PDA

View Full Version : Fan blade failure/engine certification


Jwscud
18th Apr 2018, 10:22
Thinking about the two Southwest incidents, I am aware of the certification requirement to demonstrate containment on a static test rig. There are some impressive test videos out there showing a lot of flex and distortion in the static situation. I know there are some experienced certification people posting here, and wondering if they can answer how this test data moves to failure mode effect analysis in flight.

Is data used from the static test used to model the effects under flight loads? Is it possible to effectively model the dynamic loads and vibration/resonance induced by an unbalanced rotating fan at high dynamic pressures and sideslip angles during the initial failure? How is the engine failure case applied to the airframe as a whole? Are there defined maximum vibration levels following a failure?

I am guessing the sums are done using CFD and Finite Element Analysis but my engineering and physics knowledge isnít good enough to know quite what to ask.

lomapaseo
18th Apr 2018, 18:56
For starters

The engine generated loads against it's mounts are measured and transmitted to the engine installer on a plane.

The installer than evaluates his pylon system against the aircraft certification basis.

Experience has shown that this works well in the last 50 years.

underfire
18th Apr 2018, 20:06
The mount bolts and/or mounts themselves have a shear feature. There are a few reasons for this, mostly..

1. Catastrophic Internal Failure... If this happens, usually the engine will develop a major vibration that could literally shake the airplane to pieces. Better to have an engine drop off than to have the entire airframe come apart.

2. Externally Induced Damage... If something hits the engine so hard that something has to give, it is better to have the engine break off than tear away the structure that the engine is mounted to.

The shear feature is usually built into the actual mount bolts, or into a linking feature known as "Fuse Pins". These pins and bolts are covered by Airworthiness Directives, which require inspections, torque checks, and replacement at scheduled intervals.

For some reason, I seem to remember talk about 'explosive bolts" that could shear the engine off...but dont think that is the case, or is a feature that can be commanded by the pilot or system....

ZeBedie
18th Apr 2018, 20:35
The industry is really on top of safety now, both with aircraft reliability and human factors. But I do wonder if uncontained engine failures are a few too many these days? There have been some lucky escapes and let's hope the luck continues.

pattern_is_full
19th Apr 2018, 04:30
What strikes me is the number of recent dramatic intake cowl failures. AF A380 over Greenland + the two 737 accidents. We understand that turbo/fanjets will occasionally self-destruct and if violent enough may not be contained. But I don't remember the entire cowl back to the fan face "vaporizing" so violently in past decades.

It sure makes such failures "look" worse from a PR perspective (compared to a few jagged holes in the cowl walls), and may be contributing to the amount of shrapnel unleashed.

Is this just a result of structural or material-choice decisions to reduce weight, because the intake cowls are considered "expendable" - or even (intentional design choice) because a flat fan face will produce a bit less drag than a "cupped" inlet trapping ram air?

Or did the engineers just miss a "fan failure mode" which fails these cowls ahead of the fan more commonly or severely than expected?

Or is it an unintended consequence of the never-ceasing pursuit of less weight, that has now compromised the structural integrity of the intake cowls? Should they be coming apart so often?

Seems to me the intake cowl should be able to withstand and/or contain a blade failure (or failures) or a resulting compressor stall without blowing off completely. Both in terms of internal structure, and strength of attachment to the fan containment ring and the rest of the cowling.

And now that one may have killed someone (we don't yet know what material broke the window in the most recent SW accident) - perhaps the NTSB may start to wonder what's going on as well.

With luck this will flow into the discussion of Jwscud's original question, and not hijack the thread too far afield.

Dan Winterland
19th Apr 2018, 04:42
Certification requires a containment ring which invariably is an annulus around the fan and quite narrow. This does not address the fact that under load there is a large forward force on the blade and often at altitude, the blade will go forward as well as out. If the blade strikes the cowl ahead of the containment ring, the cowl will be damaged and the Southwest and AF events are indicative of this. About five years ago, a blade separated from a Trent on an A330 in the cruise. It came out of the front of the engine and embedded itself in the belly fairing.

