Improving Air Temperature Sensor Accuracy with a Custom Aspirated Housing Unit

Many fields of research require the ability to accurately measure ambient air temperature. This task is complicated due to solar radiation which heats the sensor and causes it to record a higher temperature than is accurate. The traditional way to protect against this kind of interference is to have a radiation shield like the one depicted below, with the sensor housed inside layered shelves of material. However, as the schematic shows, solar radiation can still penetrate the shield and cause significant error in the measurements (Thomas & Smoot, 2013 and Gunawardena, 2018). To better mitigate the solar radiation effect, there needs to be a way to mechanically aspirate the sensor. In other words, a fan is needed to pull outside air over the sensor to allow it to record ambient air temperature accurately. Such aspiration units, used instead of the radiation shield, are available to buy commercially, though they are expensive and may have difficulty integrating with existing hardware and software, as is the case with the LEMS devices. The LEMS are low-cost energy management systems developed at the University of Utah as miniature data loggers (Gunawardena, 2018). They are composed of custom-made circuit boards (PCBs) that have many ports for sensors. This project is focused on developing a custom aspiration unit for use on these LEMS devices.

Improving Air Temperature Sensor Accuracy with a Custom Aspirated Housing Unit Alexis DeFord, Eric Pardyjak, Ph.D, Rob Stoll, Ph.D Department of Mechanical Engineering Introduction Many fields of research require the ability to accurately measure ambient air temperature. This task is complicated due to solar radiation which heats the sensor and causes it to record a higher temperature than is accurate. The traditional way to protect against this kind of interference is to have a radiation shield like the one depicted below, with the sensor housed inside layered shelves of material. However, as the schematic shows, solar radiation can still penetrate the shield and cause significant error in the measurements (Thomas & Smoot, 2013 and Gunawardena, 2018). To better mitigate the solar radiation effect, there needs to be a way to mechanically aspirate the sensor. In other words, a fan is needed to pull outside air over the sensor to allow it to record ambient air temperature accurately. Such aspiration units, used instead of the radiation shield, are available to buy commercially, though they are expensive and may have difficulty integrating with existing hardware and software, as is the case with the LEMS devices. The LEMS are low-cost energy management systems developed at the University of Utah as miniature data loggers (Gunawardena, 2018). They are composed of custom-made circuit boards (PCBs) that have many ports for sensors. This project is focused on developing a custom aspiration unit for use on these LEMS devices. Ambient Weather Radiation shield Experiment and Results First, the aspirated housing unit was created following the model used in the 2013 paper by Thomas and Smoot. After creating the aspirated housing unit, it was necessary to evaluate its performance. The same type of sensor, the SHT31, was put into the new unit and the radiation shield. Tests were conducted indoors at a range of elevation angles (noted in the schematic in the first column). Heat lamps simulated the sun, and the distance between them and the two SHT sensors was kept at a constant 0.5 meters. A research-grade finewire sensor was positioned behind the radiation shield to act as a standard temperature measurement. An example of the experimental set-up: this is the set-up for the 75˚ test. In the picture, the traditional radiation shield is on the right, and the new aspirated housing unit is on the left. - Research Objectives Increased accuracy ambient air temperature measurements Ease of construction Low cost to manufacture Ease of installation and integration with current LEMS hardware and software - No significant reduction in the battery life of the LEMS devices Materials Relevant LEMS Components: - Custom Circuit Board (PCB) - SHT31 Temperature Sensor (+/- 0.3˚C): $13.95 - 12V Battery Traditional Radiation Shield Component: - Ambient Weather Radiation Shield Model SRS100LX: $39.99 3D printed aspirated housing unit Aspirated Housing Unit Components: - Aspirated Housing Shield: $7.81 - Fan FAD1-04010BBLW11 (5V, 40mmx40mm): $5.83 Additional Testing Materials: - FW05 Type-E Finewire thermocouple: (+/- 0.07 ˚ C) - Campbell Scientific CR1000 Data Logger - 2 Heat Lamp hoods - 2 Capsylite Halogen Light Bulbs SHT31 Temperature Sensor (39W, 2850K) https://www.adafruit.com/product/2857 Lessons learned: - Aspirating the SHT sensor significantly increases its sensitivity to fluctuations in the ambient temperature (enough to pick up HVAC effects) - The fan selected draws current even when it is not on Value of this project: - Designed and validated a more accurate and lower cost air temperature measurement housing unit Conclusions Elevation Angle Radiation shields still allow some solar radiation to interfere with the sensor's reading. Discussion Here is an example of the raw data collected for a single elevation angle (18˚); this data was collected for a total of six angles. The statistics on the graph are all relative to the standard finewire measurements, with the line indicating perfect agreement. This graph shows the average temperature measurements for each sensor for each test. The colored bars represent the inherent error of the sensors, and the black bars represent the standard deviations of those measurements. This chart shows the mean difference between the sensors. Baseline testing showed the difference between the Aspir-Rad sensors to be 0.073 ˚C; much less than the mean difference observed across difference elevation angles. Clearly, the aspirated housing unit performed better than the radiation shield. A LEMS device deployed at the Dugway Proving Grounds in Utah's West Desert. - For the elevation angles tested, the aspirated housing unit recorded an average 0.91 ˚C lower temperature than the radiation shield, with the aspirated housing unit itself having an average reading of 0.58 ˚C higher than the finewire. So, in the indoor, windless testing environment, the aspirated unit works better than the radiation shield. - These sensors had the greatest differences at very high and very low elevation angles, coinciding with sunrise/sunset and noon. - Even the new aspirated housing unit is still affected by some radiation interference. Future Research - Improving battery life and decreasing the system's energy consumption by selectively turning the fan off/on given different environmental conditions - Investigating the relative radiation effects on different sides of the housing unit and the effect of wind speed - Outdoor testing with various other sensors to test for compatibility Acknowledgements Special thanks to Alex Bingham and Travis Morrison for their continued support throughout this project. Funding in part provided by the University of Utah Sustainable Campus Initiative Fund (SCIF), U.S. National Science Foundation grants IDR CBET-1512740 & CBET-PDM 1134580, and UROP funding from the Office of Undergraduate Research. References N Gunawardena et al (2018) Meas. Sci. Technol. 29 D Jensen et al (2015) Boundary-Layer Meteorol 567-87 C Thomas and A Smoot (2013) Journal of Atmos. and Oceanic Technol. 526-37