Open Access
Add to BibSonomy
Abstract
A new detector for solar absolute radiometers for ground-based applications was developed in a joint project by PMOD/WRC and Davos Instruments AG. New coating technology was applied and characterized, and a flat receiver was designed for the next generation of radiometers. The new detector, which is already commercially available, will also be considered and adapted for future space experiments.
© 2022 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
PMOD/WRC is one of the leading institutions in the field of solar radiometry and provides high-quality services to international and national organizations such as the WMO (World Meteorological Organization) and space agencies, in particular, ESA (European Space Agency) [1]. Davos Instruments AG is a spin-off company from the PMOD/WRC that develops high precision measurement equipment for solar radiometry and offers engineering services in related fields of application.
Absolute solar radiometers are the instruments utilized to measure Direct Normal Incidence (DNI) solar irradiance in W/m2. DNI is a measure of the solar power over the whole electromagnetic spectrum per unit area (W/m2). DNI is an important quantity in meteorology, climatology, and for solar energy applications. Absolute solar radiometers also serve as reference instruments for calibrating other solar radiometers, such as pyranometers and pyrheliometers.
Absolute solar radiometers are operated by the electrical substitution principle, which is based on the substitution of solar radiant power with electrical power [2]. The radiometer features two black sensors, often in the shape of cavities to enhance their absorptivity: a shaded reference sensor that is heated with constant electrical power and an active sensor that is electrically heated, so that the heat flux from both sensors to the common heat sink is equal. If the active sensor is exposed to solar radiation, the electrical power needed to equalize the heat flows is reduced by the amount of the radiant power. Thus, the radiant power can easily be measured as a difference in electrical power in the active sensor [3]. The design with two sensors makes the setup very robust against environmental influences.
In this work, the characterization of a new detector for solar absolute radiometers is presented, describing the changes in the design, and showing the optical performance.
2. New design
TSI receivers have been built in different shapes trying to optimize for the characteristics of the optical absorber coating. They can be found as a cylindrical cavity like DIARAD (Mekaoui et al. [4], conical cavity, applied in the PMOD series in different instruments like CLARA (Walter et al. [5]), and more recently, with the development of new technologies, flat receivers like CTIM (Compact Total Irradiance Monitor) described by Harber et al. [6]
2.1 Flat receiver
The geometry of a receiver is designed to optimally exploit the optical properties of the selected absorptive paint, while taking into account other requirements like thermostatic and thermodynamic properties, size, cost, manufacturability.
Two different designs in conical shape can be found in the PMOD instruments. As a cone (like CLARA or DARA instruments) or as an inverted cone (like PMO6 instruments) shown in Fig. 1. In both cases the cavity was designed such that the incoming beam is reflected five times for a glossy paint before it leaves the cavity again through the entrance aperture. More detailed information on the properties of these cavities can be found in the PhD thesis by Suter [7].
Fig. 1. Schematic of inverted conical and conical shape cavities. Yellow represents the incoming radiation, and the red areas indicates the position of the electrical heater foil. In the case of the PMO6 (left), part of the radiation is absorbed on the cylindrical wall, where no electrical heating occurs.
A conical or inverted conical cavity shape receiver is not an optimal geometry for a diffuse coating such as the sprayable carbon nanotubes paint from Surrey NanoSystems [8]. It was shown that the studied carbon nanotubes coating shows no measurable degradation after exposure to UV radiation equivalent to the dose expected for a space mission [9]. This property, combined with its high absorptance [8,9], makes this coating a good candidate for a new generation of TSI radiometers. Thus, a new receiver has been designed in a flat shape. This new geometry presents advantages with respect to the cavity shape, i.e., it is much simpler to manufacture, and it allows the utilization of a modern thin film technology for the heater and thermometers. The reflectance in a flat surface is higher than in the cavity but a highly reflective dome can improve this. The design of the new receiver (Fig. 2) consists of a flat surface covered with a highly reflective dome with an aperture. A labyrinth structure between the sensor area (black circle) and the outer aluminum ring serves as heat link, across which the heat flux is measured. The flat surface facilitates the spraying of the carbon nanotube coating. The coating applied in a cavity produces a concave accumulation (or meniscus) of the paint at the bottom of the cone. Avoiding this effect, the homogeneity of the absorptance is improved.
