Crashworthiness

HomeBuiltAirplanes.com

Help Support HomeBuiltAirplanes.com:

GESchwarz

Well-Known Member
Joined
Oct 23, 2007
Messages
1,179
Location
Ventura County, California, USofA.
5 point harness is the only way to go. My seats are similar to a typical ejection seat in that the seat back and the seat pan are all one solid unit forming an angle of 94 degrees. As the seat back tilts back, so does the seat pan tilt up, so the "dive" is not an issue.
 

bmcj

Well-Known Member
HBA Supporter
Joined
Apr 10, 2007
Messages
13,300
Location
Fresno, California
5 point harness is the only way to go. My seats are similar to a typical ejection seat in that the seat back and the seat pan are all one solid unit forming an angle of 94 degrees. As the seat back tilts back, so does the seat pan tilt up, so the "dive" is not an issue.

A rule to remember... if you install an ejection seat, make sure the seat belts are attached to the seat and not to the airframe! :roll:
 
Last edited:

RacerCFIIDave

Well-Known Member
Joined
Feb 8, 2008
Messages
412
Location
Asheville, NC
Thanks for the input Dave. Always looking for advancements in the area of safety. It seems to me that what this type of product does is increase load distribution across the bearing surfaces of the body.

Is this the stuff?: www.pegasusautoracing.com/foam-seat-inserts.asp

Indeed it is...the "Creafoam" bead seat is the preferred material...

Just watch any Indycar crash of the last decade...

Dave
 

GESchwarz

Well-Known Member
Joined
Oct 23, 2007
Messages
1,179
Location
Ventura County, California, USofA.
You are one sharp tack bmcj, it is always good to think about the details. :roll:

All kidding aside, I believe that the greatest sources of knowledge can be found in the failures of others. As an example just last Friday I was visiting a guy who is doing a mod on his Zenith for a known wing spar issue. The spar web actually leans forward a few degrees. Under high load the spar folds in the direction of the lean, causing total failure of the wing. :dead:

Well, my spar web leans back just like the Grumman folding wing. The Grumman has a compound spar in that it has two webs that resemble and inverted letter V. This configuration prevents the fold mode which the Zenith has demonstrated so well. I have always been aware that a sloped spar could behave in such a way, and the Zenith has confirmed it. Grumman encountered the issue 60+ years ago when they first designed the Wildcat. So it's funny how we have to learn these lessons all over again as time and lessons learned get burried six feet under.
 

PTAirco

Well-Known Member
Joined
Sep 20, 2003
Messages
3,519
Location
Corona CA
You are one sharp tack bmcj, it is always good to think about the details. :roll:

All kidding aside, I believe that the greatest sources of knowledge can be found in the failures of others. As an example just last Friday I was visiting a guy who is doing a mod on his Zenith for a known wing spar issue. The spar web actually leans forward a few degrees. Under high load the spar folds in the direction of the lean, causing total failure of the wing. :dead:

Well, my spar web leans back just like the Grumman folding wing. The Grumman has a compound spar in that it has two webs that resemble and inverted letter V. This configuration prevents the fold mode which the Zenith has demonstrated so well. I have always been aware that a sloped spar could behave in such a way, and the Zenith has confirmed it. Grumman encountered the issue 60+ years ago when they first designed the Wildcat. So it's funny how we have to learn these lessons all over again as time and lessons learned get burried six feet under.

Has that failure mode actually been seen or demonstrated in the Zenith wing?

I am not sure about this; spars and webs are never exactly vertically oriented in the direction of the bending load. If you place the spar perpendicular to the chord, it will be tilted backwards about 15 degrees or so at high angle of attacks and maybe a few degrees forward at low angles, for the same G-load. There may be an optimum angle, but it's always a compromise. With the wing skins and ribs stabilizing the spar, I can't quite see how the would be a 'folding' failure due to the web angle.

I agree the inverted V shape of the Grumman is probably a good feature for that particular application.
 

