Southwest flight 1380 makes emergency landing in Philadelphia after engine failure

[bshole]

I actually know a little something about turbine failures (in an industrial environment) and incorrectly extrapolated to an airplane gas turbine engine. Why are you being such a complete dick?

True Story: I worked with an engineer for years who had previously worked on the software for the flushing of toilets on jets. I told him that was the explanation for his shitty software. Get it....

TBH, I loved the dude. He recently retired and moved to Colorado. He is all pissed off because they opened a frack site about a mile from his house. I found it hilarious that he spent a good portion of his career developing software for fracking equipment (with me) only to end up next to a frack site. Makes me think there is a God.

PS. His son is now a principal software engineer at MicroSoft making over $200k/yr (he is only 30). Kid was writing software programs and hacking the Windows registry at the age of 12.

 

[Zorba]

FYI, here is a video to a fan blade out test that is required for engine certification. The blade is has an explosive charge placed at the root of the blade and it is released at the rotor redline speed. It is released in the most critical location, generally under the oil tank. You have to wait (15 second IIRC) before you can pull the fire handles that shut off fuel and hydraulic supply to the engine. You must demonstrate that no components exit the engine radially, that the engine mounts can react the load, that there is no post-event fire and that the oil tank remains intact.

Depending on the size of the engine, each fan blade has 60,000-80,000 pounds of force pulling it radially away from the hub. That means that each fan blade is like hanging a fully loaded semi off the root off the blade. The forces involved are insane. The fireball you can see in the video comes from the compressor stalling, which allows high pressure gas in the combustor to flow forward through the engine. I.e. when the engine stalls, combustion isn't contained in the "burner."

*Note: This is a different engine, I couldn't find a video of the CFM56-7B on Youtube, but every engine goes through this testing.

Again in this accident, it appears the fan blade was contained, but the nose cowling (non-engine part) failed and that debris is what hit the fuselage. It shouldn't happen, and hopefully the investigation can tell us why it did.

 

[feralkid]

I actually know a little something about turbine failures (in an industrial environment) and incorrectly extrapolated to an airplane gas turbine engine. Why are you being such a complete dick?

My bad...the next time you make an erroneous assumption based on zero evidence, I promised I'll just let it slide.

 

[Darwin333]

But the 737 can also fly with no hydraulics, there are manual controls for the critical flight controls, and backups to unlock the gear and power the brakes.

Now that is something that I didn't know, I am definitely going to read up on that. Thanks for the info!

Further, it is required for certification to do an analysis that shows no high energy engine component could take out any two redundant systems, i.e. two hydraulic systems, flight control cables, or electric buses. You do this analysis assuming everything within a cone is completely removed from the aircraft and you look at that cone in every angle around the engine. This is one reason the engines are in front of the wings and not under them.

That is true but it is virtually impossible to analyze everything that could possibly happen in the real world. Take United Flight 232 for instance, it had 3 independent hydraulic systems that were all taken out by the catastrophic failure of a single engine. The fact that anyone survived that flight, much less almost 2/3 of the souls on board, is a serious testament to those pilots skill, nerves of steel and huge gigantic balls. I assume that they just tuck their feet under the seat and alternate weight on each one of their balls to work the pedals for the rudder and brakes otherwise I don't know how they fit them.

 

[Darwin333]

Excellent points! I just assumed it was thrown blade. Thanks for clarifying what actually happened.

If it was the blade I'd imagine it would make it completely through the aircraft and still travel a considerable distance. Those things are spinning at crazy speeds and have a tremendous amount of energy. The skin of an aircraft isn't that strong.