Fig. 2. a) Receiver designed for the CNTs coating before (left panel) and b) after (right panel) application of the heater foil and CNT coating. Made of aluminum, 46 mm diameter, with screw holes for a dome and with a shallow 0.1 mm depression at the center of the receiver, where the heater is fixed, and the CNT coating is applied. Thanks to this depression, the coating is at the same level as the entire flat piece. The labyrinth structure between the sensor area (black circle) and the outer aluminum ring serves as heat link, across which the heat flux is measured when the radiometer is operated.
2.2 Dome
The cavities of the previous detectors were designed to optimize the absorptivity by producing multiple reflections inside themselves. With each reflection a high percentage of the incoming radiation will be absorbed by the paint, aiming to collect all the radiation on the coating.
Trying to replicate this geometrical advantage and in order to increase the absorptivity of this new flat receiver, a highly reflective dome was designed with the shape of an oblate spheroid (an ellipse rotating by its semi-minor axis, Fig. 3)
Fig. 3. Schematic representation of the dome (ellipse) to calculate the dimensions
(a = 15mm, c = 5mm, b = 14.14mm)
This shape was selected for its geometrical properties, i.e., every beam coming from the coated area will be reflected back by the dome to the coated area after a single reflection, instead of a parabola of rotation, seen in instruments like CTIM [6], which needs two reflections to get the beam back to the absorber. For the CTIM instrument the number of reflections is less of an issue because the dome is thermally connected to the receiver so that any residual optical power absorbed by the reflecting dome is thermally conducted into the receiver. Our dome is connected to the heat sink, to facilitate the assembling of the instrument, thus any radiation absorbed on the dome will not to be measured but these losses by absorption can be evaluated in advance as the dome is characterized.
The coated area, which must fit within the “focal circle” (distance between the two focal points F and F’) of the oblate spheroid, shown in Fig. 3, is a 10 mm diameter circle. In order to fulfil these characteristics within the available space, limited by the housing of the instrument, the dimensions of the ellipsoidal dome were calculated.
The dome is 15 mm for the semi-major axis, being the maximal distance allowed to fit in the radiometer. By ellipse properties this distance is equal to the distance between the focus and the intersection between the semi-minor axis and the ellipse (distance “a” in Fig. 3). “C” corresponds with the radius of our coated area, it is 5 mm. With these two parameters the semi-minor axis “b” was calculated (Eq. (1)), which is the height of our dome. The ellipsoid was generated by rotating the ellipse around this semi-minor axis. The eccentricity was also calculated (Eq. (2)).
The incident beam enters through the 8.3 mm aperture and hits the carbon nanotubes coating. This high-absorptance coating absorbs most of the incoming radiation, but a part is reflected. Because of the properties of the ellipse, every beam from one focal point, is reflected onto the other focal point. Thus, every beam from a point between these two focal points is reflected onto another point between the focal points. A “focal circle”, corresponding with the circumference of the coated area, is generated by the rotation of the ellipse to create the ellipsoid.Every beam from inside this circle will be reflected back by the ellipsoid (our dome) into the circle. The “focal circle” is the area to be coated, the absorber.
The dome (Fig. 4) was made by Fisba AG of polished aluminum (surface roughness Ra 0.1µm), selected because it has a high reflectance in the visible range and spectral flatness (Fig. 5, provided by the manufacturer), with a protective layer of silicon oxide in the inside to avoid damage or oxidation from the environment.
Fig. 5. Simulated reflectivity of the aluminum dome with protective layer provided by the manufacturer. The coating of the upper curve was selected, and it will be tested in our facilities to confirm its reflectance.
