trizjournal | On 15, Jun 2002
Abstract. The article analyzes the failure of the Russian project ‘Znamya 2’, which is better known as ‘Space Mirror’. When unfolding, the space mirror membrane caught on the antenna that projected over the space station body.
Using TRIZ, the causes of the failure were analyzed and a number of solutions for improving the reliability of the flexible membrane unfolding under weightlessness were proposed.
Introduction. The space mirror that reflects the Sun’s rays onto the nightside of our planet is one of the impressive space projects. In 1993, the spaceship ‘Progress M-15’ placed into orbit a 20-meter film mirror (the project ‘Znamya 2’). The mirror unfurled and produced a light spot that was equal in strength approximately to one full moon. A huge plash of sunlight glanced over beclouded Europe to be seen only by astronomers on the top of the Alps.
The project ‘Znamya 2.5’ stood head and shoulders above its predecessor. The mirror was expected to be perceived on the Earth as 5 to 10 full moons and it formed a trace of about 7 km in diameter which could be controlled by fixing it on one spot for a long time. The space mirror was a slightly concave membrane of 25 m in diameter made of thin film with a mirror surface, which was attached around the periphery of the station. The membrane was expected to unfold and be held unfolded by centrifugal forces (Fig.1).
However, the project was a failure. Soon after the deployment started, the membrane caught on the antenna (Fig. 2). The spaceship ‘Progress M-40 was de-orbited and buried in the ocean.
It would be interesting to consider this situation in terms of TRIZ. Why exactly hasn’t the membrane unfolded? Has everything been done for it to unfold? What can be done to make such membranes unfold under various conditions? There is a great many of similar constructions and it is not always possible to provide such unfolding conditions as for the solar mirror.
Was this failure natural? In principle, from the point of view of traditional design everything is OK. The membrane’s prototype is a parachute canopy. There is no oncoming air stream in space, so inertia forces that occur as a result of the space station rotation are used to unfold the membrane. The first membrane – that of ‘Znamya 2’ – unfolded successfully. Besides, there is a good analogue – designs of spaceships with a solar sail in which the sail is unfolded and held by a centrifugal force occurring during the spaceship rotation. It also seems correct in the context of TRIZ. The basic requirement of ideality is satisfied – the membrane itself without any additional devices moves off the station, unfolds and is held unfolded.
This, however, has not worked. The station started rotating, the membrane began moving off it and caught on the antenna. We can certainly blame the antenna. Parachute-jumpers thoroughly remove all buckles, hooks and heelplates – everything on which the parachute may caught while deploying – from their clothing and especially from the helmets and footwear. But the space station is an expensive and serious thing and is designed to solve many different problems during one flight. It is difficult to remove all devices from the station’s surface, as they are needed for solving other problems. Under such conditions, one accidental jerk of the engine can cause inertia to throw the unfolding membrane back onto the antenna.
Rational membrane-unfolding technology. Let us try to solve this problem with the aid of TRIZ and find a method for unfolding the membrane without removing the antennas and other objects projecting over the station’s surface. Unfortunately, we’ll only have to be guided by the published information, so we’ll have to think ourselves of such important ‘trifles’ as the design of membrane-holding devices, locks, etc. Or, for instance, what are the radial slits in the membrane for?
Thus, we have a packed membrane attached around the periphery of the station. It is held, for instance, by several flexible bands with locks (Fig. 3, a).
Let us recall one of the basic TRIZ laws – the law of coordination of ES parts’ interaction – and try to build a ‘desirable technology’ of membrane deployment. The deployment process falls into two distinct stages. At the first stage, the station starts rotating, locks are released and the membrane moves off the station’s surface (Fig. 3,d). The second stage includes the unfolding of the membrane, its final stretching and holding in the unfolded state. (Fig.3,c).
The first stage is apparently most complex. Since the entire mass of the packed membrane is situated near the axis of rotation, centrifugal forces are comparatively small at that moment, while the work to be done by them is considerable. At the same time, certain requirements are also imposed on the sequence of membrane deployment. It would be better for the membrane not to unfold at once throughout the entire volume, but from the center toward the edges, quickly receding from the station to the area that is free from projecting antennas, in the form of a compact torus-like packing and not as an inordinate heap.
