Wednesday, February 22, 2017

A Guest Post by Handsome Dan

Wait…You’re Calling it GOAT?
The Greenhouse, Ozone, and Atmospheric Trace gas (GOAT) project is an atmospheric experiment built on a high altitude balloon platform.  It is one of 12 payloads that will be integrated onto a larger gondola that will float at approximately 37,000 meters above ground over the state of New Mexico for approximately 15 to 30 hours.  The primary mission of GOAT is to add to the continuous study of SO2 levels in the upper atmosphere and to test the collections methods to aid in similar future high altitude balloon atmospheric studies.  Simultaneously the payload will measure the concentrations of NO2 and Ozone for real time calibration during the flight and to add to the understanding of how the three gasses interact. Though the build poses many challenges to the team, the project parallels the team’s previous HAB experience and continued interest in engineering and atmospheric experimentation.    

What’s the Point?
Studying the concentrations of SO2, NO2 and O3 in the upper atmosphere is important for learning more about atmospheric composition, how they interact and their effects on biological life.  Sulfur dioxide is a toxic, reactive gas that is emitted both naturally and from human activities such as the consumption of fossil fuels.  The gas tends to react with moisture in the atmosphere, creating sulfurous acid and sulfuric acid aerosols which contribute to acid rain.  Acid rain affects the acidity levels of soils, water supplies, organisms and their ecosystems.  Ozone is also a highly reactive gas which is important in the absorption of ultraviolet radiation.  Due to its reactivity, there is a relationship between ozone concentration levels and the amount of NO2 present.  Nitrous dioxide is also a large contributor to acid rain and the depletion of ozone in the atmosphere making it an important trace gas to study.  Much of the NO2 concentrations is also due to human activities.  Understanding the interactions of the three gasses will show us more about how our atmosphere works and how that interaction affects our lives.

I’m not Dramatic, I’m Just Sensitive
The GOAT project will measure these gasses with Arduino calibrated electrochemical SPEC sensors as well as passive Ogawa samplers.  The team will compare the collected data from the sensors to the data reported by the EPA in the geographical region.  The goal is to use the data to understand composition of the upper atmosphere and to test the accuracy of using inexpensive, consumer quality terrestrial based sensors for high altitude data collection.  The expectation is that the data the sensors collect will fall within the same levels reported by the EPA.
The chosen SPEC sensors will be paired with diode temperature sensors and are sensitive enough to collect data at the expected concentration levels.  They will be controlled by an Arduino microcontroller paired with a microprocessor that will interpret and log the data from each SPEC and temperature sensor, taking readings every 15 to 30 seconds.  The sensors operate in a variety of atmospheric pressure, humidity and temperature levels. The most restricting parameter for the sensors is their operating temperature range.  Our most challenging engineering task will be to keep the sensors at a temperature range of -25 to 50 degrees Celsius which allows a buffer against the sensors operating extremes and ensures more accurate readings during the flight. 
Before flight, the sensors will be given a known amount of chemicals to establish a baseline for measurement.  The sensor array will also be tested in a cryo-vacuum chamber after assembly in order to test their functionality in an atmosphere that mimics the possible extremes the payload may encounter.  Air flow will be directed across the sensors using fans or possibly blowers which will be decided on during testing. The sensor housing and hull will be insulated as well to help minimize the effects of extreme cold or hot conditions.  In order to calibrate the sensors for fight, they will be mounted along with the Arduino microprocessor several hours pre-flight, which will use the manufacture data to reconcile the readings the sensors take to the ambient temperature.  The team will be able to log the sensor readings against time, temperature, atmospheric pressure, and altitude. 
Under the recommendation of Dr J.J. Bang from the NCCU atmospheric lab, the payload will also carry Ogawa passive sensors to test their usefulness in collecting samples in high altitude conditions.  The samplers use a coated pad to measure concentrations of SO​2​, NO​2​, and O​3 from ambient air without the requirement of a power source and can be mounted inside the sensor housing which will have access to ample air flow.  The sampler is designed to be exposed to different weather conditions, keeping out debris.  They are unaffected by the extreme temperatures the payload will encounter and will be well protected inside the housing.  The sensor pads can be affected by moisture so silica gel packs will be installed to mitigate that risk.
The GOAT payload will fly three sensors of each type for redundancy. The electrolyte sensors will be separated into three banks that will consist of one type of each sensor.  The three banks will be mounted from the outside of a rectangular shaped tunnel with one sensor from each bank occupying one wall.  The side of the sensor that needs access to air flow will be on the inside of the tunnel.  The tunnel itself will run from the fore to the aft plate as shown in the CAD model and mounted to air flow ports on either end.  An electric fan will be mounted on both port ends facing in the same direction to push atmosphere through the tunnel and over the sensor array.  The sensors are designed to work in static conditions at ground level.  The fans will help move atmosphere over the sensors in the thin stratosphere.
The sensors will be mounted with bolts on small non-conductive standoffs.  Wiring for the sensors around the outside of the tunnel and along the inside panel wall.  The team will test anti – arcing spray and layers of anti – arcing silicon to eliminate the high risk of arcing in the upper atmosphere.
In order to keep the integrity of the measurements each bank of sensors will take their respective readings simultaneously.  The readings will be coupled with the telemetric data from HASP and stored on redundant SD card devices from the dual Arduino setup (more on that below).

