# What are these very large, square panels on Inmarsat 5?

Question: What are these very large square panels on Inmarsat 5's? They seem to each have an extremely reflective mirror-like flat/smooth tiled-appearing side, and a nearly black side with a lot of tubing and/or wiring and connections. Also they are oriented oppositely; the shiny side of one is facing the same direction as the black side of the other.

If I understand correctly, the dishes all point Earthward (nadir), and the long solar panel axis should be North-South so that they can rotate and receive fairly constant illumination, except during eclipses (similarly to the rotation described in this currently-unanswered question). So that means that these square panels I'm asking about are roughly co-planar with the Earth's equator, and so are roughly edge-on to both the Earth and the Sun.

above: Cropped, from this found at Defense Talk.

. .

above: Cropped sections from image below.

above: Inmarsat 5-F1 (F4 Identical) – Photo: Boeing. From Spaceflight 101

This screen shot from the video EUTELSAT 8 West B - Testing Solar Array Deployment shows a different satellite but with a very similar looking shiny tiled-appearing surface (on the left). Perhaps just a coincidence?

• Looks a lot like radiators to get rid of the estimated 10 kW of power the sat produces. Having them orthogonal to Sun and Earth is the best you can do to avoid additional heating by external sources. – asdfex May 15 '17 at 18:01
• Yes, knowing nothing about them, sounds like radiators. – Organic Marble May 15 '17 at 18:41
• @asdfex can you find another picture of a known radiator that looks like this - shiny on one side and black on the other? Also, these look more like wires than hoses to me. – uhoh May 15 '17 at 18:42
• The black side is facing the two side antenna arrays - a metallic surface could make some problems with reflections of signals. Also, almost half of the hemisphere facing the black side is covered by the satellite, a radiating surface would not work well because a large part of the radiation hits the satellite again – asdfex May 15 '17 at 19:05
• Could they be shunt resistors to dissipate excess current in the solar arrays? See all of the answers to the question When there is more power than needed, are satellite photovoltaics shunted or open circuit? – uhoh May 15 '17 at 20:14

The large square panels are deployable radiators. They provide extra heat rejection capability in addition to the fixed radiators that can also be seen in the photographs.

Context part 1 - Thermal design for Geo Comsats

Most geostationary communications satellites are roughly cuboid in shape with the solar arrays and antennas as appendages. The basic thermal design of a satellite of this type is to put insulating blankets on the four sides that face the sun directly during the normal day. In contrast the "North" and "South" panels are usually made of aluminium honeycomb, they are painted black on the inside and have second surface mirrors on the outside. The bulk of the heat dumped by the satelite's electronics then takes this slightly easier route to deep space than going through the blankets. The honeycomb panels usually have embedded or bolted on heat pipes in the inside face in order to spread heat as evenly as possible through the panel so as to prevent local hotspots. The orientation of the satellite to the Sun changes over the year and so at sometimes the sun can be up to 23.5 deg from edge-on. The mirrors help by rejecting optical sunlight.

Context part 2 - Second Surface Mirrors

A second surface mirror, also called an optical solar reflector (OSR), is similar in concept to an ordinary domestic mirrors. The mirror comprises a thin layer of quartz or glass with an aluminized rear surface. This allows them to have good optical reflectivity, $\epsilon_{vis}<0.1$, and high IR emissivity, $\epsilon_{IR}>=0.8$, (as the outer surface is glass).

The cuboid shape isn't universal but it is convenient for structural design and packaging into the launch vehicle fairing whilst providing some heat rejection capability. This leads to a potential problem with particularly high power satellites. As far as I am aware the Boeing 702 platform is the only one in regular production that has taken the design solution of creating additional radiator aread for cooling by deploying panels once in-orbit. Clearly, if the heat pipes could spread heat over both of the additional panels, doubling the overall area, then it would greatly increase the possible DC power of the satellite by raising the heat rejection limits. The photographs aren't especially clear though it appears that some of the hardware showing on the reverse, black painted, side of the panel could be a network of heat pipes for this purpose.

Moving further into assumption territory, I assume that the reason that the deployed panels are shown black on one side is to keep a modest environment for the antenna related equipment attached to the East-West faces.

One feature that isn't so clear to me is whether the panels are left fixed after deployment. Clearly there would be a twice daily opportunity to adjust the angle of the radiators so that they shadow the main body panel though my outsider's view is that it may not be worth the extra complexity.

Context part 3 - Heat pipes

Of the variety of things going by the name of "heat pipe" I'm only aware of two types used commonly in satellites:

i) rigid narrow tubes lined with narrow grooves. These wick the working fluid in either direction and it will even out hot spots where ever they form. Their capacity is often limited and they can be zero-g only. The grooves have to be present all the way along the tube to prevent drying out.

ii) loop heat pipes. These have a preferred direction of flow and I think use a slighty different thermodynamic principle (not my territory, that's as far as I'm going!). As I understand it they have a much higher pumping capacity than the normal type and that this enables it to incorporate non-grooved flexible sections. Assuming I have got it right, this latter feature makes them suitable for deployable radiators.

As an aside, in the description of heat pipes I have excluded mechanically pumped fluid loops.

• This is a great answer, thank you very much for putting in the time and effort to explain so clearly. – uhoh Apr 14 '18 at 11:11
• I've put a reward bounty on this answer and for parity I'll put one here as well (I'm maxed out at 3 bounties right now). I think the whole topic of heat management is fascinating, thanks for your thorough answer! – uhoh Jun 16 '18 at 2:52
• Thanks for taking the time to dig deeper and generate a broader answer that teaches more about a given subject! – uhoh Jun 16 '18 at 9:43
• @uhoh thank you for the show of appreciation! – Puffin Jun 18 '18 at 19:49

This is a partial answer. According to the video animation Inmarsat-5 F4 Satellite Launch Simulation they are radiator panels. However the question still needs to answered how they work. Are they radiating electrical power dissipated by resistors, or cooling off circulated coolants? Why are they not black on both sides so that they radiate twice as effectively?

By the way, the launch simulation is really interesting and worth watching!

• Black would be far more absorptive - not heating up due to solar irradiation is more important than being more emissive ISS radiators are white. The wiring inside suggests thermoelectric heat transfer (peltier cells); the objects sticking below at 2:51 are definitely simple emissive radiators - likely on condensation tubes for coolant. – SF. May 17 '17 at 12:07
• Radiators such as these are known as second surface mirrors, i.e. they are just like ordinary domestic mirrors. This allows them to have good optical reflectivity <0.1 and high IR emissivity >=0.8 as the outer surface is glass. – Puffin May 17 '17 at 21:09
• Also, "edge-on" only applies at equinox. At the solstice the there is 23.5 degrees between the plane of the mirror and sunlight. Hence the thermal design of geostationary comsats accounts for summer, winter and equinoxes. – Puffin May 18 '17 at 17:25
• I have assumed to date that the pivot is only for deployment to a position where they are in the same plane as the main body mounted radiator face. If the side panels could be inclined both towards and away from the East-West faces then twice each day (in 6 hour arcs) there would be an opportunity for a lower heat load though I'd be very surprised if this is the actual design. – Puffin May 23 '17 at 22:14
• @SF. Something working once is not the same as it working flawlessly 5,475 $\times$ 2 = 10,950 more times, plus have you considered all of the extra work the ADCS system needs to do twice a day for 15 years. Better to leave it alone, considering the cost and lead time to replace it. This is not a phased array satellite, those 89 individual beams are aligned to the surface of the earth, if the spacecraft rotates, everything is screwed up. Please don't speculate (or just say) that it is "robust enough" for that also. – uhoh May 24 '17 at 9:04