Ambient pressure and temperature monitoring to improve testing accuracy
For the researching community and students involved in technical analyses and experimentation, it is extremely important to ensure to remove any possible sources of errors which may affect the outcome. These errors can be various and could be caused due to instrument or environmental factors or could simply be introduced due to an assumption for specific values for calculations. Myself being a recent aeronautical engineering graduate and after carrying out several experiments, it always appears to be the case when using gauge pressure sensors and temperature sensitive equipment where we used to assume standard atmospheric values. These values which we take up as 101325 Pa for pressure and 20 degrees Celsius for temperature, do not account for altitude affects or local weather conditions. These simple assumptions could indeed introduce bias errors which may prove to be significant.
This was in particular important for my case as the laboratory in my university (Birla Institute of Technology, Mesra) was situated at about 600m above the sea level and weather conditions used to change frequently due to proximity of the lab to a river. I had carried out a simple experiment over a period of several days to identify how the ambient pressure and temperature were changing and how it could affect the experimentation which we used to carry out. We used an open circuit wind tunnel for subsonic flow testing and several gauge sensors which used the ambient pressure as a reference value to record the pressure differences over the test models. A report for the same is presented in this article, focusing majorly on pressure variations and possible interpretations of the results. There are other ways in which ambient conditions can influence the experimentation which too have briefly been discussed.
A simple setup was developed to observe the variation of pressure and temperature data over a period of 25 days. The setup was developed and installed at Aerodynamics Lab., Department of Space Engineering and Rocketry, BIT Mesra, Jharkhand where the readings were recorded. The lab is situated at an elevation of about 600m above sea level and hence the pressure values are expected to be lower than standard sea level conditions.
A simple setup was developed using a micro-controller board Arduino UNO and a breakout board with Bosch BMP180 sensor.
The BMP180 sensor from Bosch is an absolute pressure sensor. It also includes temperature measurements. The sensor has an absolute pressure range of 30kPa to 110kPa which easily covers the expected pressure variation. The accuracy of the sensor of ±12Pa is acceptable for the current application given the variations is expected in a few hundred Pascal from the standard sea level value. The temperature range for the same is from -40ᵒC to 85ᵒC with an absolute accuracy of ±1ᵒC. Further details for the sensor are provided in the datasheet in reference 2.
The sensor was interfaced with the Arduino using a jumper wires connected on a breadboard. The circuit arrangement is shown in figure 1 below for the same.
The setup was attached on the outer wall of low speed wind tunnel. It was attached on the side wall, clear of flow path which the air would follow during tunnel runs. The figure 2 below shows the mounted hardware.
The code was written in Arduino IDE using examples from Sparkfun and Adafruit libraries available. These provide functions to address the BMP180 sensor and to record and interpret the data to report it as pressure and temperature. These can then be worked upon for unit conversion or for other calculations such as for pressure altitude. The example code provided by Adafruit was used for the current setup.
The data was recorded over serial communication with a laptop using an open source, free telnet and SSH client PuTTY. The data recorded was stored on the local hard drive and the used to plot the variations in conditions.
Results and discussions
The readings were taken from 24th January 2020 till 18th February 2020, between 11 am to 4 pm. The data obtained was then imported for processing and plotted versus the dates.
Data was acquired over a period of several days. The pressure and temperature values were plotted against the dates, approximate time at which measurements were started and variations of pressure versus temperature. The pressure data was then compared with standard pressure values at known altitude for validation.
The lab is situated at an altitude close to 600m above sea level and hence is expected to have a lower absolute pressure than at sea level. The value for 600m for the standard temperature and pressure is provided in standard atmosphere table from reference 11, Appendix A. As per the table, the standard temperature at 600m is 11.11 degree Celsius (284.26 kelvin) and standard pressure at the same altitude is 943.22 hPa.
The data obtained from the setup is listed in table below. The distribution observed is further plotted and discussed.
The average absolute pressure at the aerodynamics lab is close to 949.4 hPa for the duration of readings taken. When compared to standard pressure altitude data, this value is close to the standard value of about 943.22 hPa. Also, the average temperature is higher at 18 degrees Celsius compared to 11.11 degrees Celsius for the standard value. The standard values in the standard atmosphere model are based on the hypothetical model to give a constant reference and does not account for humidity, local barometric variations due to wind conditions, variation in air composition and non-standard temperature variations. These conditions are ever changing and induce a variation in the measured value compared to what may be given in standard atmospheric model. The average increase in temperature could be a contributing factor for the measured average atmospheric pressure. However, it is observed from the data and figures further, other factors appear to be a dominating factor compared to temperature variations for localized variations in pressure values.
The recorded value ranges from 943.1 hPa to 954.1 hPa, which is close to the standard atmospheric values for the given altitude. The variations which can be seen in figure 3 are expected as there are local weather patterns which keep changing and thus will cause a slight variation in measured pressure value. The standard atmospheric value at sea level which, is used in several instances as a reference pressure, induces a biasing error during data processing. This shows that with the measured values, the error can be substantially reduced for various calculations if proper reference values are taken.
