Vertical dependence of the thermal wake of a helium balloon during ascent and descent

Research Group: Paul Idiaghe, Mark Mangino, and Helio Ramollari

Launch: Whitworth Spring 2020

''The thermal wake of an object during flight has some importance in the aeronautic domain, therefore the understanding of its nature would be useful in designing more efficient flight vessels. The research done on the topic from past years were either not successful or affected by failures. We intended to study the vertical dependence of the thermal wake of an ascending and descending helium balloon using a tail of temperature sensors on our pod and parasite sensors on the other pods to record a temperature profile with respect to height (and this time, without failures). However, due to the coronavirus outbreak, we weren't able to launch our pods or record any observations.''

Layers of the Atmosphere
The atmosphere is made up of several layers: the troposphere (8-14.5 km from the earth's surface), stratosphere (14.5 - 50 km from the earth's surface), mesosphere (50 - 85km from the earth's surface), thermosphere (85-600 km from the earth's surface), and exosphere (600-10000km from the earth's surface). The temperature drops as you go higher into the atmosphere. At the tropopause, the temperature stops getting cooler, until it resumes its increase in the stratosphere. The helium balloon wold pop 30km away from the earth's surface (which is in the middle of the stratosphere).

Thermal Wake
Thermal wake is the disturbance in temperature that results from the movement of a weather balloon through the air. As the helium balloon absorbs heat from solar radiation (during the daytime), its temperature would increase. The heat transfer between the balloon and the surrounding air affects the temperature of the stream of air within the thermal wake. As the helium balloon moves higher into regions of lower pressure, it would keep expanding, increasing the size of the thermal wake trailing behind it. Thermal wakes cause temperature measurement contamination as the air within the wake is usually artificially warmer during the day and cooler in the night. The effect increases with an increase in height because of the decrease in pressure. The farther the temperature sensors are below the balloon, the less they would be affected by these temperature disturbances from the thermal wake.

Theory
Thermal wake is like any other kind of wake in that it changes based on the speed of the object it is caused by. The purpose of this experiment was to discover how the change in altitude altered the vertical profile of the thermal wake behind our ascending weather balloon. Considering the air gets progressively more cold the higher one goes (to a point) we believed that the thermal wake would shorten, indicated by a reduction in temperature starting in the lowest temperature sensors of our sensor tail.

Mechanical
The main mechanical components of our project were going to be the pod and the tail. The fact that we decided to study the vertical dependence of the thermal wake meant that we needed to build a tail and as a result our pod was going to be the last one. Also, to compare and contrast as many data points from as many different points possible, we were going to place more temperature sensors in the other three pods. For the tail, we were going to use 20-25 DS1820 temperature sensors, and 5 TMP102 temperature sensors for the rest of the pods and the neck of the balloon.

The pod
The pod itself consists of a styrofoam box with a removable lid, wrapped in aluminum foil. We wanted to keep the mass of the pod as low as possible so it was going to be made out of styrofoam, and wrapped in aluminum foil so that its surface would be reflective. It was going to be assembled using cloth strips. These cloth strips had metallic hangers attached and that would be the point where the pod was going to be fastened to other pods.

Inside the pod, we would place the flight computer with its parts. It would be fixed using zip ties. At the bottom of the pod a hole, through which the tail would go, would be necessary. This tail would be connected to the flight computer through a breadboard already attached to the flight computer. During the flight we expected the pod to move forcefully in different directions and since the electronic connections are not good mechanical connections we thought to create a U-turn of the tail inside the pod and fix it on the sides of the pod using zip ties so that the tension was not transmitted to the electronic connections. This was never tested but seemed like a reasonable idea.

The tail
The tail consists of three different-colored stranded wires and about 25 DS1820 temperature sensors. This tail would be let to flow free under the pod. We agreed to use stranded wire, instead of solid-core ones so that it would not get disconnected anywhere in its length because of the rapid changes in the direction of the motion. Since the DS1820s have three separate pins we also needed three different-colored wires to distinguish them. From the left to the right they would be Blue(ground), Black(signal), Red(power). These wires would be twisted with each other and fixed using duct tape.

In every one foot of the length of the tail, we decided to insert a slot: a three-unit section from a female header. These slots would be where we would insert the DS1820 sensors. From the experience of the previous groups working on similar projects, we were brought in attention the need to make sure sensors do not fall out of their slots. In the beginning, we had the idea to print small capsules to place around the slot and the sensor, but later we also considered the idea of using silicone sealant. This was never tested but seemed like a reasonable idea.

Electrical
The main electrical components of our project were the tail and the sensors. We decided to use two types of sensors for reasons explained in the mechanical design section of this page.

The tail
As also described in the mechanical design section of this page the tail was made up of three different wires that were twisted around each other. In every one-foot distance along the tail, we inserted the slots by soldering them to the wires. For that, we needed to make sure that the wires were not touching each other so we placed in each of the wires heat shrink sleeves. This was done to prevent any possible short circuits in the slots.

