The upcoming SpaceX launch with the apparently first re-use of a dragon capsule will bring with it the Roll Out Solar Array (ROSA).

(See also: More science, reused Dragon capsule featured on CRS-11)

How do the photovoltaic cells roll up tightly and nicely, and then unroll without any problems with the connections to the silicon? Are these basically small, rigid photovoltaic cells and standard, soldered ohmic contacts, but flexible PC board-like interconnects, or are they made from flexible semiconductor material to begin with?

While I have seen various flexible photovoltaic technology demonstrations in photos and articles, these are mostly R&D and PR, rather than space-rated reliable.

All I know so far is that ROSA has a "innovative flexible blanket!"

above: "A large version of the scalable ROSA technology undergoes testing in 2014." Photo Credit: Deployable Space Systems. From here.

  • 1
    $\begingroup$ One of the PIs on this payload answered some questions over at nasaspaceflight in this thread: forum.nasaspaceflight.com/index.php?topic=40499.0 - it doesn't directly address this question, but it's apropos. $\endgroup$
    – Tristan
    Commented Jun 2, 2017 at 13:12
  • $\begingroup$ The upcoming SpaceX launch will feature fourth, not first re-use of a dragon capsule. $\endgroup$ Commented Jun 26, 2018 at 19:46
  • $\begingroup$ @user3715778 Hmm... the post is a year old, and the linked article states that CRS-11 was going to be the first. Let's figure this out! I've asked Reuse statistics of SpaceX Dragon capsules? $\endgroup$
    – uhoh
    Commented Jun 27, 2018 at 0:34

2 Answers 2


The private firm that developed ROSA is Deployable Space Systems. One of its principal investigators answered some questions about the system on NASASpaceFlight. Since ROSA is still in development and from a private company, some of the details are unclear at this time, but the technology's patent holds a lot of information.

The Roll-Out Solar Array technology demonstration had two main components:

Slit-tube composite booms

Anyone who is familiar with party horns will understand how these booms operate - a structure that lies flat to be rolled up in one configuration and forms a cylinder for rigidity in the other. The difference with the slit-tube booms as they use a high internal tension in the carbon fiber reinforced polymer composite to provide the force to unroll and form the cylinder, rather than air pressure.

This allows the solar array to deploy without the need for a motor - just release it and the booms will unfurl themselves (a motorised system is required to retract the array).

Note - some sources I've read suggest that the booms are made of an Elastic Memory Composite. This is apparently untrue - ROSA's booms unfurl with stored strain energy. I'm unsure of the distinction here.

Flexible Solar Blanket

The solar blanket itself uses an array of flexible cells - Thin Film Cells or Inverted Metamorphic Multijunction Cells:

Our ROSA system certainly use[s] flexible thin film technologies or more advance IMM technologies depending on the customers needs.

I assume the demonstration model used thin film cells, but future commercial products may use IMM cells instead.

These are mounted on a flexible substrate (patent item 204):

Flexible Photovoltaic (PV) blanket: A thin flexible substrate that has mounted to it an array of photovoltaic solar cells and associated wiring that can be rolled or folded into a small package for stowage; and is attached to the deployable solar array structure for unfurling into a flat, tensioned configuration during deployment.

This has a compressible backing that supports the blanket during rolling (patent item 301):

Compressible open cell foam: open-cell foam applied to the back (non-cell populated) side of the flexible PV blanket (204) in various forms such as strips, patches or continuous sheets. When the blanket is rolled for stowage, the foam is compressed to take up the differential spacing between the elastic boom roll diameter and the blanket roll diameter...

It's worth noting that during the demonstration on the ISS, the blanket array was not fully populated with cells as can be seen clearly in some images:

enter image description here

This is because the demonstration was mainly aiming to showcase the deployment technology rather than the power output of the system.

There are more prototyping images available here.


This NASA article explains that the solar arrays are made of standard solar cells on a flexible mesh backing, not unlike the current arrays on the International Space Station (only the ISS panels were stored in an accordion arrangement, not as a roll). A major problem with the arrangement is that individual cells can break while the array is being deployed. At the time they were still deciding the best form of interconnect between the cells.

The test array is also equipped with three different wiring options: open wire, ribbon wire and a combination of the two. This allows for a variety of solar cell modules to be tested on the array. “If a cell cracks or malfunctions, we want to know the best way to bypass the failed cell and keep the whole operation intact,” he says.

  • $\begingroup$ Your linked article seems only to list a bunch of different options being evaluated in early lab tests by an engineer in matching blue jeans, shirt & nitrile gloves. I'm asking about the one being sent to space on Dragon. It says "...*to determine whether certain solar cells can withstand multiple cycles*." and "The Glenn team wants to know which cells will crack or be damaged by the operation." and "The test array is also equipped with three different wiring options: open wire, ribbon wire and a combination of the two." and "... we want to know the best way to bypass the failed cell..." $\endgroup$
    – uhoh
    Commented Jun 1, 2017 at 20:26
  • $\begingroup$ It seems the date of the article is July 2015. The design must have evolved and a best technique been chosen for this deployment in the last two years. Can you find some more recent, finalized information that addresses my question more directly? Thanks! $\endgroup$
    – uhoh
    Commented Jun 1, 2017 at 20:37

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.