Further stretching of the unfolded membrane is not connected with any troubles because at that time the membrane is already far from the ‘human factor’. Here we can do with centrifugal forces.
Thus, we can formulate an important requirement for the deployment process: ‘the unfolded internal part of the membrane must be permanently stretched’.
How can we accomplish sequential unfolding of the membrane? The chances are that it is ‘concertinaed’ before the flight and then it is unfolded throughout the entire volume after releasing the locks.
Proposal 1. Leave free some of the folds adjoining the station and fasten the rest of the folds together with destroyable filaments or bands. Then peripheral tensile stress will first spread the folds near the station surface and the part of the membrane adjoining the station will be stretched. And only then the ties of the peripheral part of the membrane will break and the final deployment will start (Fig.4).
How can we accelerate the membrane deployment at the first stage?
It is obvious that we can either increase the station spinning frequency or increase the membrane mass. But carrying an extra load is not an ideal solution. Suppose we first spin up the station and then release the locks’ And suppose we place a ring charge BB under the packed membrane, then release the membrane from the locks and explode the charge! Or place small jet engines around the external periphery of the membrane’
Something will definitely fly to pieces, because the membrane is only 7 mm thick.
What are the requirements for the membrane mass in this case?
On the one hand, the membrane must be heavy to be able to unfold fast and to move off the station.
At the same time, the membrane must be lightweight to prevent the occurrence of excessive loads produced by centrifugal forces.
We can resolve this contradiction in the following way. Let the membrane be heavy at the first stage of deployment and lightweight during the rest of the time. Or rather, the unfolding force must be great at the first stage. Then it may be much smaller, but great enough to unfold and stretch the membrane.
What resources are available?
Substances: membrane, space station, vacuum, air that remained inside the membrane after packing (if it is not evacuated while packing).
Fields: electric, magnetic, Sun’s light field, heat field, space wind. The characteristics of most fields may vary within very wide ranges by turning the station relative to the Sun.
Time: unlimited before the conflict, including the time of space station preparation on the Earth.
Let us draw an abstract model of the problem (Fig. 5).
B2 is a station. It is a tool. B1 is a packed membrane. It is an object. B2 acts on B1 by the mechanical field (Fineria), produced by the inertia forces. This action is insufficient.
What do the standards recommend (1)?
Introduce one more substance and a field to control this substance.
For instance, add ferromagnetic particles while manufacturing the membrane. Then, while unfolding, the membrane may be acted upon by an annular electromagnet placed around the station body’s periphery. However, it is necessary to additionally introduce an electromagnet, which reduces the system’s ideality.
Add one more substance to one of the interacting substances.
B1 is a station. A space station is a large vessel with air. Consequently, the second substance is already available. It only remains to use it.
Proposal 2. We can provide rapid withdrawal of the membrane from the station to the distance of the non-fastened folds by using the energy of compressed air supplied under the packed membrane. In this case, as distinct from the explosion of a ring charge, the action value can be easily controlled (Fig. 6).
Transfer to a capillary-porous substance.
Make the membrane capillary-porous.
Proposal 3. For instance, paste thin pipes in the mirror membrane cloth and supply compressed air into them while unfolding. The membrane will change into a plane practically instantly and with a great force.
Eliminating possible catching. As for possible hooking of the membrane during the deployment, the following contradiction occurs.
The membrane must not catch on a hook.
Under the deployment conditions, the membrane can contact a hook.
Of course, it would be ideal to remove the conditions that cause the membrane catching. But if we cannot eliminate such a probability, it is necessary to make so that the caught membrane can be easily released. How can we resolve this contradiction?
Firstly, we can do this in time.
At the first stage of deployment, when the catching is most probable, the unfolding force should be large enough to secure tearing a piece out of the membrane cloth. The peace should be big enough only for normal unfurling and stretching of the unfurled membrane.
