The idea of crash release harnesses sounds a bit like tension release bindings on snow skis. I've escaped many knee reconstructions from my skis releasing in a crash.
I had a crash on saturday, where I hit something under the water at about 25kts and went straight over the front, hooked in and still holding on. I reckon I was hurt most by still being in the hooked in. I was flying through the air at 40-50km/h and suddenly being stopped by me harness. (A seat harness at that!)
I think if the harness lines released, I'd have cleared the rig, and only been decelerated by the water - surely a much safer ride.
Do you think that the tension during a catapault is significantly greater than at any time during normal sailing?
Any kind of metal mechanical release mechanism with moving parts is very problematic.
Sand, salt and corrosion are too severe in the marine environment.
I had a set of Look release footstraps acquired in the mid 80's. They worked fine in dry land testing but after a few sailing sessions they were seized solid. I had one on the back strap of my 14" speed board and it stayed there for many years (as a fixed strap 
) and I only took it off around 2004. That strap is still in to bottom of my junk cupboard but I recon there is still a brand new one in its box in my shed. I will try to find it and post a pic for historical interest.
The plastic 'split loop' is on the right track but I have to agree that I don't think there are any plastics resilient enough and stiff enough to hold in normal use in that shape and still be able to be banged back over the boom tube after a fall. I would love to be proved wrong........
I have a few of the old harness lines with the velcro system to attach to the boom,i'm sure that they would release in a crash.
The days of really good bone crunching crashes are long gone. I remember one event at Tarifa back in the early 1990's when 1/4 of the fleet made trips to the emergency room, broken bones, dislocated shoulders, concussions.....all manner of carnage. It was brutal and the rigs were a lot less forgiving, one moment you were in control and then next you were in hospital!
I think the forces through the rig during the catapault are significantly higher than normal sailing but probably not as high as during the in-water deceleration. I distinctly remember big G forces through the harness once before hitting the water. An eye witness to my last crash reports seeing the mast break while still in the air.
Ok , it's time to get scientific . I think there are numerous ways of engineering a simple breakaway system , but the vital piece of required input is the load at which it should break / release . For this we need some data , so I am calling for volunteers .
Somebody needs to attatch a load guage to their harness lines and sail around in normal freesailing , bump and jump conditions and register the loads . Then they need to sail flat out at a sandbar and hopefully survive the impact well enough to be able to retrive the data !!!
I,ll volunteer to design the fitting once somebody does the essential testing !!!!
If you wanted to get high tech you could use a device that measures decelleration as opposed to force , like the airback triggers on cars .
Seeya Martin
How about a trigger snap-shackle type device incorporated into the harness lines near the boom.
It would have a tension spring of suitable strength to hold your weight, and a solid link from the boom to the release trigger.
When the tension on the line exceeds the spring by a certain amount (which you could adjust by changing the length of the link if you were a heavier sailor), the load will transfer to the release pin, the shackle will pop, and one end of the harness line will breakaway.
It would be easy to re-clip the harness line into the shackle and close it ready for another go once you have your breath back and your vision has cleared.
Something like this:
I think the hardest part is finding one with the correct breaking load rating. Most appear to be above 500Kg working load. Below is the lowest rated one I have found. Cheap too. 400Kg breaking might be low enough. The impulse force could get that high. Only one way to find out for sure. Let me know how it goes. 
You could modify the pin until you get the force you want. It should also make a good safely quick release.http://www.boatingandleisureworld.com.au/xcart/product.php?productid=18094
It may be even better than that. Looking at the design, I could imagine during a catapult the loading point may move towards the clip end and away from the hinge which would lower the release load a huge amount. Almost best of both worlds, a spring release and a orientation position release. Cool for $5, gotta try this , though I haven't done a catapult at speed for years (as opposed to the slow motion kind trying to get going on a sinker in gusty conditions...)
You could load up a spare one of these, they should be fairly consistent in yield strength, and set the safety release at say 50% to start with.
Snap shackles aren't actually supposed to snap.. they're called snap shackles because they 'snap' closed!
The 4-2300 kg rating is the physical outright failure of the device - usually when the hinge pin fails in shear.
My idea is to use the force of the sailor pulling on the harness line to trigger the shackle. The trigger pulls a retaining pin (sprung loaded) and releases one side of the shackle, allowing anything in the shackle to go free - i.e. one end of the harness line.
The shackle and spring assembly stays attached to the boom, and the opposite end of the harness line also stays attached.
A normal snap shackle like nobody and yoyo have posted won't work in the way that we'd need - the force required to pull the pin is out the side (that's the ring you see on the pics), not along the axis. That would work okay if it was going to be pulled by the sailor as a bail-out mechanism. Some kite safety systems use this method (at least they used to!)
The trigger type in my earlier post work by pulling the trigger in the same axis as the load. The trigger in this case is the piece inside the middle/smaller hole in the shackle, and it just needs to be pulled a little bit to release the hinged shackle arm.
It needs to be automatic because as already mentioned there's only 0.2 of a second between thinking "Oh shi..." and needing it to have let go already, which is really not enough time to find the release and pull it.
Using a tension spring and an adjustable length link would make the breakaway strain variable, so you could dial in whatever load you wanted and have it self pop at the moment of impact (or during the catapult phase, whichever you prefer).
Yes, that makes sense I like it.
Do you have a link to where they're available???
I wouldn't mind trying it.
Why not do a shear pin set up similar to the old stump jump plow.Three loops with a pin joining them together and have a spare pin attached to the harness line for when you break it.Hardwood dowel could work as the shear pin
)
Nice description SAILQUIK ,and it got me thinking of years ago before harness lines got so stiff and strong. Today when you pull in the boom to release the harness line from the hook it pretty much stays where it is in relation to the hook .
