JP3078365U - Light pipe used in rapid prototype of lighting equipment - Google Patents

Abstract

(57) [Summary] [PROBLEMS] Conventionally, there has been a disadvantage that it takes an extremely long time to create a prototype for a lighting system. SOLUTION: The light pipe used in the rapid prototype of the lighting device of the present invention includes a plurality of metallized microprisms and a plurality of non-metallized spacers. Arranged between microprisms, the plurality of spacers and the plurality of microprisms are arranged alternately, transmission of light through the light pipe is performed by total internal reflection, and reflection of light from the output surface is This is accomplished by the plurality of metallized microprisms.

Description

[Detailed description of the invention]

[0001]

[Technical field to which the invention belongs]

 The present invention relates to a light pipe used for a prototype of a lighting device.

[0002]

[Prior art]

 Conventionally, as background lighting devices, U.S. Pat. Nos. 5,390,276, 5,359,691, 5,534,386, 5,609,939, and U.S. Pat. There is 782,962.

In general, a polygon optical device has many small optical elements. The polygonal optical device includes a synthetic mirror and a kaleidoscope. In a folded optical path device, the light beam passes through multiple directions (or a bent optical path), as opposed to a uniaxial device in which the light beam has only one fundamental direction. (Here, "bend" is used interchangeably with "multi-axis.") For example, a conventional periscope is multi-axis but not multi-plane. Background lighting devices, such as those used in laptop computer screens, are multifaceted and multiaxial.

[0004] Optical traces are known for modeling lighting devices. Numerous computer programs have been created for performing optical tracing, including the code V, OPTICAD, radiance, and the like.

When light tracing is performed for multi-axis and / or multi-sided lighting devices, a large number of rays (at least 10,000, and sometimes 1,000,000 or more) are desired to accurately trace the lighting device.

[0006] In order to perform the complicated optical tracing task, the counting time desired by the optical tracing program becomes essential. For example, to perform an optical trace containing 100,000 rays, OPTICAD requires a counting time of 1 to 10 hours using a single computer.

[0007] The counting time desired to perform optical tracing is longer than for more complex multi-axis and / or multi-sided lighting devices. Such devices include background lighting for laptop computer screens, dome lights for automobiles and torrance lighting structures used for ignition key lighting devices. All of these more complex illuminators require on the order of 1,000,000 rays to perform accurate optical tracing. Approximately one day of counting time is required to perform such an optical trace.

In each of the same industries, there are various special products that cannot be mass-produced for individual customers. For example, in a background lighting device for a laptop computer screen, each laptop computer manufacturer starts by creating a slightly different background lighting specification for each type of computer. As a result, not all types of laptop computers can be used accurately with the same background lighting device. Rather, background lighting devices vary slightly in size and / or execution parameters. Therefore, while mass production is still being carried out, it is moving away from the idea of "one size that fits all" in conventional manufacturing.

[0009] More specialized products have led to the rapid creation of prototypes. Prototypes are known as products that have been developed to test the product design before mass production. Prototypes are created to make the product acceptable and satisfy customer requirements. Traditionally, the prototype creation process has required ample time. When a product was developed, it was sold as "one size that fits all" to customers who did not want the specially developed product to be optimal for their particular needs. Spending too much time on the prototyping process for special products is no longer allowed. Product designs are changing rapidly, and the “sustainability period” for a given design tends to be shorter. From the customer's point of view, there is no benefit in purchasing a specially developed product compared to the current one, because of the time delay associated with prototyping. Manufacturers need to match their manufacturing schedules.

In the example of a lighting device, it is necessary to create a prototype of the variation of each product. The product is complex enough that a small change in one parameter will affect the entire product.

[0011] The prototyping process associated with product specialization needs to be as short as possible. Therefore, it is not desirable to use an optical tracing program that optically traces a single illuminator prototype over a day for rapid prototyping. The prototyping process to obtain a lighting fixture prototype that perfectly matches the customer's requirements typically requires about 100 iterations. If an optical trace of one prototype takes one day and the prototyping process is repeated 100 times, the prototyping process requires at least 100 days. This is inconsistent with rapid prototyping, where the final design does not exceed one month in one day or one week.

Thus, in order to rapidly prototype a multi-axis and / or multi-sided lighting device, the time required to repeat the prototyping process once needs to be reduced to the order of 100. In practice, it is desirable to reduce this since more than 100 iterations are required to produce a satisfactory prototype.

Further, there should be only one prototyping / design stage. In other words, when computer prototypes are completed, the prototypes produced should satisfy all customer requirements. Therefore, there is only one manufacturing stage here. Physical prototyping and the manufacturing process require more time and money than computer prototyping, so it is not desirable to physically manufacture prototypes to find hidden defects.

[0014] Therefore, rapid development of prototypes is not generally applied to lighting device development and special light tracing. Current optical tracing programs, such as Code V, are extremely accurate single optical tracing means. Other programs, such as optical radiance, can be used as conventional satisfactory methods (eg, Monte Carlo method) to provide a statistical multi-optical tracing algorithm. All of these programs are slow, and no programs have been implemented that transfer data in the preferred format. For example, these programs produce optical coordinates, direction, and density information, which is not translated into photometric quantities that allow the designer to easily determine running parameters for the quality of the final product. Furthermore, there is no known method for converting optical trace information into a three-dimensional visible image when a visible surface image cannot be obtained. Furthermore, there is no known method applied to a computer network for performing optical tracing to reduce the counting time.

[0015]

[Means for Solving the Problems]

 The above disadvantages of the prior art can be obviated by a prototype rapid-producing apparatus for a lighting device which is at least one of a bent path and a multi-facet whose output power is determined by the counted phase space density. Preferably, the output power is determined using light intensity (ie, flux, brightness, etc.). Unlike the amount of radiation, the photometric amount can be seen by human eyes and is preferable. Luminance is a basic photometric quantity and can be directly measured by the phase space format. The use of the phase space format has the advantage that the counting time desired to produce useful information about the lighting fixture prototype can be reduced and the prototype can be rapidly created.

The light pipe used in the prototype rapid manufacturing apparatus of the present invention is formed by a plurality of separation microprisms separated by a plurality of spacers, and the microprisms are metallized, and the plurality of spacers are not metallized. In this light pipe, the light is transmitted by total internal reflection, and the light is reflected by metallization. Therefore, the loss due to metal absorption is minimized, the shape tolerance of the microprism is maximized, and the number of device modules is minimized.

Other objects and features of the present invention will become apparent from the following description of the drawings.

[0018]

[Embodiment of the invention]

 Hereinafter, embodiments of the present invention will be described with reference to the drawings.

The phase space form is applied to optical traces that model photometric quantities such as brightness (brightness), intensity (emissivity), and flux (power). Photometric light traces facilitate the construction of incident light traces based on quasi-homogeneous light source light intensity and spatial coherence. As explained below, metering trace model spreads spatial coordinates (x, y) to the direction unit vector coordinates _{(k x, k y, k} y) four-dimensional defined by (4-D) phase space. Here is a ^{_{k x 2 + k y 2 +}} k y 2 = 1. Spatial coordinates (x, y) whereas defining a space d rear direction coordinate (k _x, k _y) is shall be determined the called Fourier (direction) area.

