The Effect of Rotorwash on Atmospheric Profiling
Ryan Martz
Author
03/30/2021
Added
12
Plays
Description
An analysis of how the rotation of rotors on an unmanned aerial vehicle affects atmospheric measurments
Searchable Transcript
Toggle between list and paragraph view.
- [00:00:01.630]Hello. My name is Ryan Martz
- [00:00:03.330]and my UCARE project this school year
- [00:00:04.980]was about Using Unmanned Aerial Vehicles,
- [00:00:07.100]which will be referred to as UAVs,
- [00:00:09.230]for Atmospheric Profiling, which is basically
- [00:00:12.020]finding how different atmospheric variables,
- [00:00:14.200]like temperature and humidity change with height.
- [00:00:16.600]And my supervisor for this project was Dr. Adam Houston.
- [00:00:20.610]So the motivation for this project
- [00:00:22.490]was each National Weather Service site
- [00:00:24.940]currently launches two weather balloons each day,
- [00:00:27.860]12 hours apart, and thousands of miles apart
- [00:00:30.620]between these sites.
- [00:00:32.420]So there's a large time gap and space gap,
- [00:00:35.030]which makes it hard to get a good depiction
- [00:00:37.210]of the atmosphere across the country.
- [00:00:39.510]And it's also wasteful as the weather balloons
- [00:00:41.710]are almost never retrieved,
- [00:00:43.180]so they need new sensor equipment
- [00:00:45.100]and new balloons every day, twice a day.
- [00:00:48.010]And so the idea is that UAVs could be used
- [00:00:50.230]to fill in the gaps in space and time,
- [00:00:52.170]and can be also be a more reusable alternative.
- [00:00:57.920]The problem with UAVs that I'll be investigating
- [00:01:00.688]is that the spinning rotors can mix up the air
- [00:01:03.810]and pull air from above and push it out below.
- [00:01:06.700]And so, as you can see on the diagram from right,
- [00:01:10.640]this can create some big impacts
- [00:01:13.780]with where you're actually measuring the air from.
- [00:01:16.230]It could be pulling air from below
- [00:01:17.820]and it could be pushing it out from,
- [00:01:19.970]pulling air from above and pushing it out below,
- [00:01:21.970]and it could recirculate back to the rotors,
- [00:01:24.380]so you don't necessarily know where exactly
- [00:01:27.180]the air is coming from that you're measuring.
- [00:01:31.800]And so the hypothesis that we found
- [00:01:33.690]through background research
- [00:01:35.230]is about a two meter downwash for our experiment.
- [00:01:38.880]And two things that would impact this would be,
- [00:01:41.780]wind, shearing the top off of this and making it smaller,
- [00:01:44.940]and also the ascent and descent speed
- [00:01:47.250]creating a vertical wind.
- [00:01:48.990]An ascent would probably
- [00:01:50.695]make the two meter downwash smaller
- [00:01:53.340]and a decent would likely make it larger.
- [00:01:55.950]And the diagram on the right shows
- [00:01:57.370]what happens to the rotor wake,
- [00:01:58.750]so this is below the rotors as opposed to above,
- [00:02:01.510]but it shows the effect that wind has on that rotor wake.
- [00:02:04.410]And the idea is that wind might have a similar effect
- [00:02:06.810]on the rotor or wash from above.
- [00:02:10.500]So for our methods, the set up for this experiment,
- [00:02:13.420]first we used eight Vaisala or six Vaisala XQ2 sensors
- [00:02:17.750]and two XQ!s for a total of eight sensors.
- [00:02:20.780]And we first did an intercomparison
- [00:02:22.450]of these sensors to calibrate them.
- [00:02:24.490]We zip tied them to a horizontal bar
- [00:02:27.940]and we let them record for about 10 minutes
- [00:02:30.080]and then compared their temperatures,
- [00:02:32.130]pressures and relative humidities
- [00:02:33.690]to make sure that there were no outliers.
