Priority

This application claims the benefit of U.S. Provisional Patent Application serial number 62/211,376, filed on August 28, 2015; U.S. Provisional Patent Application serial number 62/153,716, filed on April 28, 2015; and U.S. Provisional Patent Application serial number 62/153,723, filed on April 28, 2015, which are incorporated herein by reference in their entirety for all purposes.

Field of the Invention

The present invention is an improved mirror panel assembly used for reflecting light to a target. More specifically, this mirror panel approach is intended for use in applications related to the field of Concentrated Solar Power (CSP), such as for heliostats and solar troughs, among others.

Background

Solar power plants or other systems that collect and concentrate solar energy onto one or more centralized targets are well known in the art. The concentrated solar energy often is used to directly or indirectly produce electricity and/or heat. Direct conversion, often referred to as concentrating photovoltaics (CPV) occurs in some modes of practice when photovoltaic cells (also known as solar cells) serve as the target(s) to convert incident, concentrated solar energy into electricity using photovoltaic effects. Indirect conversion, often referred to as Concentrating Solar Power (CSP) occurs when thermal energy of the concentrated solar energy is used in some modes of practice to heat a working fluid, or sequence of working fluids, that in turn drives machinery such as a turbine system to generate electric power.

Working fluids include steam, oil, molten salt, or the like.

U.S. Pat. Nos. 8,833,076; 8,697,271; 7,726,127; 7,299,633; and U.S. Pat. Pub. No. 213/0081394 Al describe systems in which solar energy heats molten salt to store the thermal energy. The molten salt can store the heat for extended periods of time for later use on demand. The molten salt thus functions as a thermal battery in this regard that is charged by the sun. The molten salt in turn is used in illustrative modes of practice to heat steam that drives a turbine to generate electricity. After heating the steam, the molten salt cools down but is readily heated again, or re-charged with solar energy, by heating again using concentrated solar energy. Molten salt can be heated, used, and recharged this way many times without being consumed to any significant degree. Facilities that use molten salt in this fashion are projected to have lifespans extending for decades.

CSP systems typically rely on a field of reflecting devices that track, reflect, and collectively concentrate incident sunlight onto a solar receiver. Many types of reflecting devices are known. Examples include heliostats, parabolic dishes, trough concentrators, and the like. A CSP system often may use hundreds or even thousands of reflecting devices to concentrate solar energy.

Mirrors in most instances are a fundamental element of the reflecting devices used in CSP plants. The primary function of the mirrors is to reflect sunlight onto a target where the resultant concentrated sunlight can then be converted into other forms of useful energy, such as electricity or heat. Mirrors may have a variety of shapes, and many shapes are suitable to redirect sunlight onto a desired target. As examples of shapes, mirrors may be flat, curved in two dimensions, curved in three dimensions, faceted, and the like.

The mirrors often are supported by a suitable support structure so that the mirrors substantially maintain their shape without undue sagging, thermal deformation, or shape deformation as the mirrors articulate and are impacted by wind, moisture, age, temperature changes, and other surrounding factors. An important factor that affects energy delivery over time is any deviation between the actual mirror shape and the intended mirror shape, or slope error. A goal is to limit this slope error to desired tolerances. The degree to which slope errors are tolerated is referred to as the slope error budget. In order to maintain a desired shape, the mirror is generally integrated with a suitable support structure. This supporting structure, together with the mirror, comprises at least a portion of a mirror panel assembly.

A mirror panel is a component of many different types of reflector devices. A heliostat is one type of reflector device. A heliostat is a term in the art that refers to an assembly comprising one or more mirror panel assemblies, one or more drive mechanisms attached to the mirror panel to articulate the mirror panel to track the sun, and a base structure mounted to the drive mechanism to attach the heliostat to the ground, a frame, or other fixed or moveable mounting site. Trough reflectors are another type of reflecting device.

The adverse impact of slope errors becomes more pronounced with increasing distance from the target. This is less of an issue with solar trough reflectors as often these are integrated into CSP systems in which the mirror-to- target distance is relatively low and where the mirror-to-target distance is similar for all mirror panels. On the other hand, heliostats are more often used in CSP systems that typically have much longer distances between the mirror panels and the target. In some systems, this distance can be up to a mile or more. Heliostat-based CSP systems of this magnitude, therefore, are less tolerant to slope error and may experience significant losses in energy production if the slope errors are too large.

Minimizing slope errors is a key aspect of heliostat engineering. From design, through fabrication and assembly, and ultimately through the performance under operating conditions, there are a number of factors that influence the slope error characteristics of the mirror panel assembly. A key factor is the influence of temperature changes and differential thermal expansion characteristics between the glass and supporting structure.

Composite sandwich construction is well known. A composite sandwich panel assembly typically includes two stressed skins separated by and bonded to a core material. The attachment between the core and skins is usually accomplished using some type of adhesive and/or mechanical coupling. The resulting panel structure often uses materials efficiently for the stiffness and strength achieved.