The basic coupling between photometric and 4-D phase space is based on the fact that the brightness (brightness) is 4-D phase space brightness. Thus, any constraint on light passing through a given plane (x, y) can be shown as multiple points in the phase space (ie, a single light becomes identically shaped by a single point in the phase space). Shown). In particular, the Lambertian light source is indicated by a uniform 4-D distribution in the phase space.

The basic photometric physical quantity is cd / m^TwoIs the luminance within the range. The present invention provides a uniform photometric physical quantity, that is, W / m. ^Two Can be used in conjunction with the brightness (brightness) within. Similarly, luminance can be replaced by radiation intensity or flux in radiometry. The present invention will be described as a photometric unit.

The basic relationship between luminance and phase space density or 4-D light intensity can be used to evaluate photometry by numerically calculating the amount of light passing through a selected phase space domain. . For example, the luminance can be calculated as the number of lights located in the phase space elementary cell defined by the Heisenberg instability. On the other hand, the emissivity may be calculated by integrating the elementary cell through the direction vector space (k _x, k _y). As a result, the described format allows for accurate coupling of Monte-Carol optical traces for multiple lights to the basic photometric amount.

Consider a plurality of lights passing through a given surface (physical or non-physical). This light passes through the surface at different coordinates and in different directions. Each light passing through this plane, four coordinates; can be uniquely indicated by _{(x, y k x, k} y). Here x, y is the point of intersection on the surface, k _x, k _y is the direction cosine indicating the direction of the light at the intersection. The totality of these lights can be represented by the full set of coordinates (x, y; k _x , k _y ). Coordinates _{(x, y; k x,} k y) indicates a four-dimensional position space.

The directions and positions are therefore in the same four-dimensional space. This coordinate system is an abstract away from the main observations.

FIG. 1 shows the intersection of light passing through a predetermined plane. Angle when intersecting the light, the spatial coordinates x and y (This is indicates the point in the plane where the light crosses), the direction cosine unit vector k _x and k _y (this is the light intersects the plane Is shown in a special form.

This light has an angle α with respect to the x-axis, an angle β with respect to the y-axis, and an angle θ with respect to the Z-axis. The above parameters are shown in Equation 1.

[0027]

(Equation 1)

The total light passing through the plane is represented by the full set of spatial coordinates (x, y) and directional coordinates (k _x , k _y ).

The coordinate k_xAnd k_yIs a unit vector component. There are three coordinate cosine (k _x , K_y, K_z) And the sum of its squares equals 1, so there is only one degree of freedom. Therefore, the direction of the light is directional cosine coordinate k_xAnd k_yIs shown in a unique way by

The spatial coordinates alone are not enough to indicate the intersection of light. The two lights intersect the plane at the same point, but in different directions. In this example, the Cartesian coordinates are the same, but the directional coordinates are different.

FIG. 2 shows three lights that intersect the x, y plane. The first light coordinates; having _{(x 1, y 1 k x1} , k y1). The second light coordinates; having _{(x 2, y 2 k x2} , k y2). Third light coordinates; having _{(x 3, y 3 k x3} , k y3). The second and third lights have the same coordinates (ie, x ₂ = x ₃ and y ₂ = y ₃ ) but different directional coordinates. Therefore, they are at different positions in the four-dimensional phase space.

In a four-dimensional coordinate system, each light has a unique position. Even if the second and third lights intersect the x and y planes at the same position, they are at different positions in the directional coordinate space because their directional coordinate spaces are different. Only two lights with the same (same position and direction) have the same point in the 4-D phase space.

FIG. 3 shows the four-dimensional phase space diagram of FIG. 2 reduced to a two-dimensional diagram by setting y = 0 and k _y = 0 for each of the three lights. Since it is impossible to describe a four-dimensional space, the description is reduced to the two-dimensional coordinate space.

FIG. 4 shows a two-dimensional view of the light shown in FIG. The second and third lights have the same coordinates along the _x axis, but different coordinates along the k _x axis. Since the second light is tilted to the left, the second light has negative coordinates along k _x , while the third light is tilted to the right so that the third light has a k _x axis Along with positive coordinates.

Of course, in an actual lighting device, the number of lights is on the order of 10,000 to 10 million. Counting each light path individually is extremely expensive. However, the density of light (number of lights per unit area) can be determined. The total surface area of the output surface of the lighting device is divided, and the density of light per unit area is obtained from the total number and total area of light. Since the total number of light passing through the surface is proportional to the power (either the subject or the target unit), the light density in four-dimensional space can be combined with luminance as described below.

The analysis of lighting devices that use phase space for the most abundant light tracing programs to produce spatial and directional information can be easily accomplished.

However, in the absence of an optical tracing program, spatial and directional information is combined in a single coordinate device. However, phase space analysis can be facilitated in coordination with existing optical trace programs.

For the purpose of optical trace photometry, three basic photometry (flux (P) amount, luminance intensity (J), and luminance (B)) are described with respect to a four-dimensional phase space (x, y; k _x , k _y ). Determine. Here, the arrangement of light in the phase space is shown in FIG. As shown in FIG. 5 (x, y) is a directional coordinates (physical or non-physical) spatial coordinates of a predetermined plane (k _x, k _y) is the direction cosine space, the directional unit vector outer 1, the direction cosine _{(k x, k y, k} z) is represented by the number 2.

[0039]

[Outside 1]

[0040]

(Equation 2)

Using the phase space notation, the basic photometry can be determined by several equivalent equations shown in Equations 3 to 11.

[0042]

(Equation 3)

[0043]

(Equation 4)

[0044]

(Equation 5)

[0045]

(Equation 6)

[0046]

(Equation 7)

[0047]

(Equation 8)

[0048]

(Equation 9)

[0049]

(Equation 10)

[0050]

[Equation 11]

[0051] Here dω solid angle _{element, dΩ = dAdk x dk y,} dΩ is throughput (between phases empty) element, dA = dxdy.

Equation 3 is a general equation for power defined as a function of luminance and solid angle. Equation 4 determines the power of luminance as a function of phase space coordinates. Importantly, the above formula or et luminance (i.e., B (x, y; k x, k y)) is the arc seen to be equal to the phase-space density. Consequently, four-dimensional _{(dx, dy; dk x,} dk y) over the integrated the position phase space density to make the flux (power).

To aid understanding, compare Equation 4 to the more common Equation 12 for mass.

[0054]

(Equation 12)

In Equation 12, the mass density is ρ (x, y, z). The mass is obtained when this parameter is integrated with the capacitance elements dx, dy, dz.

[0056] Similarly, the number 4 dx, dy, dk _x, when integrated over dk _y, the total number of the light passing through the X-Y plane are obtained. From Equation 4, it can be seen that the number of lights is proportional to the luminance. This is a very effective means to obviate the need to count light individually. The number of lights is specified by the photometric amount according to Equation 24.

[0057] Equation 5, is equal to the number 3 except replacing the term cosθ in k _y. Since cos θ is a directional cosine in the Z direction (that is, k _z = cos θ), the above substitution can be made.

