Discussion in 'The Watering Hole' started by Winston, May 29, 2019.
good call dude!!!
Yes... And the Sensor package is named "Bill Paxton"
Nice looking rocket!
Did you post the open rocket file anywhere?
Not yet... I've got a line on someone who says that they're in touch with the builder, I'm hoping to open a dialog with him to clarify a few details before I let this one soar. The stock version is now ready and posted.
I've now got a direct line with a member of the project... No, not Reed. I'm hoping to score some more images of the rocket for documentation purposes. I'm also trying to reach the builder to get some questions answered about the rocket.
FYI, that "white" aft transition is not white, it's clear(ish), and they are using some foam inside the payload section (looks like the stuff Frank Burke uses for his gliders).
With the permission of the people who made the sensor package, I give you... "Bill Paxton"
Photo Credit: ChasinSpin
From what I can see, there appears to be a yellowish "window" cut into the nosecone. You can also see the clear aft transition. I have yet to get into direct contact with the builder to ask about the decisions made in the assembly of this sensor package.
I'm told that the multiple "spots" located around the payload section are vents to allow for airflow to reach the humidity sensor(s).
I've also been granted permission to share images of the actual sensor package...
Photo Credit: ChasinSpin
I've also been told about the waterproofing used...
The bit about salt water was a side discussion I had asking about salt water recovery and the possibility of me building a water rocket for my local flying activities.
Why pick a rocket with canards for this? Useless drag, all it does is reduce peak altitude.
Over the duration of the rocket’s upward flight, I would question whether an internal thermometer would be sensitive (fast responding enough) to plot very many TEMPERATURE data points.
There are a number of questions that I'd like to pose to the builder. Price point I'm guessing. The kit's not very expensive. I don't know if there's a thermometer in there or not. I'll ask... Later... After I get some sleep.
Hope you had a good “snooze” as my Dad says.
Post one shows a circuit board. On lower right side are two sensors, one says
And second says
Lots of discussions on TRF about size of vent holes, I guess these relate to correct altimeter readings or to prevent internal pressure from blowing out nose cones. While these holes may allow enough external air to come in for correct pressure measurements, I would doubt given the short duration of a low power rocket flight they would accurately transmit enough air to allow readings for humidity and temperature fast enough to be useful.
I was getting pretty fuzzy there at 2:30am.
I've established contact with the builder of the rocket now. He's currently busy, but we'll talk later today.
OK I'm awake again... Spent about one-and-a-half hours, yesterday, talking to the builder of the rocket, Curtis Brooks. He's the engineer for Team Dominator. He's not a rocketeer like most of us, who do it for the hobby, he was just doing what Reed asked him to do. Here's a paraphrased report on the questions I asked him and his answers.
Q: Why a cluster? A: From their experiments, single motor rockets (even with larger motors) don't seem to have the oomph to punch the rocket through the winds into the region of the tornado they need it to be in. Also, due to restrictions from state to state regarding rockets, going with LPR is less restrictive than using larger, more powerful motors.
Q: Why not a preloaded tube launched system? A: Reed wants to have be able to have 10 rockets ready to fire, but the only tube launched systems we could find use metal fins, and due to weight limitations (from the motors), that's not an option.
Q: Why all the holes in the payload section? A: They're there to improve the airflow over the sensors, they were made with a grommet hole punch. In the future, the design will likely change (perhaps with some of those mods being suggested by me... K'Tesh).
Q: What kind of launch rod? A: 1/4", the rocket is mounted on the launch rod from inside the Dominator, and carried out to the launcher, ready to fly. It now takes about 30 seconds between opening the door, and firing the rocket. On the rocket that was successfully used, there was a second launch lug installed for a smaller launch rod (this was not its first flight). That particular rocket was launched about 20 times for practice, but is considered expendable. The sensors, however, hopefully can be found and reused.
Q: Why did you reposition the smaller fins? Why have them at all? A: In the experiments done with this rocket, they found that the weathercocking (he frequently used "weathervaning" when referring to this) was more favorable with the new position. There's also some ballast in the back of the rocket to help reduce weathercocking.
Q: Why launch it at an angle to the wind direction, and not directly into it? A: Because Reed wants the rocket to enter the tornado's uptake winds.
