Steve Innova

I always brush epoxy on the part before laying down the glass. I think you get better wetting out when the epoxy soaks up through the glass. And you ensure that the glass-foam bond is fully wetted out. If you're working with UNI, that actually makes it easier to slide the UNI around. (The best way to apply UNI to large surfaces, IMHO, is to cut the required length, then roll it onto a smaller roll for easier handling, then roll it out onto the surface.)

If you do end up with extra foam, make sure to put it back into the box and keep it covered from light (any source). Over time, divinicell foam exposed to light (I assume UV) degrades. I've had to trim the edges of foam sheets, b/c they discolor and get crumbly.

I'm sure this will piss some people off, but I also did some basic calculations on the standard Cozy MkIV shoulder harness. Rather than try to guess at the sheer strenth of the shoulder harness attach point layups, I calculated the load capacity of the entire shoulder brace (if that fails, you don't care of the hardpoints shear out, right?) Here are some basic (and generous) assumptions: Rectangular cross section: 3.875" x 3.875" (it's actually a triangular cross-section, so actually less of the above, but I was generous and made it rectangular since my software doesn't calcualate those) 2 plies BID inside, 1 ply BID interior (Being generous, I used .013" / ply x3) 42" long Distributed load over 31.2", 5.4" from either end. Non buckling These assumptions result in a calculated ultimate load capacity of 1422 lbs. (I speculate, using correct triangular cross section, you end up with a lower #, but I've also ignored the role of the lower portion of the seatback below the shoulder brace, as well as the map pocket. (I don't think the map pocket's role is as simple as cutting the beam into 2 separate sections (which would significantly improve it's load capacity), but I don't have the ability to calculate that.) Granted, these calcs are very limitted, since I don't have access to (or know how to use) FEA software. But if anyone can show me tests (or even any design calculations) that suggest that the standard Cozy MkIV meets these requirements, I'll buy you a drink. Make that 10. The only way to know ultimate load for sure is to test it.

Right now I'm only redesigning the shoulder harness attach points, but may revise the attachments for the lap belts as well. To meet FAR 23 standards, the shoulder harness should be designed to meet the following static load requirements: FAR 23.785a specifies a design occupant weight of 215 lbs, with a 1.3x safety factor. FAR 23.561 specifies a static load (in the foward direction) of 9 Gs. FAR 27.785 (no mention of load distribution between shoulder/lap belts in FAR 23) recommends using 40% load for the shoulder harness load. I've heard recommendations to use a 2.0 safety factor for homebuilt composite construction. (I don't know if that replaces the FAR 23.785a 1.3x safety factor, or is in addition to it. I'll assume for the moment, that it's cumulative.) That gives me static test load of 215lbs x 1.3 safety factor x 9 Gs x .4 x2.0 = 2012.4 lbs. It's also worth noting, that my seat-belts have static load design strength of 1500 lbs. For the four belts, thats 6000 lbs. I happen to think that meeting these FAR specs should allow for some reasonable survivability. Although we don't have to follow these FARs in experimentals, I don't see how it can be a bad idea to match them. Lancair and some other kit companys have made a point of designing to FAR standards.

Actually, I'm in possession of a plans-built Cozy MkIV fuselage, in addition to my modified design. I've posted it for sale for some time, but now I'm leaning towards using it for destructive testing. If I could find local assistance, and instrumentation, it might even be possible to do dynamic testing.

I think the difference is that you didn't build the parachute yourself. Or maybe you did... in which case I would test it first.

As Marc mentioned, I'm refering to the fatal 1995 crash of Dr. Larson's Cozy MkIV. My info is coming from a message posted to the Cozy Builders email list by "TEllis (titled: Re: COZY: Seatbelt prices too high), describing the complete failure of the forward seat-belt and shoulder harness attach points in this crash. "Both the pilot and front seat passenger were thrown thru the instrument panel and died. There were two teenage boys (as I recall) in the back seats and they survived with little injury. The fact is that there was no problem with the front seat belts themselves. The sholder attachments pulled out, both outside belt attachments pulled out and the whole center heat duct seat belt attachment unit separated from the fuselage. From the front of the strakes forward, the whole fuselage disintegrated so it is extremely improbable that the front seat passengers could have had a chance in any event, but the rear seat belts and their attachments held and saved the rear passengers. Which is a miracle in itself." Of course, this accident may have been entirely unsurvivable, but it is worth noting that the rear seat belts held, and both the rear passengers survived. One thing that's certain though -- survivable crashes become very unsurvivable if the human body is not adequately restrained inside the vehicle. I recall providing first aid to a guy ejected from his vehicle in frontal crash w/roll-over. The car was destroyed, but everyone else survived with minor injuries, except for that guy who hadn't worn his seat belt.