Equation 6 defines power as a function of radiation. Direction by the number 6 (dk _x, dk _y) integration is already achieved. The radiation is lumens / m ² and is equal to the number of light passing through the unit area of the XY plane.

Equation 7 includes the integration of dx and dy and the multiplication of cos θ. Equation 7 is obtained by replacing the right side of Equation 3 with Equation 10.

The processing amount Ω of the apparatus is represented by Expression 13.

[0061]

(Equation 13)

Referring to the above comparison for mass, throughput is similar to volume unit ∫∫∫dxdydz. _{dx, dy, dk x, dk} y is an elementary phase space volume limit. (Here, "volume" is used in a mathematical sense, not a physical meaning.)

The measurement surface (x, y) is a physical or non-physical (radiation) source surface or light receiving surface, or another surface. In the example of the Lambertian plane (x, y), the brightness (B) is constant over the entire plane in space and directional coordinates.

In this case, Equations 10 and 11 decrease as shown in Equations 14 and 15 (Lambert method).

[0065]

[Equation 14]

[0066]

(Equation 15)

Here, J ₀ = B ₀ ΔA and E = BΔΩ = πB.

The total throughput is equivalent to half of the total solid angle (2π) and equal to π. Therefore, it is equal to P = BΔΩ.

From Equation 13, the basic relationship between the elementary processing amount and the solid angle is obtained as dΩ = cosθdωdA.

Therefore, there are Expression 16 and Expression 17 for determining the luminance.

[0071]

(Equation 16)

[0072]

[Equation 17]

In equation (16), cos θ is in contact with the surface element ΔA (therefore, a surface perpendicular element ΔA1 is created), whereas in equation (17), cos θ is in contact with the solid angle element. Typically, Equation 16, preferably Equation 17 is used. This is because the brightness is defined as the phase spatial density, [rho: [rho; is _{(x, y k x, k} y) from applicable to metering light trace in the most direct means or several 18.

[0074]

(Equation 18)

_{A; (k x, k y x} , y), dΩ = dx dy d k x dk y in this phase space elementary volume; [0075] where _{ρ (x, y k x,} k y) = B.

As described above, in the example of the Lambertian surface, the brightness B is constant over the entire surface in both spatial and directional coordinates. Thus, light is uniformly distributed in both spatial and directional coordinates.

All light sources except the laser are Lambertian. Therefore, if the output of the light source is known, the light distribution can be easily determined. The luminance information can easily be converted to a light distribution. The same applies to other aspects (physical or non-physical) of the lighting device. )

The definition of the elementary four-dimensional volume unit dΩ is repeated, defining a spatial and directional area. Here is a _{dΩ = dx dy dk x dk y} , dx dy = elementary space area, dk _x dk _y = elementary Fourier (directional) area.

When k _z is 0 (non-elementary), the Fourier area is maximized. In this example off Rieeria (k _x k _y) is equal to [pi. Thus, for any direction of light light coordinates (k _x, k _y) is always k _x, in the unit circle on the k _y plane. Furthermore, the light source direction coordinates of the light of all because it is La Nbaten is k _x, is equal rather distributed k _y units in the circle. This is an example of a complete run bar ten, corresponding to FIG. 6A.

An example of a restricted run bar ten is shown in FIG. 6B. In this example, the light distribution is uniform, but the area is limited because the curvature is smaller than one. In this example, the light is uniformly distributed, but only the range of directions is limited. Thus, in FIGS. 6A and 6B, the light source is isotropic and Lambertian, while in FIG. 6B, the light source is limited to Lambertian.

Examples of the anisotropy are shown in FIGS. 6C and 6D. In this example, the distribution of light directions is not the same in each direction.

As shown in FIGS. 6A to 6D, the parameters of the Fourier area are as follows.

Example of FIG. 6A: Lambertian, F ^(a) = π

Example of FIG. 6B: Limited Lambertian, F ^(b) = πsin ² α

Example of FIG. 6C: rectangle-anisotropic, F ^(c) = π sin α _x sin α _y

Example of FIG. 6D: ellipse−anisotropy, F ^(d) = π sin α _{x ′} sin α _{y ′}

Although the example of FIG. 6A can be obtained by Expression 13, the above-mentioned means can be obtained more easily.

FIG. 7 shows a 2-D phase space with a uniform luminance distribution in the linear segment of the Lambertian light source. This Lambertian light source has a length L _x , an omnidirectional angle (k _{x max} = ± 1), and an elementary cell (Δx, Δk _x ).

In the example of the Lambertian light source, the luminance is made constant in the phase space according to Expression 14. In contrast, as shown in FIG. 7, each ray is represented by a single, unique point in the 4-D phase space, and the phase space density ρ is improper for lossless optics according to Riubiru's theorem. It becomes a variable. This cannot be explained in 4-D space. The luminance point can be described using only the 2-D phase space (x, k _x ). FIG. 8A illustrates a 2-D luminance distribution in a Lambertian light source straight line. However, according to Lambert's law, this distribution must be uniform for one elementary cell (Δx, Δk _x ).

Consider the Lambertian point light source in FIG. 8A to convert a uniform phase space distribution into a regular optical trace as shown in FIG. The outer 1 distribution is uniform in the directional (k _x ) space and not uniform in the angular space (J = J ₀ cos θ). On the contrary, a uniform angular distribution as shown in FIG. 8B indicates a Lambertian light source. FIG. 8A corresponds to a Lambertian light source, and FIG. 8B corresponds to a light source having a uniform radiation intensity.

This can be understood more easily from Table 1.

[0092]

[Table 1]

As is clear from Table 1, a uniform distribution along the k _x axis does not result in a uniform angular distribution. For example, the angle difference between k _x = 0.8 and k _x = 1.0 is only 11.5 degrees. Therefore, surprisingly, the Lambertian point light source cannot obtain a uniform light distribution when plotted in spatial coordinates as shown in FIG. 8A.

The above findings can be applied to a Lambertian cylindrical light source. These light sources have many practical uses, such as fluorescent sources. In this case, the light source is locally Lambertian, and the luminance is constant in a predetermined direction despite the curved surface. (Therefore, there is no difference between a Lambertian cylindrical light source and a Lambertian flat light source). For example, the sun and moon are Lambertian spherical light sources. Therefore, both are observed as Lambertian flat plates. In practice, as shown in FIG. 9, the flux element or the luminance element in the Z direction is proportional to the surface element dA and cos θ according to equation (15).

Here, dJ ₀ = Bcos θdA, and dx = dAcos θ as shown in FIG.

Therefore, Equation 19 becomes uniform as a function of x.

[0097]

[Equation 19]

Expression 4 is quantized into Expression 20 in order to indicate the measured light amount as the light density.

[0099]

(Equation 20)

Here, B _ijkl is the average luminance value in the elementary four-dimensional cell ΔΩ, and has the form of Equation 21 and Equation 22.

[0101]

(Equation 21)

[0102]

(Equation 22)

Equation 23 is obtained by applying the quantization rule.

