Carbon Fiber Mirror building block for space based solar power

Lightweight parabolic mirrors for solar concentrators have been fabricated using carbon fiber reinforced polymer (CFRP) and a nanometer scale optical surface smoothing technique. The smoothing technique improved the surface roughness of the CFRP surface from ~3 micron root mean square (RMS) for as-cast to ~5 nm RMS after smoothing. The surfaces were then coated with metal, which retained the sub-wavelength surface roughness, to produce a high-quality specular reflector. The mirrors were tested in an 11x geometrical concentrator configuration and achieved an optical efficiency of 78% under an AM0 solar simulator. With further development, lightweight CFRP mirrors will enable dramatic improvements in the specific power, power per unit mass, achievable for concentrated photovoltaics in space.

A proof-of-concept 11x concentrator made with lightweight carbon fiber parabolic mirrors achieved up to 77.5% optical efficiency. The primary loss of efficiency was due to shape deviation from the nominal parabola, as the smoothed surface of the mirrors provided excellent specular reflectance over the visible and near-infrared wavelengths. Further optimization of this system will include improving the shape, reducing the thickness of the polymer smoothing layer, and making thinner CFRP composite for overall mass reduction. In addition, we will investigate the stability of the system in vacuum and under elevated temperatures and thermal cycling consistent with operation in space. Overall, this UV curable nano-meter scale smoothing process for making CFRP mirrors offers promising performance for an ultra-light concentrated photovoltaic system intended for space applications.

Arxiv – Lightweight Carbon Fiber Mirrors for Solar Concentrator Applications

27 thoughts on “Carbon Fiber Mirror building block for space based solar power”

  1. Odd… The POINT of using parabolic strip mirrors is to concentrate sunlight onto a likewise-strip-oriented photovoltaic converter which due to higher light intensity works at higher efficiency. For instance, “naked” PV cells, down here on Planet Dirt without much fanfare can be convinced to convert 20% of incident Noon sunlight to electricity on a cold, clear winter day. Likewise, tho’ it increases the cost fairly substantially per square meter, PV solar cells can have multiple layers overlaid to substantially increase conversion efficiency as well. The record for multilayer PV a few years back was over 33% conversion. In SPACE, this is even more highly desired, as the cells are bathed in sunlight that hasn’t travelled thru the Earth’s atmosphere, which absorbs more than 25% of it. Multilayer, with parabolic concentrators, might achieve 35% conversion. But if (per the article) the use of silver-over-substrate mirrors is only 77% efficient, well … some of that concentrator benefit is eaten up by the cure! 77% (concentrators) of 30% (multi-layer PV) is what … 22% or so overall? That isn’t appreciably better than just using large, flat, fairly cost effective non-concentrated PV from silicon at 20% conversion. Moreover, the solar concentrator method is a fabrication-and-transportation nightmare. The things I presume might be able to be rolled up (which relieves much of the nightmare), but you still need to ensure than when unrolled all the parabolas do the right reformation. And that the angle-of-incidence relative to the Sun is also very well controlled: when an array isn’t pointing directly at Sol, the concentrators still concentrate… but onto the stuff surrounding the PV cells. If they are 10× concentrators, that’s pretty intense unwanted heating of a surrounding part. Nope. Nice research, but I’ven’t seen the compelling bottom line. Just saying, GoatGuy

    Reply
  2. Odd… The POINT of using parabolic strip mirrors is to concentrate sunlight onto a likewise-strip-oriented photovoltaic converter which due to higher light intensity works at higher efficiency. For instance “naked” PV cells down here on Planet Dirt without much fanfare can be convinced to convert 20{22800fc54956079738b58e74e4dcd846757aa319aad70fcf90c97a58f3119a12} of incident Noon sunlight to electricity on a cold clear winter day. Likewise tho’ it increases the cost fairly substantially per square meter PV solar cells can have multiple layers overlaid to substantially increase conversion efficiency as well. The record for multilayer PV a few years back was over 33{22800fc54956079738b58e74e4dcd846757aa319aad70fcf90c97a58f3119a12} conversion. In SPACE this is even more highly desired as the cells are bathed in sunlight that hasn’t travelled thru the Earth’s atmosphere which absorbs more than 25{22800fc54956079738b58e74e4dcd846757aa319aad70fcf90c97a58f3119a12} of it. Multilayer with parabolic concentrators might achieve 35{22800fc54956079738b58e74e4dcd846757aa319aad70fcf90c97a58f3119a12} conversion. But if (per the article) the use of silver-over-substrate mirrors is only 77{22800fc54956079738b58e74e4dcd846757aa319aad70fcf90c97a58f3119a12} efficient well … some of that concentrator benefit is eaten up by the cure!77{22800fc54956079738b58e74e4dcd846757aa319aad70fcf90c97a58f3119a12} (concentrators) of 30{22800fc54956079738b58e74e4dcd846757aa319aad70fcf90c97a58f3119a12} (multi-layer PV) is what … 22{22800fc54956079738b58e74e4dcd846757aa319aad70fcf90c97a58f3119a12} or so overall? That isn’t appreciably better than just using large flat fairly cost effective non-concentrated PV from silicon at 20{22800fc54956079738b58e74e4dcd846757aa319aad70fcf90c97a58f3119a12} conversion. Moreover the solar concentrator method is a fabrication-and-transportation nightmare. The things I presume might be able to be rolled up (which relieves much of the nig

