I.

INTRODUCTION AND DEFINITIONS

Stray light is a significant issue in optical design and can dramatically influence the performance of the optical system. Stray light reduces the dynamic range and the SNR (signal to noise ratio) in the device. The effect is caused by light scattering, ghosting in refractive optics, diffraction and etc.

Using proper design and black surfaces are the most popular ways for stray light reduction. In this paper we will deal with the blackening solution. A lot of blackening methods are available today, such as black anodized aluminum, black paints, flock paper, Acktar black and others.

Black coatings are usually characterized by specifying the specular reflected and the diffuse scattered components. The hemispherical reflectance of a surface is defined as the ratio of the total energy reflected into the subtending hemisphere to the energy incident on the surface and hence contains both components. Direct measurement is usually performed by integrating spheres.

The specular reflectance R of a surface is defined as the specularly reflected power P_R normalized to the incident power P_i. Total backscattering TS_b is defined as the backscattered power P_s normalized to P_i, where the range of scattered radiation to be detected (range of acceptance angles) is θ_s = 2°…85° and φ_s = θ°…360° [1]. Hence, other than the hemispherical reflectance, TS_b excludes the power contained in the specular reflectance. TS_b can be determined by direct measurement or through numerical integration of angle resolved quantities as the Bidirectional Reflectance Distribution Functions (BRDF).

The BRDF is the fundamental way to describe the distribution of light scattered from an optical or non-optical surface or material [2]. BRDF data is also used in optical engineering software to model the propagation of stray light in optical systems and its impact onto the image properties during the design phase.

The BRDF is defined as the power ΔP_S scattered into the solid angle ΔΩ_S normalized to the solid angle and the incident power P_i:

θ_s and ϕ_s are the polar scatter and azimuthal scatter angles (see Fig. 1), respectively. The BRDF also depends on other parameters such as the angle of incidence θ_i, polarization, and wavelength. The cosine factor originates in the radiometric definition of the BRDF. It is sometimes omitted; the resulting function being called Angle Resolved Scattering (ARS) or cosine-corrected BRDF.

Fig. 1.

Geometry for the definition of light scattering and specular quantities: n sample normal; P_i – incident light power at the incidence angle θ_i; P_r – specular reflected light power; ΔP_s – scattered light power into detector solid angle ΔΩ.

For black coatings, the hemispherical reflectance can reach values of less than 1% for quasi normal incidence or for near VIS illumination wavelengths. The specular reflectance is relatively low and has no significant influence on the hemispherical reflectance. However, for gracing angles of incidence and for infrared illumination wavelengths, the specular reflectance increases and the surface tends to behave more as a mirror (see Fig. 2).

Fig. 2.

Left: BRDF measurements at 532 nm and 4630 nm illumination wavelength on a black coating (Acktar Magic Black™ [3]) at different incident angles (3° and 60°). TS values were determined to 1% (532 nm, quasi normal incidence) and 15% (4630 nm, quasi normal incidence). Right: Measured specular reflectance (Acktar Magic Black™) as a function of the incident angle for different illumination wavelengths.

The purpose of this paper is to describe a new black surface that will minimize the mirror-like effect of the surface, allowing to use this in star trackers and telescopes.

II.

VANES AS STANDARD STRAY LIGHT SOLUTION

Most of the star trackers and telescopes use baffles as a solution for stray light problem. The idea of this design is to allow the light pass directly to the detector, but prevent any reflections from the internal walls of the telescope tube. The baffle made from a tube, which has vanes in internal walls. The vanes protect unwanted light from reaching the detector, by reducing it intensity as much as possible.

In the first step, the field of view (FOV) of the detector placed at the end of the tube is determined. The most critical surfaces for stray light are those found outside of FOV and can be seen from the detector position or focal surface. These should be removed from the field of the detector.

In the next step, the optimal number of vanes should be calculated and placed in the inner diameter of the tube. Several issues should be considered for best performance:

The design variables for the baffle are determined by the following vane parameters: vane aperture, distance from the detector, spacing between the vanes, edge radius of vane aperture, bevel angle, angle relative to main baffle, and coating type [3], [5].

If the design has been properly chosen, stray light that enters the baffle is attenuated by several orders of magnitude before it reaches the detector.

Fig. 3.

Ray trace for determining the position of the vanes in the case of diffusely reflecting surface [4]

III.

NEW BLACK SURFACE TO MINIMIZE THE AMOUNT OF VANES

As described before, vane based design is an efficient way for stray light elimination, but often it may be not easy to implement or can be too costly. Using the same idea of vanes we will suggest an alternative design, fast and easy to implement, light weighted and with reduced cost.

Assume we can place as many vanes as we can, for example every 10 mm. This means the detector to be well protected from stray light. Now let us make the vane height very small, namely large aperture, for example 20 mm less than the tube internal diameter (10 mm vane wall height).

This design will allow the use of very thin vanes, as small forces will act on every single vane. Thus the sloping edge will be very small, depending on the ability to produce such thin vanes. In our investigation, the smallest thickness was 20 μm. This will reduce the problem of edges that reflect light directly onto the detector, as the signal from such small areas is very small.

The next step is to find the proper black coating to reduce the reflectance as much as possible. Several points should be taken into account:

Fig. 4.