[0104]

(Equation 23)

Here, N is the total number of light rays, and _nijkl is the number of light rays passing through the elementary cell ΔΩ (i, j, k, l). In other words, each light passing through a given plane (x, y) is represented by a single point in the 4-D phase space, so that _nijkl is the number of light points inside the elementary cell. Similarly, the other photometric amounts as the number of light beams can be represented as the radiation shown in Expressions 24 and 25.

[0106]

(Equation 24)

[0107]

(Equation 25)

This relationship is shown in FIG. 10 for a 2-D example. Here, the set of rays passes through the same segment, but in different directions. These rays, indicated by points in phase space, are then located in elementary cells having the same x coordinate. Next, the number of light spots is counted in each elementary cell (ijkl), and these numbers are added along the i-th column in the phase space vertical. The total number of lights in this column is equal to the number N _ij according to the _{equation (} 21). The number of light beams to be added is arbitrary, but the above number cannot be too large to avoid lengthening the counting time. Also, in order to perform smooth integration, the above number cannot be so small, and is preferably _set to _nijkl to 10.

FIG. 11 shows a background lighting device. This background illumination device is mainly used in a laptop computer, and has a light engine LE, a light pipe LP, and a screen output surface. The light engine LE has a Lambertian cylindrical light source LS and a collimator C. The light pipe LP has at its bottom a row of microprisms for directing light from the light engine LE to the output surface. At the top of the output surface is provided a spatial light modulator (SLM) that modulates light from the output surface based on signals indicating information to be presented to the user. Further, a diffuser on the top surface of the SLM, a film between the light engine LE and the light pipe LP, and a film below the SLM are provided as necessary.

In order to rapidly create a prototype, each of the modulators (modules) is separated and analyzed. Normally, one module is modulated at a time. As a result, if the design changes, it is only necessary to re-analyze the modulated module, not the unchanged module. This saves counting time and speeds up the creation of prototypes.

The above phase space format can be used to quickly predict the output power of one element of a background lighting device, for example, a light pipe. The means are as follows. First, the shape of the hypercube is determined. A hypercube is a four-dimensional cube and is an elementary cell in a four-dimensional phase space coordinate system. For determining the hypercube, x, y plane has a range defined with x values, it is assumed that both the direction cosine k _x and k _y is changed to +1 to-1. Therefore, _x, y, k x, all the expected values of k _y are known. Next, the hypercube is divided into uniform elementary cells (see Equation 21). Then assume the total number of rays and the total number of cells. The total number of rays and the total number of cells are selected such that at least some (preferably about 6-8) rays pass through each elementary cell. The total number of rays was found to be in the range of 100,000-1,000,000, and the total number of cells was in the range of about 10,000. Finally, the relationship that the luminance equals the phase space density was applied to the output power prediction. The following example illustrates the details.

(Example 1)

Equation 26 was obtained assuming the following parameters. Lamp brightness B₀= 39,000 cd / m^Two(Lambertian); light source diameter D = 2R = 2 mm; light source length L = 30 cm. Incident flux P ₀ = (Surface area of cylindrical surface) (Lamp brightness) (Fourier area for all Lambertian light sources)

[0114]

(Equation 26)

Next, each light beam was given a weight (lumen / light). If this input power is equal to a total of 100,000 rays (ie, N ₀ = 100,000), then the light density is given by Equation 27.

[0116]

[Equation 27]

A power factor is assigned to each ray according to Equation 27. Light rays travel through many interfaces as defined by conventional optical tracing programs. There is material loss such as reflection loss at each interface. However, the weight given to each ray is reduced by these losses. Statistical averages can be used for illustration.

Next, assume that the optical efficiency of the device is η = P / P ₀ = 70%.

If N = 70,000 rays pass through the end (assuming that each ray has a loss of either 100% or 0%), the output power is 0.002 lm / light and P = ηP ₀ = 162 lm Becomes

Next, it is assumed that L = W = 30 cm and the total number of elementary cells is M = 10 × 10 × 10 × 10 = 10 ⁴ .

FIG. 12 shows a 10 × 10 spatial coordinate device, and FIG. 13 shows a directional coordinate device. Each of the 10 × 10 elementary cell shown in FIG. 1 2 has its own k _x k _y plot (Figure 13).

In this example, it is assumed that the shapes of the elementary cells are equally distributed over the directional coordinates and the space coordinates. In other words, 10 × 10 cells are shown in directional coordinates and 10 × 10 cells are shown in spatial coordinates. However, the means of allocating cells between spatial and directional coordinates is flexible and can therefore be different. For example, the following allocation may be used: M = 20 × 20 × 5 × 5 (ie, 20 × 20 cells shown in the directional coordinate space and 5 × 5 cells shown in the spatial coordinate space). This will increase the accuracy in the directional coordinate system. The range between k _x = 0.8 and k _x = 1.0 gives a difference of 37 degrees, which is often quite large.

Based on M = 10 × 10 × 10 × 10, the phase space elementary cell Ω ₀ is Ω ₀ = 0.2 × 0.2 ster × 3 cm × 3 cm, and the average number of light rays passing through this elementary cell <n > Is <n> 70,000 rays / 10,000 cells = 7 rays / cell.

Therefore, the average flux P _{B for} each cell is P _B = 7 × 0.00231 lm / light = 0.02 lm, and the average output luminance <B> is as shown in Expression 28.

[0125]

[Equation 28]

The luminance efficiency x is X = 555 / 39,000 = 1.5%.

In this example, this output is extremely dark and the viewing angle is unnecessarily large.

(Example 2)

In this example, the same parameters as in Example 1 are assumed. However, in this case, the output light is collimated so that the average of the output light is limited to within the same y angle α _y = ± 11.5 ° as the x angle α _x = ± 11.5 °. As a result, (SIN ^-1 (11.5) = 0.2, since the light is seen emitted from the cell with k _y = -0.2,0 or +0.2 with k _x = -0.2,0 or +0.2 ) The number of occupied elementary cells is reduced to 5 × 5 = 25-factor, and the average output luminance is <B ₂ > = 555 × 25 = 13,875 cd / m ² .

In this example, the output is extremely bright and the viewing angle is extremely small.

(Example 3)

In this example, the same parameters as in Example 1 are assumed. However, in this case, the output light is collimated so that the average of the output light is limited within the same y angle α _y = ± 30 ° as the x angle α _x = ± 30 °. As a result, the average output luminance is <B ₃ > = 555 × 4 = 2,220 cd / m ² . This matches the typical brightness values for laptop computer screens.

The spatial and directional light information organized according to the phase space coordinate device as shown in FIGS. 14A, 14B and 15 is used to create a two-dimensional graph showing the performance characteristics of the lighting device.

FIG. 14A is an example of a background lighting device having output surfaces x and y. The phase space coordinate system for background lighting is organized in a M = 20 × 20 × 5 × 5 coordinate system. Therefore, the spatial coordinate space consumes 5 × 5 cells (ie, the output plane x, y is divided into 5 × 5 = 25 elementary cell output planes as described above), and the directional coordinate space is Consumes 20 x 20 elementary cells. Phase space coordinates B for a set of 10,000 elementary cell _{(x, y; k x,} k y) complete set of is obtained for background illumination device. As mentioned above, this information is obtained using a conventional optical tracing program (even though the conventional optical tracing program is not organized as phase space coordinate system information). For example, assuming that it is also desirable to know the output characteristics of the exemplary cell x ₄ y _3, preferably to know a change state of the luminance for cell x ₄ y ₃ different output directions of excitation light the alpha ( The angle α is defined in relation to the x-axis as described above, which is a changing parameter k _x ).

