SOURCE https://www.eng-tips.com/viewthread.cfm?qid=444763
QUESTION
Sparweb
There’s something I have noticed in the current crop of all-composite aircraft (Diamond DA-40+20 is an example) where the seats are not independent units, but fastened directly to the airframe, in fact they are molded into it to a large extent, though still removable with discrete fasteners. Instead of moving the seat to the controls, these aircraft move the controls to the pilot. They deal with adjustment issues by keeping the center column very close to the pilot and moving the foot pedals instead. In the grand scheme of things, I kind-of like this (having flown a DA-20 it’s easy to get used to) and some certification issues like pilot field of view may be simplified.
So now I’m wondering how this kind of seat design would be dealt with for compliance with FAR 23.562, Emergency Landing Dynamic Conditions?
Would the aircraft manufacturer have to replicate some of the airframe in the test apparatus in order to provide suitable attachments, and reactions that respond in force and deflection that corresponds to the actual thing? How much of the airframe, and how much fidelity does it need? Could the interface loads be determined in the same way as a 4-leg seat, with load transducers?
My recent reading of the AC guidance on seat certification treats the seat as a “unit”, but this assumption doesn’t seem to apply to the all-composite molded-in seat pan. All the figures in the AC show seats with legs that adjust back and forth on floor tracks. Again, more assumptions that aren’t the case for the molded-in seat. AC 23.562-1 provides in Paragraph 9 some descriptions of alternative test fixtures for seats that don’t incorporate 4 typical legs, but avoids describing what is expected. I should also note that on this size of aircraft, some of the structure supporting this seat is the spar-carrythrough, and the landing gear attachments are not far away either. There’s a lot going on in that one small area of the fuselage.
This is coming up because I see a project on the horizon where I will may to understand the certification of this kind of seat, and possibly need to figure out the interface loads at the attachment points.
Berkshire
@Sparweb
Most of thes composite aircraft coming out of europe started as motor gliders certificated and licensed under JAR 22 this was later changed to EASA
The certification basis of the Basic Model DA 40 is JAR-23 issued 11-Mar-1994, and JAR-1, at Change 5 issued 15-Jul-1996, plus elect-to-comply with NPA 23-3 ACJ Material and 23-6 Printing errors. For the DA 40 D this was upgraded to JAR-23 including Amendment 1. This is an acceptable certification basis in accordance with NZCAR ( New Zealand Civil Aviation Regulations.) Part 21B Para §21.41 and Advisory Circular 21-1A, because JAR-23 is accepted as an equivalent to FAA FAR 23, the basic standard for Normal Category Airplanes called up under Part 21 Appendix C. For the DA 40 and DA 40 F there were four Special Conditions, while for the DA 40 D six Special Conditions were complied with and there were five findings of equivalent safety. These have been reviewed and accepted.
For the most part the practice of most of the aircraft that started out under JAR22 has been to use a fixed seat pan with Adjustable rudderpedals and adjustable seat backs. this is typical of the Grob 109, the Hoffmann Dimona, the Valentin Taifun.
When working on the C70 fiberglass sailplane. The first fiberglass sailplane produced in serial production 44 years ago, we ran foul of the FAA on structural seat pans, The pan started life molded in with a fiberglass seat belt attach made from rovings glassed to the fuselage skin following european practice with a small acces door to get at the controls. First FAA question " will that seat belt anchor stay attached to the fuselage in a crash"? answer " Yes", FAA " Prove it.". At that time there was very little body of knowlege on what Fiberlass /epoxy resin could or could not do. Most of it was closely held by the bigger companies,or classified.At that point the owner decided to fit a steel cage into the forward section and mount the hardware on it , simply because the engineers at the time could handle a clip welded to a chrome moly tube, and stress it accurately which they could not do in fiberglass, next command make the access hole bigger this reduced the strength of the seat pan I found this out the hard way when I put my ass through the bottom of the pan in a hard landing on a test flight.
As far as sliding seat rails go people like Cirrus are just using part 23. They have an aluminum seat rail bolted through the cabin floor into the spar tunnel upon which the seat slides.
RB1957
You mention cockpit seats. I’d’ve thought that 23.562 was directed towards passenger seats.
What you describe sounds “odd” … the seats are molded together with the airframe but removable (attached with screws)
Do they mold the structure then separate the seats ?