 

[BUTCH1]

Hmm, seems Southwest wanted "more time" to inspect these engines after a similar incident last year, (no injuries). CFM, (engine manufacturer), recommended blade inspections finished within 12 months, Southwest and others wanted more time to complete them. The labor costs are nil for this but the blades themselves cost $50K for new, $30K refurbished, each CFM56 engine has 24 blades. One wonders if these airlines are afraid of finding lots of iffy blades and were trying to spread the cost over a longer operating period. https://www.apnews.com/7b50c05eac7440f1945dfc1a5bf48fbe

 

[Indus]

Hmm, seems Southwest wanted "more time" to inspect these engines after a similar incident last year, (no injuries). CFM, (engine manufacturer), recommended blade inspections finished within 12 months, Southwest and others wanted more time to complete them. The labor costs are nil for this but the blades themselves cost $50K for new, $30K refurbished, each CFM56 engine has 24 blades. One wonders if these airlines are afraid of finding lots of iffy blades and were trying to spread the cost over a longer operating period. https://www.apnews.com/7b50c05eac7440f1945dfc1a5bf48fbe

Well a new engine would cost $10 million and considering the damage to the plane, a plane would cost a lot more but they're just trying to maximize profits at the cost of accidents which they hope won't kill anyone.

 

[BUTCH1]

Well a new engine would cost $10 million and considering the damage to the plane, a plane would cost a lot more but they're just trying to maximize profits at the cost of accidents which they hope won't kill anyone.

Sadly, someone already had died, I'd bet those inspections get finished very quickly now.

 

[Puffnstuff]

Might be time for the fed to lean on CFM for internal engine data including their validation testing.

 

[zerocool84]

I flew twice these past two days on Southwest. I like living on the wild side.

 

[Zorba]

Hmm, seems Southwest wanted "more time" to inspect these engines after a similar incident last year, (no injuries). CFM, (engine manufacturer), recommended blade inspections finished within 12 months, Southwest and others wanted more time to complete them. The labor costs are nil for this but the blades themselves cost $50K for new, $30K refurbished, each CFM56 engine has 24 blades. One wonders if these airlines are afraid of finding lots of iffy blades and were trying to spread the cost over a longer operating period. https://www.apnews.com/7b50c05eac7440f1945dfc1a5bf48fbe

Airlines always ask for more time on NPRMs and to withdrawl the AD altogether. This is almost always to allow more flexibility on when the inspections get scheduled. A 737 goes through a C-check about once every two years, which would be the most convenient time to do these inspections. It also gives the airlines flexibility if they run out of spare parts (likely with this AD) that they can stop their inspections for a while until they get more parts, without grounding aircraft.

The airline I used to work for would've started inspecting as soon as the original CFM bulletin came out, and would've done it to that schedule, but we would've still asked for more time, because we always did, and they did ask for more time on this NPRM as well.

The FAA calculates probability of failure, and that is how they base their compliance times, so the additional time is never granted and is a waste of time to ask for, though.

Might be time for the fed to lean on CFM for internal engine data including their validation testing.

The FAA already has all the data on these fan blades. All analysis methods must be completely FAA approved. All analysis is 100% FAA approved. All test plans are 100% FAA approved, and all test results are 100% approved. The FAA has full access to any and all technical, design, test, or operational data at CFM and any airline any time they want it.

This isn't the self driving car market, nothing gets on an airplane without full FAA approval. The CFR for engine certification is very long and an egine company has to prove compliance to every single line of it. Some individual lines in the CFR have 100 page documents that explain how you actually have to comply with the one sentence in the regs.

 

[BUTCH1]

Airlines always ask for more time on NPRMs and to withdrawl the AD altogether. This is almost always to allow more flexibility on when the inspections get scheduled. A 737 goes through a C-check about once every two years, which would be the most convenient time to do these inspections. It also gives the airlines flexibility if they run out of spare parts (likely with this AD) that they can stop their inspections for a while until they get more parts, without grounding aircraft.

The airline I used to work for would've started inspecting as soon as the original CFM bulletin came out, and would've done it to that schedule, but we would've still asked for more time, because we always did, and they did ask for more time on this NPRM as well.

The FAA calculates probability of failure, and that is how they base their compliance times, so the additional time is never granted and is a waste of time to ask for, though.



The FAA already has all the data on these fan blades. All analysis methods must be completely FAA approved. All analysis is 100% FAA approved. All test plans are 100% FAA approved, and all test results are 100% approved. The FAA has full access to any and all technical, design, test, or operational data at CFM and any airline any time they want it.