Ongoing efforts to implement a composite sandwich panel as a mirror panel structure have been attempted. Exemplary composite sandwich structures are described in US 8,132,391 B2 and US 8,327,604 B2. Instead of making the core from a separate piece of material, the core structure in these designs is formed as tabs that are an integral part of one of the skins. This is achieved by perforating a metal sheet at regular intervals and folding up "riser elements" perpendicular to the parent material of the skin. The tips of the riser elements are folded over to create tabs that are bonded to the back side of the top sheet of skin material to form the composite structure. Additional composite sandwich structures are also shown in Assignee's copending U.S. provisional applications having U.S. Ser. Nos. 62/153,716 and 62/153,723, wherein the respective entirety of each of these provisional applications is incorporated herein by reference for all purposes.

In the past, this structure has been used in parabolic troughs but is now being considered for heliostat applications. One particular configuration under development uses a backer sheet with integral riser elements that are bonded to the back side of another continuous sheet of the same material to form the sandwich panel. A reflective film is adhered to the front side of the continuous sheet to create a mirror surface.

Other heliostat strategies involve using mirror panel assemblies in which a reflector in the form of a mirror panel is supported by an underlying frame structure. One of the potential challenges associated with this strategy is differential thermal expansion between the reflector and the supporting frame structure. This effect could result in undue slope error issues with a mirror if steps are not taken to accommodate the relative movement attributed to thermal expansion differences. Without being able to more effectively accommodate thermal expansion effects, it may be difficult for such frame and panel designs to be as effective as might be desired when used as mirror panels of heliostat-based systems.

Summary of the Invention

The present invention provides strategies for reducing the harmful effects of differential thermal expansion in mirror panel structures. The principles of the present invention are particularly useful in the field of concentrating solar power. The principles of the present invention can be used in CSP applications to make mirror panel structures with improved characteristics for accommodating differential thermal expansion between components of the structures. Significantly, the mirror structures can still be securely attached to other heliostat components such as a drive mechanism while still having the ability to accommodate differential thermal expansion between the mirror panel and a supporting frame structure. This helps to limit energy losses due to slope errors.

The mirror panel can be supported in a way to be stiffer near one or more reference locations, such as near an attachment site to other components, while also being less stiff farther from those one or more locations. For example, the stiffness can be tailored to gradually decrease with increasing distance from an attachment site or other designated reference location(s). One purpose of modulating the shear stiffness in this manner is to more closely correlate the stiffness to the differential thermal expansion motion of the mirror panel relative to the frame structure to help minimize panel stresses that could cause slope errors. In some embodiments, a reference location may be selected to coincide with an attachment site, but in other embodiments the designated reference location and an attachment site may be at different locations on the mirror panel.

For example, at a location near the attachment region where a mirror panel is attached to another heliostat component, thermal stresses due to temperature changes tend to be low. The panel can be quite stiff proximal to such attachment point without the high stiffness leading to slope errors when temperature changes. However, further away from the attachment point, the thermal stresses tend to be greater when temperature changes. If the shear stiffness remained constant throughout the panel, the resultant thermal stresses farther away from the attachment point would be greater. Fabricating the panel assembly so that it is less stiff at those greater distances reduces the stresses. Modulating the shear stiffness to provide a desired shear stiffness profile for the panel helps to mitigate the impact of thermal expansion more effectively, helping to reduce slope errors.

As another advantage, the principles of the present invention effectively separate structural functionality from the function of accommodating differential thermal expansion. This is accomplished by using flexures to couple one or more mirror panels to a strong supporting framework. The framework helps provide a mirror panel assembly with structural integrity while the flexures allow the framework and mirror panel to respond differently to temperature changes without undue stress build up.

In one aspect, the present invention relates to an articulating heliostat, comprising:

a) a mirror panel comprising a top, reflective surface and a bottom surface; b) a support frame comprising i) a plurality of tubular support members that are spaced apart from the mirror panel, each tubular support member having a first surface facing toward the mirror panel and a second surface that is further away from the mirror panel than the first surface; and

ii) a plurality of cross-members that are stacked on and interconnect the tubular support members in a manner such that the cross-members are positioned between the mirror panel and the tubular support members and such that the cross-members are spaced apart from the mirror panel; c) a plurality of flexures that couple the mirror panel to the tubular support members, each flexure having an active length of flexure that extends from a first attachment site proximal to the bottom surface of the mirror panel and a second attachment site proximal to a corresponding tubular support member, wherein the second attachment site is farther from the bottom surface of the mirror panel than the first surface of the corresponding support member, wherein said flexure is coupled to the second attachment site such that the flexure passes through an aperture in the first surface of the corresponding tubular support member; and

d) a drive mechanism coupled to the support frame in a manner effective to articulate the mirror panel on first and second axes.

In another aspect, the present invention relates to a mirror panel assembly, comprising:

a) a mirror panel comprising a top, reflective surface and a bottom surface; b) a support frame comprising

(i) a plurality of tubular support members that are spaced apart from the mirror panel, each tubular support member having a first surface facing toward the mirror panel and a second surface that is further away from the mirror panel than the first surface; and

(ii) a plurality of cross-members that are stacked on and interconnect the tubular support members in a manner such that the cross-members are positioned between the mirror panel and the tubular support members and such that the cross-members are spaced apart from the mirror panel; and c) a plurality of flexures that couple the mirror panel to the tubular support members, each flexure having an active length of flexure that extends from a first attachment site proximal to the bottom surface of the mirror panel and a second attachment site proximal to a corresponding tubular support member, wherein the second attachment site is farther from the bottom surface of the mirror panel than the first surface of the corresponding support member, wherein said flexure being coupled to the second attachment site such that the flexure passes through an aperture in the first surface of the

corresponding tubular support member.

