Friday, November 02, 2018

All the things.....
A stream of thoughts on the task at hand...

Starting at the front of the vehicle....

In front of the vehicle, aka airspace...

To start with FAR has a standing waiver for all class 2 rockets to at least 50Kft.

Once its time to exceed that altitude I strongly suspect that waivers requested from the FAA
with only the required class 2 information will be rejected as this vehicle will have performance capabilities that exceed what was envisioned for class 2. The next launch location after FAR will depend heavily on what the FAA is going to want for a waiver.
Possible locations: Black Rock, Space port in NM, offshore either in the pacific or in the gulf.

The nose cone
Before I shut down this effort last time, I made some 4" nose cone patterns and John Newman made some molds. So he has all the tooling to make a nice light mixed glass/carbon  nose cone for this.
The nose cone has dual requirements, it needs to handle the peak impact temperature and also be radio transparent for GPS and telemetry.

The primary electronics GNC board.
NetBurner is about to release a new module based on the Microchip arm Cortex M7. I will repackage the core electronics from our new module into a GNC/autopilot.   Basic features will be:
  • Cortex M7 cpu with hardware floating point.
  • GPS, will use Ublox until we exceed the ITAR limits.After the ITAR limits become a problem I may switch to the SDR GPS I've test flown, or one of several other GPS solutions I've got working beyond ITAR on the GPS simulator.(I have a LabSat GNSS simulator)
  • Integrated IMU 
  • Barometer
  • Pyro drive electronics.
  • Telemetry Radio (Probably an xbee socket)
  • LIN communications channel.
  • CAN communications channel.
  • I2C communications channel to talk to :Digikey 223-1507-5-ND pressure sensor.
  • SDIO/SPI interface to SD card interface.
Black Box
A package containing a micro SD card in foam inside some kind of crush resistant vault.
I expect the Telemetry data to be limited and this will be the primary "Product" from the flights.

Recovery Tracking:
The simplistic solution is to add a simple beeper transmitter....
The more complex solution is to use telemetry...(Probably means putting a 2nd tiny GPS on back side of nose cone so it can receive positions on the way down).

For flights with recovery I'll use a gopro session.
For flights where recovery is not expected I'll use  a 70 cm ATV TX and a atenna with gain on the ground. To simplify things, telemetry may be on the ATV audio channel.

Recovery:This vehicle is lighter when depleted that the HPR I tested all the GPS modules on.
For FAR area flights where I'm altitude limited and well withing the atmosphere I'll include the HPR recovery system I've used reliably.  This system weighs as much as the rest of the vehicle, so its not a choice for high altitude flights.

For high altitude flights I'll probably just eject  the nose cone at altitude  with a simple kevlar strap/streamer to orient the nose cone. Basically no recovery.

TanksThe notional OTRAG vehicle uses the carbon tanks that John Newman and I developed three years ago. That tanks uses a carbon sleeve with a poly bag liner and machined aluminum oring sealed end caps.   This vehicle will use 0.035 wall 6061 T6 4" tube.
I've ordered 24 ft of this tube (arrives on 11/6)
The only thing remaining to decide here is tanks closure. I have two competing concepts....
This will be the very firts hardware I build and hydro test...
In the next month I'm going to make a few 12" long tank sections to evaluate weight and closure...

Machined domes with oring groves retained by a collar screwed or flush riveted in behind the dome.

Hydro formed domes welded into the tank ends. This will probably not work as it weakens the underlying material too much, and its very hard to heat treat thin wall tanks.

Hydro formed domes welded into a collar that slips into the tank end and is itself welded to the tank.
This would involve building a heat sink that clamps on the aluminum tube trying to minimize the heat effected zone. The length of the collar would be whatever is necessary to move the weld beyond the HAZ. On a pressurized tubular tank the lengthwise stress is half the maximum stress in the other direction.   So if you can move the HAZ to the point it only effects lengthwise stress its a win.

Mid Tank Coupler.
The mid tank coupler is going to be a some what complicated piece, it will have the roll control nozzles and solenoids built in. It will also structurally join the propellant tanks to the blow down tank on the bottom of the vehicle.   The module will have its own local electronics with rate sensor and solenoid drivers. It will communicate with the primary controller on  the single LIN backbone.
Its possible that it might also be the fill and drain interface for everything.

Main propellant Valves.
Three choices here:
OTS aluminum ball valves with external drop off actuators.
This has the possible benefit that one could mount a monprop engine and do hovering stability tests under the fork lift.

Some version of my Spectra melted string actuators.