This isn't the self driving car market, nothing gets on an airplane without full FAA approval. The CFR for engine certification is very long and an egine company has to prove compliance to every single line of it. Some individual lines in the CFR have 100 page documents that explain how you actually have to comply with the one sentence in the regs.

I agree and take your word for the lengthy FAA approval process, that being said, even those stringent regulations/testing seems to be unable to replicate real world wear+tear. One could argue that with thousands of the CFM56 in service for as long as they have, a few failures might be expected but when someone dies and the plane in possible peril perhaps the FAA might not want to wait when this happens to another engine, CFM or otherwise.

 

[Puffnstuff]

The FAA already has all the data on these fan blades. All analysis methods must be completely FAA approved. All analysis is 100% FAA approved. All test plans are 100% FAA approved, and all test results are 100% approved. The FAA has full access to any and all technical, design, test, or operational data at CFM and any airline any time they want it.

This isn't the self driving car market, nothing gets on an airplane without full FAA approval. The CFR for engine certification is very long and an egine company has to prove compliance to every single line of it. Some individual lines in the CFR have 100 page documents that explain how you actually have to comply with the one sentence in the regs.

I'll bet they missed something which will force them to make some changes which is normal once you discover a problem.

 

[Zorba]

I agree and take your word for the lengthy FAA approval process, that being said, even those stringent regulations/testing seems to be unable to replicate real world wear+tear. One could argue that with thousands of the CFM56 in service for as long as they have, a few failures might be expected but when someone dies and the plane in possible peril perhaps the FAA might not want to wait when this happens to another engine, CFM or otherwise.

I agree with you that I think the FAA took too long to issue a final AD in this case, but unfortunately it is fairly typical. To release an emergency rule (which is what they've now done), you basically have to show an immediate catastrophic threat. Loss of an engine and loss of pressure generally aren't catastrophic (technically, this event wasn't either per the FAA definition), which is why it was in the normal rule making process.

I'll be interested to see what the root cause of the cracking is, most engine components failures are typically traced back to manufacturing mistakes or process changes. With titanium blades there is always a risk of internal inclusion fatigue from the grain structure, which is highly dependent on the manufacturing process to eliminate.

There is no life limit on fan blades, but the engineering is highly conservative. First, you assume absolute worst case material properties. Then figure out how long it'll take to crack, using conservative loads/cycles, and then how long it will take for the crack to cause a failure. Then you require inspections that will give you at least two opportunities to find the crack before that crack would cause a failure. Considering there are close to 10 billion flight hours on this specific blade type, I am leaning to a manufacturing issue rather than a missed design.

The real issue in this case though is to figure out why the inlet cowl is coming apart during the fan blade out, which shouldn't be happening, and I don't think has been figured out yet. The inlet cowl is made/designed by Roar, not CFM, FYI. Generally fan blade failures are more of an economic issue than a safety issue, but generally the cowlings don't fail either (and fan blade failure are very rare).

I'll bet they missed something which will force them to make some changes which is normal once you discover a problem.

You are hardcore moving the goalposts. You said the Fed needed to lean on CFM to get internal data, implying they didn't already have it. I was just saying that they already have full access to all the data. Just because everything is approved, doesn't mean something can't happen.

Obviously there is an ongoing investigation which will figure out why the fan blades are failing and why the inlet cowls are coming apart. CFM has already released multiple bulletins for inspections on the fan blades, one of which the FAA has already mandated, and the FAA is working to mandate the others. Once the issues with the inlet cowl are figured out, Roar will likely issue their own bulletins that the FAA will likely mandate.

Aircraft are extremely complex and are held to the highest standards you'll find in any industry, when issues are uncovered there are always service bulletins issued, and if the FAA considers them a risk to safety of flight they will mandate those actions. The vast majority of issues are discovered by routine maintenance or additional analysis and not an actual incident.