Q: Why the clear aft transition? A: No special reason, it was something that arrived when we ordered parts. However, according to ChasinSpin, to get black plastic black, requires some amount of carbon. Carbon may interfere with the electronics's ability to transmit, and we may be switching to a clear Polycarbonate payload section (ditching the cardboard altogether). These will be vacuuformed using molds of the originals.
Q: What paint did you use? A: Krylon
There were also a lot of questions about the construction of the rocket, and I offered a few building tips for him. He's got Dorothy's clone at home, and it'll be a few days before he returns there. I'll be getting more measurements from him later
Here's some more images shared by ChasinSpin of the "Bill Paxton" sensor package. The part that I thought was a window into the payload section turned out to be an illusion from the tape that they use to hold the nosecone to the payload section's body tube. However, that may have inspired a new way of making the the vents needed for the sensors. Cutting slots into the nosecone (kinda like a shark's gills), then covering that with a shield made from packing tape stuck to itself. The shield is taped in place, and held against the nosecone on the up phase, but then flops out of the way for the sensor reading phase of the launch.
Photos by ChasinSpin
What?! Aerotech F32-X just as one example.
What?! You don't need folding swing-down fins as used with military 2.75" FFARs which might need to be metal although there is a not very well done 3D printed example designed by someone else that I've posted a link to in the 3D printing forum. I'd use a 3D printed ABS fixed and swept 6 or 8 fin base fin can within 2.5" or 3" PVC pipe. Mid-section 3D printed ABS, highly swept, tiny chord "fins" (almost whiskers) for maintaining centering in the tube with virtually no effect of moving CP forward. There are several ways I can think of to keep the tube sealed with failsafe removal even in high winds before launch, the best method being determined experimentally.
Just got off the horn with Curtis and I posed these questions to him.
He said that ChasinSpin's sensors have G limits, and the sensors are not cheap. Breaking them as they come off the launch rod would be costly and counterproductive.
We also have to remember that the opportunities to safely test these things in those conditions are few and far between. There is also the issue of Canadian laws/licensing to remember, Curtis is in Canada, and is reliant on items he can source there.
Looks like you got him there... They were thinking of rockets with fins that wrap around the body tubes inside of the rocket and deploy at launch. Not like most of us would, where the fins are fixed to the body tube. One thing to keep in mind is that the space on the top of Dominator II is limited due to the Gull Wing doors that they now use. I suggested a Colonial Marines-like-turret (from James Cameron's ALIENS), but they need to access the back of Dominator II for their ground sensors and other supplies. Still, perhaps 4 tubes could be mounted up there (not 10 like Reed was thinking). There is a need to keep them sealed because of all the rain and hail, but one small change here can cause big changes elsewhere. The Dominator is as wide as it can possibly be, so no clamping them to the side. I would also guess they need all the visibility they can get, so I suspect the hood is off limits too.
More photos are pending. Also, the photos that I posted earlier from inside the Dominator appear to be old... Curtis said that rocket #3 and #4 were lost, and it looks like #3 is in that photo of the men holding the rockets in there. It was likely taken when they were doing some launching with a Japanese film crew last year. He also said that somehow the payload sections and launch bodies got mixed up... The sensor package launched and recovered was #2, the rocket was #1. Both rockets are identical. So far, rocket #1 has not been recovered.
Yes, with measuring G-limits perhaps too low to monitor rocket G-forces while under thrusting of a high average thrust motor, but not even remotely close to the level of damaging levels of G-forces of any MEMS accelerometer I've ever seen specs for. The accelerometer may saturate, but only for a very short period of time and it will not be damaged in any way. If I had the total mass of their electronics I could sim it to see if the Gs even reach the 16G limit of many common MEMs accelerometers (I'm not looking at the specific one they're using while typing this, so it may have another common limit - 8Gs). If MEMs accelerometers were that fragile, imagine what would happen to the one in your smart phone with probably a 2G limit if you dropped it.
Aerotech retailer in Canada:
Yeah, but I'll be the first to admit that a sealed, weatherproof, field swappable, tube launched system adds a large number of complexities and issues that need to be addressed versus the much greater simplicity of the usual launch rod method. It just makes me cringe to think of two sharp metal rods pointed skyward on top of a vehicle in close proximity to a violent storm. If they don't have one already mounted on top of the Dominator, they should install an atmospheric electrometer to possibly see if a strike is imminent. It could serve another very useful scientific purpose if its data is logged with their GPS position and the time: electrostatic field data under the storm. Also, it could serve as a warning alarm when they're outside the vehicle. The professional versions require a very good earth ground, but that's not possible. However, some amateur DIY ones not meant for extreme accuracy that I've seen on-line do not.
