pchapman

(Last photo to finish off the series -- oops, out of order for the jump sequence.)

There's a series of 11 photos that I downloaded but I can't figure out how I found them on Google / Life, since the photos are all so poorly labelled and don't show up together under a few obvious searches like 'parachute' or 'Eppridge'. ...So here they all are, slightly out of order. I'll let others look more at the cars and figure out if it was 1972, or sometime in the 60s. Either way, the aircraft, parachute, and type of jump were all quite unusual. (One more pic coming in my next post as there's a limit to attachments.)

From my post in www.dropzone.com/cgi-bin/forum/gforum.cgi?post=3399343:

No way that's a Cessna - everything is a little different when one looks closely. Someone should be able to get it if they recognize lesser known light aircraft (especially those lesser known in the USA). Not me in this case.

That's ok as long as you are well aware of the time and altitude remaining, which can be the tricky bit. Do reserves fail? Yes unfortunately they do. There is some residual risk that we aren't able to get rid of. But usually the choice of going to one's reserve is pretty simple and makes complete sense. How likely is it for the reserve to fail? That can be interpreted either as reserves failing "on their own" or in conjunction with other factors. While the issue was in the title of the thread it really hasn't been addressed much in the thread because statistics are hard to come by. There is no perfect definition of what failures to include on some accident list like this, what counts as valid vs. not really what we care about at this time. One can always chose to exclude "people not like oneself", such as students, or tandems, or reserve failures that happened to men, but at some point that's starting to ignore cases that do have at least some relevance. One of the worst reserve failures, is the reserve not clearing the freebag. I recall reading in the USPA magazine years ago of it happening. (To someone trying out for AFF instructor rating. One never knows, but without other evidence one normally assumes someone experienced will be stable for reserve deployment.) There have been other cases where it happened, but where it clearly related to the bag getting tumbled out of the reserve tray and then wrapping in its own reserve lines in some way. (E.g., due to a simultaneous main canopy opening shock or something similar. Happened to a student in Canada 5-10 years back and a tandem in the US more recently.) In earlier years we had a couple skydivers blow up their squre reserves, but that tended to involve issues of high speed and excessive weight. What are the chances of a reserve failing? (In some undefined 'serious' manner.) I tend to throw out a number like 1 in 10,000 on the guess that reserve mals are an order of magnitude less likely than main mals, for which a number like 1 in 1000 is often thrown about. That would be due to more conservative design, very careful packing, lack of wear, and some other details that are less prone to problems than with main canopies 1 in 10,000 doesn't sound great. It sure isn't "1 in a million". If it helps, the maternal death rate when giving birth is about 1 in 10,000 last I heard in the USA. Let's try to look at some numbers for reserve failures in the USA: Let's look only at reserve related fatalities, as finding out about other reserve failures is harder to do. I looked at accidents listed in the Skydiving Fatalities part of the Safety section of dz.com. For 2004 through present (June '09) there are 19 fatalities listed under "reserve problems". DZ.com's fatality list isn't perfect but is reasonable close to USPAs (which