In another aspect, the present invention relates to a solar power system, comprising:

a) a central tower comprising a target; and

b) a plurality of heliostats that reflect and concentrate sunlight onto the central target, wherein at least one of the heliostats comprises a mirror panel assembly as described herein, such as is recited in Claim 2, below.

In another aspect, the present invention relates to a method of making a mirror panel assembly, comprising the steps of:

a) providing a mirror panel comprising a top, reflective surface and a bottom surface;

b) providing a support frame comprising:

(i) a plurality of tubular support members that are spaced apart from the mirror panel, each tubular support member having a first surface facing toward the mirror panel and a second surface that is further away from the mirror panel than the first surface; and

(ii) a plurality of cross-members that are stacked on and interconnect the tubular support members in a manner such that the cross-members are positioned between the mirror panel and the tubular support members and such that the cross-members are spaced apart from the mirror panel; and

c) providing a plurality of flexures that couple the mirror panel to the tubular support members, each flexure having an active length of flexure that extends from a first attachment site proximal to the bottom surface of the mirror panel and a second attachment site proximal to a corresponding tubular support member, wherein the second attachment site is farther from the bottom surface of the mirror panel than the first surface of the corresponding support member, wherein said flexure is coupled to the second attachment site such that the flexure passes through an aperture in the first surface of the corresponding tubular support member.

Brief Description of the Drawings

Fig. 1 is a schematic illustration of a concentrated solar power system incorporating principles of the present invention.

Fig. 2A schematically illustrates a heliostat used in the power system of Fig.

1, wherein the heliostat incorporates a mirror panel assembly of the present invention.

Fig. 2B is a bottom perspective view of a portion of the heliostat shown in Fig. 2A.

Fig. 3 is a top, isometric view of the mirror panel assembly of Fig. 2.

Fig. 4 is a bottom, isometric view of the mirror panel assembly of Fig. 2.

Fig. 5 is a bottom, plan view of the mirror panel assembly of Fig. 2.

Fig. 6 is a side view of the mirror panel assembly of Fig. 2.

Fig. 7 shows a flexure attached to a mounting pad that is used as components in the mirror panel assembly of Fig. 2.

Fig. 8 shows the flexure and mounting pad of Fig. 7 coupled to a mirror panel and a tubular support member, wherein only a corner portion of the mirror panel is shown and wherein an end portion of the tubular support member is shown.

Fig. 9 is an isometric top view of the end portion of the tubular support member of Fig 8.

Fig. 10 is a top perspective view of a stacked and brazed frame structure used in the mirror panel assembly of Fig. 2.

Fig. 11 is a side view of the frame structure shown in Fig. 10.

Fig. 12 is a side cross section view of a portion of the mirror panel assembly of Fig. 2.

Fig. 13 schematically shows how mounting pads and flexures used in the mirror panel assembly of Fig. 2 can be deployed on radial deployment lines in a manner so that the flexures are able to individually flex in a radial direction relative to a common reference site when accommodating differential thermal expansion and contraction between the mirror panel and its frame structure.

Fig. 14 is a side cross section view of a portion of the mirror panel assembly of Fig. 2 showing how a flexure can be attached to a tubular support member and a mounting pad in a manner to support a mirror panel (not shown) on the pad in a relatively low position.

Fig. 15 is a side cross section view of a portion of the mirror panel assembly of Fig. 2 showing how a flexure can be attached to a tubular support member and a mounting pad in a manner to support a mirror panel (not shown) on the pad in a relatively high position.

Fig. 16 schematically shows how mounting pads and flexures used in the mirror panel assembly of Fig. 2 can be deployed on concentric, circular reference guides to help provide a shear stiffness profile that decreases with increasing distance from the reference center of the guides.

Fig. 17 is a cross-section of a mounting pad, flexure, and tubular support member showing how the active length of the flexure can be shortened to increase stiffness by attaching the flexure to the bottom of the tubular support member and a tight fit around the flexure at the top of the tubular support member.

Fig. 18 is a bottom isometric view of a portion of the mirror panel assembly of Fig. 2 incorporating an alternative, cylindrical embodiment of a flexure.

Detailed Description of Presently Preferred Embodiments The present invention will now be further described with reference to the following illustrative embodiments. The embodiments of the present invention described below are not intended to be exhaustive or to limit the invention to the precise forms disclosed in the following detailed description. Rather a purpose of the embodiments chosen and described is so that the appreciation and understanding by others skilled in the art of the principles and practices of the present invention can be facilitated.