- [00:02:35.710]And if there were, to figure out how to correct for them
- [00:02:38.740]and adjust for their biases.
- [00:02:41.320]After we did this, we tied them
- [00:02:45.130]to a tether connected to a weather balloon,
- [00:02:47.727]and the tether was about 50 meters long,
- [00:02:50.580]and we put sensors at 10, 20, 35
- [00:02:53.640]and 50 meters along this tether.
- [00:02:56.130]The 10 and 35 meter sensors were the two XQ1s
- [00:03:00.137]and they were both out of battery
- [00:03:01.640]by the time that the flights began.
- [00:03:03.400]So we only had sensors from XQ2 sensors
- [00:03:06.530]at 20 meters and 50 meters.
- [00:03:09.090]And that spacing, the 20 and 50 meters is a longer rope,
- [00:03:12.940]but as you can see from the right in that picture,
- [00:03:15.670]that wasn't necessarily reflective
- [00:03:17.280]of the height above ground level.
- [00:03:19.810]Those dots represent where the sensors were
- [00:03:22.219]along the tether, and it became very slanted,
- [00:03:27.670]so that 20 meters and 50 meters
- [00:03:29.440]was not necessarily 20 and 50 meters above the ground
- [00:03:32.560]because the buoyancy from the balloon
- [00:03:34.350]was too weak to lift it,
- [00:03:35.860]and there was also wind that was pushing it off to the side.
- [00:03:40.040]In addition, there was a GPS unit at 50 meters
- [00:03:43.250]along the tether just beneath the balloon,
- [00:03:45.480]in addition to the GPS unit on the drone
- [00:03:49.130]and two XQ2 sensors attached to the drone.
- [00:03:53.210]So those sensors I was talking about
- [00:03:55.120]are located inside those yellow boxes there,
- [00:03:57.960]just underneath the trumpets,
- [00:03:59.300]and the trumpets that I'll be referring to
- [00:04:01.040]are those white cones on either side of the drone.
- [00:04:04.370]So for experiment one, we did a stepped hovering,
- [00:04:06.990]which basically, we flew to 20 meters,
- [00:04:10.370]we hovered for 10 seconds and then stepped
- [00:04:12.610]up one meter and repeated this process
- [00:04:15.140]of stepping up and hovering until we reached 30 meters,
- [00:04:18.180]which is the approximate altitude
- [00:04:20.240]of that highest sensor above the ground.
- [00:04:22.590]We repeated this process three times
- [00:04:24.490]and then we rotated the drone
- [00:04:26.710]so that the left trumpet was facing into the wind,
- [00:04:28.710]and we repeated it another three times.
- [00:04:32.220]For experiment two, we did a mini profile
- [00:04:35.000]where we flew up to four meters and then stabilize ourselves
- [00:04:38.900]and ascended at one meter per second to 50 meters,
- [00:04:42.010]waited for 10 seconds and then descended back
- [00:04:44.380]to four meters at one meter per second again,
- [00:04:46.990]we repeated that process three times
- [00:04:49.300]and then repeated it three more times
- [00:04:51.000]after rotating the drone 90 degrees like in experiment one.
- [00:04:55.740]To analyze this, I use potential temperature
- [00:04:58.650]because it's basically a pressure corrective temperature.
- [00:05:02.070]And then this potential temperature became constant
- [00:05:04.600]approaching and above 30 meters, which is where that second
- [00:05:08.290]or that higher sensor was on the tether.
- [00:05:11.960]And so that first experiment
- [00:05:13.870]where we were stepping up to that altitude
- [00:05:15.720]to find where the air was being taken from,
- [00:05:18.000]was pretty difficult to analyze,
- [00:05:19.450]and we weren't really able to get results from it,
- [00:05:21.240]due to that constant potential temperature.
- [00:05:24.210]The altitude information
- [00:05:25.300]was also unreliable from the sensors
- [00:05:27.410]and the balloon mounted GPS system was also broken.
