Basejumper.com - archive

I spend some time and quit a few jumps measuring / directly comparing pilot chute pullforces to staticline pull forces applied on the bridle attachment point on the canopy.

Conclusion:
On static line jumps done with a loop of standard 80lbs break cord the forces applied to the bridle attachment point are bigger than on any type of free fall jump.

Byproduct of the tests I did with the setup: (all values above should be understood as approximations)

* Slider down, PC size appropriate for delay, ~35 kg of pull force on bridle and attachment point are unlikely to be exceeded, although on long delays with an over sized PC this can happen.
* Slider up, PC size appropriate for delay, ~35 kg of pull force on bridle and attachment point are exceeded (jump with a small wingsuit).
* Never exceeded ~75 kg of pull force on bridle and attachment point. Slider up & down, terminal with wingsuit, tracksuit, slick.

Test setup and all information I got can be found in this pdf: http://www.watchthybridle.com/...0/PC_pull_forces.pdf (just too big to upload as attachment here)

ciao!
hirschi

Interesting :)

You could use a mechanical dynamometer (the most basic spring type would probably suffice) attached in the same way as you attached the binary sensor/break chord thingy to get a max value.

Now, if you would hook that up to a linear pot and read that with an ADC a couple of times a sec, you might get a cool graph. The best thing is, it's very cheap to make and it can be made very small and snag-free. All you need is a mechanical spring, a linear pot, a micro controller with an ADC (pretty much all of 'em), some memory and a power source. Also a case for it. The spring and the case would have to be strong. Maybe a case made of a steel pipe. Let's say half a mentos tube in size for a prototype. The values would be somewhat inaccurate (the spring experiences mechanical fatigue, etc), but it would be very easy to calibrate and it wouldn't make a huge difference. The resolution would be low if the span is high (from 0-100 kg e.g.) too. You could transfer the force over several springs of course so you wouldn't need one very solid spring.

Very inspiring to say the least.

EDIT: Found this: http://en.wikipedia.org/wiki/Load_cell There seems to be some viable options out there on the market :)

EDIT2: I did some research into the world of springs, and... Well, forget everything I said about springs :)

One thing here is not very clear to me. Your 80lbs break chord lasts 76 kg which is 168lbs. How did you measure it?

Are you using a loop of break cord or a single strand?

From the photos it is hard to tell. Is that a loop on the static line setup?

Why are you using loops instead of single strands? Are you intentionally doing that to increase the required breaking force?

yes, i am using loops with suregeon knot to close it.

yes, i am intentionally using a loop since i found out that break load deviates less than with a single strand. (may be the fact that the weak point is always the knot and the single strand would have 2 of them, so when i did my pulltests the loop version deviated less than the single strand version.)

i use the loop in my static line setup aswell, and yes, i do this on purpose since i think that the bridle attachment point on a base canopy can take that load without getting any damage.

btw, a single strand of 80lbs breakcord with 2 palsteks to close them seems to (as average) hold a bit more than a loop of my other (red 50lbs) breakcord, although deviating much more.

i measured by putting a loop of 80lbs breakcord onto the same connectors (ring and rapide link) on a spring scale and slowly adding weight to it.
you can reconstruct this quit easy by just taking a normal scale putting the staticline somewhere on your ceiling and try to do pullups on it and see how much of your weight you can put on it.

this is of course static load, i think it will break a bit earlier when its dynamically pulled because if the knot is pulled tighter at very high speeds it maybe damages the breakcord more due to friction, but thats just a theorie, however, i wanted to get maximum values, and its not going to hold more than that when dynamic load is applied.

i thought about building something myself but initially i was thinking about using a strain gauge and something (eg. metal strip) that would elongate linear in that area.

but when going a bit into detailed planning i dumped that idea (adc would have to be very quick (high frequency) to get load spikes), and all in all i thought it would be still rather bulky and i would have to open my rig before every jump to turn it on.

i already found something about load cells but a quick search on the web didn't show anything that was build to be small and mobile including power source and logging, maybe i was searching for the wrong thing so i dropped the idea. if you found something, please let me know.

im pretty sure that its not going to be cheaper than my setup, but if someone likes to sponsor me to build a load cell for that, please step forward.

Let me just clarify that I am in no way an engineer, so things I say might not be correct. Also, I found out about the existence of load cells yesterday, so... Yeah... But I would love to see a load graph of the bridle attachment point.

FUTEK seems to have some mini tension load cells here: http://www.futek.com/miniaturetension

the specs look pretty good. As for price, well... I found a Finnish company selling the LCM200-FSH01934 (250 lbs version) for 656 EUR. An american company selling them for 644 USD.

That's just an example. China usually sells really good stuff (I'm serious) for low prices. Check this out: http://www.alibaba.com/products/tension_load_cell/--523.html

Some of the products there are down to like 40 USD. When it comes to electronics, it's easy to make good deals if you're prepared to buy from china or used. I know a guy who got a hold of some quite expensive FPGA chips for free. The price for single units are really high, but if a company orders a large batch for their product and some are left over, they're not worth anything to that company.

Anyway, I think it's possible to get a good load cell to a decent price.

The next problem identified is the ADC. What sample rate is necessary? I have really no idea. Maybe you or someone else does, though.

The load cell might set the limit for the ADC, but if it doesn't I wouldn't expect the load spikes to the bridle attachment point to be that spiky. Now I'm just guessing, but it's not shockwaves propagating through metal we're measuring, it's stretching nylon. I would start with an ARM Cortex M3 chip from NXP (there's lots of 'em, all with different specs), they're pretty inexpensive and commonly available. Also, since they're MCU's, they have all kinds of cool stuff/functionality to play with during prototyping. I have an LPC1343 that can handle up to 400 kilosamples/s, and there's way better stuff out there too. Again, I'm just guessing, but I wouldn't expect a shock load in a long piece of stretcy nylon to last less than 1/400000th of a second.

As for the functionality, turn it on and forget about it, for many jumps. Just have the device throw away values below a certain threshold.The passage of that threshold signals start and stop. The memory is the limit, and SD memory is pretty cheap :) Of course, if we're taking 400.000 samples/s, and for the sake of simplicity we're assuming an 8 bit sample, it fills up pretty quickly. But anyway, that's a later matter. Running it at 400.000 samples/s would probably not give the MCU time for processing etc, so an LPC1343 might need to work at a lower sample rate.

Power source... Clock batteries would probably do fine, or a small, high cap Ni-MH package. Shouldn't be a problem.

Packaging... the load cell and the power source are the biggest components. The load cell is pretty small, an inch or so in diameter.

It would be a pretty big project though, everything from manufacturing the physical parts (the case, which has to be balanced at the attachment points, etc), the electronics (though there's a lot of shops printing circuit boards on the Internet) and the software. A lot of man hours and a lot of cash (a couple of hundred EUR for parts maybe). But you'd get a pretty cool sensor package.

Some time ago i heard that some paraglider manufacturer had some sensor system for test flights, so i will ask around there.

i also found this here:
http://www.industrologic.com/llconst.htm
thats pretty much the thing i would need...