According to ChasinSpin, the sensors (w/battery) is 17.7 grams
My latest PM to Curtis (paraphrased):
Winston, I also passed along your info about the motors, and "G loads".
Looks like ChasinSpin's a new member of the Forum... I'm now in direct communication with Reed as well... Let's see if we can help them get
ChasinSpin's still getting used to how things work around here... So, he asked me to share this with you folks.
That's the launch phase and the data
So I'm going to attempt to answer some of the sensor questions posed in this thread. (I'm not the rocket designer however, but where I know I'll try my best)
That's correct, I wasn't concerned about the G-limits of the Mems sensor. It's rated to +-8G for data, and 2000G (<1.0mS). So although it doesn't completely quantify the launch phase (forces are slightly over 8G), it's purpose is to measure forces in the Tornado in flight (which it's within spec for).
In general, we're concerned about G forces however as any mass of electronics / battery can have failure scenarios under high G, we mitigate these,
but limiting G forces is a prudent practice. Additionally we don't require (or want) the rocket to go high, but we need it to get there fast.
We also had limitations on the Inertial Management Units we could choose from given the speed we were traveling at the quality of data we required and size/weight constraints. The first prototype, testing showed that sensor couldn't keep up with the speed and resolution we needed, so we had to replace with a better sensor. (the one we launched with).
All data is sampled at 10Hz, except GPS which peaks out at 5Hz, and Battery (1Hz)
A few comments about the data to bear in mind when you take a look at it (I've only included the launch data which would be of interest to you folks).
GPS Altitude will take some time to be accurate, this is normal, GPS favors 2D positioning over 3D. During launch, the GPS is subject to greater than 4G, so it will be inaccurate until the acceleration is more reasonable and a good fix has been restored.
Pressure data will look inconsistent compared to the normal data you see when launching rockets.... the reason is that the Vortex of the Tornado was entered during the flight phase, so pressures are unusually low for the altitude.
Kartman: Countdown: We were inside the vehicle and were aware (camera was outside in wind).
Winston: It’s possible Curtis didn’t see your email at the time, or he was working on something.
Sooner Boomer: Temp/Humidity. I purposefully didn’t put sensors outside the payload bay due to being subject to a Tornado environment although that was the original plan. The concern being thin wire sensors and impact debris, so opted for air flow instead. The sensors are quite sensitive (even when mounted on the board), and we’ve post processed the data to adjust for the response time using similar techniques that are used for Radiosondes. That said, I’ll be adding an external sensor on the next board spin, but still keeping the board sensor as a fail safe in case the external sensor is damaged in flight, as we learnt from the data that we were very close to lagrangian flight so impact to the sensor package was minimal and not as much of a concern as I thought originally. Also confirmed from the lack of damage to the payload bay. We also have completed a thermal study on the board / payload bay. Transmitter is very low power, so RF doesn’t couple into the sensors enough to corrupt the reading (if that’s what you wondering about), and you are correct, higher powered transmitters would require a different layout.
Jiffycoil: Highest altitude reached by the sensor (parachute attached to Payload Bay from rocket containing sensor) was 11,914.40m (39,089 ft)… which is from the GPS. This was due to the tornado and updraft within the storm. The rocket was purposefully designed to not go to the altitudes you guys are competing for, because we wanted the sensor to follow with the tornado so we could document wind speeds, position etc., rather than have it fly through the Tornado at high speed which would be of less scientific use to us. I.E. Lagrangian flight. (although it would be cool to launch a rocket to 39,000 ft too, but we wouldn’t be doing that from Dom 3 because of safety)
Winston: Chute/Parachute. Yes there’s an attachment point on the PCB. For this flight it was in in the Payload Bay, but now we’ve proved Lagrangian flight, we considering using that attachment point. Concern here would be damage from debris, but I’m reasonably sure that most of the time this wouldn’t happen, 3D printing a minimal cover maybe an option too.