FIG. 14B shows an exemplary cell x_Foury_ThreeCoordinate space B (x_Four, Y_Three; K_x, K_y ). (Directional coordinate space B (x_Four, Y_Three; K_x, K_y) Is cell x_Foury_Three, But for each other cell x_i, Y_iIs a similar directional coordinate space B (x_i, Y_i; K _x , K_y)). Since it is only desirable to know the degree of change in luminance for different output directions α (as opposed to different output directions α and β), only a two-dimensional graph is desired and the directional coordinates k_yIs kept constant. Value k for the purpose of selecting an intermediate range value_y= 0 is selected. k_x= Coordinate B (x for −1.0 to +1.0_Four, Y_Three; K_x, 0) is the highlighted portion in FIG. 14B.

FIG. 15 shows a two-dimensional graph of B (x ₄ , x ₃ ; k _x , 0). This graph shows the luminance values for each of the 20 values of k _x . These data points have been interpolated to get a smooth curve.

FIG. 15 shows a cross section obtained at the point of change of luminance k _y = 0 when the parameter k _x changes from −1.0 to +1.0. made a similar graph for each of the 20 different values of k _y, three-dimensional graph of the change in luminance as a function of k _x and k _y for cell x ₄ y ₃ by combining the graphs of these 20 Can be obtained. Interpolate the data points again to form a smooth surface. Similar two-dimensional and / or three-dimensional luminance graphs can be obtained to leave the 24 cells that define the output surface of the described background lighting device.

The following preferred means can be used to obtain photometric light trace information. First, catalog type data (eg, light source shape, luminance, absorption constant, refractive index, micro prism shape, etc.) is selected for light source display. Next, a Lambertian distribution of light is created on the light source surface using the phase space method. Next, an optical trace rule is created for each signal light using a computer program. Next, let us assume the number of light inputs and the number of output elementary cells, assuming that there are several rays per cell. The weight per light or power reduction is then counted. Next, a photometric data acquisition device is determined, and the phase space coordinates of the output light are converted into a cell arrangement. In other words, output to the optical coordinate _{(x i, y i; k} xk, k yl) gave, assign light to special elementary cell by the device. The total weight of light for each cell is then counted to determine the amount of output power per elementary cell. Next, for a given cell, the output power of all cells is divided by the throughput of the cell (see equation 28). The result of this division is the average luminance for a given cell. Next, this operation is repeated for all cells to obtain a luminance distribution per cell for the range of cells (see Equation 23). For example, when the number of cells is 10,000, there is a value of 10,000 indicating the luminance output. Eventually, as described above, the data is presented in the form of a two-dimensional or three-dimensional graph.

The spatial and directional light information organized according to the phase space coordinate system is used together with a cost function. For example, the least mean square error (MMSE) shown in Expression 29 is used.

[0140]

(Equation 29)

[0141] Here, B _MAX (i, j) is the maximum luminance output for a given cell for all values of k _x and _{k y (x = i, y} = j), B AVG total The maximum average brightness output for all cells. This function is used to make the luminance distribution on the output surface uniform. If the luminance output is uniform, then in all cases B _MAX (i, j) = B _AVG and the output C of the cost function is equal to zero. If the luminance outputs are not entirely uniform, then B _MAX will not be all equal and the cost function output will not be zero. By minimizing the output C, a uniform luminance output can be obtained.

FIG. 16 shows an example of parameters optimized by using the above cost function. FIG. 16 shows a microprism (groove) MPM of a light pipe. Ideally, the microprism is triangular. However, manufacturing imperfections round one of the corners. The degree of the roundness is measured by the height h with respect to the distance g. The other two corners are the corner α ₁ And α_TwoMeasure with Parameters h, g, α₁, Α_TwoBy changing the above and adding the above cost function, the performance of the light pipe is optimized.

The photometric light trace can be applied to geometrically-optically related problems. However, photometric traces can be applied by other means, such as areas where physically-optically relevant phenomena are revealed by soft means. In such an example, the optical interference (and diffraction) capability is very low, while the optical intensity should change gradually with wavelength. Such a weak spatial coherence source is called a quasi-homogeneous source. These are very common in nature and industry, and include all heat sources, fluorescent sources, LEDs, most of semiconductor lasers, plasma sources, and the like.

FIG. 17 shows a non-Lambertian diffuser with angular properties different from those defined by Lambertian's law. This non-Lambertian diffuser is used to scatter light. Thus, unlike other optical devices where the number of rays is doubled or tripled, a non-Lambertian diffuser increases the number of rays by about 100 times.

Conventional optical tracing techniques have been applied to non-Lambertian diffusers ignoring scattering effects. However, the complexity of the optical tracing problem increases surprisingly because the number of rays increases surprisingly. For example, a trace of 100 million rays would be required instead of a trace of one million rays. That is, the counting time becomes longer.

However, if, for example, the input surface of the diffuser is optically coupled to the output surface of the spatial light modulator (as described below), the spatial light modulator is coupled to the output surface of the light pipe. Another approach may be used if the light pipe has an input face coupled to the output end of the light engine. Generally, this different approach involves optical tracing from the light source to the output of the spatial light modulator. Integral equations can be applied to the output of the spatial light modulator. These equations are linked to empirically defined information, which greatly speeds up the modeling of illuminators by eliminating the need for optical tracing through diffusers. This increase in speed increases the speed of prototyping.

_{Passing; (k x, k y x} , y) in the form of a (diffuser; [0147] B 'and _{(x, y k x, k} y) of the form (without diffuser) output brightness distribution, B Consider the final luminance distribution). According to the homogenous linear device theory, the relationship between these two luminances is a spiral like Equation 30.

[0148]

[Equation 30]

Here, h (k _x , k _y ) is the point response of the device and does not depend on the position (x, y) for a homogeneous diffuser. The number 30 is related to the luminance output of the diffuser with respect to the luminance input as a function of _{h (k x, k y)} . Function h (k _x, k _y) is determined experimentally. Therefore, the luminance output when a luminance input is applied to the diffuser can be determined using Equation 30.

The point response characteristic of the function (h) can be easily changed by replacing the Dirac delta function distribution with B′-luminance as shown in Expression 31.

[0151]

(Equation 31)

Here, δ (...) is a Dirac delta, and F (x, y) is a smoothing function. By substituting Equation 31 for Equation 30, Equation 32 is obtained.

[0153]

(Equation 32)

Accordingly, the h function is actually a device point response. From the law of conservation of energy for a lossless diffuser (ignoring Fresnel losses), a uniform radiation is obtained in equations 33-35.

[0155]

[Equation 33]

[0156]

(Equation 34)

[0157]

(Equation 35)

By substituting equation (30) for equation (34) and using equations (33) and (35), a standardized relationship shown in equation (36) is obtained.