    Reply
  3. Isn’t that where the 11x concentrator comes into play? It’s not to work at higher efficiency levels, it’s to put more sunlight on each photovoltaic cell, which is generally the expensive part. With this, you only need 1/11 th of the PVC surface area to generate the electricity, so even losing 22.5% of the energy in the reflection process would give you about a 8.5x improvement in electricity (per PVC) over a standard flat panel. If I did the math right…

    Reply
  4. improved the surface roughness of the CFRP surface from ~3 micron root mean square (RMS) for as-cast to ~5 nm RMS after smoothing.” In case you’d forgotten, like me, a micron is also called a micrometer, or 10^-6 meters, and of course nm is a nanometer, or 10^-9 (and an Angstrom is 10^-10). So it’s roughly 600x smoother.

    Reply
  5. Isn’t that where the 11x concentrator comes into play? It’s not to work at higher efficiency levels it’s to put more sunlight on each photovoltaic cell which is generally the expensive part. With this you only need 1/11 th of the PVC surface area to generate the electricity so even losing 22.5{22800fc54956079738b58e74e4dcd846757aa319aad70fcf90c97a58f3119a12} of the energy in the reflection process would give you about a 8.5x improvement in electricity (per PVC) over a standard flat panel. If I did the math right…

    Reply
  6. improved the surface roughness of the CFRP surface from ~3 micron root mean square (RMS) for as-cast to ~5 nm RMS after smoothing.””In case you’d forgotten”” like me a micron is also called a micrometer or 10^-6 meters and of course nm is a nanometer”” or 10^-9 (and an Angstrom is 10^-10). So it’s roughly 600x smoother.”””

    Reply
  7. The record for multilayer PV a few years back was over 33% conversion.” I remember close to 50%. I have bookmarked article about 46% efficiency, though that’s also a concentrator. I also have an NBF article bookmarked about a triple junction cell achieving 41%. (Though I can only dig it up from the web archive)

    Reply
  8. The record for multilayer PV a few years back was over 33{22800fc54956079738b58e74e4dcd846757aa319aad70fcf90c97a58f3119a12} conversion.””I remember close to 50{22800fc54956079738b58e74e4dcd846757aa319aad70fcf90c97a58f3119a12}. I have bookmarked article about 46{22800fc54956079738b58e74e4dcd846757aa319aad70fcf90c97a58f3119a12} efficiency”””” though that’s also a concentrator.I also have an NBF article bookmarked about a triple junction cell achieving 41{22800fc54956079738b58e74e4dcd846757aa319aad70fcf90c97a58f3119a12}. (Though I can only dig it up from the web archive)”””

    Reply
  9. I have never understood why fossil fuel power plant don’t use parabolic solar heater to preheat water. I do understand that the water is still pretty hot after the steam condenses but even a few hundred degrees can still save some fuel.