may not be perfect either). (E.g., 2004-2008 USPA shows 117 fatalities, DZ.com 120.) Looking at the 19 fatalities, taking only ones in the USA, there were 7 incidents (8 fatalities since one was a tandem pair). I then eliminated regular main-reserve entanglements, where the problem occurred because the jumper did or had to fire their reserve into a main that wasn't cleared. They're just as dead but we'll say for our purposes that it wasn't really the reserve's fault. That left 2 that could have fairly purely been a reserve failure: -- tension knots on a reserve (the reserve may have been wet, but there's little consensus on how that actually may have changed things, other than a convenient thing to point to) -- a reserve that just didn't fully inflate (Or was it tumbling related? The reports are very unclear, so putting it into this category is uncertain.) Then there were 3 more where the reserve failed because of some temporary or permanent interference with the jumper's body, so it wasn't necessarily the canopy's fault, but the whole deployment sequence: -- a student who entangled with the reserve system when unstable after cutaway (accidentally pulled cutaway handle just as the main snapped open) -- another student who entangled with the reserve system (tumbling, AAD fire) -- a tandem where the reserve had so many line twists the canopy couldn't inflate much beyond a snivel (The problem likely started with a temporary horseshoe of the reserve PC or bridle caught on the jumpers, with the bag likely spinning during that time. One could remove the tandem from the list because they have more stability problems on cutaways, but I'll leave it in since unstable reserve deployments are not unique to them.) So we have 2 fatalities (or maybe just 1) out of 124 that related to the reserve not working on its own, and 5 of reserves failing where one also includes interference with a clean deployment of the reserve system. The numbers are small but not negligible. Can we assign a "1 in X" value to these reserve failures? I'll try it but it is all pretty speculative due to not knowing how many reserve uses there are, and how rarely reserves fail (so that 1 incident more or less changes the stats radically). UPSA estimates about 2.1 to 2.2 million non-student jumps per year from '04 on. [USPA 2008 fatality report]. Let's say it was 2.1 million a year to be conservative. Main canopies malfunction say 1 in 1000 uses. That would give 2100 reserve rides a year. But maybe the main mal rate really is higher when one includes every factor preventing the main from opening (even if it isn't the canopy's fault), because there are enough dumb little things going on out there like poorly packed pilot chutes, chopping from a popped toggle, etc. So lets try 1 in 700 is a mal, which would give 3000 reserve uses per year. Since this is for non-students only, remove the students from our reserve failure list: That leaves 2 (or maybe just 1) simple reserve failure, and 3 if one includes interference with the deployment. All happened in '04 through '08 so that’s compared to 4 years of maybe 3000 reserve deployments a year, or 12,000 total (if believing the 1 in 700 mal rate). So are the numbers? 1 in 12,000 maybe, or 2 in 12,000, or 3 in 12,000 depending on the data chosen. (For non-students in the USA 2004 through 2008.) Well, those chances are higher than I'd like to see, but as I said, the data supporting that is vague, with any one incident greatly affecting the ratio. For lack of better data, I think I'll still just quote 1 in 10,000 if someone asks about a pure reserve failure, but the likelihood could easily be at least doubled when one includes interference with a clean reserve deployment.