Pressure and temperature variations shown at different times (figure 4 and 6) are also taken on different days. Local bunching of data is observed in figure 4, between 948.3 hPa to 953.5 hPa for the readings taken between 10:30 am to 12:30 pm spread over several days. This close bunching could be an indicator suggesting nearly consistent readings for a given time of the day. However, any weather variation such as rain, overcast or storm weather could have a significant effect on the local pressure values. Measurement over longer period of times could provide further insight for the same. The plot for temperature with time, as depicted in figure 6, indicates a variation from 15.8 degrees Celsius to 20.4 degrees Celsius. There is no bunching of data as observed for pressure and the points are observed have a higher scatter. This is probably due to the local weather conditions affecting the temperature of the region. This variation however suggests that the variation in temperature does not appear to have a significant effect on the pressure and other environmental factors dominate over such effects. The combined plot of absolute pressure and temperature vs time of day in figure 8 depicts the same. Compared to the sea level values taken, these values of pressure and temperature are better indicative of the actual ambient values. This would reduce the error incurred with assuming a standard value of 1 atm for several pressure calculations. The distributions over different days also indicate minor variations in pressure values as observed in figure 9. The temperature variation over different days is shown in figure 5 and has a similar reason for variations in values obtained as for differences in time dependent values.
Figure 7 shows variation of pressure with temperature. The plot indicates there being no relation which could be derived from the sample size obtained. Also, the variations in temperature is only about 4 degrees over different days and other factors such as local density, humidity and wind are not accounted for which could potentially have a higher effect on pressure than the variations observed in temperature data. A comparison for the pressure and temperature against time of day and for different days is plotted in figures 8 and 9 respectively. For the sample size and without accounting for other factors, no conclusive results can be obtained from the said plot. However, since the major focus of the current report is to allow for reduction in bias error introduced in pressure calculations, the values for pressure obtained show a significant shift from standard sea level value of 101325 Pa. This data would still prove to be useful in reducing errors for absolute pressure calculations. The absolute value of temperature could be taken into account for applications which are sensitive to ambient temperature. This can improve the obtained accuracy and repeatability of data (assuming no instrument error) if accounted for the same.
The data is recorded in the month of January and February and is expected to vary with the variation in climatic conditions. The sample size can be increased and data reported to get a timely average which can be taken up for pressure calculations in case of unavailability of the device. An increase in sample size would also be necessary to draw better conclusions of pressure and temperature variations, their interdependence and average variations with respect to major climatic changes.
The variations obtained for pressure values are still small with maximum variation close to 11 hPa (or about 0.16 PSI). The values can be recorded close to the time of experiment run to reduce any error due to local conditions for ambient values.
A system for monitoring ambient pressure and temperature data was developed and tested. Following were the characteristics of the system and some of the observations made:
· A separate dedicated system was required to display and record data
· System is simple and compact, making it portable
· The demo codes for the sensor is available from different manufacturers and is customized to record the required information
· The system is mounted on the side of the settling chamber of the low subsonic wind tunnel at Aerodynamics laboratory, Department of Space Engineering and Rocketry, BIT Mesra
· The data recorded was stable for an extended period of time per data point collected
· The pressure value on average show about 6% variation from the standard sea level value of 1 atm (or about 101325Pa). This is close to what is expected at the altitude at which the laboratory is situated.
· There were fluctuations observed from day to day variations and during different times which can be attributed to local weather variations
· The system was successfully tested and limited sample data collected to identify how the variations occurred over the sampling duration
· The data can be used to improve accuracy of calculation of absolute pressure and for calibration of gauge sensors. The data could be recorded as soon as the device is turned on as there are negligible effects of duration of turn on time on the readings obtained.
The setup functioned as expected and was useful in providing ambient environment data for use in calculations for data recorded from gauge pressure sensors. Further analyses could help identify other regions where the data recorded could be useful such as in hot wire anemometry.
Future work and possible modifications
Further there are some modifications which can be carried out to better meet the requirements and the way the data is delivered. These include but are not limited to
· Attachment of a display indicating the measured values of pressure and temperature.
· Addition of moisture sensor for further investigation of variation of pressure and temperature variations with relative humidity
· Independent power source for standalone operation of the device
· Data recording using a memory card to record data for further processing
The data obtained was for a small sample size and indicative of proper functioning of the device. The methodologies outlined in the current report are some of the ways how the data can be processed and utilized to predict how the variations might be in the future. A larger data set over a period of different months would be required to further investigate the variations that occur. The data obtained from the device can prove to be useful for experiments which utilize or are sensitive to ambient pressure and temperature.
2. Bosch, BMP180 Digital pressure sensor data sheet, Document number BST-BMP180-DS000–09, revision-2.5
4. Atmel, ATmega328P Datasheet, Revision 7810D-AVR-01/15
5. Sparkfun BMP180 breakout Arduino library, https://github.com/sparkfun/BMP180_Breakout_Arduino_Library/
6. Unified sensor driver for Adafruit’s BMP085 & BMP180 breakouts, https://github.com/adafruit/Adafruit_BMP085_Unified
8. Atmospheric pressure, Encyclopædia Britannica, Published by Encyclopædia Britannica, inc. on 13th February 2019, https://www.britannica.com/science/atmospheric-pressure [Accessed 29th April 2020]
9. Engineering ToolBox, (2003). Altitude above Sea Level and Air Pressure. https://www.engineeringtoolbox.com/air-altitude-pressure-d_462.html
10. Fritzing app, https://fritzing.org/home/
11. John D. Anderson Jr., Aircraft performance and design, Mc Graw Hill, pg. 546
Sample of raw data readings
26th January 2020