GESchwarz

Well-Known Member
Joined
Oct 23, 2007
Messages
1,179
Location
Ventura County, California, USofA.
The FAA has grounded that Zenith model until the spars have been retrofitted. Many have been killed as a result of this folding spar. The Zenith owner is the one who told me about the failure mode and it all makes perfect sense to me. A common spar only has any real strength when the load is applied parallel to the spar web. As the load deviates from that parallelism, as at high AoA, the integrity of the wing relies more on the surrounding structure to maintain stiffness. When the surrounding structure fails to measure up, the spar folds across the web.

Two contributing factors to the Zenith wing failure is that it has a slightly forward sweep; due to the forward slant of the spar web when dihedral is added, the forward sweep is a result. The other is that the ailerons are not balanced, so it is believed that aileron flutter may be contributing to the overstress at the wing root.

By the way, the main spar web of the Grumman folding wing is sloped back at about 26 degrees. The secondary web joins the primary on the aft side at the top cap, and it is fairly perpindicular to the chord line. It is the secondary that keeps the primary from buckling aft.
 
Last edited:

berridos

Well-Known Member
Joined
Oct 10, 2009
Messages
904
Location
madrid
Don`t know if it was already mentioned, but B Carmichael recommends an energy absorbing foam called Ensolite
 

GESchwarz

Well-Known Member
Joined
Oct 23, 2007
Messages
1,179
Location
Ventura County, California, USofA.
Limiting g's on you spine all boils down to the simple formula for acceleration: a=(v1-v0)/t. The means of deceleration is not as important as what goes into the formula. The only variable we have control over as designers is "t" time, which can be increased by increasing "d", distance. Distance can only be increased by increasing the depth of travel of the human body during the period of decelleration. That can only be accomplished by increasing the depth of the "cushion", in whatever form that takes...Any of a variety of foams, metered shock struts, collapsable structure, etc. What is key is that this cushion must collapse at as constant a rate of decelleration as possible. It also must not waste that distance on g loads that are well within the spine survivability range, which I believe is under 10g's...So any cushion that yields at all below 10g's is wasting precious "d" that cannot be used to absorb greater g's that may be on their way in the crash sequence. Wasting "d" results in higher "a" in the "d" that remains in the cushion.

The only way to maximize "d" is with a stroker seat in a cockpit designed deep enough to allow sufficient travel. There are no kit planes designed like this and that's one of the reasons I'm not building a kit plane. I plan to test various types and configurations of foam blocks and I will share my results in this forum.
 

JimCovington

Well-Known Member
Joined
Mar 30, 2009
Messages
217
Location
Burlington, VT
Actually, we can take a lot more than 10 - we can *sustain* 10 gs. Military pilots sustain 9 frequently. IIRC, the number the FAA uses for crashworthiness is 21 gs parallel to the spine. Stapp's tests showed that 35+ gs perpindicular to the spine produced loads you could walk away from, and he sustained over 45 during testing.

"There are no kit planes designed like this"

Wow, that's a strong statement. Although I can't point to one that is designed that way, are you sure there are *none*?
 

GESchwarz

Well-Known Member
Joined
Oct 23, 2007
Messages
1,179
Location
Ventura County, California, USofA.
Jim,

Most of us are quite a bit older than "military" pilots and thus our tolerance for parallel g's on our spine is considerably less. Limiting perpindicular g's is outside the scope and capability of any seat cushion.

I meant "none" in the sense that I am unlikely to find a design that I really like that also has room for a stroker seat...and that is one reason why I'm designing my own.
 

JimCovington

Well-Known Member
Joined
Mar 30, 2009
Messages
217
Location
Burlington, VT
Yeah, I'm getting that way also - plus I have previous damage. I took an impact load parallel to my spine somewhere in the neighborhood of 15 Gs several years ago and my toes still tingle. Not something I want to do again.
 

RacerCFIIDave

Well-Known Member
Joined
Feb 8, 2008
Messages
412
Location
Asheville, NC
NASA has stuff on human limits and you can search ntrs.nasa.gov.
For example
"Objective Measurement of Human Tolerance to +GZ Acceleration Stress"
http://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/19800010433_1980010433.pdf
Well...I can add this...