Berkshire
@RB1957,
There are several different methods according to the company making the aircraft, Some single seat and tandem 2 seat companies mold the basic fuselage shell then put in a composite scaffold which holds the steering gear. Then glue in a separately molded seat pan which forms part of the structure of the aircraft. Others install re-enforced side rails as part of the structure then the seat pan is screwed to those side rails. This makes the seat pan removable which is much more convenient for maintenance. The side by side two seaters tend to use the side rail molded flanges allowing the seat pans to come out for maintenance. There is usually a lot under there. The landing gear attach brackets, steering gear, electrical, and the mainspar carry through. Depending on the type the spars may overlap with through pins on the neutral axis of the spar. or the two spars may completely overlap with spigots on the spar ends that plug into sockets on the root rib of the opposing wing, with just a safety needle to prevent them coming out. Or in the case of Grob spigots on the end of the spars with phallic looking retainer pins held with multi ball sliding collars.
@Sparweb,
Here is a cross reference chart for new and old sections of part 23. It may or may not help.
Sparweb
Thank you Berkshire for the detailed descriptions. This gives me some leads for making comparisons.
Complicating matters even further is that the FAA has gone and re-written FAR 23 in its entirety. Nobody knows how to do anything any more.
The switch to eCFR has made it harder to see the rulemaking that used to be available. In the absence of a comprehensive set of AC’s for all the new rules, your opinion is probably right - who knows how compliance is expected to be shown? There are miscellaneous comments that the “old way still works” but I haven’t seen it on a compliance plan, yet.
It took a lot of clicking to get there, so here’s the final rule you describe:
It does have the right cross-references. Also of interest is the Small Airplane Issues List (SAIL) on the RGL webpage and other stuff such as FAA Accepted Means of Compliance for Part 23 Airplanes (Amendment 23-64). So there is guidance for these aircraft, it all comes from ASTM, now (no surprise but interesting how the SAE didn’t get to keep their monopoly).
Berkshire
The nemesis of certification with the FAA is uncertainty . If they have doubts about your FEA program or strength of materials choices, their fall back position is " Prove it.", this often means tests to destruction, or at least mean failure of the component involved, this can get expensive in a hurry. It is better to under design on the initial proto type than over design. You can always add stuff to structure. the Feds are very reluctant to let you take anything off.
Other countries are sometimes a little more forgiving, that is why several American designs have originated here as prototypes, then been taken to eastern Europe for quantity production, and type certification, then licensed here under a reciprocal agreement.
Here are a few links
https://www.fire.tc.faa.gov/2007conference/files/Crashworthiness/
Also look at the Diamond Aircraft website on crash testing. You can see their crash testing for seats.
I am finding that in fact there is very little on line about this subject, and most of the paper publications on the subject that
I have date back to the 1990s.
This news article may change the certification direction of some part 23 manufacturers.
Apparently the news article from Avweb contained some mis-information here is their correction.
Sparweb
What a moving target “light aircraft” is becoming!
Thanks for the scuttlebutt, BE. The NIAR presentation is helpful too.
Seems corrugations are a great energy absorber in both metals and composites.
DGApilot
The new Part 23 is a real cluster. Nobody knows how it is supposed to work. They were supposed to be working on ASTM standards much like the Light Sport rule, but to date no standards have been written. Everything is supposed to be “Risk” based now, but no guidance on how to establish a risk based certification plan. I’ve asked my ACO advisor, asked at DER recurrent sessions and still blank stares. Thank goodness I’m only a Vintage DER, at least I understand CAR 3, Car 4 and Aero Bulletin 7A.
Sparweb
A little note to “close” the subject, concerning the seats I was going to look at:
In the end, the designer chose to elevate the seat on legs after all. They invited me to “advise” and suggest improvements, and were not going to call on me to do more or to take responsibility. That was a relief. My inspection revealed some serious issues integrating the seat to the floor.
The nature of the floor structure under the seat legs in their aircraft, once they do get the details sorted out, will offer some energy absorption of its own. The seat pan is also extra deep, allowing for a soft cushion and a crushable block underneath it. The combined energy absorption will come from the combined crush of the block, legs, and floor beams. I do not believe the guidance gives credit for crushing of the floor structure itself, as a contribution to the crash energy absorption.
A final “thank-you” to everyone!