 

[Zorba]

Here is the link to the emergency AD the FAA released on Friday: http://rgl.faa.gov/Regulatory_and_Guidance_Library/rgad.nsf/0/6DE148FE35DDEDDB8625827500759AA4

Here is their description of the unsafe condition:

This AD was prompted by a recent event involving an engine failure, resulting in the engine inlet cowl disintegrating, debris penetrating the fuselage causing a loss of pressurization and prompting an emergency descent. There was one passenger fatality as a result of the event. We are issuing this AD to address fan blade failure due to cracking, which could result in an engine in-flight shutdown (IFSD), uncontained release of debris, damage to the engine, damage to the airplane, and possible airplane decompression.

Here are the required actions, for any engine with more than 30,000 cycles:

(1) Within 20 days after receipt of this AD, perform a one-time ultrasonic inspection (USI) of all 24 fan blade dovetail concave and convex sides to detect cracking.
(2) Use the Accomplishment Instructions, paragraphs 3.A.(3)(a) through (i), of CFM Service Bulletin CFM56-7B S/B 72-1033, dated April 20, 2018, to perform the inspection required by paragraph (g)(1) of this AD.

The biggest issue with this action is that fan blades don't have to be tracked and they can be moved between engines, so it is possible that you have a blade with more than 30,000 cycles on an engine with less than 30,000 cycles.

 

[BUTCH1]

Here is the link to the emergency AD the FAA released on Friday: http://rgl.faa.gov/Regulatory_and_Guidance_Library/rgad.nsf/0/6DE148FE35DDEDDB8625827500759AA4

Here is their description of the unsafe condition:



Here are the required actions, for any engine with more than 30,000 cycles:



The biggest issue with this action is that fan blades don't have to be tracked and they can be moved between engines, so it is possible that you have a blade with more than 30,000 cycles on an engine with less than 30,000 cycles.

This is the main fan at the front that is supplying all the bypass air (and generating thrust), correct?, curious as to why they would let blades go from engine to engine like that. In that scenario a plane with almost enough cycles for a C check could possibly donate a blade to a different plane that had just undergone a C check and said blade possibly go unchecked, guess I'm playing devils advocate here. I tend to agree with a manufacturing issue, this is probably the most widely used engine in aviation today, I find it difficult to believe a sub-par design goes this long before issues would arise.

 

[Fenixgoon]

I agree with you that I think the FAA took too long to issue a final AD in this case, but unfortunately it is fairly typical. To release an emergency rule (which is what they've now done), you basically have to show an immediate catastrophic threat. Loss of an engine and loss of pressure generally aren't catastrophic (technically, this event wasn't either per the FAA definition), which is why it was in the normal rule making process.

I'll be interested to see what the root cause of the cracking is, most engine components failures are typically traced back to manufacturing mistakes or process changes. With titanium blades there is always a risk of internal inclusion fatigue from the grain structure, which is highly dependent on the manufacturing process to eliminate.

There is no life limit on fan blades, but the engineering is highly conservative. First, you assume absolute worst case material properties. Then figure out how long it'll take to crack, using conservative loads/cycles, and then how long it will take for the crack to cause a failure. Then you require inspections that will give you at least two opportunities to find the crack before that crack would cause a failure. Considering there are close to 10 billion flight hours on this specific blade type, I am leaning to a manufacturing issue rather than a missed design.

The real issue in this case though is to figure out why the inlet cowl is coming apart during the fan blade out, which shouldn't be happening, and I don't think has been figured out yet. The inlet cowl is made/designed by Roar, not CFM, FYI. Generally fan blade failures are more of an economic issue than a safety issue, but generally the cowlings don't fail either (and fan blade failure are very rare).



You are hardcore moving the goalposts. You said the Fed needed to lean on CFM to get internal data, implying they didn't already have it. I was just saying that they already have full access to all the data. Just because everything is approved, doesn't mean something can't happen.

Obviously there is an ongoing investigation which will figure out why the fan blades are failing and why the inlet cowls are coming apart. CFM has already released multiple bulletins for inspections on the fan blades, one of which the FAA has already mandated, and the FAA is working to mandate the others. Once the issues with the inlet cowl are figured out, Roar will likely issue their own bulletins that the FAA will likely mandate.