This is achieved by introducing additional forces at the first stage of deployment (Proposals 2 and 3) to help the peripheral centrifugal force.
Secondly, we can do this in space.
For this purpose we use a resource: the membrane material.
We can formulate the following requirement for this material. Let the membrane have rigid surface in close proximity to the station and let the surface become flexible and elastic as the membrane moves off the station.
Proposal 4. The peripheral part of the membrane moves off the station in the form of compact torus-like packing. The packed membrane may be placed inside a rigid thin-walled body with an open bottom (Fig. 7). Because during the deployment the circumference of the packed membrane constantly increases, the body must be corrugated and must extend like accordion. Or it must be composed of several sections that move relative to one another. The use of such a body provides a good protection of the packed membrane against hooking and allows for preserving its compact form without bonding or binding.
The membrane is likely to have equal strength in all directions. This is quite correct for that of its parts, which is situated comparatively far from the station. At the same time, the requirements for the central part of the membrane are contradictory:
On the one hand, the membrane must not be durable so that it can tear and unhook.
At the same time, the membrane must be durable so that is does not fail while unfolding and in a working position.
To satisfy these contradictory requirements, we can structure the material of the central part of the membrane.
Radial filaments must be comparatively durable to provide reliable attachment of the membrane to the station body. Tangential filaments that are perpendicular to those radial must not be durable so that they easily break in case of hooking. For peripheral forces not to break the membrane at the interface of its central and peripheral parts, a ring band can be used here for reinforcement.
In principle, since the area of the membrane’s central part is comparatively small, tangential filaments may be removed at all.
Proposal 5. Make the membrane in the form of a wide ring attached to the station body with bands or cables. In packing, such bands are easy to ‘concertina’ like shroud lines. Each folded band can be put into a covering attached to the membrane. This will ensure well-ordered unfolding of the central part of the membrane and will eliminate hooking (Fig. 8).
Final solution concept. To build a final solution concept, let us take proposal 3 as a basis that provides energetic withdrawal of the membrane from the station as envisaged in our ‘desirable technology of membrane deployment’. This proposal is well supplemented by proposal 1 that ensures the membrane moving off the station in a compact packed state and by proposal 5 that excludes the catching of the middle part of the membrane on the projecting parts of the station’s surface. In this case proposal 3 and proposal 5 are just combined. As for proposal 1, we only use the function ‘to hold the membrane packed’ which is performed with the aid of resources available in the system (pipes with compressed air).
To provide fast and energetic unfolding of the middle part of the membrane, it is necessary to use, instead of a one-piece film, thin radially positioned pipes that connect the packed peripheral part of the membrane with the station. By supplying compressed air into the pipes we’ll practically instantly obtain a rigid frame that will ensure energetic unfolding of the membrane at the first stage.
To hold the membrane in a compact packed state until the full unfolding of its middle part, it is proposed to make each of the membrane-holding clamps in the form of two pipes that encompass the membrane on its sides. The pipes’ end faces should be closed and connected with the rest of the frame. The pipes’ ends should be connected with each other with a break-off band or filament. After the full unfolding of the frame of the membrane’s middle part, compressed air starts going into the clamps’ pipes. The pipes straighten with a great force, break the bands and release the membrane (Fig. 9).
Proposal 2 is partially realized by using compressed air.
Intensified solution. If it is still impossible to avoid the contact of the membrane, which moves off the station, with hooks, it is expedient to additionally use proposal 4.
The packed membrane is placed in a rigid multi-section thin-walled body, which is discarded after complete deployment of the membrane.
Conclusion. No doubt that using only open sources of information we missed out many circumstances. But I hope that this attempt to analyze the problem will be to a certain extent useful to space mirror designers, and not only to them. We are sure that all the errors will be corrected and after the next attempt we’ll see a hand-made sun in the night sky.
- Altshuller, Genrikh. ‘The Innovative Algorithm. TRIZ, Systematic Innovation and Technical Creativity.’ Technical Innovation Center, INC. Worcester, MA. 1999.
The pictures by Elena Novitskaja.