Referring to the figures and in particular Fig. 1, which schematically illustrates a concentrating solar energy system 10 that incorporates principles of the present invention. System 10 includes a central tower 12 including a mast 14 and a target region 16 at the top of the mast. A field of heliostats 20 is deployed around central tower 12. The heliostats 20 redirect and concentrate incident sunlight onto target region 16. If system 10 embodies a photovoltaic solar power system (a.k.a. concentrating photovoltaics, or CPV), target region 16 generally would include solar cells (not shown) that absorb the concentrated light and generate electricity that could then be stored for later use or distributed to one or more users or a power grid or the like. If system 10 embodies a concentrating solar power (CSP) system, used to convert thermal energy into electricity or mechanical energy (not shown), then the thermal energy generated on target region 16 may be used to heat a working fluid. The thermal energy in the heated fluid may then be used directly or indirectly to generate electricity or pressure. A CSP embodiment of system 10 is particularly useful in molten salt-based power systems such as those described in U.S. Pat. Nos. 8,833,076; 8,697,271; 7,726, 127; 7,299,633; and U.S. Pat. Pub. No. 2013/0081394 Al.

Figs. 2A and 2B schematically illustrate an exemplary embodiment of heliostat 20 used in system 10 of Fig. 1. Heliostat 20 incorporates an array 21 of mirror panel assemblies 36. Each mirror panel assembly 36 is shown schematically in Fig. 2A. Each mirror panel assembly 36 is shown in more detail in Fig. 2B as a mirror panel 38, a support frame 46 including tubular support members 48 and cross members 72, mounting pads 92, and flexures (not shown in Figs. 2A and 2B but described in more detail as flexures 100 below). The mirror panels 38 are each coupled to the underlying frame structure 46 but are not coupled to each other in this embodiment. The separation between mirror panels is shown schematically by boundary lines 35 in Fig. 2A. In Fig. 2B, the sky can be seen between panels 38, which are spaced apart by a small gap from each other.

For purposes of illustration, array 21 includes a 4 x 6 array of panels 38. Many other array sizes may be used. For example, exemplary arrays include m x n arrays where m is 1 to 20 or more and n is 1 to 20 or more.

Heliostat 20 includes pedestal 22 attaching heliostat 20 to the ground or other supporting surface. Rotatable vertical shaft 24 is rotatably housed inside pedestal 22. Shaft 24 rotates about a vertical, or azimuth, axis 19. Rotation of vertical shaft 24 causes array 21 to be rotatably driven about axis 19. A motor (not shown) is housed inside pedestal 22 in order to rotatably drive shaft 24. Port 25 provides access to the motor for installation, service, repair, replacement, or the like.

Yoke 26 is fixedly attached to drive shaft 24. Thus, rotation of shaft 24 causes yoke 26 to rotate about axis 19. A rotatable elevation tube 27 (also referred to as a torque tube) is rotatably mounted in bearings 28. Bearings 28, in turn, are fixedly connected to yoke 26. Lever arm 29 is connected to linear actuator 30. Linear actuator 30 drives lever arm 29 up or down as desired. Raising and lowering arm 29 rotatably drives elevation tube 27 about the elevation axis 31. Rotation of tube 27 causes array 21 to be rotatably driven about axis 31. Controller 33 is coupled by wiring 34 to motors that rotatably drive shaft 24 or actuate linear actuator 30.

Elevation tube 27 is fixedly mounted to trusses 32. Trusses 32 are further coupled to spar tubes 122. The spar tubes 122, in turn, are coupled to the cross- members 72 of support frame 46.

Figs. 3 to 17 show mirror panel assembly 36 of Figs. 2A and 2B in more detail. In use, mirror panel assembly 36 can be articulated to be in a wide range of orientations to track the sun, for storage, for service, to avoid storm damage, or otherwise. Such articulations can cause the major plane of mirror panel 36 to be horizontal, vertical, or otherwise oriented with respect to the ground. Regardless of how mirror panel 36 is oriented, the top or upward direction will be taken to be along the direction of normal vector 68 that projects outward from mirror panel assembly 36, as shown in Fig. 6. Similarly, the bottom or downward direction will be taken to be along the direction of normal vector 70 that projects outward from mirror panel assembly 36 in a direction opposite from vector 68.

Accordingly, mirror panel assembly 36 includes mirror panel 38 having top reflective surface 40, bottom surface 42, and a central region 44 on bottom surface 42 that serves in one aspect as a reference for aligning features of mirror panel 36 as described further below in order to help accommodate thermal stresses, gravity loads, and the like.

A plurality of mounting pads 92 are attached to the bottom surface 42 of mirror panel 38. Pads 92 are deployed in a rectangular array, but other deployment strategies can be used. For example, the rows and columns of pads 92 can zig zag to distribute mounting stresses to panel 38 more randomly. As another example, the density of pads 92 per unit area of bottom surface 42 can be relatively higher proximal to central region 44 but relatively lower with increasing distance from central region 44. This can help to provide strong support in central region 44 to handle mounting and load stresses while allowing the support frame 46 to accommodate thermal stresses in regions distal from central region 44 where thermal stresses tend to be greater.

Mounting pads 92 can be attached to mirror panel 38 using a variety of different strategies such as screws, bolts, welding, brazing, adhesives, rivets, snap fit engagement, threaded engagement, combinations of these, and the like. As shown, pads 92 are bonded to bottom surface 42 using adhesive 96.

