Recover harness - Matched components or overengineer?

I'm working on a combined Level I and II project. An anti-zipper kit-bash of the Loc Fantom 438 EXL with DD.
The vendor was on site this weekend and I bought a bunch of recovery goodies. I'm good with numbers, formulas and am using OpenRocket. I've been reading as much opinion as possible in the past 2 weeks and now I have a question.

Consider the following:
A 1000# rated swivel attached to a 1000 lb test TN harness attached to 200# rated quick links attached to 60 lb rated turned eye bolt. The entire rocket is say 7 lbs. If the max g-load on the recovery system is say 8Gs, then the max force on the recovery system is around 65 lbs force.

By the example above, we could EASILY get away with smaller, lighter materials.

My question is:
Do you guys tend to do these types of analysis and choose hardware fitting to the task, or just throw the biggest stuff you can fit into the project? Are there factors I'm not recognizing?

Thanks!

I would love to see a chart with those kinds of figures. But I just go with the flow, most of the time. I use what comes with the kit. My X-Celerator was my first kit with no recovery included, so I asked Teddy (OneBadHawk) what I needed. Honestly, if your harness is long enough, you aren't going to stress your metal.

plan for the worst and hope for the best. I've heard that one should plan for a 20G load minimum.
Rex

For the shock cord, what are the knots rated at? Different knots place stress on the line. I had some 200lb line which would break at significantly less force because a double overhand knot was actually cutting the line as it tightened up under load. The easy answer is to use a long enough line to reduce the load.

Probably the only time you would do a real analysis to optimize your recovery system for minimum weight, but adequate strength would be of you are very concerned about performance. I'm a sport flyer, so I'm not that concerned about it, and I tend to pick materials based on how easy they are to work with. So some of my typical recovery set up might be a little overkill. Maybe I might use a 2" swivel instead of a 1.25" even though the larger one is overkill in terms of weight it can bear --- I'm picking it because the size is easier to handle in my hands, not because I'm expecting a large load.

You can calculate all those number, but I think ThirstyBarbarian is right. Most of us choose material that are easy to work with and have a large margin. For my rocket in the < 6 lb range, I use 1/8" tubular Kevlar. It's strong and packs in a small space. For my heavier 4" rocket, I use 9/16 tubular nylon. Calculating for a 20G load is probably a good idea.

The advice about longer shock cords reducing stress is correct for the initial event, but can cause a lot more stress on the system later in the flight, especially on DD flights. My advice is don't over do the ejection charges and keep the shock cords shorter. IMHO, "Blow it out or blow it up" is one of the stupidest ideas going.

If you fly your DD drogueless, sooner or later you'll get a ballistic recovery where the payload section gets nose down and pulls the fin can behind it. It doesn't come in as fast as it would if the two pieces were together, but it's still pretty fast. When the main deploys at the higher speeds it usually opens pretty quick. The payload will fall to the end of its shock cord while the main chute sits almost motionless with no weight on it. You get a pretty good shock to the recovery system when the payload section hits the end it's shock cord. As for the fin can, if it didn't hit and foul the main chute on its way down, it will continue falling a distance equal to both shock cords, until it hits the end of it's shock cord. That's when you get the largest load on the system.

Most of the time those flights will work out just fine. Nothing breaks, the crowd claps for a successful DD flight, the flier is congratulated, and everyone is happy. When I see a flight like that, I think of Molly Hatchet's "Flirtin' With Disaster".

You can go smaller and lighter. Quicklinks are unnecessary (use a knot), and eyebolts can be replaced with loops of kevlar in your bulkheads. Some guys think overbuilding is a badge of honor and makes them good engineers. That may be true for building bridges, but not so much for vehicles. Efficiency is the name of the game if you are are doing serious design stuff. More output for less input.

While we're on the subject of shock load and what kind of G force multiplier to rate recovery at, does anyone have any actual data? I'd really love to see acceleration plots from onboard avionics, assuming their sample rate is fast enough to catch this kind of shock.

A rule of thumb that I use is 30G shock at main deployment. So, for a 7# rocket, I want the attachment point to be rated for 210# minimum, and prefer a 50% margin on that. So, say, 300# rated eyebolt or U-bolt, harness, any swivels and quick links, that sort of thing.

Loads are low when things work well. Loads may not be low when they don't. I had a flight on an L which angled off the pad more than expected, then hit some wind shear and angled a fair bit more. It deployed at about 8kft doing a few hundred mph at apogee. It shredded the booster but everything held together. That was a 20# rocket. If I hadn't "overbuilt" to handle the sub-optimal situations, then it could have been dangerous. As it was, my only real complaint was that the onboard video couldn't take the G's and shut off leaving me nothing but corrupted video.

Design and plan for things going wrong, particularly for usual modes of things going wrong! Don't design for everything going right. The latter comes down in ballistic pieces when things go wrong.

Gerald

Bringing up some good points here. Thanks everyone.

I tend to COMPLETELY overdo recovery. I use 1/4 inch quick links, 7/16" 25' shock cords, and that is only on a 4 pound rocket.

You'll hear all sides. My 438EXL has flown many times with no recovery issues, and fairly heavy build. I used 1/4" tubular kevlar, 15' on drogue, 20' on main, Giant Leap Fireball on drogue cord, 3/16" quicklinks, and standard LOC-included non-welded eyebolts. I did upsize the main chute several times (still not big enough) to compensate for our higher elevation.
Just my $0.02, but it works.
-Ken