Some version of  a pyro valve.

Pyro or spectra shut off valve.

Gimbals and TVC actuators.I have a design for a gimbal with built in fluid channels. This is an intricate little piece 3D printed and post machined. All in all its probably 40 hours of fabrication, each....
Its light, compact and easy to gimbal.

The alternative is a simple rod end bearing assembly like the LLC vehicle used with flexible hoses.  This is easy to fab, but I'm concerned about its impact on the size/wt of the gimbal actuators
and actuator batteries. Pressurized flexible hoses are stiff and require strong actuators.

For actuators I'm going to use OTS brushless servos.

Rear End electronics:
Another custom electronics board.
Two axis rate sensors.
Actuator driver logic. (Probably servo or RS-485)
Pyro drive channels (at least 2)
LIN communication channel.
System simulation will determine if LIN is fast enough to communicate from main GNC to the TVC actuators, it the TVC actuators have local rate sensing. It this is too slow then inter module comm will be upgraded to CAN.

Thrust Chamber
The final thrust chamber here will be smaller and lighter than the 3D printed aluminum chambers I was testing when I stopped rockets 3 years ago.  The will otherwise be very similar to the previous efforts.  I may build a cat packed   based mono prop chamber to allow tethered hover testing of the GNC without risking the vehcile

Rear Fill and Drain
The  full OTRAG would require rear access fill and drain.
For the single tube test article mid point or top side fill and drain might be advantagous from a getting the weight forward stability stand point.

The current plan is to have multiple batteries local to each of the the electronics bay, forward/nose cone(GNC And telemetry), mid bay (Roll control) , rear (Main Valves and TVC).
It may prove to be lighter to have a single battery and power distribution, but getting power from the front to the back in an aerodynamically clean way may be more weight than individual batteries.

Ground support equipment:The necessary transfer tanks, pumps, valves, regulators etc,,,,, to fill drain and pressurize remotely.

Tracking telemetry antenna.

Spar buoy to support offshore launches.

Simulation System...
I'd planned on using JSB sim when I had other people helping me with things...
I'm currently trying to get that environment running in a way that is useful to me...
BTW Did I ever tell you I hate XML.

That's enough for one day...

A list of random TLA's I might have used...

CAN -Controller area network,  initially developed for automotive communications
GNC - Guidance and navigation control, aka where are we pointing.
HAZ - Heat affected zone, ie a welding term
IMU - Inertial measurement unit
ITAR-International Trade in Arms Regulations aka don't sell ballistic missiles to bad guys.
LIN - Local interconnect Network. A single wire asynchronous communications protocol.
LLC- Lunar lander challenge aka the contest that started Unreasonable Rocket.
OTS- Off the Shelf,ie purchased and unmodified.
SDR- Software defined Radio
TVC- Thrust vector control.
TLA Three letter acronym.


Jimmy Beal said...

Paul, so glad you're back at it! I think you flew some MEMS IMU experiments on your HPR? So will you go with an IMU of your own "design" using one of those inexpensive MEMS chips, or will you purchase an IMU? I recall you used the Microstrain IMU for Blue Ball?

Paul Breed said...

I'm going to use one of the IMU chips, probably MPU-9000

Jimmy Beal said...

Sweet. Will you need to fuse the accelerometer, magnetometer, and gyro readings together yourself? I've tinkered with sensor fusion on the IMU chips, but never got really rock solid orientations out of it.

Lars Osborne said...


I definitely got to come out strong on the side of machined end caps with o-rings. While a welding solution without heat treatment could work, when producing the vehicles at scale adds a lot of complexity to the inspection process. Attaching the end caps with a bunch of screws can be a PITA, so I have a couple alternatives to consider:

* Consider match-drilling the holes and using shear fasteners with an interference fit, or just using shear pins.

* If you are considering riveting them in place anyway, thinking about using an epoxy - being non-cryo this could be a viable option.

Why not rely on the SDR GPS 100%? There must be some drawbacks I do not understand.

Paul Breed said...

SDR GPS:, power, mechanical reliability and weight...
The SDR GPS weighs 4X and uses 10X as much power as the entire rest of the GNC system.

Jonathan said...

Simulation: what are you estimating for your aero coefficients, and are you doing it fixed, linear, multi dimensional (mach, aoa)? Also, is your sim for 2d (downrange,alt) 3d (including pitch rotation), or 3DOF as a point mass (no attitude simulated)? I am more than willing to help with sim work, structural calcs, CAD, etc. Im learning fusion 360 (started earlier this year for a small uav project) as well.