ExPAC ver. 2 - Active Stabilization and Steerable Recovery Test Vehicle

The Rocketry Forum

Help Support The Rocketry Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

What colors should I paint ExPAC? (See post #22)

  • Blue & Silver

  • Orange & Green

  • Teal & Watermellon

  • Teal & Purple

  • Red & Blue

  • Red & Siver


Results are only viewable after voting.

Ravenex

Well-Known Member
TRF Sponsor
Joined
Aug 13, 2014
Messages
930
Reaction score
45
After the largely successful testing of my first steerable recovery prototype I have decided it is necessary to rebuild my test vehicle ExPAC (Experimetal Platform for Advanced Concepts). This rocket will be used to develop my project into a commercially viable recovery and control kit that builders and kit manufacturers can incorporate into their designs to add stabilization and steerable recovery functions. This thread will be used to document the vehicle build and the development of the advanced control systems.

For a little background on this project see these threads:
https://www.rocketryforum.com/showt...or-Advanced-Concepts-quot-design-build-thread
https://www.rocketryforum.com/showt...light-computer-development-C-Developer-Needed

This new rocket will be a 4 in x 8.5 ft fiberglass rocket with 1/8" fiberglass thru the wall fins and a filament wound ogive cone. It will have a 75mm motor mount with room for a 4G CTI.

The rocket will be brought down under a 2.5 sq. m Ragallo wing type kite with the ability to steer and control decent rate. There will also be a 54" Spherachute as a backup in case there is a failure of the kite.

As you can imagine the electronics for this project are a bit complex. It will have two av-bays. The first a traditional bay to hold standard altimeters for deployment and to deploy the backup chute in the case of a failure of the steerable chute. The other is a control bay that will house my flight system and the recovery steering servos. The control bay will also hold two cameras for recording the recovery, one forward facing, and one upward facing. The rocket will also contain a canard bay near the nose that will have fours servo driven canards for stabilization. Initially I will be using Jim Jarvis' electronics but will eventually drive them from my flight system. There will be a third camera above the canards to record the canards during the boost and separation.

The flight control system will include a flight computer with barometer, 3 axis accelerometer, 3 axis gyro, compass, 250mw 900Mhz telemetry, advanced GPS, and power conditioning and monitoring. I plan to supplement it further with power distribution, expandable deployment electronics, e-match resistance / current monitoring, and a single axis high G accelerometer.

This rocket is intended to last the duration of the product development as well as demonstration fights so it will be a bit over built. It will feature cnc cut phenolic bulkheads, threaded inserts for all screws, a custom Aeropack tail cone, and all stainless hardware.

In the morning I will post screen captures of the design models and pictures of the collected parts.
 
I am posting so screen captures from the Solidworks design. I am pretty happy with the overall design of the rocket.
Exterior.jpg

The canard section is particularly pretty:
Canard.PNG
Here is the internal cross section of the rocket from tail to nose: custom Aeropack tailcone, CTI 75mm 4G, lower bulkhead, 54in Spherachute back-up chute, secondary av-bay, ejection seal plate, 2.5 sqm kite, control bay, canard bay, ogive cone.
Cross Section.jpg

Electronics cross section
Electronics.jpg

Control bay (with transparent walls)
Av-Bay.PNG
 
Awesome renderings.

For the canards would it make more sense to have the fins themselves be balanced CP wise to reduce servo strain? Basically if the fin rotates around the CP (of the fin itself, not the rocket) there would be no torque on the servo, just down force.

This could be accomplished by changing the servo attachment point on the existing fins, or a new fin geometry that is symmetrical (equally triangle or such, with central servo attachment).

Of course your servo choice might well accept the limited effect of air friction induced torque.

This is just my reasoning for my active control, and maybe I'm over thinking :)
 
The pivot point of the fin it in line with the fins center of area, I figured this would be good enough. If anything I would want the pivot slightly forward so the fins would pivot to a neutral position if I servo broke.
 
The pivot point of the fin it in line with the fins center of area, I figured this would be good enough. If anything I would want the pivot slightly forward so the fins would pivot to a neutral position if I servo broke.

Awesome, the renderings look a little ahead the CP, but like you said it would bring the canard back to neutral if a servo fails, thank you for the explanation.

Again, great looking setup, I look forward to seeing it all together!
 
The design looks really good. One suggestion is to avoid making the canards too small. You want to be able to make trajectory corrections at low angles of attack to maximize authority and minimize drag and avoid stalling.

Bob
 
The design looks really good. One suggestion is to avoid making the canards too small. You want to be able to make trajectory corrections at low angles of attack to maximize authority and minimize drag and avoid stalling.

Bob

Thank you very much. In Jim Jarvis's thread there was some worry about making the canard to large which caused me to shy away from making them as big as my gut told me to. They will be easy to swap out and I will be milling a few sets of base plates anyway so I will be able to try different sizes if needed. To start out I may increase the size a bit more.
 