Ryan Hull, it’s pronounced “O-G-A-W-A.”
The pads will arrive in a refrigerated storage container. Before flight the pads will be removed from the container with tweezers and placed into their respective housing.  A secondary pad will be taken out at the same time and kept isolated as a blank control.  The time and date will be noted.  The pads will be reacting as soon as they are added to the payload until they can be removed after landing.  As soon as the payload is recovered the pads will be removed from their housing with tweezers and placed into glass vial purposely used for their storage.  Telemetry and atmospheric readings from the payload will guide analysis in order to have the most accurate results.  Dr. Bang will be consulted to help with analysis. 

Structure Heard
The hull will be made from 6061 aluminum paneling, bolted together with stainless steel interior corner brackets.  Due to their durability steel fasteners have been chosen.  Thermal expansion and contraction of the aluminum and steel is negligible in reference to fastener tolerances.  Due to the relatively short time frame the payload will be assembled we can also neglect the steel and aluminum mismatch.  The hull is a rectangular prism shape with fore, aft, front, back, top and bottom panels.  The fore, aft, top and bottom panels will be permanently mounted leaving the top, front and back panels removable for assembly and trouble shooting.  Joints for permanent mounting will be sealed with a silicon caulk cured in a vacuum chamber.  The bottom of the hull will mount to the provided PVC mounting plate with no modification to the plate needed save the holes for the through bolts.  Fasteners for the bottom plate will be counter bored and fastened with flush mount bolts to ensure a good fit to the PVC plate without any major modifications required.  Any internal struts or shelves will be made from aluminum sheet, aluminum angle, or PVC board depending on the requirements and restrictions.  All custom parts will be 3D printed with ABS plastic or made by our team machinist.

It’s Getting Hot in Here….
Thermal control for the payload will primarily come from the 15W power supplied by HASP and the sun.  The temperature range will be controlled by the Arduino between -25 and 50 degree Celsius since all materials and sensors in the payload are rated within that range.  Testing will determine the exact equipment and methods that will be used for thermal control as the assembly comes together.  For testing the team will use one of their several contacts in order to source a cryo-vacuum chamber that can be used to mimic the diminished atmosphere and expected temperature fluctuations as outlined.
The plan is to fly Peltier-effect thermoelectric units for each SO2, NO2 and O3 sensors to generate or bleed thermal energy to or from the individual sensor as needed.  The exact amount of heating and cooling that will be required is unknown.  And due to the time lapse between the ground and the payload, control will have to be automated from the Arduino.  The coldest portion of the flight will be when the payload travels through the troposphere.  During float, however, the challenge in thermal regulation will come from needing to cool the payload.   The hull will be painted with a reflective aircraft grade paint and heat generating components connected to the hull with copper heat straps which use the hull as a heat sink.  Testing is needed to confirm that these options (or others) do not chemically interfere with the sensor array.
GOAT will fly two AM2315’s with one capturing external atmospheric temperature and barometric pressure conditions and the other will be inside the hull taking measurements of the internal conditions.  In addition, there will be two BME280’s inside the sensor tunnel reading the temperature of the air as it passes over the sensors. As referenced above each sensor will be fitted with a diode temperature sensor that will be read individually by the Arduino to tell if temperature regulation is needed and for which of the individual electrolyte sensors. 

Dual Arduino Action
The Goat will fly two Arduino Megas in a master and slave configuration.  Due to the number of sensors and sensor management devices, two Arduinos are needed.  An oscillator will be used for accurate time keeping and two SD storage cards will be used for redundant data storage.  The master Arduino will control the overall operations of the payload and delegate tasks to the slave Arduino that it cannot complete.  To complete tasks efficiently, each Arduino will have individual tasks assigned to them to complete at set intervals.  The master Arduino will be responsible for the serial uplink and downlink.  Plans also allow for real time data to be sent to the ground and read through a web based interface that will be provided for HASP.  This gives the team information relatively close to real time and the ability to send commands in the uplink to the Arduino pair.
Once power is provided to GOAT, the Arduinos will cycle through their startup processes and the fans will start. After receiving the first batch of telemetry from HASP the oscillator clock will be configured with the timestamp GOAT received.  At this point the Arduino programs will run the majority of the payload’s operations.   The dual set up is designed to be dynamic and adaptive to any in-flight situations that may occur and will be able to receive serial uplink commands from the on ground interface if needed.
The program for the dual set up will be created by the software team.  Program operation will consist of two stages.  The first stage will be sets of constant priority tasks such as electrolyte sensor temperature readings and serial uplink commands.  Those priority tasks will inform what will happen in the second stage.  Based on the information from the first stage, the second stage will be able to do several other operations per cycle.  Options for the second stage include heating, cooling, serial downlinking, and sensor reading.