At Niagara you're launching from a floor of cushions around the air column, right? The airflow tends to be faster in the center relative to the edges, much more so than in the "real" tunnels. That makes it harder to stay in the flow. At Montreal and all other "real" skydiving tunnels, you enter from a doorway and the air flow goes right to the walls of the chamber. In the "real" tunnels, you want to drop forward into the airflow without a big push, otherwise you tend to get too much momentum going and sail off to hit the opposite wall.

I thought it was a good read. It is a mix of history with his later personal recollections, focusing on Britain and Europe. He goes beyond the usual Garnerin, Cocking, etc, to discuss some of the stunt jumpers between the wars, the development of British WWI parachutes, etc. After being a paratrooper in WWII "Dumbo" Willans went on to be a military test jumper and civilian jumper in Britain, exploring things like the dreaded spin during long freefalls. The perspective from the early days is interesting, where for example showing up at a world championships with a blank gore parachute was a huge innovation in maneuverability.

I was thinking of an instantaneous "sharp edged" gust, which of course is an approximation. The question here is whether that's a good enough approximation. You're right that the dynamics are a factor too. There's inertia to deal with, much more so for a skydiver than a dandelion seed. If gusts hit instantly, then the angle of attack factor says that faster = less angle of attack change. If vertical winds change very very slowly, then at any speed the canopy has time to adjust back to equilibrium flight. So a question is, where are we typically in between? Look at one 'in between' situation: Let's use that example where the gust ramps from 0 vertically to 10 knots down over 20 ft. Let's pretend the canopy doesn't have time to react much in half a second. The fast canopy gets to the full gust strength in half a second, and therefore has its angle of attack lowered by "x" degrees. The slow canopy in that same half-second only reaches the point where it is at half gust strength. It is flying half as fast, getting half the gust strength, so the angle of attack change is the same, x. Neither canopy has any advantage in this case. (Naturally I'm making trigonometric simplifications, like using the small angle approximation and pretending the flight is horizontal. But that should be OK.) Then at the end of the second half-second the slow canopy arrives at the point of full gust strength. If the slow canopy (with the jumper's mass too of course) essentially had "no time to react" due to inertia, angle of attack would go down by a total of 2x -- that's the scenario I had given as an example, with the fast canopy doing much better. But allowing for some reaction from the canopy, it would pitch nose down due to natural stability to reduce the angle of attack. How fast can it compensate? If it can pitch down by less than x it will end up with a pitch down of more than x. (Worse than the fast canopy at x degrees down.) If it can pitch down by x then it will end up at x degrees down. (Same as the fast canopy at x degrees down.) If it can pitch down by more than x then it will end up with a pitch down of less than x. (Better than the fast canopy at x degrees down.) So we have a situation where the end result depends on the relative strength of the two factors: 1) the pure angle of attack change caused by a gust that depends on the canopy's speed (as I mentioned), which favours a faster canopy, and 2) the time taken for the canopy & jumper system to adjust to changing angles of attack (depending on inertia & canopy pitch stability) (as you mentioned), which favours a canopy that goes through the gusts slower. Both factors need to be mentioned. How does that sound? So which factor is more important in practice? I still kind of think that a lot of gusts we hit are in effect 'sudden', before the canopy has time to react much, so I'll still favour factor #1 and say in general that more speed is better in preventing a collapse (even if more dangerous when one happens). But I'm open to other ideas. Here's another example of how the tradeoffs can be impossible to know unless one actually knows the particular flight characteristics of a given canopy, and what the gust shapes actually look like that it has to fly through: Say that there's a maximum vertical just strength, and it is such that if the gusts are hit suddenly (reaching full strength over very little distance), then a particular slower canopy could just get to the point of having the nose fold down, losing lift & dropping the jumper until full reinflation occurs. Then if one had a bit faster canopy, it would never collapse in those conditions. And the slower canopy could collapse if the gusts are hit suddenly, but would not if they ramped up to full strength only over a larger distance, which gave the canopy (at that slow speed) time to adjust to the changing air direction. In this example, the faster canopy is by far superior. But that depends entirely on how the gust profiles match up with the speeds and angles of attack the canopies are flying at. If the numbers change, the conclusions can change about which canopy situation is better.

I'm concerned with the angle of attack change. A downgust of a certain speed creates less of an angle of attack change for a canopy moving quickly than a canopy moving slowly. That is one reason to prefer to be going fast rather than slow. (Or, similarly, a turbulence induced horizontal gust of a certain speed will cause a smaller percentage change in the speed of a faster canopy.) (But one can still argue about different angles of attack to begin with, whether one is comparing the same canopy fast vs slow or a big canopy vs a small canopy. If you speed up a given canopy you might get a smaller angle of attack change, but be flying at a smaller angle of attack to begin with, so I'm not sure what the trade off becomes. Things get messier.) Sure when going faster one will hit the turbulence quicker but does that matter on the timescales and frequencies we deal with? A fast canopy feels some more sudden hits by turbulence; a slow canopy wallows more, but neither is of the type of G loading that is going to be a problem to the pilot (as opposed to the canopy). Would I rather be fast or slow if I somehow have a collapse? Well, yes I'd rather be slow if about to hit the ground. But on a given canopy, I'd want to have some decent speed going, as too slow would leave too little energy to allow for pulling out of a dive, flaring etc. But you were likely just arguing that you don't want too fast, eg, not in a front riser dive. Agreed. A soft canopy with big nose openings might breathe more and lose some efficiency in turbulence, but it is angle of attack that is what would cause a canopy to collapse, not the canopy being "too soft". Referring to mdrejhon's post, a sudden change towards a lower angle of attack or lower line tension, as with letting a canopy pitch forward and then hauling down on front risers, yes that's more dangerous as it gets more turbulent. So I figure any transition like that should be done slower and more smoothly if there's turbulence.