The static tests detailed in the FARs.

This arguement is common and tired in experimental aviation. It can be summarized as: "If you don't make any mistakes, you won't crash, and you won't need any safety equipment. Therefore the only worthwhile investment in safety is in improvements that reduce the likelyhood of a crash." It's not surprising to hear this arguement, as pilots are some of the more self confident (arrogant?) individuals, most of whome, upon reading an accident report, are quick to point out how they would never have let those circumstances happen to them in the first place, by virtue of their superior decision-making and pilotting skills. Sure, whatever. If that's your position, I'll wish you luck, not that you'll ever need it, of course... I, for one, happen to believe that most light aircraft accidents (which happen, even to the best pilots, BTW), are in fact survivable. A 2 lb investment in an improved shoulder harness attachment seems like a small price to pay.

Perhaps a more gramatically correct statement would have been: In a Cozy MkIV crash, the shoulder harnesses failed, contributing to the deaths of the pilot and front seat passenger. The two rear seat passengers survived. The crash wasn't caused by the shoulder harness, it was caused by striking a wire on approach.

The shoulder harness attach points are structually deficient, IMHO, and directly contributed to the loss of two lives in a crash. I intend to test these to failure, and demonstrate their deficient design.

This arguement comes up every few weeks. There seems to be line of reasoning in this hobby, that states to the effect of: No need to verify information with testing, just follow the plans. As long as you don't crash, the safety equipment isn't neccessary. Ok.

Re-do. I'm not convinced the designed seatbelt attach points meet aviation standards for safety anway -- certainly wouldn't want to compromise there.

The strake is an airfoil, just not an optimal shape. At positive angles of attack it produces lift.

The "HSS" stands for "high strength steel." Obviously $8 doesn't buy very high strength...

But will the retractable oleo nose strut beat the Infinity Sport plane? How's that for a question!

True. The CF UNI (Hexcel 716) on AC Spruce is still 4.7 oz, only a 2.3 oz weight saving over 7 oz fiberglass UNI (you'll use the same weight epoxy), plus you have to add a ply of fiberglass (BID?) to cover the AL parts. Not worth it, in my opinion. I've used 5.8 oz Hexcel 282 (Carbon Fiber BID) and 4.7 oz Hexcel 716 (Carbon UNI) and have had no problems wetting out the plys. While it's true that CF isn't translucent like fiberglass when wet, with these styles of CF, it's pretty apparent when it's wetted out. You can see the sheen change as the epoxy soaks through. One technique is to apply the epoxy to the object surface, then lay down the carbon fiber ply, and use a hairdryer & squeegee until the epoxy soaks up through the ply. Then you're sure the CF is thoroughly wetted out. I generally brush on extra epoxy when working with CF, and then use a vacuum bag to remove any excess. I can't speak for any other styles of CF, other than the long tape, which I don't recommend -- very hard to wett out.

What about GPS antenna reception? That'll likely be in the nose section of the aircraft. If the roll-bar is carbon fiber, would that blank the antenna in some orientations?

I'm using some carbon fiber strategically, to reinforce small parts/locations, such as the engine mounts, seat-harness anchors, canopy surround, gull-wing door frame, etc... How will these small carbon fiber fuselage parts effect the reception of hte antenna located in the wing, winglets, and canard?

One way to track epoxy quality: Take a 2x4 and drill 1/2" deep, .75" holes all down the board, spaced out every 2". Then every time you do a layup, poor the extra epoxy into the hole and record the date on the board with a sharpy. Now you have record of every batch of epoxy. Probably overboard (i don't do this), but a big confidence builder.

To this end, you would probably be better off buying a functional Veri-EZ or Long-EZ, and then build modifications, testing each modification incrementally until you had explored the performance parameters of your new concept. Then you could execute the concept on a new aircraft. The risk you run with building a new aircraft implimenting new aerodynamic concepts is that you spend 5 years and $70k, and it doesn't work or it sucks.

Yeah not stirring your epoxy is a pretty good way to waste the $130 / gallon MGS. Cheaper to use elmer's glue if you don't want to follow mfg instructions -- there for a reason.

Another thing you should banish from your shop is WD-40. Or pretty much any aerosol lubricant or cleaning agent. It'll drift and contaminate any surface in your shop. If you have a compressor, that often will expell oil into the air too. I keep all those things outside of the shop.

This picture should also help explain what I'm doing. The left "A" frame. The cross-beam at the longerons will come later.

Here are some diagrams, may help explain the above picture better.