed by a German firm. Another contender was from a US firm which proposed the use of an assembly of Nickel-Metal Hydride cells for the main battery and a more conventional Nickel-Hydrogen stack for the auxiliary. As in the case of the solar cells, cost was an important factor in reaching this decision.

The Main Battery is composed of 20 cells of 40 Ahr capacity, for a 22-28 VDC supply. These rectangular cells are from a terrestrial application, but have been very well characterized for space service. They are contained in three subassemblies, two of seven cells and one of six cells. The seven cell assembly is 11.5kg mass and the respective assemblies will be mounted to reinforced Divider Panels, as previously discussed. The selection of the location of the lighter six cell subassembly will give us an option to use in achieving spacecraft balance.

For the Auxiliary Battery, a relatively new 10 Ahr cylindrical cell has been selected. A total of 40 of these cells will be mounted in two 20-cell parallel strings. The entire group is divided into four, ten cell subassemblies, mounted to the remaining three Divider Panels. As one panel will need to mount two of these subassemblies, we are again provided with a tool for spacecraft balancing.


Providing the 'containment' for all of the systems and experiments for the Phase 3D spacecraft has provided an interesting engineering experience for quite a number of us on the team. We are confident that this effort will translate into a very long lived mission for us to all enjoy in the communications afforded by Phase 3D.
Last updated: Feb 15, 1996
by Ralf Zimmermann,