Mounting pads 92 may include features to assist in the manufacture of mirror panel assembly 36. As shown, mounting pads 92 include alignment features in the form of notches 94. The use of notches 94 to assist with manufacture alignment is described further below with respect to Figs. 7, 8 and 13.

Support frame 46 includes a plurality of tubular support members 48 and a plurality of cross members 72 as main components. Tubular support members 48 are generally in the same plane and are deployed parallel to each other in this embodiment. Mirror panel 38 is mounted to the tubular support members 48 using a plurality of flexures 100. Tubular support members 48 are spaced apart from mirror panel 38 by the distance 66 so that mirror panel is suspended above cross members 72 by flexures 100. In this embodiment, four tubular support members 48 are shown. This number of members 48 would be suitable to support many different sizes of mirror panels 38. For example, in one mode of practice, using four members 48 that are 2.2 m long would be suitable for supporting a mirror panel that is 2.5 m long by 1.5 m wide. A greater or lesser number of tubular support members 48 can be used depending on factors such as the weight of mirror panel 38, the size of mirror panel 38, the stiffness of mirror panel 38, the environment in which mirror panel 38 is used, the thickness of the mirror panel 38, the stiffness and strength of support frame 46, and the like.

Each tubular support member 48 extends from a first end 50 to a second end 52. The length of each support member 48 between ends 50 and 52 generally is long enough to provide a desired degree of support for mirror panel 38. As illustrated, the support members 48 do not extend fully to the edges of mirror panel 38 but rather are slightly shorter so that edge portions of mirror panel 38 overhang slightly. The overhang of mirror panel 38 beyond ends 50 and 52 desirably is short enough to avoid sagging. In one mode of practice, using members 48 with a length of 2.2 m is suitable for supporting a mirror panel 38 that overhangs members 48 by 0.05 to 0.3 m, more preferably 0.15 m.

In the illustrative embodiment, each tubular support member 48 has a generally rectangular cross section with walls 54, 56, and 58. Regardless of orientation of mirror panel assembly 36, the portions of members 48 that face towards mirror panel 38 in the direction of upward vector 68 are deemed to be the top surfaces. Similarly, the portions of members 48 that face away from mirror panel 38 in the direction of downward vector 70 are deemed to be the bottom surfaces. Accordingly, wall 54 that faces toward mirror panel 38 in the direction of upward vector 68 is referred to as top wall 54 in this embodiment. Regardless of orientation of mirror panel assembly 36, wall 56 that faces away from mirror panel 38 in the direction of downward vector 70 is referred to as bottom wall 56 in this embodiment. Walls 58 thus are sidewalls to interconnect top wall 54 and bottom wall 56. The walls 54, 56, and 58 define an internal volume 64 inside each tubular member 48.

Although tubular support members 48 are shown with a rectangular cross section, a wide variety of other cross sections may be used. For example, tubular support members may be round, square, ellipsoid, solid members, members with surfaces including T-slots or other features to facilitate alignment or attachment to other components, or the like.

Each tubular member 48 includes an array of enlarged apertures 60 in top wall 54 and a corresponding array of apertures 62 in bottom wall 56. The apertures 60 and 62 are aligned to match corresponding mounting pads 92 so that flexures 100 may be used to couple members 48 to the mounting pads 92. Attachment strategies using flexures 100 are seen best in Figs. 8, 9, 12, 13, 14, 15, 17 and 18 (shows attachment strategy using alternative embodiment in the form of flexures 200). Cross members 72 are stacked on and coupled to tubular support members 48 in a manner effective so that support frame 46 provides a rigid framework to support mirror panel 38. Cross members 72 are generally in the same plane and are deployed parallel to each other in this embodiment. Cross members 72 also are spaced apart from the bottom surface 42 of mirror panel 38 so that mirror panel 38 is suspended above both members 48 and 72 by flexures 100 by gap 90.

In this embodiment, two cross members 72 are shown. This number of would be suitable to support many different sizes of mirror panels 38. For example, in one mode of practice, using two cross members 72 that are 1.44 m long would be suitable for supporting a mirror panel that is 2.5 m long by 1.5 m wide. A greater or lesser number of cross members 72 can be used depending on factors such as the weight of mirror panel 38, the size of mirror panel 38, the stiffness of mirror panel 38, the stiffness and strength of support frame 46, the environment in which mirror panel 38 is used, and the like.

Each cross members 72 extends from a first end 74 to a second end 76. The length of each cross members 72 between ends 74 and 76 generally is long enough to interconnect and brace at least two tubular support members 48. In this embodiment, each cross member 72 is long enough to span and interconnect the full array of tubular support members 48. As illustrated, the cross members 72 do not extend fully to the edges of mirror panel 38 but rather are slightly shorter so that edge portions of mirror panel 38 overhang slightly. The overhang of mirror panel 38 beyond ends 74 and 76 desirably is short enough to avoid sagging, although the length of cross members 72 has a much lower impact on preventing sagging as compared to tubular support members 48, spar tubes 122, and trusses 32. In one mode of practice, using cross members 72 with a length of 1.44 m is suitable for supporting a mirror panel 38 that overhangs cross members 72 by 0.05 to 0.3 m, more preferably 0.08 m.