    Reply
  10. The record for multilayer PV a few years back was over 33% conversion.” I remember close to 50%. I have bookmarked article about 46% efficiency, though that’s also a concentrator. I also have an NBF article bookmarked about a triple junction cell achieving 41%. (Though I can only dig it up from the web archive)

    Reply
  11. The record for multilayer PV a few years back was over 33{22800fc54956079738b58e74e4dcd846757aa319aad70fcf90c97a58f3119a12} conversion.””I remember close to 50{22800fc54956079738b58e74e4dcd846757aa319aad70fcf90c97a58f3119a12}. I have bookmarked article about 46{22800fc54956079738b58e74e4dcd846757aa319aad70fcf90c97a58f3119a12} efficiency”””” though that’s also a concentrator.I also have an NBF article bookmarked about a triple junction cell achieving 41{22800fc54956079738b58e74e4dcd846757aa319aad70fcf90c97a58f3119a12}. (Though I can only dig it up from the web archive)”””

    Reply
  12. “The record for multilayer PV a few years back was over 33% conversion.”

    I remember close to 50%. I have bookmarked article about 46% efficiency, though that’s also a concentrator.
    I also have an NBF article bookmarked about a triple junction cell achieving 41%. (Though I can only dig it up from the web archive)

    Reply
  13. Isn’t that where the 11x concentrator comes into play? It’s not to work at higher efficiency levels, it’s to put more sunlight on each photovoltaic cell, which is generally the expensive part. With this, you only need 1/11 th of the PVC surface area to generate the electricity, so even losing 22.5% of the energy in the reflection process would give you about a 8.5x improvement in electricity (per PVC) over a standard flat panel. If I did the math right…

    Reply
  14. Isn’t that where the 11x concentrator comes into play? It’s not to work at higher efficiency levels it’s to put more sunlight on each photovoltaic cell which is generally the expensive part. With this you only need 1/11 th of the PVC surface area to generate the electricity so even losing 22.5{22800fc54956079738b58e74e4dcd846757aa319aad70fcf90c97a58f3119a12} of the energy in the reflection process would give you about a 8.5x improvement in electricity (per PVC) over a standard flat panel. If I did the math right…

    Reply
  15. improved the surface roughness of the CFRP surface from ~3 micron root mean square (RMS) for as-cast to ~5 nm RMS after smoothing.” In case you’d forgotten, like me, a micron is also called a micrometer, or 10^-6 meters, and of course nm is a nanometer, or 10^-9 (and an Angstrom is 10^-10). So it’s roughly 600x smoother.

    Reply
  16. improved the surface roughness of the CFRP surface from ~3 micron root mean square (RMS) for as-cast to ~5 nm RMS after smoothing.””In case you’d forgotten”” like me a micron is also called a micrometer or 10^-6 meters and of course nm is a nanometer”” or 10^-9 (and an Angstrom is 10^-10). So it’s roughly 600x smoother.”””

    Reply
  17. Odd… The POINT of using parabolic strip mirrors is to concentrate sunlight onto a likewise-strip-oriented photovoltaic converter which due to higher light intensity works at higher efficiency. For instance, “naked” PV cells, down here on Planet Dirt without much fanfare can be convinced to convert 20% of incident Noon sunlight to electricity on a cold, clear winter day. Likewise, tho’ it increases the cost fairly substantially per square meter, PV solar cells can have multiple layers overlaid to substantially increase conversion efficiency as well. The record for multilayer PV a few years back was over 33% conversion. In SPACE, this is even more highly desired, as the cells are bathed in sunlight that hasn’t travelled thru the Earth’s atmosphere, which absorbs more than 25% of it. Multilayer, with parabolic concentrators, might achieve 35% conversion. But if (per the article) the use of silver-over-substrate mirrors is only 77% efficient, well … some of that concentrator benefit is eaten up by the cure! 77% (concentrators) of 30% (multi-layer PV) is what … 22% or so overall? That isn’t appreciably better than just using large, flat, fairly cost effective non-concentrated PV from silicon at 20% conversion. Moreover, the solar concentrator method is a fabrication-and-transportation nightmare. The things I presume might be able to be rolled up (which relieves much of the nightmare), but you still need to ensure than when unrolled all the parabolas do the right reformation. And that the angle-of-incidence relative to the Sun is also very well controlled: when an array isn’t pointing directly at Sol, the concentrators still concentrate… but onto the stuff surrounding the PV cells. If they are 10× concentrators, that’s pretty intense unwanted heating of a surrounding part. Nope. Nice research, but I’ven’t seen the compelling bottom line. Just saying, GoatGuy