To my mind, it's a clear case that the design certification is based on ground testing and not the operating environment.

tdracer
19th Apr 2018, 06:42
What strikes me is the number of recent dramatic intake cowl failures. AF A380 over Greenland + the two 737 accidents. We understand that turbo/fanjets will occasionally self-destruct and if violent enough may not be contained. But I don't remember the entire cowl back to the fan face "vaporizing" so violently in past decades.
Pattern, the A380 lost the entire fan section - not just a fan blade (still waiting to hear how or why that could have happened) so it's not really a comparable event.
However there was a PW4000/777 that lost the entire inlet a couple months ago after a fan blade release which would be somewhat comparable to the recent Southwest events. Based on the photos I've seen of the 777 event, it appears the inlet separated at the flange which could be indicative of a relatively straightforward failure of the attach bolts (that's what happened when they did the FBO test on the GE90 and the inlet came off). The good news is would mean it's an easy fix to install stronger/more attachment bolts.
As I noted on the related R&N thread, I've been involved in the investigations of several fan blade out events over the years (I was on the Boeing Propulsion Safety Review Board for over twenty years before my retirement). While we often saw 'low' energy debris damage forward and aft of the fan containment ring, we never saw 'high' energy debris damage outside the containment ring (there is a great deal of analysis that goes into how for forward (and aft) the containment ring needs to extend to contain any high energy debris). In most cases the low energy debris didn't even make it through the acoustic material.
Now there is a first time for everything, and perhaps there is something unique about the CFM56 that is causing a FBO to send high energy debris forward of the containment ring. However my gut feel is that it's not high energy debris - it's some unanticipated interaction with the combination of aero forces and the massive G-forces associated with the FBO event that's causing the inlet to fail.
Now, I don't have any direct knowledge of any of these events (and if I did, I wouldn't be permitted to discuss them) - the 2016 event occurred right before I retired. So we'll all have to wait for the NTSB to make their report. But I'm guessing there is some serious overtime being used in the 737 nacelle structures group at the moment...

Jwscud
19th Apr 2018, 09:19
Pattern is full has got more what I was leaning towards - a fairly benign engine failure has led to some significant bits departing the aircraft and damaging both wing and fuselage on the way. The ground test rig photos and videos I’ve seen don’t consist of a representative installation on a wing, just the engine itself.

Am I right in thinking the engine manufacturer just supplies the engine and specs the connections, and the fairing and inlet cowl shape are designed by the aircraft manufacturer to meet mass flow and incidence requirements, or is the whole unit designed by CFM, GE &c and just hooked up to the wing pylon?

If the former, it’s easy to see where developmental cracks could fall between two design teams. A certified installation that allows large parts to depart following a simple blade failure causing downstream structural damage shouldn’t be there. Have we been lucky or unlucky so far with debris paths? What if a large section of the cowling had impacted the tail for example? I am guessing having seen these incidents, people will be rushing to evaluate if their engine installation is vulnerable to this sort of failure.

PEI_3721
19th Apr 2018, 10:22
JW, this line of thought could include a generic situation where a cowl failure leads to engine damage and possible blade shedding, i.e. like a bird strike of greater mass than that assumed in certification design and testing.

lomapaseo
19th Apr 2018, 17:08
Historically there have been many cowl failures, including liberations from seemingly contained (by the engine case) fan blade failures.

In the earliest CF6 installations the commonality was the myriad of secondary blade tip fragments e.g. a gun-drilled fan blades. Of course this latest event does not have that extra.

Considering other inlet cowl problems like internal fires and or extensive delamination of the inner surface, the major risk is plain old in-flight aerodynamic cowl loading . This loading is typically uni-directional by design choice. The common load path to carry such loads is both the box structure that includes the outer skin (with the paint on it) and the front flange of the engine case structure. Taken together it is a very robust installation.

However if you were to remove a large circumferential section of the inner load path than the cowl will buckle and shear tear itself away from the engine attachments.

It shouldn't be too difficult for the designer to assess what's happening in this cowl design.

But this to me is not the end-all cause in the chain.

I still want to know why the outboard fan cowl has left the piece of itself standing straight up at the top of the engine?

2016parks
19th Apr 2018, 18:09
In the latest NTSB video, Sumwalt says that the fan blade fractured at two places—one at the root, and one part way out to the outer end. He said they have the inboard piece (so it must have been contained within the engine?). They do not have the outboard piece. Also, he said there is red and blue paint transfer on the wing leading edge (so it must have been one or more cowl pieces that struck there?) And they are already recovering cowl pieces that floated away, that witnesses have found on the ground.

tdracer
19th Apr 2018, 20:22
Am I right in thinking the engine manufacturer just supplies the engine and specs the connections, and the fairing and inlet cowl shape are designed by the aircraft manufacturer to meet mass flow and incidence requirements, or is the whole unit designed by CFM, GE &c and just hooked up to the wing pylon?