As is the case with tubular support members 48, the portions of cross members 72 that face towards mirror panel 38 in the direction of upward vector 68 is deemed to be the top surface regardless of the orientation of mirror panel assembly 36. Similarly, the portions of cross members 72 that face away from mirror panel 38 in the direction of downward vector 70 is deemed to be the top surface regardless of the orientation of mirror panel assembly 36. Accordingly, for each cross member 72, top wall 78 faces toward mirror panel 38 in the direction of upward vector 68. Bottom wall 80 faces away from mirror panel 38 in the direction of downward vector. Sidewalls 82 interconnect top wall 78 and bottom wall 80. The walls 78, 80, and 82 define an internal volume 84 inside each cross member 72.

The cross members 72 are shown as having a rectangular cross section. Other cross sections may be used, as desired. Examples include cross sections that are round, square, ellipsoid, solid members, members with surfaces including T-slots or other features to facilitate alignment or attachment to other components, or the like. Cross members 72 are hollow tubes, but solid bars of any desired cross section may be used if desired.

Using a stacked orientation in which cross members 72 are stacked on and then coupled to the tubular support members 48 provides many advantages. All of the components used for members 48 and 72 may be straight without bends or angled cuts. This allows many off the shelf tubes or bars to be used without having to use custom components. Very little machining is required. Components can simply be drilled to provide the desired array of apertures for coupling to flexures 100 and then cut to length. Many component materials, such as pre-galvanized components, need no further finishing after assembly.

A stacked orientation refers to a layout in which the tubular support members

48 are in a first plane while the cross members 72 are in one or more other planes. In the illustrated embodiment, cross members 72 are in a second plane that is above the co-planar tubular support members 48, wherein the second plane is between the first plane and the mirror panel 38. Using a stacked frame layout uses less parts than if all of the members 48 and 72 were to be in the same plane (e.g., similar to a rail, stile, and mullion layout widely used to make frames in furniture construction). This stacking layout increases the structural strength and integrity while reducing the number of joints to be secured as compared to an in-plane layout. Also, the length of each tubular support member 48 and cross member 72 is less critical than in an in-plane layout where cutting similar pieces to precise, common lengths is more important for a frame to be flat and square. The stacked nature of the frame layout is best seen in Figs. 4, 6, 10, 11, 12, 14 and 15. As best seen from the side view of Figs. 8 and 12, the flexures 100 couple mounting pads 92 to the bottom wall 56 of members 48. Advantageously, this allows the center of gravity of the mirror panel assembly 36 to be moved closer to the spars 122. In actual practice, the attachment sites 116 of the flexures 100 to the tubular support members 48 can actually be below the top surface of the spar tubes 122. This means that the center of gravity is closer to the elevation pivot axis in this design. This reduces the gravity moment as compared to an assembly in which the members 48 and 72 are in the same plane in which the attachment site to the frame is above the top surface of the spar tubes 122. In one embodiment, the stacking strategy allows a 7.5% reduction in hinge moment about the elevation axis. For an embodiment of heliostat array 21 with mirror panel assemblies 36 having a total surface area of 96 m ² , this corresponds to a hinge moment reduction of 400 Nm when the heliostat array 21 is oriented vertically.

The stacked layout also makes it easier to handle mirror panel assembly 36. The tubular support members 48 are spaced apart from mirror panel 38 by a distance that is at least the height of the upper cross members 72. This provides clearance in the form of gap 66 for one or more persons or for a mechanized handler to easily grip the lower tubular support members 48 for handling.

Mounting studs 88 project downward from cross members 72. Mounting studs 88 are used to attach mirror panel assembly 36 to additional componentry of heliostat 20. In this embodiment, connecting plates 124 (see Fig. 6) couple studs 88 to spar tubes 122. Spar tubes 122 are in turn coupled to components that are used to articulate mirror panel assembly 36 to desired orientations.

Cross members 72 are connected to tubular support members 48 so that the resultant support frame 46 provides a suitably rigid and strong framework to support mirror panel 38. Many different strategies may be used for this connection.

Examples include screws, bolts, welding, brazing, adhesives, rivets, snap fit engagement, threaded engagement, combinations of these, and the like.

Brazing is a preferred technique for connecting cross members 72 to tubular support members 48 at junctures 86 (See Figs. 4, 6, 10, 12, 14 and 15). Brazing provides many advantages. Brazing involves substantially less heat as compared to welding. This means that brazing causes much less frame distortion of brazed assemblies. This is a very important advantage for support frame 46, which is intended to support and maintain mirror panel 38 in a flat condition. Brazing can be used on metal components that are already galvanized. This eliminates the need for post-plating or post-painting, which provides considerable manufacture savings. Hot-dip galvanizing of welded frames can induce frame distortion, which also is avoided by brazing components that are already galvanized. Brazing can also be used for dissimilar metals as well as metals that are difficult to weld without distortion or other degradation. Close tolerances between components is not required, as brazing can fill gaps and still provide strong connections. Assembly is easy because all brazing joints can be accessed from one side of the pre-positioned tubular support members 48 and cross members 72.