Winston: Dropouts / Not Impressive. Hardware is designed to do the job in a very lightweight / small package with a complete sensor suite onboard, transmit that over radio and be recovered on a minimum power budget. It’s less than 1x3” (including battery and Antenna), and transmits realtime data @ 1Hz for upto 15 miles, and permits further recovery of 10Hz data remotely or on remote recovery of the sensor. Dropouts with Radio are to be expected, as no radio transmission is 100% reliable and also, we had drebris…. Dominator 3 contains a lot of high carbon steel for protection and at some points we had debris flying. I had to bring my arm in and roll up the bullet proof windows a few times (to avoid losing it) whilst moving after deployment which then acts as a faraday cage. Also, naturally towards the end of the range reception will be spotty. What was interesting is the sensor also listens to the the receiver and the sensor transmission range was very close to the receiver transmission range, which means antenna performance at both ends was similar. In future we’ll be mounting an antenna on top of the Dominator 3, this was the first deployment into a Tornado of the system. I came to the project late and we had 2 months to come up with a solution…. unfortunately the tornado season doesn’t wait. We ticked off what was important to get done first and I didn’t have time for an external antenna and also we didn’t want to modify Dom 3 until we’d deployed and tested. I’ll be tuning that antenna to compensate for all that steel ! Nothing like 1/4 inch high carbon steel to interfere with a radio signal. I'm estimating the range will likely be 2-3X further. (we reached 10km)
Winston: Electronics Guy/Rocket Guy. Yes we’re not the same person. I’m the electronics guy, Curtis is the rocket guy. Curtis designed the launcher / rockets, he’s also the engineer that keeps the Dominator running along with Sean. I designed and built the hardware / firmware and software and I also storm chase.
Boatgeek: You’re correct, it was deliberately designed to get low but fast. We went with parachute, and would likely do that again given the performance rather than streamer although we were initially considering both options. There maybe a case for parachute and streamer launch at the same time too.
Matt_The_Rocketman: The electronics survived a fairly harsh environment (EMP’s from lightning, cold temps etc.). That said, it’s been designed to be sacrificed, i.e. if the electronics is destroyed, we still have data till that point. Satellites and Doppler radars have their limits unfortunately. Satellite can’t see what’s happening inside the tornado,and Doppler has issues like “Folding”, “Resolution” and “Beam Angle” that limit it’s usefulness. For example, unless the Tornado is on top of the radar site, your standard weather radar can’t directly see a Tornado on the ground. All techniques have their uses. Ultimately we’ll be launching multiple sensors at once. The rocket can be found by driving the path of the storm looking for a signal. We fully intend to recover every rocket where possible. In a tornado there are many projectiles (like 2x4’s, roofing, vehicles) that are considerably more dangerous.
Additionally, the deployment is deliberately at altitude and when it’s ejected from the storm, it’s just floating down. The flight of the sensor demonstrates that it wasn't a projectile risk.
Sooner Boomer: Humidity at 100% constantly. In practice, this isn’t generally the case, dry air is also involved in the process.
Nytrunner: Correct, altitude from pressure is not representative in a Tornado (at least if you’re in the Vortex), so GPS was included. We found that after it exited the boundary layer, then pressure was more indicative of altitude, but below that there was a 115hPa diff due to the vortex.
K’Tesh: "I wonder what took them so long to try that out”: The big issue to date has been radio, gps, size, battery life, deployment mechanism and a team that can get in close. Until recently, sensors were too big and power hungry requiring large batteries. But the sensor is only one small part of the puzzle, expertise in forecasting and engineering was required too. Everything had to come together for this to work and each Team member had their job and brought different skillsets to the table. As a new addition, I’ve found it quite amazing to be involved in a team that functions so well, it’s rare.
Barbar: Temp/humidity/baro & Temp: Yes there are 2 temperature sensors on board. I liked the performance of the humidity/baro sensor, but I added a 2nd temp sensor for increased resolution/performance. It also serves as a backup and surface area is larger improving response time.
K’Tesh: Black paint/Curtis discussion. Just to clarify this, black paint often contains carbon. Carbon will interfere with radio transmission. Other plastics can contain carbon too, so anything black is avoided where possible, especially large areas of it around the antenna.
K’Tesh: Yellow Tape. The yellow tape was just Kapton tape I had on hand for attaching the batteries to the sensors, I couldn’t find the electrical tape, so grabbed some Kapton to attach the nose cone to the tube after inserting the sensor. When I got the sensor back, I just peeled of the tape and stuck on the nose cone (for no particular reason.)