After a bit more finalization of some design elements I have begun the build on this project. I purchased and cut to length all of the airframe and couplers. I also purchased the nose cone; custom Aeropack tail cone; received materials for the bulk plates, centerings rings, bolt rings, electronics bays (linen penolic of various thicknesses); all stainless steel hardware; and helical threaded inserts. I will be reusing the kite, backup chute, and most of the electronics from the previous version.

IMAG0165.jpg

I also borrowed my ex-employer's Mastercam license and started cutting pieces on the Techno CNC router at my work. I was able to cut my fins from 1/8" fiberglass and all of the 1/2" phenolic pieces for the structure and electronics bay.

IMAG0161.jpg

Lastly I began assembly of the various components. I epoxied the nose tip and motor retainer with JB Weld; installed the bolt rings,which will carry threaded inserts for assembling the airframe, with West Systems 5 min, and built the fin can with Proline 4500.

IMAG0166.jpg IMAG0167.jpg

This week I plan to cut the remaining phenolic pieces (3/8" & 3/16"). I will then finish assembly of the tail section and install the various couplers with their hardware in there respective tubes. That will complete the structural build leaving the remaining weeks to build the two av-bays, the canard assembly, and the nose camera assembly. With any luck I should be on track to for a first launch at QCRS on March 19th. Possibly with no paint.
 
Last edited:
Just an FYI - Subsonic symmetric airfoil fins (in general; there can be some rare exceptions) have a CP which is at 1/4 MAC, not center of area (This location does shift in the transonic to supersonic range). If you set the pivot near 1/2 MAC, the servo is going to get beat to death holding against the load trying to pivot the canards sideways as soon as it is not aligned with the freestream.

Gerald
 
Just an FYI - Subsonic symmetric airfoil fins (in general; there can be some rare exceptions) have a CP which is at 1/4 MAC, not center of area (This location does shift in the transonic to supersonic range). If you set the pivot near 1/2 MAC, the servo is going to get beat to death holding against the load trying to pivot the canards sideways as soon as it is not aligned with the freestream.

Gerald

I took a look at the thesis / technical document by Sampo Niskanen on the development of open rocket to better understand what you were getting at. As the MAC is parallel to the airframe I was a bit confused. My research found that the fin CP is located at the intersection of the MAC (Mean Aerodynamic Chord) and the quarter chord as shown in the graphic from his document I have attached. I have plotted this on my fin (approximating it as a true trapozoid) and will make some adjustments accordingly, thanks. He also notes that above M0.5 the fin CP moves rearward but the math is more than I want to go through. For the record though the point is only .8" ahead of the current pivot, I'm not sure how much torque the fin will generate around this radius but my servo will have 10 in-lb available.
Trapazoidal CP.PNG
 
Last edited:
So it's been nearly 3 weeks since my last update and I have gotten a ton done, unfortunately a series of events have prevented me from finding time to post an update, so this may take a few post. I am extremely pleased with the way everything is turning out so far.

Firstly, I forgot to post this photo earlier of my fin beveling setup, I machined a 12 degree surface into a block of aluminum and mounted it to my belt sander. I used a caliper to scribe a line on the fin where the bevel was going to end and sand until I reached the line. Worked wonderfully.

IMAG0176.jpg

I installed the phenolic bolt rings in the couplers for the nose and body tubes and using printed drill guides and scribed lines I drilled through the canard bay tube and into the couplers. I then installed helicoil thread inserts in the coupler bolt rings for some very durable 10-32 threads that will last a long time. Conveniently the helicoil tap drill is the same size as a close clearance screw drill so there is good alignment and almost no play in the joints.

IMAG0168.jpg IMAG0171.jpg IMAG0173.jpg

I did the same for the bulkhead and coupler for the tail section. I was then able to test fit all of the airframe components, the cad models don't really convey the size of this rocket, I was a little surprised. My 4 year old even helped file the fin slots while we fit everything together.

IMAG0174.jpg IMAG0175.jpg

Next, with the help of my father and older daughter I glued the fin can into the tail tube. I also waxed the inside and end of the lower payload tub and bolted it onto the tail coupler, which allowed me to epoxy the coupler into the tail section.

IMAG0179.jpg IMAG0178.jpg
 
I also had a chance to work on my steering bay and canard bay, which required a good bit of time on the CNC router. I made fixtures for for slitting 4" coupler and 54mm airframe as needed to make the pieces for the steering bay and deployment guides. They were made out of some scrap 2" thick PVC I had at work and made an amazing mess. I also made the 3/16" and 3/8" phenolic components for the two bays, and did a test fit of some of the parts.

IMAG0187.jpg IMAG0189.jpg IMAG0191.jpg IMAG0192.jpg IMAG0194.jpg

Then I used my dads router with my fixtures to slit the 4" coupler into the needed pieces. I was amazed how much the tubing collapsed once it had been slit.