[0159]

[Equation 36]

Accordingly, the h-point function is standardized to a unit value. Because it restricted to the integral homogeneous changes in the number 34, namely, uniformity when _{^{k x 1 + k y 2 ≦}} 1 is maintained. Therefore, the integration operation extends indefinitely, and the shift of the discussion (k _x → k _x -k _{x '} ) can be ignored.

The function h (k _x , k _y ) as a point response in the angular space (k _x , k _y ) is represented by the angle characteristic of light having an incident surface wave passing through a homogeneous diffuser as shown in FIG. It can be obtained experimentally by measuring. This characteristic can be standardized by Equation 36.

The light source luminance for the quasi-homogeneous source has the form of Equations 37 and 38.

[0163]

(37)

[0164]

(38)

I ₀ is the optical intensity, ω is the angular frequency, c is the speed of light in vacuum, 2 is the 2-D Fourier transform of the spatial coherence complexity, and 3 is the form of Equation 39.

[0166]

[Outside 2]

[0167]

[Outside 3]

[0168]

[Equation 39]

Here, d ² r = dxdy. Expression 40 is obtained by comparing Expression 37 with the luminance intensity definition.

[0170]

(Equation 40)

For example, if B = constant (Lambert's method), J becomes cos θ and B becomes 1 / cos θ according to Expression 38. Generally, if B is cos ⁿ θ, as shown in Table 2, J is cos ^{(n + 1)} θ, and outer 2 is cos ^{(n + 1)} θ.

[0172]

[Table 2]

Table 2 shows the Fourier transform of the coherence (μ) complexity and typical angular properties for luminance (B) and luminance intensity (J).

A typical light source is a Lambertian light source, but a typical luminance surface is non-Lambertian because its angular characteristic is narrower than that of a Lambertian light source.

As a result, their spatial coherence radius is usually higher than λ. Heisenberg's uncertainty principle for conjugate dispersion coefficients (ρ and Δρ) indicates this as ρ × Δρρλ.

Here, ρ is the spatial coherence radius. If Δρ = 1 (in the case of non-coherence), ρ to λ are obtained. That is, the spatial coherence radius is equal to the wavelength. Consider generally narrower light, ie, Δρ = 1 / ε; ε ≧ 1.

Ρ = ελ is obtained from Expression 39. That is, narrow light has high spatial coherence and ρ> λ.

For rapid prototyping, it is important to eliminate unwanted artifacts by computer modeling that appears in the lighting device being created. The design process should be completed when computer modeling is complete. This design process should not be integrated into the manufacturing process because the process of making a single prototype is time consuming and expensive (on the order of $ 50,000). The manufacturing process should be only the manufacturing process of the lighting device to be sold to the customer. 18 and 19 show examples of undesirable artifacts. FIG. 18 shows a background lighting device physically coupled by the plane CDE-FC-D'-E'-F '. The output surface of the background device (screen) is defined by surfaces EFE'-F '. Surface C -D-C'-D placed 'microprism structure (not shown) reflects light rays from the light engine LE consisting tubular light source having a luminance of about 40,000 cd / m ^2. This light beam is directed to an output surface EFE'-F 'which is used, for example, in conjunction with a laptop computer screen. The microprism structure creates an optical illusion such that the light source extends in the plane ABA'-B ', and is consequently partitioned by the plane ABCDA-B'-C'-D'. Create a virtual space.

A visible inclined plane BDBD that can be seen by the user through the output plane EF-E'-F 'is formed in the virtual space. This focal plane is an undesirable image artifact. This artifact is tilted because the microprism is closer on the left side than on the right side of plane CD-C'-D '. FIG. 19 shows an arbitrary cross section Ky-K'-y 'of the lighting device shown in FIG. A series of virtual light sources (from the multiple reflection effect of the microprism structure) extend from point K to point H. This virtual light source creates a focal line, which is combined with another focal line to create a focal plane, ie, an image artifact. Image artifacts can be shown experimentally by marking the black ring on the light source LS, which is a segment on the darkened surface BD-B'-D '.

It is difficult to detect artifacts as shown in FIGS. 18 and 19 without looking at the illumination structure three-dimensionally. However, this approach is not acceptable for rapid prototype creation for the reasons mentioned above. It is preferable to detect such artifacts by modeling the lighting device with a computer. That is, it is preferable to use a computer to three-dimensionally show a lighting structure exhibiting such artifacts.

In order to perform three-dimensional display using an automatic computer tool, photometric light tracing in three dimensions is performed. During this process, focal points, focal lines, and focal planes that can use the optical trace are detected. These are the boundaries of the image artifacts.

However, the photometric light trace described above is based on a cross section of light passing through a predetermined two-dimensional (x, y) plane. Therefore, it is necessary to make a plurality of two-dimensional cross sections to make a three-dimensional visible image.

In order to create a plurality of two-dimensional cross sections, it is necessary to select a plurality of Z planes in cooperation with the photometric trace. Parameters _{(x, y; k x,} k y) characterizes the set of light passing through the predetermined Z plane, thus these parameters depend on the Z plane is selected. However, image artifacts are usually only seen in a limited range of oblique angles, and it is not possible to know in advance which Z-plane will be chosen to create objectionable image artifacts.

The following approach is made to eliminate this difficulty. First, the virtual space is divided into cubes. The cube is then sliced in parallel and photometric traces are made for each slice. These steps are repeated for all different observation points in the direction defined by the Z axis. In general, the lighting structure is only visible from a limited viewing range, for example ± 30 °. Therefore, these steps need not be performed over the entire 360 °, but only over a limited viewing range, eg, 60 °. Next, the individual monoscopic orientations are rearranged into pairs to create a stereoscopic view (pairs of monoscopic orientations organized by the stereoscopic method). Therefore, only these high-brightness stereoscopic views are selected for optical noise cancellation. This is done using thresholds, levels, and discards all fields of view having a brightness value of, for example, 3 nits (nit = cd / m ² ) or less.

(Here, it is assumed that an arbitrary artifact has a luminance value of 3 nits or more). All of these steps can be performed automatically by a computer.

The above direct method of applying photometric traces to create a three-dimensional visible image can be replaced by tomography. This approach presupposes the existence of some focal plane (or focal plane) and is based on combining this preliminary information into the geometric location of the visible light cross section.

When the data has been aggregated, the individual designs of the lighting device are shown. These data are preferably in the stereoscopic view of the illuminator, ie in three-dimensional form. Thus, this data provides a format that is easily recognizable by humans. The proposed format by individual comparison with raw data on the spatial and directional coordinates of all the light-carrying routes is not general (because the lighting structure is taken out only from a specific field of view) and has a lot of information Does not have. However, it can be more easily determined and thus provide information in a more useful format. This format produces data that looks at the actual lighting structure in the same way, ie one angle at a time.

There are a variety of conventional computer graphing techniques for presenting data in a three-dimensional format. According to this direct approach, each cube defines a local part of a local surface. A coupling operation such as a marching cube method is employed. This results in a focal plane showing unwanted visible artifacts that must be eliminated by the appropriate design of the lighting device after the end of the coupling operation. Other conventional volume representation methods can also be used.