Non sharp sloping edges - painted aluminum sheet [5]

Considering all the specifics discussed above, we developed a unique honeycomb structured surface made of thin aluminum foil with an Acktar Magic Black™ absorbing coating.

Fig 5.

Path of the stray light as reflected from the honeycomb structured surface. Path (1) will lose most of the power after 4 reflection from honeycomb wall (reflection from bottom plane not counted as not coated with Acktar Magic Blac™). Path (2) will lose 0.01 of the power and reflected directly to the detector.

The unique properties of the black surface:

Fig. 6.

Cross section of a 25 μm thick aluminum foil, coated on both sides by Acktar black, magnification of x100

Fig. 7.

Wall tops of the black surface, magnification x100

IV.

BRDF MEASUREMENTS

BRDF measurements on high-end absorbing surfaces like Acktar black coatings require scatterometers with capabilities far beyond conventional photometers. Only few instruments exist that meet these demands. The measurements described in this section were performed using the ALBATROSS scatterometer developed at Fraunhofer IOF in Jena [6], [9]. The instrument is located in a clean room (ISO 7) under laminar flow boxes (effective ISO 5) and capable to perform measurements ranging between 325 nm and 10600 nm illumination wavelength. The scatterometer achieves the sensitivities and dynamic ranges necessary to characterize high end optical components, materials, and black coatings [8].

Characterization of the honeycomb structure (see Fig. 8) is especially challenging as (i) the light scattering distribution is expected to be anisotropic, which would require angle resolved measurements in the full reflection hemisphere. However, hemispherical measurements are especially comprehensive for infrared wavelengths. Moreover, (ii) the lateral dimension of the honeycomb cell size is larger than the illumination spot diameter. In order to get sample representative measurement data with reasonable effort, BRDF and reflectance was averaged from measurements performed at 5 different locations on the sample. An illumination wavelength of 4.6 μm was chosen, with 45° linear illumination polarization, and with an illumination spot diameter of about 3 mm.

Fig. 8.

The black surface on an aluminum substrate (for illustration contrast and brightness of the picture were enhanced)

Fig. 9 displays the BRDF and the reflectance measurements that were performed at 5 different measurement positions on the sample, respectively. The missing data points at θ_i ≈ θ_s are a result of the detector blocking the incident beam. It can be observed that for different measurements positions the BRDF changes especially for scattering angles between 30° to 60°. The reflectance measurements at different positions show highest variation also for incident angles of about 30° to 60°. These effects can probably be related to changing obscuration conditions inside the honeycomb structure.

Fig. 9.

BRDF (left, 10° angle of incidence) and reflectance (right) measurements performed at 5 different measurement positions, respectively

Fig. 10 displays the averaged data from the measurements described above. Different azimuthal orientations would probably lead to different obscuration conditions; however, both BRDF and reflectance now represent a much closer approximation to the overall performance of the honeycomb structure.

Fig 10.

BRDF (left) and reflectance (right) averaged from 5 different measurement positions, respectively

Compared to conventional black coated surfaces, the honeycomb structure is capable to effectively reduce specular reflectance at gracing angles of incidence even for infrared illumination wavelengths.

V.

SIMULATION OF THE DESIGN USING OPTICAL DESIGN SOFTWARE

Two simulations were performed using the optical design software - TracePro. The goal of the simulations was to compare a standard design of a baffle with vanes to a design of a baffle with the surface presented in this paper.

a.

Simulation of the baffle with Acktar proposed surface

A tube with the suggested material in the inner surface was tested. At the entrance a lens doublet with EFL=150mm was placed. For the light source, a collimated beam out of the FOV (9° off axis) was used. The result is shown in Fig 11. The ray tracing simulation shows that only a very small number of the input rays reaches the detector. The total rejection ratio of the system was 7.9·10^-7.

Fig 11.

Ray tracing simulation (left) - A lens doublet was placed in the left side of the tube, the detector was placed in the right side. Total irradiance map of the incident flux on the detector (right).

b.

Simulation of standard baffle with vanes

A similar simulation was made for a standard design of a baffle with vanes. Five vanes were placed inside the same tube to achieve the same FOV as before. The result is shown in Fig 12. The ray tracing simulation shows as before, that only a very small number of the input rays reaches the detector. The total system rejection was 3.3 · 10^-5, concentrated in the arch shape at the middle. Therefore, Acktar proposed surface improved the rejection to stray light by 2 orders of magnitude compared to the vane design demonstrated here.

Fig 12.

Ray tracing simulation (left) - five vanes were placed in the tube to achieve the same FOV as in the first simulation. The detector was placed in the right side. Total irradiance map of the incident flux on the detector (right).

VI.

CONCLUSIONS

In this paper we presented a novel black absorbing surface revealing reduced specular reflectance at high angles of incidence. It has been shown to perform at least comparably to vanes in star trackers and telescopes. In some cases it will show enhanced performance of stray light reduction as compared to common vanes.

As a conclusion of this study, we suggest this new development as a replacement of a part of the vanes in a variety of space designs and even all the vanes in some of the systems. This will reduce costs, weight and add flexibility to many designs.

Future work will include further investigation in order to determine the best choice of the honeycomb dimensions (wall height and cell size) for different types of tubes.