Crashworthiness

HomeBuiltAirplanes.com

Help Support HomeBuiltAirplanes.com:

henryk

Well-Known Member
Joined
Mar 8, 2010
Messages
5,282
Location
krakow,poland
Ever seen those water-filled barrels in front of freeway overpasses? Rupture and blow the water out of one and you've expelled more energy before you hit a non-yielding object like a Jersey barrier or the bridge abutments than you would destroying the barrel itself. Same principle although it in theory would impart more load on the fuselage at initial contact in an aircraft crash.
https://www.youtube.com/watch?v=Le-ksimMkJ4

=classic,STATIC energy absorber...

https://www.youtube.com/watch?v=bDsGUNRkTSQ

-and DYNAMIC absorber,s=0.15 m
 

SVSUSteve

Well-Known Member
Joined
Aug 20, 2007
Messages
3,894
Location
Evansville, Indiana
And if you want to read about how the design process went creating the Texas A&M Ag-1, pretty much the granddaddy of all modern agricultural aircraft, read From the Ground Up by the late, great Fred Weick (who also came up with the NACA cowling and the Ercoupe) or if you're at the Smithsonian, you could read the original interview transcripts on which the book was based.

From Aerofiles.com...

View attachment 45646

Texas

Texas A&M College Aircraft Research Center.
Texas Ag-1 [N222]

Ag-1 aka Weick Ag-1, -2, -3 1950 = Agricultural plane. 1pOlwM; 225hp Continental E-225; span: 39'0" length: 29'8" load: 1200# v: 115/100/45 range: 400; ff: 1/x/50 (p: Ted von Rosenberg). Hugh DeHaven, Fred Weick. Might be considered as the mother of all modern ag sprayers in the world. POP: 1 [N222]. There followed Ag-2 with 450hp P&W R-985 in 1956, redesigned by George Wing at Transland Company (qv), and the 1958, Wieck-designed Ag-3, which became the prototype Piper Pawnee.
My wife has ordered me a copy of From the Ground Up as a birthday present. I am really looking forward to reading it.
 

TFF

Well-Known Member
Joined
Apr 28, 2010
Messages
13,049
Location
Memphis, TN
The AG-1 was first iteration to what became the Pawnee. Fred Weick had ties to Piper. Did the Cherokee 140 with John Thorp.
 

Pops

Well-Known Member
Log Member
Joined
Jan 1, 2013
Messages
8,162
Location
USA.
The AG-1 was first iteration to what became the Pawnee. Fred Weick had ties to Piper. Did the Cherokee 140 with John Thorp.
Yes, and I think that Fred Weick and Bob Hoover were two of the nicest people in aviation.

Dan
 

mcrae0104

Well-Known Member
HBA Supporter
Log Member
Joined
Oct 27, 2009
Messages
3,321
Location
BJC
Fred Weick: [video]http://eaavideo.org/video.aspx?v=754056783001[/video]
 

GESchwarz

Well-Known Member
Joined
Oct 23, 2007
Messages
1,179
Location
Ventura County, California, USofA.
Regarding a bursting water vessel, that would be only as good as the ability to meter the escape of water from the vessel in a relatively constant rate, in order to achieve a relatively constant g-level through the duration of the stroke. Only by using a burst diaphragm and a flow control orifice could you achieve the required flow control, and lots of testing to dial in the performance characteristics of these two components.

I vote for polystyrene foam. A single block of foam is actually a bit too firm for what we are wanting to achieve. Through testing, I found that if the foam is made up in layers, each of which has holes bored through it, the assembly can crush down as the pieces break apart and fill the voids, all the while absorbing energy as the material particles shear and compress against one another. Again, only by testing can you learn first-hand how your test samples will behave.
 

Vigilant1

Well-Known Member
Lifetime Supporter
Joined
Jan 24, 2011
Messages
4,961
Location
US
I vote for polystyrene foam. A single block of foam is actually a bit too firm for what we are wanting to achieve. Through testing, I found that if the foam is made up in layers, each of which has holes bored through it, the assembly can crush down as the pieces break apart and fill the voids, all the while absorbing energy as the material particles shear and compress against one another. Again, only by testing can you learn first-hand how your test samples will behave.
Can you provide a little more info on your setup and what you learned? For example, if each layer is 1" thick foam, did you bore the holes through the sheets perpendicular to the face, then stack the sheets (so the straight sides of the cylindrical voids were all oriented up and down)? Did you make an effort to stagger the holes (so the holes in one sheet didn't line up with the ones above/below it)? Any observations on the impact of various void percentages/hole sizes? What density of foam did you use, and what type (XPS? EPS?)
Sorry for all the questions, but this would seem to have a lot of potential, and is an ideal project for some home testing. "Typical" XPS foam from Home Depot has a rating of about 25 PSI, but that just indicates the pressure that causes it to deform by 10%, so it doesn't tell us anything about its ability to smoothly and evenly compress over time and distance with a given load.
It would also be useful to find readily available foams with good aging properties and which don't support combustion. FWIW, EPS foam ("beadboard") will absorb a small amount of some liquids, and it might be possible to use borate solutions or other means to reduce the flammability of the stuff.
 