Bill Paxton/Dorothy: During the design, many commenters on twitter/facebook mentioned Dorothy / Bill Paxton, so we thought it was appropriate to name the first sensor and rocket after them. It was sensor number “8” which was launched, which was 8 of the first 12, that I built.
Power: One of the challenges of the design was power usage / budget. The size (1” in width) constrained the battery to 1” width and there were
a number of power hungry components on board. (GPS / Radio / Processor / Flash). The battery constraints meant 350mAh was the largest I could go in a
easily acquirable battery form factor. Sleep current is 100uA. Power usage is upto 200mA peak. It can record for 2.5 hours, and run for 2 days in recovery mode. Also, batteries don’t do well at cold temperatures.
Radio: The radio protocol handles many sensors at once using LoRa which enables us to receive data below the noise level, which allows for low power / long range transmission. It’s uses a Time Division Technique. The amount of data, multiple sensors and regulations limit the rate at which we can send the data, so in flight, we used 1Hz as the sensor rate, although internally we record at 10X that rate, which can be downloaded later.
Some of the design / data we may keep back for now , but any other questions, design decisions you have I’ll attempt to answer. Ideas are also welcome, we may not follow every one (often there are cost/complexity or operational issues concerned with how the Dominator operates, the workflows or practicality), but certainly we’ll consider every one.
We all love K’Tesh’s model, it’s impressive the work that he’s done. We’re pleased that K’Tesh reached out to us so we could answer some of the questions you all have.
WOW! Thanks for taking the time to answer all those questions (and all the questions I've already passed on to you). Thanks for the kind words too!
Well guys, I think that's cleared the questions we've all had. Now, let's see about brain storming some ideas that could help them out. Perhaps getting these guys certified would be a good idea. While I doubt that they'd actually launch a HPR into a tornado, it'd fun to see them show up at a club launch during the down season.
For the chute, have you considered switching to a tangleless pull-down apex / toroidal similar in shape to a squirrel.ws snatch pilot?
There are several manufacturers on this forum that could turn one out, and it should greatly reduce the chance of a fouling.
Another thought about tube launching. Foam (pool noodle) sabots have been demonstrated in the model rocket community. A forward sabot could provide both travel aerodynamics and weather protection.
I've been thinking about your answers about the power of the motor. I think there's something you didn't factor on... The motors only fire for a limited amount of time, and then the rockets coast before the ejection charge deploys the recovery device (presuming you're not using a zero delay motor for a two stage rocket). However, the rockets are extremely light weight for their size. Once that motor stops thrusting, it's going to slow greatly down in a high wind situation. Imagine the difference of throwing a crumpled piece of paper, compared to one that is tightly folded (like the triangular football origami fold). Even though the mass would be identical, and given the same amount of force to "launch" them, the crumpled paper will decelerate rapidly, and fall to the ground much closer than the folded one. Back to rockets, even if you managed to be so close to a tornado as to be able to punch through the downdraft, and into the "eye" of the storm, unless you're still under power, there's no way you'll punch through the other side of the storm. And even if you did, the winds on the backside of the tornado would still pick up your sensors.
I'm also predicting that your sensors that are on smaller parachutes or streamers will be more likely to be damaged or destroyed, as they'll be coming down in higher wind velocities than their larger parachute equipped counterparts.
Based upon the very high Gs experienced by commercially available sensors with SMD electronics and lipo packs long-used in HPR, you have nothing to worry about for properly connected electronics intended for high Gs. If you want it there fast, use a high average thrust motor like, as one example, an Aerotech F32-X. Also, if you are wanting to aim for a specific target, study rocket design for minimum weathercocking and greatest stability in high winds. What is being used isn't that.
EDIT: I see elsewhere in this thread that what is being used is "two Aerotech D21-7s", so high average thrust isn't an issue for you. I'd seen 18mm BP motors being previously used in early photos, but those were test flights. Still, a single, 24mm composite motor like the F32 wouldn't lose you much and would provide higher ignition reliability than a composite cluster.
All of that said, if it's doing what you need as is and no one minds the roof-top lightning strike hazard, there's no need to fix what's not broken.
Separate names with a comma.