Early concentrators

Concentration Photovoltaic is all about concentrating, cheaply, a lot of solar radiation on expensive and high efficiency cells and, possibly, as uniform as possible. But, to make some real power, you need large collectors… and they mush by reflective (or transparent), cheap and reliable over 30 years.

I did this by properly assembling multiple flat mirrors on properly designed support structures to achieve the desired illumination profile.

QUESTION) How do you build a large solar collector with uniform illumination of your receiver and optimal energy transfer?

Parabolic concentrators are expensive, do not provide uniform illumination, and deliver sub-optimal energy transfer for fundamental theoretical reasons (essentially, designed to transfer images, they sacrifice energy transfer efficiency to fulfill they primary task).

Free-form optical design, can do this but, then, mold realization with large aspherical surfaces is a complex task.

Moreover, once you have the mold, large plastic surfaces, obtained by injection, need to be treated to attain mirror reflectivity and, usually, lack proper optical grade surface finishing.  Last but not least injection molds are quite expensive anyway!
Metal impact shaping does not preserve mirror like surface finish and does not allow to create 3d shapes starting from flat foils without adding cuts or high strain regions.
3d shaping of large glass mirror is an entire class of nightmares by itself.

ANSWER) What about assembling several pre-cut flat (aka cheap) rigid mirrors on a suitably shaped support structure (or possibly inserting them in the injection process)?

The surface finishing of the support structure becomes irrelevant as the optical properties are given by the flat mirrors.

Concentration is achieved by the superposition of the illuminated areas of the flat mirrors. No hot spots can be generated and, if proper mirror positioning is achieved, uniform target illumination can be attained. Also Etandue is conserved thus providing optimal energy transfer.

The position, orientation and shape of each individual mirror are easily defined with a simple linear set of equations, given the target edges and the direction of the incoming (quasi) collimated solar radiation.

They also conserve global Etandue (thus achieving optimal energy transfer).

EXECUTION) In the years I designed multiple versions and evolution of this concept, starting with a large tenso-structure (adapted from the design of a third party!) carrying large truck rear-view mirrors (Figure 1). I then moved, with my group, to glass fiber structures incorporating plastic flat mirrors (Pictures 2 and 3) and finally, while in the start-up Cpower, I attempted direct thermoforming of reflective plastic sheets into a flat faceted concentrator (Picture 4) with triangular facets.

The design were all realized by matlab (or C++) numerical codes directly interfacing with CAD/CAM packages (I used Autocad, Rhino, Feeecad and lately Sketchup) while performance simulations were done by automatically creating a model in ray-tracing packages (Tracepro). Automated optimization based on DDE based data exchange between matlab and tracepro was also implemented.

Progressive improvements in the energy transfer and illumination uniformity were achieved to the advantage of the concentrator converter performance.