proppastie

Well-Known Member
Log Member
Joined
Feb 19, 2012
Messages
4,563
Location
NJ
R Again, only by testing can you learn first-hand how your test samples will behave.
can you describe or pictures of the testing,.....Should your results be applicable to standard body weight pilots?
 

bmcj

Well-Known Member
HBA Supporter
Joined
Apr 10, 2007
Messages
13,392
Location
Fresno, California
Regarding a bursting water vessel, that would be only as good as the ability to meter the escape of water from the vessel in a relatively constant rate, in order to achieve a relatively constant g-level through the duration of the stroke. Only by using a burst diaphragm and a flow control orifice could you achieve the required flow control, and lots of testing to dial in the performance characteristics of these two components.
I don't think you need a metering device. The timeframe of an impact is so short that the water's mass changing direction will perform its own metering function.
 

NoStealth

Well-Known Member
Joined
Feb 5, 2014
Messages
176
Location
NA
Regarding a bursting water vessel, that would be only as good as the ability to meter the escape of water from the vessel in a relatively constant rate, in order to achieve a relatively constant g-level through the duration of the stroke. Only by using a burst diaphragm and a flow control orifice could you achieve the required flow control, and lots of testing to dial in the performance characteristics of these two components.

I vote for polystyrene foam. A single block of foam is actually a bit too firm for what we are wanting to achieve. Through testing, I found that if the foam is made up in layers, each of which has holes bored through it, the assembly can crush down as the pieces break apart and fill the voids, all the while absorbing energy as the material particles shear and compress against one another. Again, only by testing can you learn first-hand how your test samples will behave.
I think GESchwarz is bang on. Tuned hydraulic flow control devices are in use - a burst disc and adjustable flow control device for occupant weight arrangement might be possible after lots of testing. I'm surprised it hasn't been done already but other designs are probably lighter.
I'm sure lots of designs and foams can work, the links I listed the other day gave some study details.
Vigilant, it is way easier and cheaper for the homebuilders to test today.
Flammability is another whole topic.
 

GESchwarz

Well-Known Member
Joined
Oct 23, 2007
Messages
1,179
Location
Ventura County, California, USofA.
I just used the common white expanded polystyrene. I used 2" thick sheets. I used about a 2 1/2" hole saw to cut the holes perpendicular through the sheets, leaving a web between the adjacent holes of, maybe an inch or so. The holes were staggered. You can do it any way that you like. The only thing that matters is the results that you get. I used a stack of steel weights as the load and I dropped them onto the foam and measured the amount of crush. All of this should be controlled so that it all hangs together so that your tests are are repeatable, so that you know that the data is reliable. I scaled everything down and extrapolated the data to figure how a full size adult would deform a full size "engineered" stack of foam blocks. So, were talking about doing a little science here.

I figured that I would want the block to begin to deform at about 9 Gs.
 

proppastie

Well-Known Member
Log Member
Joined
Feb 19, 2012
Messages
4,563
Location
NJ

Vigilant1

Well-Known Member
Lifetime Supporter
Joined
Jan 24, 2011
Messages
4,961
Location
US
I just used the common white expanded polystyrene. I used 2" thick sheets. I used about a 2 1/2" hole saw to cut the holes perpendicular through the sheets, leaving a web between the adjacent holes of, maybe an inch or so. The holes were staggered. You can do it any way that you like. The only thing that matters is the results that you get. I used a stack of steel weights as the load and I dropped them onto the foam and measured the amount of crush. All of this should be controlled so that it all hangs together so that your tests are are repeatable, so that you know that the data is reliable. I scaled everything down and extrapolated the data to figure how a full size adult would deform a full size "engineered" stack of foam blocks. So, were talking about doing a little science here.

I figured that I would want the block to begin to deform at about 9 Gs.
Thanks. My seat pan (in a Sonex) is riveted AL, the weight is supported at the top of the seat back. I'll need to figure out estimated tear-out/rivet failure forces based on the geometry, then remove enough rivets so that the seat pan "lets go" and starts its downward stroke at some threshold force (low enough so back injury is improbable, but high enough so the seat won't tear out if I step into the airplane with too much gusto (i.e. it will be the whole body weight over a few sq inches, so a more concentrated load than when seated). Next--experiment with foam to limit (minimize?) peak experienced vertical acceleration over the stroking distance up to some defined maximum sink rate. Hmmm-- two variables at work= two solutions: Option A: go easy on the back with a "soft" deflection rate (minimizing risk of spinal injury in moderate and more likely "pancakes") or go with Option B: a "firmer" rate and increase chances of survival at higher sink rates, but increase chances of spinal damage in moderate crashes (compared to Option A). I'll probably go with Option A. I have more reading to do . . .
 