FIG. 20 shows an inverted light trace used in the present invention to further reduce the counting time. As far as conceivable, the above optical trace data was obtained by tracing and collecting light from a light source and determining the combined output power at the output surface of the lighting device. However, alternatively, light may be traced from one of the human eyes to the lighting structure and returned to the light source (inverted light trace).

Inverted light traces are computationally efficient, thus reducing counting time and speeding up prototype production. According to this inverted light trace, it is assumed that a human sees the optical structure from a predetermined angle. Light is traced from one eye to the lighting structure and turned into a light source. The obtained data is combined with data obtained by repeating the same process for other eyes to obtain a three-dimensional image. Assuming that humans see through the lighting structure, a very high percentage of the light being traced produces useful information, thus increasing computer efficiency. The purpose of the light pipe is to direct the light from the light source. Therefore, the probability that the light emitted from the eyes hits the light source when the eyes are viewed through the light pipe is extremely high.

On the other hand, if the trace is started with a light source, most of the light will not reach the human eye. Therefore, when the inverted light trace is not used, the counting time for the light trace that does not reach the human eye is wasted.

The first step in performing an inverted light trace is to trace the set of light emerging from the eye as if the light source were at the focus of the eye. The eye's field of view is limited by the light pipe. Light passes through the field of view of the eye and is controlled by the same light tracing rules / formulas that control the tracing of light starting from the light pipe when the light hits the surface of the light pipe. The formula for controlling the optical trace applies whether the optical trace is made forward or backward. Therefore, the above-mentioned phase space technique can be similarly applied to the case of inversion.

Light from the eye is traced back through the light pipe. When light hits the light source, it is considered valid light. Thus, the percentage of light reflected at each reflection can be determined and the power of the light determined. FIG. 21 shows the light traced from the eye to the surface of the light source. Light is traced from the eye to the face of the light source as shown in FIG. The light then reflects several times while traveling to Luminaire. Luminaire is a model of the light engine LE. (Because the light engine LE is a complex structure, the interface is modeled by a light pipe. This model is luminaire and is formed by a set of spatial and directional information that defines the light from the light engine LE). In addition, the amount of power lost as each light is reflected is known. Thus, the luminaire can initially power the light, and then determine the power of the light at the output surface of the optical structure by working backward from the luminaire, taking into account the losses that are lost at each reflection. Thus, the brightness of the light falling on the eye is known. This can be combined with other light information to form an image of the light structure.

Using the above approach, a sequence of three-dimensional representations of a lighting device can be obtained with a relatively short counting time. For example, choose a small cube size (of the order of 1 mm ³ ) to see a non-planar surface. Image space of virtual is, for example, 10cm × 10cm × 10 cm, is there by using the 1mm cubic cube has a cube of 10 ^6. If the viewing range is 30 °, 1,000 orientations can be taken within this range. Use 100 lights per slice. As a result, there are 1,000 points per cube per orientation, and 10 ¹² data points need to be defined for three-dimensional display. Using a high-speed computer (which can handle operations on the order of 10 ¹⁰ per second), the counting time is on the order of seconds. Thus, the entire stereoscopic field of view of 30 ° can be made in minutes according to the invention.

Thus, a preferred three-dimensional visualization process can rapidly prototype and detect and eliminate hidden artifacts with computational efficiency.

The time required to computer model a lighting device can be reduced using the computer structure described below. In general, the lighting device modeling process involves a homogeneous task but a different set of computer loads. Homogeneous problems, For example B; is an object of the individual light traces through a lighting device to reach the _{(x, y k x, k} y) data points. There are a number of issues here (on the order of a thousand to 100 million). However, these issues are independent of each other in that the optical path of one light does not affect the optical path of the other light.

FIGS. 22 to 25 show a background lighting device having a light pipe LP including an enlarged microprism array MPA. Background lighting devices that provide background lighting for various things, such as laptop computers, as described above, are known.

The background lighting device includes a light source LS, a light pipe LP, and an output surface. The light pipe LP includes a micro prism array MPA. The microprism reflects the light from the light source LS to the output surface of the light pipe LP. The microprisms MPM are arranged such that some light rays always point upwards, while some light rays are transmitted through the light pipe LP.

It is ideal that the brightness of the light directed upward is uniform. The microprisms MPM are coarsely arranged at the end near the light pipe P (thus reducing the proportion of the large amount of light redirected upwards) and densely arranged towards the far end of the light pipe ( Therefore, the proportion of the small amount of light that is redirected upward is increased). At the far end of the light pipe LP, the duty cycle of the microprism reaches 100%. Either way, at each reflection, the light goes to the output surface of the light pipe LP or to the far end of the light pipe LP.

[0200] Conventionally, there are two types of light pipes, those that are metallized and those that are not. In non-metallized light pipes, there is no metal on the bottom of the light pipe. As a result, all light has the benefit of internal reflection. This total internal reflection is preferable because there is no loss and no absorption. Thus, when light passes through a non-metallized light pipe, all of the light is reflected (assuming the angle of incidence to the reflecting surface below the critical angle) and no light is absorbed by the reflecting surface.

A disadvantage of light pipes whose base is not metallized is that it limits the shape of the microprism structure. That is, to reflect light vertically upward, β = 45 ° −α / 2 (where β and α are shown in FIGS. 22 and 23), and α + β ≦ _IT (where _IT is total internal reflection). Critical angle). If β is outside this range, light incident on the surface of the microprism structure will be transmitted instead of reflected.

[0202] In the example of a metalized light pipe, the bottom surface of the microprism row is metallized. In this example, light transmission on the surface of the microprism row is blocked. As a result, β can be set to any value, and there is no light transmission.

[0203] However, metallized light pipes have the disadvantage that the metallized portion absorbs light and therefore does not result in total internal reflection. The absorption of light by metal is about 15%. Thus, when light is transmitted toward the far end of the light pipe, there is some power loss each time the light is reflected. The absorption loss due to metallization is on the order of 15-20% for each reflection. Therefore, if the loss is 15% and the light is reflected three times before the light exits, the emitted light will be approximately 60% of the initial power of the light pipe. It is desirable to reduce the metal thickness to reduce the amount of light absorption. However, in this example, increased transmission and reduced metal thickness are not a satisfactory solution to this problem.

To solve these problems, the light pipe in the present invention controls the transmitted light by total internal reflection, and controls the light reflection by the microprism MPM by metallization. As shown in FIGS. 24 to 26, the light pipe LP has only metallized microprisms, and the spacers between the microprism rows are not metallized. The spacers between the rows of microprisms transmit light toward the far end of the light pipe LP. The light hits the surface many times before reaching the far end of the light pipe LP. However, since these surfaces are not metallized, total internal reflection is achieved without loss. Since the microprism MPM is metallized, light only reflects off the microprism surface before exiting. Although there is some loss associated with the reflection from the surface of the microprism, only this type of reflection occurs and therefore this loss is acceptable. Therefore, the light pipe LP of the present invention is preferably combined with the better properties of metallized and non-metallized light pipes.