Aircraft are extremely complex and are held to the highest standards you'll find in any industry, when issues are uncovered there are always service bulletins issued, and if the FAA considers them a risk to safety of flight they will mandate those actions. The vast majority of issues are discovered by routine maintenance or additional analysis and not an actual incident.

gotta agree with your assessment here. most likely a manufacturing issue - or i would also say some sort of FOD damage that was just enough to act as a crack initiator, but not enough to be flagged by any visual/NDI inspection. admittedly, that's a pretty small window given the amount of inspection aerospace parts get.

i'm not sure how commercial aviation is certified, but back when i worked for the navy, primary structure for the aircraft had to show 4x life by analysis and 2x life by testing - with no credit taken for life-improvement processing (surface blasting, shot peening, etc.). not sure if engine parts are treated/analyzed in a similar manner, or what kind of treatment fan blades get, but if there were anything that would enhance the fatigue life, that only makes this failure all the more interesting from an engineering perspective.

for sure it will be gone over with a fine-tooth comb to find out the proximate cause of failure (e.g. FOD damage, manufacturing defect) as well as root cause.

 

[Zorba]

This is the main fan at the front that is supplying all the bypass air (and generating thrust), correct?, curious as to why they would let blades go from engine to engine like that. In that scenario a plane with almost enough cycles for a C check could possibly donate a blade to a different plane that had just undergone a C check and said blade possibly go unchecked, guess I'm playing devils advocate here. I tend to agree with a manufacturing issue, this is probably the most widely used engine in aviation today, I find it difficult to believe a sub-par design goes this long before issues would arise.

You are right on the function of these specific blades. Well, engines don't really follow the same maintenance schedules as the aircraft, so they don't really line up with C-checks. But yes, the fan blades could move from one engine to another. What would normally happen is if a fan blade gets damaged, a spare blade would be pulled from stock. If that is a used blade it would've gone through the shop and inspected before being put on the shelf. But it is possible to cannibalize a blade from another aircraft. Generally this would only be done if there were no spare blades and there was another aircraft that was out of service.

Engines do have scheduled on-wing inspections, usually visual inspection or borescope visual inspections, but they don't have scheduled checks like the airframes do. Engines will stay on wing until, 1) There is a problem (high vibes or exhaust temp), 2) The engine reaches the life limit of a rotor or other life limited part. Then they will go to the shop an get overhauled.

The thing I think is really weird about the 30K cycles based on the engine, is I think that exceeds the life limits of the rotors, which means the engine will have been overhauled at least once. When it goes into overhaul the fan blades will get pulled off and processed independently, there is no requirement that they be returned to the same engine they could go anywhere. I think I posted a link to the previous NPRM, you can read the airline's comments to that NPRM and you can see they raise these concerns. Basically for any airline that doesn't track their blades, they are going to end up inspecting all of their blades. IIRC American said they had 2400 directly effected blades, but they were going to have to inspect 15,000 since they didn't actually know where those 2400 were.

BTW: I don't remember the exact count, but I am pretty sure there over 1000 blades in a CFM56-7B, a liberation at the root of any of them will take out the entire engine and none of them are individually tracked and only get borescope inspections in the field. Fan blades just have the most energy, so they create the most violent failure (but it is supposed to not do any damage outside of the engine itself).

 

[Zorba]

gotta agree with your assessment here. most likely a manufacturing issue - or i would also say some sort of FOD damage that was just enough to act as a crack initiator, but not enough to be flagged by any visual/NDI inspection. admittedly, that's a pretty small window given the amount of inspection aerospace parts get.

i'm not sure how commercial aviation is certified, but back when i worked for the navy, primary structure for the aircraft had to show 4x life by analysis and 2x life by testing - with no credit taken for life-improvement processing (surface blasting, shot peening, etc.). not sure if engine parts are treated/analyzed in a similar manner, or what kind of treatment fan blades get, but if there were anything that would enhance the fatigue life, that only makes this failure all the more interesting from an engineering perspective.

for sure it will be gone over with a fine-tooth comb to find out the proximate cause of failure (e.g. FOD damage, manufacturing defect) as well as root cause.