I'm not Brian but my take on canopies in turbulence in general is this: -- Don't react to every little bump. Turbulence is typically bouncing the canopy around pretty fast. Reacting too much may get one get out of sync with the movement of the canopy or just be slowing the canopy in brakes if stabbing a little (more) brake every time there's a jolt. -- If the canopy dives a whole lot, then stop it from pitching very far forward with brake. -- If the canopy pitches back a whole lot, make sure there's zero brake so the canopy gets the chance to pitch forward, -- If the canopy turns off to the side a whole lot, oppose that with opposite brake, to get back straight or to the original heading for landing or whatever is appropriate. One can be forceful but unless there's a compelling reason (eg forced towards an obstacle when about to land), then don't get so deep in brake that the canopy slows too much compared to normal flight. -- If part of the canopy folds under, causing a sudden turn and dive, oppose it with opposite toggle to fly straight, and pump the collapsed side brake once or twice through a fairly long toggle stroke to help it reinflate. -- If the whole canopy stalls (whether due to a downwards leading edge 'frontal' collapse or a traditional too high angle of attack and/or too pitched up and slow situation) then use moderate to heavy brakes to avoid too much pitch forward and to help reinflation. (But one would wait until the canopy actually pitches forward or drops - - one wouldn't want to get on the brakes too fast and turn a minor stall into a more major one by adding a lot more drag when it is about to stall.) The exact amount of brakes to use will depend on what's happening -- one wants to use brakes to stop an abrupt dive and promote opening, but also not overdo their use if the canopy is slow and mushing along and still in turbulence. I could see that the brake point might be somewhere between the pure stall recovery point slightly above the stall point, and the brake set point (which varies a lot but is suitable for normal canopy inflation). How often really extreme turbulence events occur is unclear to me. Often one sees stuff happening to canopies that really don't take much input because there's nothing bad enough yet to deal with. The canopy breathes and bucks, and maybe a pressure wave goes through an end cell and it snaps almost shut and then open again in a fraction of a second. But unless the canopy really dove or pitched up, there's nothing the jumper could really have done, even if it is uncomfortable to be in the situation. Much of that sort of canopy behaviour is in itself fine until one is actually about to land, where the reduced efficiency or changed flightpath or low speed or increased descent rate can cause problems with getting a decent flare to a good landing. Associated with turbulence there's sometimes wind shear, which is technically separate but can also cause landing problems. I think all this applies well enough to canopies both big & small / old & modern / low & high wing loading. There's a lack of good turbulence related video out there, that would help people prepare for dealing with turbulence. I'd also like to hear what Brian or others think.

To clarify -- the original poster is talking about rounds that are square in shape, a.k.a. cruciform canopies. Not ram air canopies. I don't know the answer but I recall the Soviets were big on cruciform canopies, generally seeming to apply them for more applications than we saw in the west. It's an interesting possibility. Did the video seem to match up or just seem to be pasted together from different sources? Later ejection seats certainly used round rounds, although of the slotted style. (eg the K-36 seat used at the 1989 Paris airshow or whenever that was).

My impression from the flap design is that at least the Racer is one of the better rigs for avoiding line snags on flaps. (Having low aspect ratio flaps, or soft flaps without a hard lump of plastic the end.)

yeah I was actualy laughing out loud just picturing some skydiver trying to do RW with one of the falling passengers (On the less serious side of this thread:) In that aircraft breakup situation maybe I'd put on the best track I've ever done, track like hell, and go in miles away from everything else. Show that one was able to think straight until the end. The investigators would be scratching their heads, while working through the aircraft breakup sequence and thinking about ballistic coefficients. Figure that one out! But the skydivers would know. We'd all like to be remembered for something, and I'd hope some that skydivers, hoisting a beer around the campfire, would remember and admit, "That was one awesome track he had going!".

I don't know the answers but I can ask the questions: A big one is the chicken or egg question: Did the leg straps cause the body position, or did the body position change, allowing the leg straps to slip down and perhaps make it more difficult for the body position to improve? Is there a butt bungee to help keep the leg straps in place, on that rig with hip rings? (A Y-strap is another option but I can't comment on precisely how useful that would be.) Is the harness the correct size and were the leg straps tight? If a harness is short, then the hip rings are high and leg straps have to 'hang down' further, rather than just going around one's thigh. Then when one dearches they get loose. (But that is most significant if leg straps are tightened when the body is straight, such as for non-wheelies standing up.)

Hey Captain_Stan, If it helps, I should mention I recently taught a couple long seminars that I created on canopy flight & control, that make use of my aerospace engineering background -- so I'm quite aware of the lack of clarity of some terminology in skydiving, and the differences between technical aerospace meanings of terms and their common usage. So I could be pretty self-righteous and annoying too, if I set my mind to it! That sounds good to me. Nice to see something that we could take to be conciliatory. John Rich may have started a couple threads on terminology because previously you seemed imply your interpretation was the only possible one and that everyone should have anticipated your interpretation.