There are at least three ways to create a partially metallized structure. First, use a mask to cover the spacers between the grooves and leave the grooves for metallization open. This can be done using photolithographic techniques. Immediately, a photoresist is deposited on the entire bottom surface. Next, the photoresist is exposed to expose only the spacer through the main mask. Next, if the photoresist is developed, the photoresist between the grooves remains. Metalize the entire structure, including the photoresist. This results in the metal covering the photoresist and trench. If the remaining photoresist is removed together with the metal parts, the surface between the grooves becomes non-metallized.

Second, a cover is used to cover the spacers between the grooves and often keep the grooves open for metallization. The second means for this is the same as the first. However, in this example, a metal foil mask is made that covers the spacers between the grooves and keeps the grooves open. This mask is applied to groove formation and metallized. If the surface of the optical pipe LP is heated, the metal penetrates and is metallized.

Directional deposition, or more preferably oblique deposition, is used to take advantage of the shadow effect provided by the grooves. According to this approach, a metal vapor source is directed into the groove from one side. The grooves create shadows because the metal is directed at an angle. The surface in the shadow, which is the surface between the grooves, is not covered by the metal vapor when it hits the surface of the light pipe LP.

FIGS. 27 to 29 are graphs showing the output power over the range of x and y positions on the output surface of the light pipe shown in FIGS. These graphs were created in a general way.

FIG. 27 shows the output power at various points along the x-axis. FIG. 28 shows the output power at various points along the y-axis. FIG. 29 is similar to FIG. 28 except that differently arranged elementary cells are used. In FIG. 28, only 10 elementary cells are used for the y-axis, while in FIG. 29, 40 elementary cells are used. Because of this, FIG. 29 contains a higher density of data points and is therefore more bumpy.

The present invention can be variously modified within the spirit of the present invention, and is not limited to the above embodiments.

[Brief description of the drawings]

FIG. 1 is an explanatory diagram of an intersection between a two-dimensional plane and a light ray.

FIG. 2 is an explanatory diagram of an intersection between a two-dimensional plane and a plurality of light rays.

FIG. 3 is an explanatory diagram of an intersection between a line and a plurality of light rays.

FIG. 4 shows a two-dimensional figure of the light shown in FIG.

FIG. 5 is an explanatory diagram of a phase of a light ray in a phase space.

FIG. 6A is an explanatory diagram of luminance characteristics in a phase space for various types of light sources.

FIG. 6B is a diagram illustrating luminance characteristics of a phase space for various types of light sources.

FIG. 6C is an explanatory diagram of luminance characteristics of a phase space for various types of light sources.

FIG. 6D is an explanatory diagram of luminance characteristics in a phase space for various types of light sources.

FIG. 7 is an explanatory diagram of a uniform luminance distribution in a linear segment of a Lambertian light source in a two-dimensional phase space.

FIG. 8A is a comparison diagram of a uniform point light source using a photometric light trace and a Lambertian point light source.

FIG. 8B is a comparison diagram of a uniform point light source using a photometric light trace and a Lambertian point light source.

FIG. 9 is an explanatory diagram of a Lambertian cylindrical light source using photometric light traces.

FIG. 10 is an explanatory diagram of a set of light in a two-dimensional phase space and light of various direction cosine.

FIG. 11 is an illustration of a background illumination device having a Lambertian light source, non-image optics, microprism array and output.

FIG. 12 is an explanatory diagram of a spatial coordinate device.

FIG. 13 is an illustration of a directional coordinate device associated with one of the cells of the spatial coordinate device of FIG.

FIG. 14A is an explanatory diagram of a background lighting device having an output surface.

14B is an explanatory diagram of a directional coordinate space for a predetermined cell of the background device shown in FIG. 14A.

FIG. 15 is an explanatory graph of luminance values for the cell of FIG. 14B, where one of the directional coordinates changes and the other is held constant.

FIG. 16 is an illustration of a microprism of a light pipe having parameters that can be minimized using a cost function.

FIG. 17 is an explanatory diagram of the angle characteristics of the Lambertian diffuser.

FIG. 18 is an explanatory diagram of an example of an unwanted artifact.

FIG. 19 is an explanatory diagram of an example of an unwanted artifact.

FIG. 20 is an explanatory diagram of an opposite light trace traced from the eye to the light source.

FIG. 21 is an explanatory diagram of an opposite light trace of a light pipe.

FIG. 22 is a detailed illustration of a background illumination device having a light pipe with an enlarged array of microprisms.

FIG. 23 is a detailed illustration of a background lighting device having a light pipe with an enlarged array of microprisms.

FIG. 24 is a detailed illustration of a background illumination device having a light pipe with an enlarged array of microprisms.

FIG. 25 is a detailed illustration of a background lighting device having a light pipe with an enlarged array of microprisms.

FIG. 26 is a graph of a position range and an output on an output surface of a light pipe.

FIG. 27 is a graph of a position range and an output on an output surface of a light pipe.

FIG. 28 is a graph of a position range and an output on the output surface of the light pipe.

[Explanation of symbols]

LE light engine LP light pipe Ω ₀ phase space elementary cell I ₀ optical intensity ω angular frequency c speed of light

 ────────────────────────────────────────────────── ─── Continued on front page (72) Inventor Coupick Stephen A. United States 90503 Torrance Garnet 202 No. 3725 (72) Inventor Costerouski Andrew United States of America 92641 Garden Groove Grouse Coat 9816 (72) Inventor Rad Mike United States of America 91304 Sherman Oaks Magnolia Ballbird 305 No. 15222 (72) Inventor Tengala Indra United States of America 91765 Diamond Bar East Barker Drive 24207 (72) Inventor Basiliyev Anatoly United States of America 90505 Torrance Ocean Avenue 203 No. 23939 (72) 72) Device Kim Jengdal United States of America California 90036 Los Angeles Burnside Avenue 359

Claims (5)

[Utility model registration claims] 1. A light pipe for receiving light from a Lambertian cylindrical surface, comprising: an input surface, an output surface, a plurality of metallized microprisms, and a plurality of non-metallized spacers. The plurality of spacers are disposed between the plurality of microprisms, and the plurality of spacers and the plurality of microprisms are alternately arranged.Transmission of light through the light pipe is performed by total internal reflection, A light pipe in which reflection of light from an output surface is provided by the plurality of metallized microprisms. 2. The microprism is metallized by a photolithographic technique using a mask attached to the plurality of microprisms, wherein the mask comprises (1) the plurality of spacers, and (2) one of the plurality of microprisms. 2. The light pipe of claim 1, wherein the light pipe covers one of the plurality of spacers and (2) the rest of the plurality of microprisms. 3. The micro-prism is metallized using a metal foil mask, wherein the mask includes (1) the plurality of spacers and (2) one of the plurality of micro-prisms.
And (2) not covering the rest of the plurality of spacers and the plurality of microprisms, and the metal foil mask penetrates the surface of the light pipe when the surface is heated. The light pipe according to claim 1. 4. The light pipe of claim 1, wherein said microprisms are metallized using a directional deposition process. 5. The light pipe according to claim 1, wherein said microprisms are separated by a plurality of spacers.