DS1820 Thermal Sensors


In the slots along the tail we were going to insert the DS1820 thermal sensors. The DS18S20 digital thermometer provides 9-bit Celsius temperature measurements and has an alarm function with nonvolatile user-programmable upper and lower trigger points. The DS18S20 communicates over a 1-Wire bus that by definition requires only one data line (and ground) for communication with a central microprocessor. Each DS18S20 has a unique 64-bit serial code, which allows up to 2^64 DS18S20s to function on the same 1-Wire bus. Thus, it is simple to use one microprocessor to control many DS18S20s distributed over a large area. The protocol used by the DS1820s requires a pull-up resistor. A "pull-up" resistor is a resistor that is inserted between the signal and power lines. An appropriate value for the resistor that goes with the DS1820 is 4.7 kΩ.

The sensors were going to be inserted into the slots in the following order : left pin = ground, middle pin = digital signal, right pin = power.

TMP102 Thermal Sensors
We planned to connect 4 of these sensors in the 4 pods and one of them in the neck of the balloon. The TMP102 device is a digital temperature sensor ideal for NTC/PTC thermistor replacement where high accuracy is required. The device offers an accuracy of ±0.5°C without requiring calibration or external component signal conditioning. Device temperature sensors are highly linear and do not require complex calculations or lookup tables to derive the temperature. The on-chip 12-bit ADC offers resolutions down to 0.0625°C.It communicates using the I2C protocol.

The connection of the sensors in the board would have been as following: VCC=p40 (via power bus), GND=p1 (via ground bus), SDA=p9, SCL=p10, ALT (alert)=p1 (via ground bus), ADD0 (address)=no connection.

Software
The code that we had written for the Mbed depended on the DS1820 library, the extended timer library, and the SD file system library. It also made use of the standard Mbed library. The flight computer's code consisted of two parts; one was written to log messages into the SD card and the other was written to log data along with time into the SD card on the board. The former code was written to ensure error messages could be saved into the SD card so that we could detect whatever happened along the way, while the latter code was written to collect data from whatever sensors we would have used along with the time.

The sequence of this code goes as thus: First, four timers are enabled to keep track of the different tasks, then the software attempts to mount the SD card. If that fails, it exits the program and returns an error message. If it doesn't fail, the code attempts to open the file on the SD card, if that fails, it returns an error message, else, it goes on into a loop, collecting data from the sensors for 120 seconds and writing their data every 10 seconds. Also, their data is flushed into the SD card every 60 seconds (the flush function ensured that data is collected at intermediate times in case of power failure or other unexpected circumstances).

We were yet to determine if we would collect temperature values from our tail and the other pods either concurrently or consecutively.

The code goes as thus;

Calibration and Testing
The Validation and Calibration process for our project was going to be very important as we were trying to study the differences in the temperature values in different points. To have a common reference system we decided to calibrate all our sensors in accordance with the Fluke Thermometer model 52 II. We agreed on having the Fluke thermometer as our reference scale because if all the sensors were calibrated in agreement with one common reference point, the differences between different points were still going to be the same even if the Fluke was not displaying the true temperature.

After we confirmed that all the sensors were working and displaying a reasonable value for the room temperature, we started recording different data points of temperatures from the freezer. We planned to do the same process in room temperature and warmer environments, but the time did not allow us. The collection of data from environments with great temperature differences is important to see the behavior of the ratio of real temperature and the one displayed by the sensors. If this behavior is linear, it means that the sensors are in a good working state. Otherwise, it means that there is something wrong with their sensitivity.

We started recording data for all the sensors and the Fluke thermometer when the temperature was no higher than -20oC. While the temperature lowered further we recorded the temperature displayed from the Fluke thermometer and the sensors. After we recorded at least 9 different temperatures, we built a scatterplot and a trendline using a spreadsheet application to see the relation between the values recorded. As it can be seen from the graph below used for calibration of one of the TMP102 sensors, the relation was linear. After this was confirmed, using the LINEST function from the spreadsheet application we recorded a value for the slope of the line and an intercept point value. The respective values for this sensor in particular were slope :1.39 and the intercept: 11.35. By multiplying the data points from the sensors with the value of slope and adding the intercept value to this product, we were able to convert the points from the sensors to the common reference system. This process was done in exactly the same way for all the 25 DS1820 sensors and the 5 TMP102 sensors. After recording a slope and intercept value for all the sensors, we planned to finish the calibration of DS1820 sensors after the collection of data (and before analyzing them since they were all going to be working from the same source code). By manipulating the code that was going to be executed for each TMP102 sensor, we could have finished the calibration before the launch.



The new code used for the already calibrated TMP102 sensors is as follows:

Data and Analysis
Due to the disruption of the semester caused by the Covid-19 pandemic of spring 2020, this experiment was never completed or sent to the stratosphere.