3. Optical characterization
The optical characterization of the receiver with and without dome was performed, to compare and evaluate the absorptance gain provided by the dome. Possible light losses through the labyrinth structure were determined by transmittance measurements.
Finally, the reproducibility of the assembled detector was tested by comparison of two different detectors with two different domes.
The integrating-sphere-setup used for the reflectance measurements, described in [7] (Fig. 6), was the same for both cases with and without dome. It consisted of a 1 mm-width beam from a laser source (at three different wavelengths: 375 nm, 532 nm, and 633 nm) passing through a chopper at 113 Hz (this prime-number frequency was selected to avoid a possible noise produced by a multiple value of the 50 Hz of the mains frequency). The beam is divided in two parts by a wedge-shaped beam splitter. The reflected part of the beam is aimed at a monitor diode that monitors the stability of the laser, while the main beam traversed a lens and iris aperture and entered into the integrating sphere. A silicon photo diode detector measured the remaining intensity after reflection of the sample located at the far end of the sphere.
An integrating sphere is the most common method for reflectivity measurements. The theory behind this instrument is that the light inside the sphere is equally distributed and any infinitesimal area of the integrating sphere exchanges the same radiation with any other infinitesimal area [8]. The integrating sphere was mounted on a motorized platform that could be moved vertically and horizontally, perpendicular to the laser beam to produce 2D resolved reflectance maps of each sample.
For the determination of potential radiation leaking through the labyrinth (transmittance), the setup was also utilized but with the samples located differently. Due to the geometrical properties of the ellipsoidal dome, all the reflected beams from the coated area should return to the coated area, and theoretically losses through the labyrinth are not expected. Only stray light due to scattering off the imperfect inner surface of the dome could possibly leak through the labyrinth. Transmittance measurements were performed to quantify the quality of the dome’s inner surface geometry. We located the sensor with dome at the entrance port (12 mm radius) of the integrating sphere and covered the exit port with a white plug (painted with the same high reflective BaSO4 paint as the integrating sphere). Any stray light that was transmitted through the labyrinth entered the sphere and was measured by the same silicon diode as before. Instead of comparing the samples with a reflectance reference calibrated sample, we compared it against a neutral-density filter foil (OD 5.0) with 0.001% transmittance locating this foil as well at the entrance port.
The optical characterization of the coated receiver (without dome) by reflectance measurements is shown in Table 1 and Fig. 7. The result of every measurement performed, the average of them, and their standard deviation are shown.
Table 1. Mean value and standard deviation of the reflectance maps for the new receiver without dome
It is shown in Fig. 7 that the reflectivity increases towards shorter wavelengths. As expected, our absorptivity values are higher than the described by the manufacturer [9] due to a second layer of carbon nanotubes coating applied [10].
The optical characterization of the assembled receiver (flat receiver with dome) was performed by reflectance and transmittance measurements shown in Table 2 and Fig. 8 and fig. 9.
Fig. 9. Example of one of the four transmittance maps performed at each wavelength (a) 375 nm (top), b) 532 nm (center) and c) 633 nm (bottom)).
Table 2. Reflectance measurements for receiver with dome and transmittance measurements
The heritage conical cavity improves the absorptivity of the carbon nanotubes by a factor of 4 [11], while the dome and the flat receiver of the new detector improve the absorptivity in average a factor of 6. This number is obtained by comparison of the reflectance without dome against the reflectance with dome plus the transmittance.
The transmittance measured was lower than 0.001% at the center area of the receiver (Fig. 9), where the illumination hits directly, increasing closer to the edges, and never higher than 0,0035% at any point at any wavelength. The level of irradiance is orders of magnitude higher on the directly illuminated area of the sensor than other peripheral regions. This mitigates any effect form sightly increased transmittance towards the edge of the sensor area.
As we discussed before, radiation absorbed by the dome will not be measured by the instrument. Thus, losses by absorptance at the dome must be quantified by studying the reflectance of the dome.