It would be an interesting discussion about the reliability of freepacked canopies at terminal, as is done for BASE. I'm used to the extra staging of a bag for skydives where the occasional failure of that component (eg, a bag lock) isn't fatal. Having a strap is somewhere in between having a bag and going completely bagless. The line coiling thing went away after they killed people from wrapping around main container flaps. At least that's the history I've read. (E.g., CSPA Technical Bulletin #3, 1980, reported on suspension line entanglements being discussed at the then PEIA meeting in the USA. Bulletin #4 of the same year started off with a report of the 2nd fatal accident within 10 days to a Canadian jumper due to free stowing line. So at least here in Canada that's when it became officially unacceptable.)

Well I hope you can tell us something soon. I thought a lot of people had been interested in copies of Ripcord but nobody had ever mentioned actually knowing where any existed!

Not the way it stalls and collapses! You try it sometime, I have. But anyway, from looking at other threads I will agree with what I expect to be John Rich's unstated premise: Captain_Stan likes to use words to mean things in ways slightly different than most people do, and strongly defends his interpretation to the exclusion of other possible meanings. I'll just move along now...

No problem Chris. 10lbs of lead on a 115lb guy might well be be a worthless attempt that just adds to landing risk for the student, or it might be enough to help lighter instructors in big suits work with the student. And as you said a good arch would help a lot too, but that can't be relied on for a student. I don't have the experience to know for sure. But you did the PFF -- how did it work, instructors trying to stay down with you? (The thread started about weights on light students under canopy but there has been other talk about light students in freefall too.) Accuracy competition limits are 8m/s, basically 18 mph, at canopy level. So if the air isn't too turbulent and gusty, one can be running accuracy with some fairly strong winds up high. It is an old issue, the one of light students only having the standard huge canopies to jump...

Thanks Peek for the video. Less blocky would be nice though. You'll know all this, but there aren't many stall videos out there to use for a skills course. Stalls are not exactly risk free for line twists, if one doesn't manage to exit symmetrically. (I got a video off skydivingmovies to demonstrate that.) And a video doesn't replicate how sudden and violent a stall seems if one isn't used to it. There just isn't much formal instruction on stalling a canopy, other than usually a bit during student days where they might not even go past a little rocking with arms fully down on a docile canopy. So it is easy to see why people are so wary of them. If someone has never really stalled their canopy, it gets awkward to teach if they are now on a small crossbraced canopy. They should have learned while they were still on something moderately sized.

All that looks reasonable to me Chris. Weights for a student doing AFF isn't desirable at all but I can understand it being useful and being done that way. Under canopy though, 10+ pounds of lead will be really awkward for the light student yet not make much difference in canopy loading and speed. Just not worth it for that. Better to drag out the training a bit and wait for lower winds. Solos on tandem canopies? Shouldn't be a big problem -- doesn't happen a lot and the jumpers are somewhat experienced, although I could see the occasional tandem course being held up for a day. Big canopies at low wing loadings? A 0.5 loading is getting really light but I'm used to F-111 accuracy canopies at .6 or so. It isn't that big a deal to jump when one is backing up at altitude and then coming down vertically to a landing. It may be very intimidating to people not used to it, but as with other things in skydiving, when one builds some experience, basic techniques make it quite doable. Big canopies aren't that bad in high winds. It's just not something one wants students to deal with. As for actual collapses, again I'm not familiar with 0.5 loadings. But for myself or students I've seen at at 0.6, it may be a bouncy ride, but I don't see things like collapses seeming to be imminent unless there is unusually stormy weather involved or there are a lot of upwind obstacles. (It might be worse at some desert DZ's with gusty winds, however.)