In the 2009 Indy 500, Tony Kanaan hit the wall at nearly 230 MPH... According to the accelerometers in the car...he pulled over 50g,if only for milliseconds...
Tony got out of the car and wobbled away!

Of course, there was regular Physical Therapy to keep him in the car for the rest of the season!

Dave
 

GESchwarz

Well-Known Member
Joined
Oct 23, 2007
Messages
1,179
Location
Ventura County, California, USofA.

GESchwarz

Well-Known Member
Joined
Oct 23, 2007
Messages
1,179
Location
Ventura County, California, USofA.
Well...I can add this...

In the 2009 Indy 500, Tony Kanaan hit the wall at nearly 230 MPH... According to the accelerometers in the car...he pulled over 50g,if only for milliseconds...
Tony got out of the car and wobbled away!

Of course, there was regular Physical Therapy to keep him in the car for the rest of the season!

Dave
I'll bet he wobbled away! I once dove onto a 1.5" thick dense foam mattress floating on the surface of the water. I don't believe that that foam deflected downward more than an inch or two.

Just out of curiosity, I did the calculation on the g's I experienced in that landing. Here's the result... If my vertical drop was a whopping 2 feet and the distance of deceleration was 1.5" on the foam, I went from a vertical velocity of 11.2 fps to 0.0 fps in .0111 seconds. This works out to 32.36 g's!!! It felt like a lot and now I know why. It felt like every organ in my torso was bruised. I wasn't too active for a few hours after that.

I've attached the spreadsheet that I put together to make such calculations. You're going to have to make heads and tails of it on your own.
 

Attachments

JimCovington

Well-Known Member
Joined
Mar 30, 2009
Messages
217
Location
Burlington, VT
Your spreadsheet can be GREATLY simplified with a few acceleration formulas:

Acceleration Equations Formulas Calculator Velocity Given Initial Constant Acceleration Time

WikiAnswers - What is the formula for calculating acceleration

I think I found an error in E20/E22 - you need to double the deceleration time. If you do that, you'll find it's 16 Gs, not 32. Not coincidentally, this is the ratio between 24" (your acceleration distance) and 1.5" (your deceleration distance.)

In other words, to find the Gs sustained in stopping from a fall:

Gforce = height of fall/distance to stop.

It's that simple.

(Your spreadsheet is still useful for the variable velocity/30 degree impact scenario.)

I doubt you stopped in 1.5" on water - in my accident, I could have sworn I stopped in 3" but photographic evidence showed it to be at least 10". It sure felt like hitting concrete, though.
 

lr27

Well-Known Member
Joined
Nov 3, 2007
Messages
3,822
Not that simple. Very unusual to decelerate at constant g's, I bet.
 

GESchwarz

Well-Known Member
Joined
Oct 23, 2007
Messages
1,179
Location
Ventura County, California, USofA.
I think you’re right Jim!

Thank you for pointing out the error in E20 and E22. This value is supose to represent the velocity during the deceleration period. My original formula used the INITIAL velocity. My error was not using the AVERAGE velocity during the deceleration period which can be obtained by averaging the velocity between V (11.2 fps) and Vo (0.0 fps), which in this case is 5.6 fps. Because the purpose of my formula is to calculate decelerations to 0.0 fps, I obtained the average velocity simply by multiplying the initial velocity by .5, assuming a constant, linear rate of deceleration.

That's pretty slick about the ratio. Thanks for pointing that out.

When I hit that pad it felt like I had landed on cement.

The links you provided are the same ones I used to make my spreadsheet, and I’ve added the freefall table which is valuable for my drop tests. I created that myself.

There is nothing I appreciate more than someone setting me straight. I just wish my two teenage daughters were of the same attitude!!!

Thanks Jim

The attached file in this post contains the correction.
 

Attachments

Last edited:
2
Top