    Reply
  18. Odd… The POINT of using parabolic strip mirrors is to concentrate sunlight onto a likewise-strip-oriented photovoltaic converter which due to higher light intensity works at higher efficiency. For instance “naked” PV cells down here on Planet Dirt without much fanfare can be convinced to convert 20{22800fc54956079738b58e74e4dcd846757aa319aad70fcf90c97a58f3119a12} of incident Noon sunlight to electricity on a cold clear winter day. Likewise tho’ it increases the cost fairly substantially per square meter PV solar cells can have multiple layers overlaid to substantially increase conversion efficiency as well. The record for multilayer PV a few years back was over 33{22800fc54956079738b58e74e4dcd846757aa319aad70fcf90c97a58f3119a12} conversion. In SPACE this is even more highly desired as the cells are bathed in sunlight that hasn’t travelled thru the Earth’s atmosphere which absorbs more than 25{22800fc54956079738b58e74e4dcd846757aa319aad70fcf90c97a58f3119a12} of it. Multilayer with parabolic concentrators might achieve 35{22800fc54956079738b58e74e4dcd846757aa319aad70fcf90c97a58f3119a12} conversion. But if (per the article) the use of silver-over-substrate mirrors is only 77{22800fc54956079738b58e74e4dcd846757aa319aad70fcf90c97a58f3119a12} efficient well … some of that concentrator benefit is eaten up by the cure!77{22800fc54956079738b58e74e4dcd846757aa319aad70fcf90c97a58f3119a12} (concentrators) of 30{22800fc54956079738b58e74e4dcd846757aa319aad70fcf90c97a58f3119a12} (multi-layer PV) is what … 22{22800fc54956079738b58e74e4dcd846757aa319aad70fcf90c97a58f3119a12} or so overall? That isn’t appreciably better than just using large flat fairly cost effective non-concentrated PV from silicon at 20{22800fc54956079738b58e74e4dcd846757aa319aad70fcf90c97a58f3119a12} conversion. Moreover the solar concentrator method is a fabrication-and-transportation nightmare. The things I presume might be able to be rolled up (which relieves much of the nig

    Reply
  19. Isn’t that where the 11x concentrator comes into play? It’s not to work at higher efficiency levels, it’s to put more sunlight on each photovoltaic cell, which is generally the expensive part.

    With this, you only need 1/11 th of the PVC surface area to generate the electricity, so even losing 22.5% of the energy in the reflection process would give you about a 8.5x improvement in electricity (per PVC) over a standard flat panel. If I did the math right…

    Reply
  20. “improved the surface roughness of the CFRP surface from ~3 micron root mean square (RMS) for as-cast to ~5 nm RMS after smoothing.”

    In case you’d forgotten, like me, a micron is also called a micrometer, or 10^-6 meters, and of course nm is a nanometer, or 10^-9 (and an Angstrom is 10^-10). So it’s roughly 600x smoother.

    Reply
  21. Odd…

    The POINT of using parabolic strip mirrors is to concentrate sunlight onto a likewise-strip-oriented photovoltaic converter which due to higher light intensity works at higher efficiency. For instance, “naked” PV cells, down here on Planet Dirt without much fanfare can be convinced to convert 20% of incident Noon sunlight to electricity on a cold, clear winter day.

    Likewise, tho’ it increases the cost fairly substantially per square meter, PV solar cells can have multiple layers overlaid to substantially increase conversion efficiency as well. The record for multilayer PV a few years back was over 33% conversion. In SPACE, this is even more highly desired, as the cells are bathed in sunlight that hasn’t travelled thru the Earth’s atmosphere, which absorbs more than 25% of it. Multilayer, with parabolic concentrators, might achieve 35% conversion.

    But if (per the article) the use of silver-over-substrate mirrors is only 77% efficient, well … some of that concentrator benefit is eaten up by the cure!

    77% (concentrators) of 30% (multi-layer PV) is what … 22% or so overall?

    That isn’t appreciably better than just using large, flat, fairly cost effective non-concentrated PV from silicon at 20% conversion.

    Moreover, the solar concentrator method is a fabrication-and-transportation nightmare. The things I presume might be able to be rolled up (which relieves much of the nightmare), but you still need to ensure than when unrolled all the parabolas do the right reformation. And that the angle-of-incidence relative to the Sun is also very well controlled: when an array isn’t pointing directly at Sol, the concentrators still concentrate… but onto the stuff surrounding the PV cells. If they are 10× concentrators, that’s pretty intense unwanted heating of a surrounding part.

    Nope. Nice research, but I’ven’t seen the compelling bottom line.

    Just saying,
    GoatGuy

    Reply

Leave a Comment