A plurality of flexures 100 couple mirror panel 38 to the tubular support members 48. Each flexure 100 extends from a first post end 102 with shoulder 104 to a second post end 106 with shoulder 108. Flattened body 110 having major surfaces 111 (see Figs. 7 and 8) and edges 112 extends between shoulders 104 and 108.

Figs. 7, 8, 9, 12, 14, and 15 show best how flexures are used to couple mirror panel 38 to tubular support members 48. The coupling strategy will be described with respect to a single flexure 100 but can be implement with all or a portion of the flexures 100. First post end 102 of flexure 100 is attached (such as by brazing or welding or any other suitable attachment technique) to a central region 98 of a corresponding mounting pad 92 to provide a first attachment site 114 proximal to the bottom surface 42 of mirror panel 38. The second post end 106 fits into aperture 62 on bottom wall 56 with a close fit to provide a second attachment site 116 proximal to the tube member 48. Flexure 100 passes through relatively large aperture 60 in top wall 54. Aperture 60 is enlarged relative to flexure 100 so that flexure 100 does not contact the aperture edges 61 at top wall 54 when flexure 100 is in an unflexed state. Thus, flexure 100 is spaced apart from edges 61 by gap 118. Gap 118 provides room for flexure 100 to flex to absorb thermal stresses that otherwise might cause mirror panel 38 to distort.

Flexure 100 has an active length of flexure 120 that extends between the first attachment site 114 and the second attachment site 116. In other words, the full length of flexure 100 between sites 114 and 116 is available to help absorb thermal stresses and to help avoid distortion of mirror panel 38. Note that the second attachment site 116 is farther from the bottom surface 42 of mirror panel 38 than the is top wall 54 so that the active length of flexure 120 is longer than the width of gap 66 between tubular member 48 and bottom surface 42. By attaching flexure 100 in this manner, the length of flexure 100 can be longer than if attachment site were more proximal to top wall 54. This connection strategy helps to lower the center of gravity without having to increase the width of gap 66 (i.e., the spacing between mirror panel 38 and member 48). Additionally, the increased length of flexure 100 allows the use of a more robust flexure design (e.g, large cross section) for enhanced strength and durability while still being able to provide the desired flexibility to help accommodate thermal stresses to avoid mirror distortion. A larger flexure 100 also is easier to attach to pads 92 and members 48 using a variety of attachment techniques, such as welding or brazing, preferably brazing.

Fig. 13 shows how flexures 100 can be radially oriented relative to the central region 44 of mirror panel 38 Radial orientation of the flattened flexures allow compliance in the radial direction, while still providing good rigidity against the effects of gravity, for example when the mirror panel 38 is oriented vertically. In some embodiments, the flexures 100 can be deployed in a manner to provide a stiffness gradient in which the flexures provide greater stiffness proximal to center region 44 and less stiffness distal from center region 44. This can be done by making flexures 100 thicker for more stiffness or thinner for more flexibility.

Alternatively, the flattened flexures 100 could be indexed as a function of radial location such that the major surfaces 111 are more parallel to the radial deployment lines 128 (see Figure 13) closer to the facet center 44 and gradually turn more perpendicular as the radial distance increases. As another alternative, the flexures 100 can be very stiff in one direction, to resist gravity, and very flexible in a radial direction, to accommodate thermal expansion. This would allow a common flexure design to be used throughout, without having to implement a flexure stiffness gradient.

This is beneficial because thermal stresses are greater with increasing distance from central region 44. Fig. 13 shows a portion of mirror panel 38 and a portion of support frame 46 and the corresponding mounting pads 92 that are adhesively bonded to panel 38 (not shown in Fig. 13). Radial deployment lines 128 are schematically shown as projecting radially outward from central region 44 through mounting pads 92. To accomplish radial deployment of the flexures 100, the flexures 100 are attached to mounting pads 92 and to tubes 48 in a manner so that the flattened bodies 110 of flexures 100 are perpendicular to the radial deployment lines 128 and face central region 44.

Although this approach requires that flexures 100 be deployed with a radial orientation, this is easily achieved with the help of notches 94. As seen best in Figs. 7 and 8, flexures 100 can be attached to pads 92 in a manner such that the flattened bodies 110 have a known alignment to notches 94. In this particular embodiment, the face of flattened bodies 110 is oriented perpendicular to a reference line extending between notches 94 and center 44. With this approach, if both notches 94 of a pad 92 are placed on the desired radial deployment line 128, then the flexure mounted to that pad 92 will have the desired radial orientation. This orientation can be easily achieved by using a brazing or welding jig that references notches 94 and holds flexure 100 in the proper alignment when flexure 100 is attached to pad 92 by a technique such as brazing or welding.

Figs. 14 and 15 show how the flexures 100 also allow easy adjustment of the height of the mounting pads 92 so that they are all substantially coplanar (or positioned accurately relative to a desired surface profile) after brazing flexures 100 to the bottom wall 56 at second attachment site 116 and the first attachment site 114. This helps to ensure that the adhesive bond line thickness between the pads 92 and mirror panel 38 is substantially consistent. Advantageously, the adhesive bond line thickness can be made to be thicker or thinner as needed among an array of pads 92 to accommodate some degree of error between the desired and actual pad locations. This eases the need to have extremely strict tolerances on the height placement among pads 92.