A calibrated reference sample with nominal reflectivity of 99% from Labsphere was used to compare simultaneously its reflectance against the reflectance of the dome. At 633 nm, the calibrated sample had a reflectance provided on request by the manufacturer of 98.9%, 98.8% reflectance at 532 nm, and 98.6% reflectance at 375 nm. We located the inner face of the dome in the sample port (Fig. 6). The reference sample was located at the aperture of the dome and utilized to be compared against the dome at the same time without changing the setup. A dark scan was not needed as we performed a relative simultaneous measurement of the dome and the reference sample, and the reflectance signals were much larger than the effect that might be produced by the stray light of the entrance port.
We analyzed the reflectance at the three available wavelengths (633 nm, 532 nm, and 375 nm).
Figure 10 and Fig. 11 show the reflectance of the reference sample (the inner circle of the reflectance maps), followed by the reflectance of the concave interior face of the dome (highly reflective annulus around the calibrated sample) and at the corner of the reflectance maps the reflectance of flat surface where the dome is attached to the receiver is found (blueish-green colors).
Fig. 10. (a) Reflectance of the dome and SRS 99 calibrated sample at 633 nm (top left panel), b) 532 nm (top right panel), and c) 375 nm (bottom panel).
Fig. 11. Description of the different areas shown in the reflectance measurements of the inner surface of the dome.
The transition area between the reference sample and the dome was not considered for the analysis of the dome's reflectance, because the change of the geometry might produce scattering and would not contribute to the real characterization of the inner surface of the dome.
The corner regions, corresponding to the flat surface of the dome for attachment to the receiver does not contribute to the reflectance of the dome, thus was not considered neither.
The average reflectance values measured of the dome were compared to the reflectance calibrated sample. The reflectance average of the area of the calibrated sample corresponds to the value of the nominal reflectance at the wavelength measured and it is compared with the average value measured of the area of the dome. The reflectance of the dime is higher than the reflectance of the calibrated sample at all wavelengths (Fig. 10). This relation provides an average value of the reflectance of the dome of 99.80% (+0.2% - 2%).
3.4 Reproducibility
As a final test, the reproducibility of the instrument was studied by analyzing a second receiver (produced in the same set as the first one) at 633 nm. This property is important for future commercial implications of the prototype.
The comparison between the previously described receiver (detector 1) and the new one (detector 2) are found in Table 3. The overall results are in excellent agreement.
Table 3. Comparison between two detectors at 633 nm
3.5 Uncertainty budget
The uncertainty budget for the reflectance measurements (with and without dome) and the transmission measurement were evaluated as the contribution of the following effects (Table 4, Table 5, Table 6). They were selected and calculated by Dr. Suter [7] following the guidance of BSRN Uncertainty Report [12], and adapted to the new paint and samples.
- • cut out uncertainty: the sensitivity of the mean reflectivity to the center of the sample is evaluated. It is an uncertainty associated to the homogeneity of the sample by comparison of the mean value and the value at the center of the sample.
- • losses through the entrance: an uncertainty is associated to the correction for the losses through the aperture of the integrating sphere determined by Suter [7] with the same setup.
- • reference uncertainty: provided by the manufacturer of the calibrated reference sample.
- • repeatability: determined by comparison of the multiple measurement of the same sample, by taking half of the difference between the higher and lower reflectance value.
- • spectral flatness: the reflectance measurements are performed at three wavelengths. An uncertainty is associated when these three points are extrapolated to the whole spectra. It is assumed that the deviation at other wavelengths is in the order of the variance of the measurements at 375 nm, 532 nm, and 633 nm.
Reflectance without dome:
Table 4. Uncertainty budget over 1 for reflectivity measurements without dome
Reflectance with dome:
Table 5. Uncertainty budget over 1 for reflectivity measurements with dome
Transmittance without dome:
Table 6. Uncertainty budget over 1 for transmission measurements with dome
4. Conclusions
A new sensor for TSI radiometer has been successfully developed and characterized. The new coating applied in the new flat-shape receiver provides a higher absorptance coefficient in a simplified mechanical design than the previous cavity design with the glossy paint. This very low reflectance combined with the absence of measurable degradation due to UV irradiation showed in [10] will allow us to achieve higher accuracy for long time series measurements in space applications.