Now that I posted it I'll comment on it: Looks good. I've been advocating the "slight brake" idea for some time, to get the feel of what the canopy is doing, NOT to actually brake the canopy. That's likely one reason why people sometimes were more comfortable adding a little brake in turbulence, even if they couldn't articulate why it might be beneficial (rather than causing problems by slowing the canopy). [Edited to add:] Mind you the paragliding folks do tend to use a little more brake than that, actually applying a couple inches of brake as Brian's article alludes to. Although the canopies are quite different, they do this despite having a very low wing loading. It has pluses and minuses - the canopy is already slower and closer to the stall, but by letting brake up one can faster recover from any swinging back of the canopy induced by a gust. It also dampens the movement of canopy in general, perhaps more of a factor with highly aerodynamic paraglider wings despite their larger size and entrained air mass. While that's all something interesting to discuss, I still go along with the recommendation to not brake while flying a parachute in turbulence. (Other than to take up brake line slack.) (This rule can be relaxed somewhat for big old F-111 canopies where the tradition had been to use a little brake.) One could emphasize the idea of keeping the canopy in a "window" over one's head a little more, although Brian has written about that before. Both the window and the tension on the brakes to feel the canopy are part of the paragliding "active flying" style. One need not over react to every little bounce one gets in turbulence, but quick and precise action on the toggles can reduce the chance of the canopy getting too far away from normal flight. I've always liked Brian's idea of putting on a little extra G on approach in turbulence, as it is one way in skydiving to "keep the speed up on approach" as one does in airplanes in turbulence. It's just that with parachutes there are fewer ways to add speed safely in turbulence. Straight front risering isn't all that good for canopy stability, so a gentle carving turn works. I figure a harness turn would be the best way to do the turn. (As opposed to too much front riser or toggle input that could remove some of the margin of canopy stability that's useful to have in turbulence? However those inputs tend to be more abrupt only at the beginning of a turning maneuver.) It's a good point that in practice it might be harder to keep such a turn going long enough and from high enough to be practical with many medium sized canopies. Still, it might be useful to just do a little harness turn acceleration for the final 90 degree turn to final. Anyone practice the adding-G idea with canopies on the light side of medium wing loading?

Brian Germain just posted an article in the Safety section, that seems not to have been put in the Forums. I think it is worth getting into the forums too, to be found in searches or discussed. http://www.dropzone.com/cgi-bin/safety/detail_page.cgi?ID=718

To clarify, I tested the lines with 20 twists of 360 degrees. However when we look at lines we tend to say, "there are two twists in that line" if we see two 180 degree rotations.

There are various reasons discussed why brake lines shouldn't be left to get too twisted. Just how true or important the different factors are, is not generally clear. Opinions vary, and little actual data is available to the average jumper. Issues that are suggested include: - greater wear along edges of the line from slider grommets - greater shrinkage of Spectra line due to increased friction and thus heating by the slider grommets moving along twisted line - possible loss of strength from twisting (but the filaments are already twisted in creating a woven line) - possible increased chance of tension knots due to the tendency of a twisted line that is not under much tension to tangle up - shortening of the brake lines due to the twists (affecting openings, especially if one side is twisted more than the other) Various threads can be found debating the issues, by searching for "twisted brake lines" for example. I was feeling a little skeptical about the idea that brake lines get significantly shortened by twisting, and it was easy enough to actually collect some data. While it is only one set of simple tests, the results suggest that brake line twists DO NOT tend to shorten the brake lines all that much. Still, it might have some effect on openings, whether symmetrically applied on both brakes, or asymmetrically between a twisted and untwisted brake. It can be debated how much twisting is seen in practice and what the effect on openings would be. The tests: I twisted up a few lines that happened to be 68" to 80" in length, comparable to lower steering lines from eye to cascade for quite small canopies. (E.g., lower steering lines on a Stiletto 97 are 78" long; an FX88 ~79" long.) All were measured when under 20 lbs of load. (Under more load, such as 30 lbs, the difference between twisted and untwisted length tended to be a little less.) A 1000lb Spectra brake line lost about 1/2" in length when given 20 full twists. (0.7% of its original length) A much thicker 600 lb Dacron line lost more in the same situation, about 7/8". (1.2%) A Vectran suspension line, thinner than a typical brake line, demonstrated how a thinner line will be less affected as one might suspect. It lost only 1/4" (or 0.3%). Twenty full twists seemed like a reasonable simulation of a badly twisted brake line. One tends to notice 180 degrees of twist as as a distinct twist in a line. Those lines with 20 full twists had a half twist slightly more than once per two inches. The testing was hardly exhaustive but I did want to provide some data where I hadn't seen any before.