ExPAC ver 3.x - Active Stabilization and Steerable Recovery Test Vehicle

This thread is long overdue since the partial destruction of ExPAC ver 2. It’s thread can be found here: https://www.rocketryforum.com/showthread.php?130992

ExPAC ver 3.0 is almost complete, and I am aiming to fly it at MWP 15 on Nov.4^th. For the rebuild I have made a number of significant improvements to the vehicle and the experimental product it’s designed to carry. In the next few posts I will post all the details and pictures.

The av-bay was redesigned to be much more robust and more professional. The old av-bay was very utilitarian, components were held down with zip ties and foam tape, the wire routing was an afterthought. Also the backup charge was handled by a raven with an output channel that turned on at apogee and latched on and a switch in the off position triggered remotely. In my final flight when I should have fired the backup chute to save the tail of the rocket there was too much confusion and I failed to do so.

In the new av-bay everything was carefully designed and laid out in CAD. It has 4 10-32 stainless threaded rods with threaded couplers loctited on each end. On one end the av-bay cap is permanently attached with loctited fasteners, on the other the cap is secured by 4 button head screws. The caps are prevented from rotating by 1/8” dowels keyed into the coupler. The sled has two FG plates, one fixed and one removable, mounted to three phenolic cross bars. The altimeters are mounted on the outside of the fixed plate with holes by each screw terminal for the wires to go through. The batteries are captured between the plates with foam padding. A screw switch is mounter on each end of the center cross bar, which makes them easily accessible right beneath the vent holes. All wiring is contained between the plates, with holes in the crossbars for routing. Each end cap of the bay has a 4 pole screw terminal epoxied to the outside with color-coded 24awg high flex silicone wire through the cap and the removable end cap is connected with a 4 pole micro deans connector.



The rebuilt tail section of the rocket has now has a removable bulk plate at the bottom of the backup parachute bay. This parachute is only used in the case of a failure of the steerable recovery system to prevent damage to the rocket. Once I have built up confidence in the steerable kite I can removable bulk plate to allow me to fly longer 75mm motors. I have also decided to add a seal plate to charge section of the tail like I have in all of my other ejection spaces. The tail still has two partially cracked fillets, which will be repaired later, but are not structurally necessary.

One of the major goals of v.3 is to move closer to a production ready version of the steerable recovery system. The newly designed bay has a bunch of improvements that I feel make it much more durable and easier to assemble and fly.

First, I removed the need for a hard to build part of the tube assembly, the c shape coupler section and rails, which prevented the bay from rotating. I decided using a “remove before launch” pin or screw to keep the bay oriented was preferable. The bay orientation is only necessary so that I can locate the magnetic switch to turn the bay on. This allowed me to replaces the split coupler with a complete section of coupler for the bay to slide in. This change also makes it much less likely the burrito or bay will snag in the tube.

Next, I updated the way I attached the shroud lines to the steering bay. In the previous version the end caps of the bay each had two inline u-bolts, the upper u-bolt on each end attached to the kite with a quicklink, and the lower u-bolt on each end attached via quicklink to the y-harness that went down to the swivel and main shroud. The tension between the upper and lower u-bolts and the spacing between end caps provided for the proper positioning of the bay. This approach had a number of downsides. The force being transmitted from one ubolt to the other through the end cap, the four ubolts and four quicklinks were bulky and heavy, and at deployment the ubolts would need to move into the correct orientation under tension without snagging or damaging anything. I replaced all of that hardware with a single quicklink clamped to each end cap of the bay. All of the recovery forces are now transmitted through the quicklink and the bay is essentially clamped to the shroud line. This fixes all of the problems with the old design and results in a reduced part count. I also have decided to make the connection to the servo lines with a quick-release used for key rings which are plenty strong and are much less bulky than the three 1/8" quicklinks I was using before.

Lastly, I made some changes to the way some components fit together to protect potential weak spots and make assembly more consistent. I added shoulders to the end caps to protect the ends of the coupler tubes from impact. In a previous flight I had one of these tubes break away after it impacted another part of the rocket or hardware edge on. I have also designed tabs and grooves that will key together some of the pieces to help with alignment during assembly, but those changes will not be implemented until I begin CNC cutting the side plates on the next iteration.