- [00:05:30.520]And so I had to use a recorded pressure
- [00:05:32.570]to compute an approximate altitude above sea level,
- [00:05:34.950]as opposed to using actual altitude measurements.
- [00:05:38.500]To analyze this, I split the data,
- [00:05:40.920]to analyze it I split the data into four groups,
- [00:05:43.350]ascent and descent for profile one,
- [00:05:45.800]the trumpet perpendicular to the wind and profile two,
- [00:05:48.240]the trumpet into the wind.
- [00:05:50.320]And for each instance where the potential temperature
- [00:05:52.670]was within 0.02 degrees Celsius of the potential temperature
- [00:05:55.980]at the lower sensor, so at about 20 meters,
- [00:05:58.920]I subtracted the altitude of the drone
- [00:06:01.130]from the altitude of the tether to try to find a rotor wash.
- [00:06:05.290]So this graph at the bottom shows approximately
- [00:06:07.700]what we were trying to analyze,
- [00:06:08.830]those black dots show where the M600 left and right sensors
- [00:06:12.850]intersected with the 20 meter XQ2 sensor,
- [00:06:17.160]and so we can find the time
- [00:06:18.510]where they had the same potential temperature,
- [00:06:20.830]and then having that time
- [00:06:22.040]find the altitudes of each of them.
- [00:06:24.330]And so I found that the air was taken from below the drone
- [00:06:28.000]by about 6-10 meters, but with a large standard deviation.
- [00:06:31.940]And this was largely due
- [00:06:33.240]to the highly variable data that we had,
- [00:06:35.650]and also the limited number of data points
- [00:06:38.190]as we were only able to analyze experiment two,
- [00:06:40.800]we only had six flights to analyze,
- [00:06:43.100]and which each flight, there were only a few times
- [00:06:45.316]where the potential temperatures were actually equal
- [00:06:49.456]between the drone and the sensor.
- [00:06:52.150]So that led to the large standard deviations.
- [00:06:56.440]So these were only preliminary flights
- [00:06:58.300]and a more robust experiment is taking place in Oklahoma,
- [00:07:01.220]the week of March 28th.
- [00:07:03.070]Some things that we'll be correcting
- [00:07:04.490]is using a second balloon to provide a stronger force uplift
- [00:07:08.527]to make the tether hopefully more vertical.
- [00:07:12.030]In addition, there'll be an onsite tower
- [00:07:13.770]with attached sensors and that'll provide constant altitude
- [00:07:16.710]a more reliable data.
- [00:07:18.460]And we've also learned what went wrong with the GPS set up
- [00:07:20.910]and should have more accurate altitude data
- [00:07:22.860]from that with the balloon in the next experiment.
- [00:07:26.700]Some acknowledgements and references to Dr. Adam Houston
- [00:07:29.730]for guiding me through each step
- [00:07:31.210]and knowing what to do in the research.
- [00:07:33.110]Ashraful Islam and Daniel Rico for operating the drones
- [00:07:36.338]and helping with planning and setup of the experiment.
- [00:07:39.510]And Dr. Carrick Detweiler also
- [00:07:41.340]with helping to set up the experiment and also
- [00:07:43.470]taking the pictures that were used in this presentation.
- [00:07:46.640]Thank you for your time.
The screen size you are trying to search captions on is too small!
You can always jump over to MediaHub and check it out there.
Log in to post comments
Embed
Copy the following code into your page
HTML
<div style="padding-top: 56.25%; overflow: hidden; position:relative; -webkit-box-flex: 1; flex-grow: 1;"> <iframe style="bottom: 0; left: 0; position: absolute; right: 0; top: 0; border: 0; height: 100%; width: 100%;" src="https://mediahub.unl.edu/media/16110?format=iframe&autoplay=0" title="Video Player: The Effect of Rotorwash on Atmospheric Profiling" allowfullscreen ></iframe> </div>
Comments
0 Comments