But years ago it was'nt that way as the lines were threaded thru a webbing (like CHOCO shows) which had velcro to attach to the boom ,and this rope was often very soft and thick with a bit of plastic pipe slid over the top. The harness would drop quickly under it's own weight iff the boom was pulled in suddenly.
Kato , A sheer pin is the way to do it . I think a shackle with a plastic pin would be so simple . You carry a spare one for if and when it breaks . You could get nylon pins machined up so easily. The question is what diameter and material spec do they need to be , so we need to know what load they have to break at .
Who is volunteering to be the crash test dummy ?
I prefer this type because, as you say, they fall out of the hook when pressure is released. Don't you just hate that when you think you have unhooked and do a gybe to suddenly find you are still hooked in as you flip the rig?
Once I decided to make my own harness line using downhaul thickness line. Or maybe it was a bit thinner, whatever the breaking strain was several body weights so I thought it would be fine. I didn't put any plastic tubing on which made it a pain to hook in as it was too light to hang but also I found I only got about a kilometre and a half over the Swan River chop before it broke from abrasion /point shear on the harness hook. Anyway it was strong for a while.
My idea is to have a boom connection ( velcro or hard plastic) with 3 breakable (and easily replacable) cord loops a couple of centremetres long. The harness is normal except on one end instead of the velcro boom attachment you have a G clip/hook which you then clip on to one of the breakable strings/cord.
In fact this system would allow you to maKe different cord lengths so you would be able to have an adjustable line length instead of fixed as a bonus.
I fact, I've got some fairly new harness lines which were too short so I will try this and let you know if it works.
I found this on a climbing supply site..
"The breaking strain is given in kiloNewtons which is a measurement of force. To give a rough equivalent in weight, add two zeros to the figure to give an approximation in Kilograms (eg 5.5kN is approximately 550kg).
Mammut 3mm Cord x 1 metre
Static. Breaking Strain when new: 1.8kN approx. Weight per metre: 4g. Colours vary. Priced per metre.
I think that is polyester, so 3mm supports 180kg , seems a good starting point ( 2mm only supports 85kg)
Later..
Actually I forgot one end is only supportting 1/2 the weight so maybe 2.5mm would be better.
I have attached a very rough sketch incorporating the sheer pin idea proposed earlier. Of course none of the details have been thought through so please feel free to criticise without fear of hurting my feelings.
What about an off the shelf adjustable harness line like the Prolimit that takes the rope through a cringle. (That is its means of adjustment.) Instead of tying the rope through the cringle, you could poke a bight through and put a shear pin through the bight.
For shear pins you could try different gauges of aluminium wire.
tried my take on the subject yesterday at goolwa,6mm rope x 1250long inside 8mm plastic tube clove hitch backed off with half hitch on front end and knot in other end looped around boom and tied together with 5mm cable tie.the knot in the end stops the rope from just sliding through.got hooked in and all went well for 100 mtrs.next attempt 8-10mm cable tie.
my idea is to replace the cable tie with a high tensile wire loop,thread the line through tie the knot and your set,if it were to be manufactored then you buy a set of lines with say 3 sets of wire loops of different tensile, 1 for 70-90kg,1 for80-100kg,1for 90-110kg sailors.after this little experiment not sure if i will ever buy a set of harnes lines again......2.5mts line=$5;00,1mtr plastic tube $1;60 =2harness lines clove hitch adjustable or i have also worked out how to set up with a clam cleat for adjustment.
Does a harness line have equal loading on both ends?
My guess is not too often..and this is one of the main problems with designing a breakable link on one end of a line!
Firstly you have point loading of hook on line(not necessarily centre) and then there is metal hook to plastic friction as you move(or are propelled) one way or the other.
So which end do you put the link,let alone find a consistent force??
I'd put the breakable link on the rearward harness line. Whenever I catapult I always seem to end up near the mast (and usually twisted up so it's hard to unhook), so I think there would be more force on the rearward line.
exactly nebbian ,i think it is the logical end,mind you the force to break a boom ,sail ,mast or bones is probably going to break either end.
I have been thinking about a different approach to this speedsailing safety concern:
The opening hook.
This is a hook that has a pivot at the tightest point and can be sprung to release under extreme pressure.
First problem is it would have to have moving parts - the pivot,
and a spring that will not corrode.
I would like to approach this with a good stainless pin drilled straight through the pivot point so the nose of the hook will open -
and a rubber universal tendon as a spring mounted over the top of the pivot with two plates welded to the hook either side of the pivot(one to the nose and one to the arm of the hook) to hold the tendon in and taught.
End result is that on impact hook opens(maybe 30 degrees) and releases victim!
ps.I'm not really all that concerned about safety,so if anyone wants to try it be my guest.
I have a couple of questions on this along the lines of what happens to the rig with no harness pressure.
When I have crashed I have always had the boom in my hands and been able to manoeuvre it a little to save myself and the rig has come with me at a similar speed.
If I was to lose the boom with my feet still in the straps what happens to the rig?
Is it possible to fall in front of the rig and get hammered?
I kinda like the thought that I know that I either go through the rig with some control or next to it in the water.
One of my greatest fears is having the whole kit follow me at speed after I hit the water stationary and get impacted with no control.
Sorta like bailing out of a double mast jump and having the whole rig nail you on the water, broke three ribs and got concussed, maybe that's why I never let go.
Just looking at the dark side.
..how can you crash on flat water