I have some additional plans for a future v3.1. Including replacing the stabilization bay with a production version with small changes to allow it to have its own switch and battery (it’s currently powered by the steering bay), replacing the prototype 808 camera shroud in the nose with a production Mobius Mini shroud, and possibly moving the ejection electronics into the steering bay.

I will post another update after the upcoming flights at MWP.

Brian, if you get this perfected, it will revolutionize the hobby. My hat's off to you. I wish I could buy stock!

Glad you finally shortened your signature graphic! NOW we're making progress!

I did forget to mention a very important change in ver 3.0, though I did comment on it in the previous version thread, this version has new kites.

Ravenex said:

... Further analysis of the flight data confirmed my suspicion that the kite I have been using is grossly oversized (2.5 sq m.) for the rocket resulting in the aggressive openings that have been plaguing the project. Despite having a projected area equivalent to a 70” round chute and a it descends at only 5.5 ft/sec with a load of 19.7lb. Even when deflated the decent rate was only 19 ft/sec. With a fixed glide ratio this slow decent rate, I’d like to see 18 ft/sec, results in a forward penetration of 30% of what was expected. I ordered, and have since received, two smaller custom kites for testing a 1.4 sq m. and a 0.8 sq m.

Click to expand...

So far the only flight in which I was able to overcome the wind and land the rocket nearby was my flight on 4-17-16 in Michianna. In this flight the backup chute came out do to a deployment issue and trailed behind the kite and rocket. This had the effect of pitching the kite downward and increasing the decent rate from a typical 6-9 ft/sec to 18.3 ft/sec. The increase in decent also increased the glide.

As mentioned the old kite was a 2.5 sq. m kite with the equivalent area of a 70" round chute. Looking back at my data I figure the kite has an equivalent Cd of between 5.0 and 6.2 which based on my research is not unreasonable for a Rogallo type kite that produces lift. This suggests that to achieve 18-22 ft/s decent rate I want a kite with an area of 0.52-0.98 sq. m, or an equivalent diameter of between 32” and 44”.

The new custom kites I have purchased are 1.4 sq. m and 0.8 sq. m, and while the numbers suggest I should fly the smaller one it’s hard to believe the tiny 0.8 sq. m kite would handle a 20 lb rocket. There is also a lot of potential for error in my decent rate measurements and calculations so I have decided to fly the 1.4 sq. m first, and if the decent rates check out switch to the 0.8 sq. m.

So many shroud lines...

I use these in both a 2.6 inch and 4 inch airframes respectively:

https://www.amazon.com/dp/B000XG48EA/?tag=skimlinks_replacement-20
https://www.amazon.com/dp/B006FTUR3K/?tag=skimlinks_replacement-20

https://prismkites.com/prism-kites-company/

CORZERO said:

So many shroud lines...

I use these in both a 2.6 inch and 4 inch airframes respectively:

https://www.amazon.com/dp/B000XG48EA/?tag=skimlinks_replacement-20
https://www.amazon.com/dp/B006FTUR3K/?tag=skimlinks_replacement-20

https://prismkites.com/prism-kites-company/

Click to expand...

To be honest I've never had an issue with tangled shroud lines in any of my flights or flying the kite on the ground. Careful packing helps a lot, but I've seen these kites launch by kite surfers by wadding them into a ball and throwing them into the air.

I will admit though that my first consideration was a parafoil style kite like that because of the reduced number of lines and wide availability in a range of sizes. Many designs though have rigid frame members which would make packing into a rocket impossible, and many are quite expensive. After some searching I eventually contacted an expert in power kites and r/c paragliders, he steered me away from a parafoil towards a Ragollo wing. He was concerned that with the cross section of a rocket having more drag than the kite in the direction of motion, a parafoil might overtake the rocket too much and nose down, making for an unstable flight.

My kite also has a third "depower" line. Pulling this line pulls in the nose and partially collapses the kite. Releasing the line allows the kite to fully recover. This will allow me to make the kite act like its own drogue so I can depower it to drop through higher wind layers at altitude.

All that said these two kites look really promising and I will likely have to try them out at some point. Do you have any video posted from your flights with them? How about flight data / decent rates? How small do they pack? My 1.4 sq. m. kite fits easily into a 4” diameter 8” long bay with nomex and lines.