Last edited:

GESchwarz

Well-Known Member
Joined
Oct 23, 2007
Messages
1,179
Location
Ventura County, California, USofA.
Unfortunately, most kit designs have no travel space between the seat pan and the floor, or the wing spar.

It is the controlled deceleration within that distance that will save your spine. Actually you only get about two thirds of the distance for travel...When it crushes, that foam has no where to go, except into the voids I spoke of, within that space. The perforated foam block will crush down to about one third of it's original height. Various configuration may certainly yield better results, or rather, a greater percentage of stroke distance to the total space available.
 
Last edited:

mcrae0104

Well-Known Member
HBA Supporter
Log Member
Joined
Oct 27, 2009
Messages
3,321
Location
BJC
Clearly, the material we need is reverse Oobleck. (Oobleck, for the uninitiated, is a mixture of corn starch and water which flows like a liquid but is firm when pressure is applied).

[video]https://youtu.be/cn0mIj0_4ws?t=64[/video]
 

SVSUSteve

Well-Known Member
Joined
Aug 20, 2007
Messages
3,894
Location
Evansville, Indiana
GESchwarz said:
I vote for polystyrene foam. A single block of foam is actually a bit too firm for what we are wanting to achieve. Through testing, I found that if the foam is made up in layers, each of which has holes bored through it, the assembly can crush down as the pieces break apart and fill the voids, all the while absorbing energy as the material particles shear and compress against one another. Again, only by testing can you learn first-hand how your test samples will behave.
If given my choice, I would go for something else. Rohacell isn't prohibitively expensive (especially for the small amount fitting under a seat) and has a well documented testing history for this exact application. The tendency of homebuilders to go for whatever is at hand or whatever Aircraft Spruce stocks is a hard one to overcome. Styrofoam has one or, depending upon its specific makeup, two less than desirable traits. The first is that it can produce some nasty smoke even without direct flame impingement (you have the issue of it being under the seat and in a lot of aircraft there is wiring passing just below where it would be located so you have another potential fire hazard). The second is that some forms of it have a pronounced rebound tendency because of the air in the cells being surrounded by a material that doesn't resist recoil. Step on a piece of common grade Styrofoam and watch how it reacts when you reduce the load. Plus, compared to the other options out there, polystyrene isn't going to attenuate the loads as well.

If you want something that is nearly ideal, my suggestion would be one of the expanded metal foams that were developed for automotive bumpers. I've got a piece of it around here somewhere that was provided as a test sample and it's not appreciably heavier than polystyrene. It also will crush down to a much greater degree without any substantial rebound. Depending upon the makeup (which can be specifically ordered to deliver a target response), it could approach or exceed 75% of the original height before bottoming out. It would be my first choice with Rohacell as a second option.

The other option would be to have something that is going to breakup and can be forced out of the way as it compresses beyond its maximal attenuation. That would allow- theoretically- for a greater attenuation in a given stroke distance because it would maximize the functional stroke distance.

I figured that I would want the block to begin to deform at about 9 Gs.
Remember that for spinal loads, it's not really the G load you have to design around. You have to take into account the load at the lumbar spine (usually measured in kilonewtons; a load of 3.5 kN carries a roughly 50% chance of spinal injury for a middle aged adult (Yoganandan et al, 2013)) rather than thinking of it in terms of the floor acceleration. Another example: a load applied across a 300 mm square area of the buttocks and the back of the thighs would produce a wedge fracture at L1 (first lumbar vertebra) in the range
of 8.92-10.6m/s (or 1750-2475J impact energy) according to Zhang and Zhao.

By the time you stroke the seat at 9 G, in concert with the flexion and/or rotational loads in most crashes, you might- dependent upon other factors related to the seat, underlying structure and the crash- well have exceeded the capacity for the vertebral bodies or the associated ligaments to withstand the load. It's one reason why it's so important to think of the cockpit as a system rather than viewing the components in isolation. The example I normally use to explain it to aviation folks is that you can't design the wings and tail independently because a change in one will dictate a change in another.

This is one of those things that you can't just pick a G load and call it a day. The calculations can make blood shoot out of your nose if you're not into math but it's a necessary evil. I mean....you CAN say "Stroke at X G...moving on" but it doesn't mean it's ideal especially if you're trying to get a crashworthy design as both you and I are doing. Hopefully you don't take any of this as anything but constructive criticism because I am exceptionally proud of the amount of effort you're putting into your project.
 
Top