Based solely on the fact that they are calling out ultrasound, I am leaning toward beta grain causing internal fatigue cracks. By the time you'd see them at the surface it would likely be too late. Once the NTSB releases their report for the 2016 incident we should know. On a different engine, all of a sudden they started having fan blade failures, after the fourth one they were able to trace it back changing the diameter of the bar stock used for the forging, that mistake shelled out at least 5 engines and never made the news (and cost the company tens of millions).

4x analysis and 2x by test is similar for commercial airframes. When I designed blades back in the day, we had to show 2x the rotor life by analysis. I know we tested all the way up to the rotor life limits, but I can't recall if we tested beyond that. However, the rotor life limits were "Scored," where you took big knock downs based on actual experience, quality of material data, and validation of loads. So you the published life limits were much lower than analysis until you gained experience.

 

[Puffnstuff]

You are hardcore moving the goalposts. You said the Fed needed to lean on CFM to get internal data, implying they didn't already have it. I was just saying that they already have full access to all the data. Just because everything is approved, doesn't mean something can't happen.

Companies falsify data to gain approval all the time which works until they get caught....just ask VW.

 

[Zorba]

Companies falsify data to gain approval all the time which works until they get caught....just ask VW.

It is a completely different industry, not at all comparable. Again though, you are moving the goalposts. GE has been making engines since the 40s, link to any examples of them falsifying certification data. This is also the most used jet engine of all time, if there was an inherant and known design issue it would've likely shown itself before now.

FAA representatives literally sit next to designers at these companies and have full network access. The FAA witnesses all design reviews and testing. Further, the FAA could pull their type certification or their manufacturing certification which would basically put them out of business.

People that haven't worked in this industry just don't understand how much oversight the FAA provides. Of course there can be misses and mistakes and unknowns that get you, but that is a lot different than saying a company is purposely falsifying data.

So far you've said 1) CFM wasn't providing data and the feds should make them (not true), 2) Just because it's approved doesn't mean it can have issue (true), 3) CFM lied or falsified data to certify this engine (absolutely no evidence of this).

 

[Puffnstuff]

So far you've said 1) CFM wasn't providing data and the feds should make them (not true), 2) Just because it's approved doesn't mean it can have issue (true), 3) CFM lied or falsified data to certify this engine (absolutely no evidence of this).

No that's your assumption and since you profess to be all knowing you should know what is happening right now.

 

[Zorba]

No that's your assumption and since you profess to be all knowing you should know what is happening right now.

What is my assumption? That there is no evidence that CFM falsified data and hide it from the FAA and EASA? Feel free to provide some evidence of you think it exists.

I've just been through this process scores of times, I assume you haven't. I have also read all the relevant materials that have been published on both incidents, including the NPRM comments.

There is obviously an issue, and it is being addressed through inspections. Once they figure out why there was an issue that will be addressed as well. There is no need for the conspiracy bullshit when there is literally no evidence of such.

 

[BUTCH1]

Companies falsify data to gain approval all the time which works until they get caught....just ask VW.

VW didn't just "falsify data", they created PCM software that knew when the engine was being tested and then it went into "cheat" mode, I've never heard of an aviation engine manufacturer pulling anything like that.

 

[Fenixgoon]

What is my assumption? That there is no evidence that CFM falsified data and hide it from the FAA and EASA? Feel free to provide some evidence of you think it exists.

I've just been through this process scores of times, I assume you haven't. I have also read all the relevant materials that have been published on both incidents, including the NPRM comments.

There is obviously an issue, and it is being addressed through inspections. Once they figure out why there was an issue that will be addressed as well. There is no need for the conspiracy bullshit when there is literally no evidence of such.

i don't think most people appreciate how much design, analysis, and testing goes into aircraft overall, and the regulatory structure that controls that design. there's a reason there's only been 1 fatality in a literal decade of commercial aviation. can we do better? sure. but that's pretty goddamn good if you ask me.