Fig. 14 shows how flexure 100 penetrates relatively deeply through member

48 to position mounting pad 92, and hence mirror panel 38 (not shown) relatively close to member 48. This position would be secured by attaching flexure 100 in this position at attachment site 116 using a desired technique such as brazing. In comparison, Fig. 15 shows how flexure 100 penetrates less deeply through member 48 to position mounting pad 92, and hence mirror panel 38 (not shown), relatively far from member 48. This position is secured by brazing flexure 100 in this position at attachment site 116 using a desired technique such as brazing.

Fig. 13 shows a flexure deployment in which all flexures 100 have a radial deployment relative to central region 44. However, the flexures 100 can be deployed or have other geometries to help control stiffness of the flexures 100 to optimize the flatness of mirror panel 38, to help hold mirror panel assembly 36 more securely, or to achieve other objectives. For example, when mirror panel 38 is closer to vertical, its weight causes flexures 100 to bend. This could result in gravity-related slope errors. This makes it desirable to use stiffer flexures 100 closer to central region 100. In the meantime, thermal stresses can cause thermal-related slope errors farther from central region 100 if the more distal flexures 100 are too stiff. This makes it desirable to use more flexible flexures farther from central region 44.

Fig. 16 shows how the array of flexures 100 can be visualized as being deployed on reference circles 130, 132, 134, 136, and 138 that are concentric with respect to central region 44. Based on the discussion above, it is desirable that the flexures are stiffer on the smaller reference circles to resist gravity loads while more flexible flexures are used on the larger reference circles to accommodate thermal stresses.

Optimization of flexure stiffness depends on the characteristics of the flexures. Depending on the radial position of flexures 100, flexures 100 are able to have different stiffness values without negatively impacting slope errors due to thermal effects. When mirrors are closer to a vertical orientation, its weight causes flexures 100 to bend, which can cause gravity-related slop errors. Therefore, it may be desirable to optimize the flexure stiffness to reduce thermal and gravity slope error effects. This may be corrected by using stiffer flexures 100 closer to the mirror center to resist gravity loads, and less stiff flexures 100 further from the center.

There are a variety of strategies that may be used to tune flexure stiffness. Maximum flexibility for flexures of a given geometry in a radial deployment occurs when the major surfaces 111 face central region. Stiffness increases when the edges 112 are increasingly presented toward central region. Accordingly, flexures 100 on the innermost circle 130 can be deployed with edges 112 facing central region 44. Flexures 100 on the outermost circle 138 can be deployed with the major faces 111 facing central region 44. The flexures 100 on the other circles can be deployed with intermediate radial orientation so that stiffness decreases in a desired fashion with increasing distance from central region 44. Another strategy for tuning stiffness is to use thicker flexures 100 to provide greater stiffness or thinner flexures 100 to provide more flexibility. With regard to the round-bodied flexures 200 shown in Fig. 18 below, the diameters of the round bodies can be increased for more stiffness or reduced for more flexibility.

Still yet another strategy to increase flexure stiffness is shown in Fig. 17 when using a round-bodied flexure such as flexure 200 of Fig. 18. Fig. 17 shows how reducing the active length of a flexure 200 increases its stiffness. In Fig. 17, this is achieved by fitting flexure 200 through top wall 54 of member 48 (third attachment site 117) with a snug fit while attaching flexure 200 at bottom wall 56 (second attachment site 116) with a more secure attachment such as by brazing. Now the active length of flexure is reduced to the distance between first attachment site 114 and the third attachment site 117. Where greater stiffness is desired closer to central region 44, this connection strategy can be used. Farther from central region 44, flexure 200 would be coupled to member 48 at attachment 116 but not at attachment site 117. As an additional advantage, using a braze connection on only one side of tube 48 makes it easier to access all the braze joints at attachment sites 116 between flexures 200 and tubes 48 from one side of frame structure 46 without flipping the assembly over in order to gain access to the other side to accomplish brazing on that side at the attachment sites 117.

An alternative embodiment of a flexure 200 incorporated into mirror panel assembly 36 is shown in Fig. 18. In contrast to flexure 100, flexure 200 has a cylindrical body 202 that extends from first end 204 to second end 206. First end 204 is attached to pad 92. Second end 206 is attached to bottom wall 56. Annular gap 208 surrounds flexure 200 at aperture 60 to create room for flexure 200 to flex. The stiffness of this kind of flexure shape can be tuned by increasing the diameter for more stiffness or by reducing the diameter for more flexibility. Stiffness can also be increased by further attaching flexure 200 to member 48 at aperture 60. The cylindrical flexures 200 do not require radial or other alignment when being installed since their stiffness is the same in all directions.

All patents, patent applications, and publications cited herein are incorporated by reference in their respective entireties for all purposes. The foregoing detailed description has been given for clarity of understanding only. No unnecessary limitations are to be understood therefrom. The invention is not limited to the exact details shown and described, for variations obvious to one skilled in the art will be included within the invention defined by the claims.