It depends - sometimes the engine manufacture provides a complete package - inlet, cowl, reverser, nozzles that just bolts on to the strut (the CF6-80C2 was that way, also some versions of the RB211), sometimes the airframer does everything except the actual engine, and often it's some combination of the two. Boeing has gone through phases - on the 777 the nacelles for all three engines were done almost entirely in-house, on the 787 nearly everything was outsourced. Obviously regardless of who designs or builds the nacelle bits, a great deal of coordination is necessary between the airframer and engine manufacture (and often a third party supplier).
I never worked the 737 NG, but my memory says the inlet is made by Spirit to Boeing drawings (however I wouldn't swear to it).

lomapaseo
19th Apr 2018, 21:34
jwscud

regardless of who makes the inlet, cowlings or reverser to be used on an aircraft, the certificate holder for the aircraft must show that it meets the specified regulations.

Also, in spite of the regulations covering it's manufacture and installation, the installer must ensure that any unsafe conditions that pop up in-service must be addressed under Continued Airworthiness.

In the most recent event you can expect the NTSB to revisit any loose ends in this regard and to direct the regulator to address them with the parties. If it's cut and dry (obvious) it will be covered in only the final report. If it turns out to be complex (who's in charge) there may be a public hearing with testimony from the parties.

Jwscud
20th Apr 2018, 09:28
I was wondering mainly because many different failure events have come at the interface between different design teams, due to lack of communication, miscommunication, configuration control, different design assumptions and so on.

Mainly, I’m interested in learning as much as I can about how the aircraft I fly are designed and certified because firstly I’m a bit of a geek and secondly that knowledge can help fill in the large blank areas in today’s appallingly thin FCOMs.

The Boeing AERO mag used to have a lot of really interesting stuff which is still there in the archives but has I believe been discontinued.

vapilot2004
21st Apr 2018, 20:55
It's useful to remember that turbine engine cowls are not designed to protect from high energy bits from inside the core or fan. The engine and fan cases are.

The major strength specifications are for the nose cowl, to protect against ramp rash, and the clamshell sections, to not flutter, and the heat considerations for the reverser slots. Otherwise, the cowling is wholly designed for aerodynamic, deicing, and noise suppression considerations.

tdracer
21st Apr 2018, 21:26
The major strength specifications are for the nose cowl, to protect against ramp rash, and the clamshell sections, to not flutter, and the heat considerations for the reverser slots. Otherwise, the cowling is wholly designed for aerodynamic, deicing, and noise suppression considerations.
There are some other considerations for the cowling - such has fire wall/fire block - but overall that's correct. If the nacelle prevents engine bits from departing, that's a bonus - it's not a design consideration.

vapilot2004
22nd Apr 2018, 12:20
Thank you, Tdracer.

Providing that fire wall between the engine case and aircraft structure comes with the added bonus (perhaps by design?) of containment of the extinguisher agent.

Jwscud
23rd Apr 2018, 15:18
I wouldn't expect the cowl to contain anything coming off the rotating part of the engine. It is however a large surprise to see the whole thing depart the front of the engine due to (presumably) vibration and aerodynamic forces after a foreseeable failure. Bits falling off aircraft and hitting other bits are not ideal. I know it's pretty much thin skinned aluminium and low density honeycomb material, but if it has a low ballistic coefficient and the aircraft has a high TAS, if it hits something important it could do so with a lot of kinetic energy.

Damage to either the horizontal or vertical stabilisers would certainly keep me awake at night.

lomapaseo
23rd Apr 2018, 17:36
Jwscud
I know it's pretty much thin skinned aluminium and low density honeycomb material, but if it has a low ballistic coefficient and the aircraft has a high TAS

Agree ... but only small re-ingested pieces attack it ballistically and are easily stopped by the multiple layers of skin.

The major blade piece is mostly sliding into the inlet cowl and is being attenuated by sliding friction and slicing over the circumferential feet of travel.

Jwscud
23rd Apr 2018, 23:50
I was thinking more of large departing chunks of the cowling heading towards the empennage.

tdracer
24th Apr 2018, 02:32
I was thinking more of large departing chunks of the cowling heading towards the empennage.
That's why this event (and the previous 2016 Southwest event) will probably end up with a two major corrective actions. First of course is to prevent the fan blades from failing - first step in that process is the inspection AD that is already out. I'd expect eventually there will be a design change to the blade to alleviate the fatigue cracking.
The design intent is that the inlet (and the rest of the cowling) remain intact during a fan blade out event. These events demonstrate that there is some shortcoming in the CFM56-7 inlet design since it's failed - twice. So I have to assume that some of my buddies in Renton are busy figuring out why it failed and how to fix it. I have little doubt that there will be a Boeing Alert Service Bulletin in the future to strengthen the inlet so it doesn't come apart if/when there is a future CFM56-7 fan blade out event.