The transmittance of this new geometry is at every measured wavelength very low, around 0.001% is leaked through the labyrinth.
The dome reflects 99.80% of the non-absorbed radiation back to the coating. Around 0.2% of the incoming radiation is not absorbed by the carbon nanotubes at first incidence. From these 2000 ppm the 99.80% of them will be reflected back to the black absorber by the dome. Thus, only 4 ppm of the incoming radiation are absorbed by the dome.
The dome improves the absorptance of the coating from 99.8% to 99.97% with the oblate spheroid design.
The dome’s q factor (ratio between reflectance of the coating in a flat surface with and without dome) is at the same order of magnitude than other diffuse paints sprayed in cavities. This design, specially thought for this new coating, is optimized for the optical properties of the carbon nanotubes which behaves as a lambertian radiator.
The sensor characteristics is highly reproducible, the standard deviation at each wavelength is low, allowing a reliable calibration.
Currently, a prototype has been successfully produced as part of an Innosuisse project in collaboration with Davos Instruments AG [13] and it is commercially available. The new sensor will be considered for future research experiments to measure the Total Solar Irradiance (TSI) and other components of the Earth Radiation Budget (ERB) from satellites.
Disclosures
The authors declare no conflicts of interest.
Data availability
Data underlying the results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request.
References
1. Annual Report 2018, PMOD/WRC (2018).
2. R. W. Brusa and C. Fröhlich, “Absolute radiometers (PMO6) and their experimental characterization,” Appl. Opt. 25(22), 4173–4180 (1986). [CrossRef]
3. K. Ångström, “Über absolute Bestimmungen der Wärmestrahlung mit dem elektrischen Compensationspyrheliometer, nebst einigen Beispielen der Anwendung dieses Instrumentes,” Ann. Phys. 303(3), 633–648 (1899). [CrossRef]
4. S. Mekaoui, S. Dewitte, D. Crommelynck, A. Chevalier, C. Conscience, and A. Joukoff, “Absolute accuracy and repeatability of the RMIB radiometers for TSI measurements,” Physics Solar. 224, 237–246 (2003). [CrossRef]
5. B. Walter, P-L. Levesque, G. Kopp, B. Andersen, I. Beck, W. Finsterle, M. Gyo, K. Heuerman, S. Koller, N. Mingard, A. Remesal Oliva, D. Pfiffner, R. Soder, M. Spescha, M. Suter, and W. Schmutz, “The CLARA/NORSAT-1 solar absolute radiometer: Instrument design, characterization, and calibration,” Metrologia 54(5), 674–682 (2017). [CrossRef]
6. D. Harber, Z. Castleman, G. Drake, S. Van Dreser, N. Farber, M. Miller, J. Rutkowski, A. Sims, J. Sprunck, C- Straatsma, I. Wanamaker, W. Zheng, G. Kopp, E. Richard, P. Pilewskie, N. Tomlin, M. Stephens, C. Yung, M. White, and J. Lehman, “Compact Total Irradiance Monitor flight demonstration,” Proc. SPIE. 11131, 111310D (2019).
7. M. Suter, “Advances in Solar Radiometry,” Thesis UZH PhD. (2015).
8. Labsphere, 2017, Technical Guide: Integrating sphere theory and applications
9. https://www.surreynanosystems.com/
10. A. Remesal Oliva, W. Finsterle, and P. Martín, “Degradation process on TSI radiometers due to UV radiation,” (2022, to be submitted).
11. A. Remesal Oliva, W. Finsterle, B. Walter, and W. Schmutz, “Characterisation of a new carbon nanotube detector coating for solar absolute radiometers,” J. Phys.: Conf. Ser. 972(1), 012007 (2018). [CrossRef]
12. N. Hyett, BSRN Uncertainty Report. (2000).