IMAG0198.jpg IMAG0199.jpg

We then assembled the steering bay. Most of the parts were keyed together and the assembly went quickly and everything aligned well.

IMAG0202.jpg IMAG0203.jpg
 
Finally we started to assemble the canard bay. Unfortunately the 6-32 STI tap broke half way through so it is only half assembled, but it is still very rigid. I decided to go with metal case servos to increase the rigidity of the output shaft, and they fit closely into the housing which improves the rigidity further. The electronics board is the one being developed by Jim Jarvis which I will continue to use until I can make progress on the firmware for my own controller. I also made a phenolic piece that can be installed to act as a cutting guide for routing the body tube (not pictured).

IMAG0204.jpg IMAG0205.jpg IMAG0206.jpg IMAG0207.jpg
 
Well I've made a lot of progress since the last post and will do a full write up later today after I take some pictures. One of the things I did this week is work on the paint design. I have come up with a layout I really like, and the original design and color (orange and green) were inspired by the kite I will be flying it under. However, I have also tried a handful of other color options and decided to do a poll. I forgot to put an other option, if you have another suggestion just don't vote and please post it.

Blue & Silver
Blue & Silver.jpg

Orange & Green
Green & Orange.PNG

Teal & Watermelon (My wedding colors)
Pink & Teal.PNG

Teal & Purple
Purple & Teal.PNG

Red & Blue
Red & Blue.PNG

Red & Silver
Red & Silver.PNG

Thanks!
 
Last edited:
It's been about a month since I've had time to put together an update but I have been getting a lot done during that time.

I was able to finish all of the CNC router work. Shown here: the canards, av-bay sled, canard base plates, and ejection seal plate. I also cut the steering bay covers.

IMAG0223.jpgIMAG0220.jpg

I masked and pulled the fillets on the tail section

IMAG0230.jpgIMAG0229.jpg

I assembled and filleted the canards and fitted them to the canard bay for testing.

IMAG0260.jpgIMAG0236.jpg

Next, I cut holes for the body tube section for the canards and applied small 3D printed shrouds around the hole. These covered any mistakes in the cutting and made a little aerodynamic riser for the canards. After some filling and primer they look awesome.

IMAG0264.jpgIMAG0288.jpgIMAG0289.jpg
 
I also worked a lot on the steerable recover bay and airframe section. The bay has to be able to slide out of the rocket and prevented from rotating. To do this I slit a piece of coupler into a C-shaped section and epoxied on some strips that act as anti-rotation slides.

IMAG0225.jpg IMAG0226.jpg

We then had to install the coupler tightly to the inside airframe so we brushed the outside of the coupler with laminating epoxy and vacuum bagged the assembly inside and out. This pulled the coupler tight to the airframe but it did make positioning it difficult.

IMAG0228.jpg IMAG0227.jpg

My dad also started filling / sanding / priming the airframe which we assembled to drill shear pin holes and alignment marks.

IMAG0291.jpg

I finalized the paint colors we mixed some air brush paints to match.

Final.jpg IMAG0301.jpg

I have also gotten some work done on the Steering and avionics bays, pictures of the avionics bay will be posted when I get the av-bay coupler back from my dad. The the GPS, telemetry, and R/C units are housed under the smaller covers on either side allowing them to be moved as far as possible from the rest of the electronics. In the last picture you can also see the servos drums for the steering control lines.

IMAG0303.jpg IMAG0302.jpg IMAG0292.jpg
 
For the canard control are you using a PID system (and if so how will you tune the loop?), or straight gyro control from an RC system etc?
 
For the canard control are you using a PID system (and if so how will you tune the loop?), or straight gyro control from an RC system etc?

Currently I am using the canard system being developed by Jim Jarvis and Bill Premerlani so I am not versed in the internal workings of the control algorithms. The system is base on the UDB5 board sold by spark fun and a branch of the MatrixPilot firmware. I'm sure with a little research on this firmware you should be able to determine the control scheme. Eventually I will likely move to my own control system and will be able to address these questions more directly.
 
Currently I am using the canard system being developed by Jim Jarvis and Bill Premerlani so I am not versed in the internal workings of the control algorithms. The system is base on the UDB5 board sold by spark fun and a branch of the MatrixPilot firmware. I'm sure with a little research on this firmware you should be able to determine the control scheme. Eventually I will likely move to my own control system and will be able to address these questions more directly.

Currently, the control scheme is proportional only. Brian will likely be using a new version of the program, which will make it easier for him to center his servos. It has a few changes to the original, including separate initiation of roll and pitch/yaw control and some calculations that limit the maximum canard deflection and even out the deflections of the yaw and pitch canards. These changes are to reduce the chances of stalling the canards due to too high of an angle of attack. I guess it's time to let him know how it all works :blush:

Jim
 
Back
Top