Serial Downlinking and Uplinking (couldn’t think of anything fun)
The serial downlinking will be set to send packets of information to our web based interface at a rate of approximately 1 stop bit per 8 bits of information with a maximum packet size of 1200 bits.  That is enough to provide memory for the packet’s timestamp, checksum, data, and recommended terminating bit.  This means that our downlink rates may be as low as one packet of information per second.  The data portion of the packet will allow for up to 4 sensor readings to be transmitted per packet which allows for possible redundancy in the data sent.  The downlink process will have three modes that have respective corruption risks associated with each. 
The payload will only be using the ‘power on/off” discrete command as outlined by HASP.  There are 256 commands that can be sent up to the Arduinos that range from “request” to “force” commands that can be used depending on the situation read from the downlink information.

GOAT Power!
GOAT power will be provided by the HASP gondola at 30V and .5A.  This gives the payload a total of 15W available to power all of the systems.  Voltage regulators will be used to step down the voltage to an amount that is usable for the payload.  The payload has some systems such as the fans and Arduinos that require continuous power and other systems such as the SD cards and thermoelectric module that will use power intermittently.  The Dongles and the oscillator will be powered directly from the Arduinos. 
The thermoelectric modules draw amperage well above .5 when they are all in use.  To circumvent this problem relay switches will be used to selectively control the modules and regulate voltage flow.  The relays will be controlled by the Arduinos and will allow the modules to either heat or cool the sensors depending on the voltage flow.  Weight mitigation due to the amount of wiring required for each of the payload components is being explored by the Electrical Engineering team.

Show Me the Money!
The GOAT project has already received $5,000 from NC Spacegrant and may be able to secure additional funding from industry partners, Durham Tech, and the NCCU atmospheric lab.  The team is well rounded and resourceful.  Tool and material procurement, shop space, and machinery can be easily sourced.  Enthusiasm for the project by the team is high and a detailed schedule for testing and completion of the project has been cast. 

2 Cents

The challenges in thermo control, programing, proper testing, and electrical engineering are formidable for an amateur team of students, but so were tUR-1 and tUR-2.  I think we can pull this off and do it well.

Friday, February 10, 2017


Fig. 8: A cross-section of GOAT
A: Ozone Sensor array, B: NO2 Sensor array, C: SO2 Sensor array (mostly occluded), D: BME Sensor array, E: Ogawa Canisters, F: 40 mm Fans
G: Heatsinks (rudimentary copper strapping; mostly not shown) with attached Peltier thermoelectric heater/coolers
H: 2 Arduino Megas
I: AM2315 (Internal, mounted on clip)
J: AM2315 (in exterior facing alcove and housing)
K: EDAC and DB9 Connector slots
The black handle (not labeled) remains permanently affixed to the top hatch in the center of the top panel. The top hatch and handle are designed to be repeatedly attached and detached; the top panel is a permanent fixture.

Thursday, February 9, 2017

Revisionist History

We are working hard on tackling the comments from reviewers so that we can keep a seat on HASP. At first blush the comments looked pretty simple and some of them were even complimentary! However, now that we're up to our elbows in this, it's a different story. 

We had to remove our biologic experiment and cut out anything related to that. This meant that the biggest engineering challenge of our payload was also sliced out. We also had a request to add some new gas sensors which has opened up another set of tasks to program and wire those. While we aren't entirely rewriting the application from scratch, we're changing so much that it almost seems like it would be easier to do it that way.

Ian and Emes have succumbed to the flu and Gramma has been diagnosed with walking pneumonia so Abandoned is...once up to his nickname. 

We have decided to reward ourselves with skeeball if this is ever finished.

 On Tuesday Memes and I were starving to death, delirious with hunger, mumbling and grunting to each other and yelling at Jimmy about wiring over Jimmy sent a pizza to us in the lab and was basically a hero. It was a glorious moment when Dan and I realized we could use complete sentences again.

The geology lab has seen some stuff over the years! It makes me happy when all of the kids from the different NASA projects wander in and set up their work. I like having a space with huge tables that is an inviting place for folks to work.

Yesterday I went to the atmo lab at NCCU to help Jimmy measure Ogawa Samplers.

And then Seth came by to show off something he had 3D printed for the FLOW project.

Last night we had a wonderful conference call with the Wizard Kings all clustered in the NCSU library, icons coming and going in the app, Gchat off the chain, and everyone working to refine what we have. It's a lot of work but it is so cool to see it coming together. I was lucky that Dr. Easson, Dr. Panhorst and Dr. Augenbaugh found opportunities for me to participate in undergraduate research and I'm happy to see what a positive impact it is having on all of these students.