Korean Ferry Disaster

[Clark1961]

When you said; "point of design weakness" it's one of those engineer-speak terms that non-engineers use to deflect blame from the operator, like the ham-footed pilot.

I disagree entirely. It has nothing to do with blame and has everything to do with an evaluation of a design. Every design has points of weakness. It can be helpful to acknowledge what those points are. Saying the operator shouldn't do something is fine. Refusing to acknowledge a structure's weak point is not so fine. Refusing to discuss a structure's weak points might border on negligence depending on application.

One extreme example of the level of analysis of a problem is the age-old argument between metallurgist and engineer. The metallurgist says "it broke here because of this inclusion or that crystal structure and the engineer says that it broke because they pulled too hard on it. In other words, the part was in spec but the load was out of spec.

In order to improve the design we have to know whether we exceeded the limits or whether the materials were bad. If we exceeded the limits then we have to build it stronger or prevent exceeding the limits (or some combination).

 

[Greg Bockelman]

In order to improve the design we have to know whether we exceeded the limits or whether the materials were bad. If we exceeded the limits then we have to build it stronger or prevent exceeding the limits (or some combination).

Ok, but how far do you want to take that? Your airplane has all sorts of limits that can be exceeded. You have an airspeed redline that can be exceeded. Beyond a certain point structural failure can occur. You have a stall speed limit. Go below that and you can fall out of the sky with drastic results. Oil temperature limits. Pressure limits. It goes on and on.

Sure, you can design something that will not exceed those limits, but at the expense of systems complexity and in a lot of ways, weight. At the same time, you design the operator out of the equation. An example of that is the Airbus product line. Airbus did their best to design an airplane in which a lot of limits could not be exceeded. And people complained that the pilot didn't have ultimate control over the airplane. Sort of a no win situation.

Bottom line is that most designs are a compromise. A lot of those design decisions are based on a human being being smart enough to not exceed the limits that are placed on the design. If we are not going to trust the operator with making the right decisions then we must accept the fact that in the future the human operator can be designed out of the machine. Do we really want to go there?

 

[denverpilot]

I doubt those commenting are military pilots on the front lines. But it could be they are blowing off steam after dealing with the stress of flying a Cherokee 140, an activity fraught with danger.



Or maybe your comment is just nonsense.

It's an Internet forum. Of course it's nonsense.

Transporting mass humans on water involves drowning a few once in a while when someone screws the pooch. It happens, it'll continue to happen.

And the jokes will always be a part of it, whether you like it or not.

Captain Ho Le Fuk left his job at the airline and joined the ferry drivers, apparently.

 

[Fearless Tower]

I disagree entirely. It has nothing to do with blame and has everything to do with an evaluation of a design. Every design has points of weakness. It can be helpful to acknowledge what those points are. Saying the operator shouldn't do something is fine. Refusing to acknowledge a structure's weak point is not so fine. Refusing to discuss a structure's weak points might border on negligence depending on application.

One extreme example of the level of analysis of a problem is the age-old argument between metallurgist and engineer. The metallurgist says "it broke here because of this inclusion or that crystal structure and the engineer says that it broke because they pulled too hard on it. In other words, the part was in spec but the load was out of spec.

In order to improve the design we have to know whether we exceeded the limits or whether the materials were bad. If we exceeded the limits then we have to build it stronger or prevent exceeding the limits (or some combination).

Perhaps I missed something, but where has it been established that the ship capsized because of the amount of rudder used?

I can't say one doesn't exist, but I am not familiar with any case involving a ship (not counting small boats) where the vessel capsized due to excessive use of rudder. Plenty of cases of ships rolling over due to improper cargo loading and/or securing exacerbated by rudder or sea state, but the primary factor was always improper loading, not improper use of rudder.

 

[Clark1961]

Ok, but how far do you want to take that? Your airplane has all sorts of limits that can be exceeded. You have an airspeed redline that can be exceeded. Beyond a certain point structural failure can occur. You have a stall speed limit. Go below that and you can fall out of the sky with drastic results. Oil temperature limits. Pressure limits. It goes on and on.

Sure, you can design something that will not exceed those limits, but at the expense of systems complexity and in a lot of ways, weight. At the same time, you design the operator out of the equation. An example of that is the Airbus product line. Airbus did their best to design an airplane in which a lot of limits could not be exceeded. And people complained that the pilot didn't have ultimate control over the airplane. Sort of a no win situation.

Bottom line is that most designs are a compromise. A lot of those design decisions are based on a human being being smart enough to not exceed the limits that are placed on the design. If we are not going to trust the operator with making the right decisions then we must accept the fact that in the future the human operator can be designed out of the machine. Do we really want to go there?

It isn't that extreme. We have to figure out what the limits are and we have to communicate the limits. Some like Vne are easy to communicate and perhaps even put control limits on. Others, like control reversals are not so easy to implement control limits on and folks obviously failed to communicate them in an effective manner.

Anyway, at the heart things is a need to know and communicate what will break the system. Then we can look at improved design and when that isn't possible we can look at effective training. Airlines have certainly gone that way by training crews to use automation standard procedures and there are still failures there.

Anyway, none of that is solved by denying a structure has a weak point. All structures have weak points...

 

[weilke]

IF you go hard rudder and immediately shift it, not much will happen in a carrier. Newton stuff. Lot of mass to move and the ship takes a little while to respond to the rudder. It is actually a material check that is required for INSURV (congressionally mandated ship inspection done roughly every 5 years).

The thread has wandered off into the tall weeds a bit.

I believe the term I used was 'air carrier aircraft' as in part25 certified 'transport category' aicraft. As the AAL Airbus over Queens demonstrated, there are limitations on how hard you can stomp on the rudder in one of these. Same with things like gust limits near Vne etc.

I dont know what if any operating limitations this particular ferry had and whether the crew exceeded them. I just responded to the suprise that a hard-over could cause a ship to capsize. I dont think an empty ferry or freighter will capsize, but add shifting or sloshing cargo and history has shown that it can happen.

 

[Fearless Tower]

I just responded to the suprise that a hard-over could cause a ship to capsize. I dont think an empty ferry or freighter will capsize, but add shifting or sloshing cargo and history has shown that it can happen.

Understand. My point simply being that hard over rudder alone cannot capsize a ship without improper loading/shifting of said load.

 

[Sac Arrow]

Understand. My point simply being that hard over rudder alone cannot capsize a ship without improper loading/shifting of said load.

Nobody's established what happened but barring rough seas or hitting something, it appears that a hard steering input may have at least indirectly caused the boat to capsize. A statement had been made previously that several ferry ships of the same type had capsized, which led me to speculate that there might be a flaw in the design of the ship.

If you can load all the trucks on one side of ferry "A" and capsize it, and do the same thing on ferry "B" and it won't, I would say ferry "B" is a better design.

But if ferry's C, D, E, F, G, H, I and J won't capsize either similarly loaded, I would argue ferry A is not designed to the industry standard of care.

 

[Everskyward]

We have to figure out what the limits are and we have to communicate the limits. Some like Vne are easy to communicate and perhaps even put control limits on. Others, like control reversals are not so easy to implement control limits on and folks obviously failed to communicate them in an effective manner.

This was the real problem. Back before that accident pilots of all experience levels generally thought that you couldn't break the airplane with full control deflections below maneuvering speed. No one emphasized that repeated deflections had not been tested.

Now it's taught differently.

The design maneuvering speed (VA) is the speed below which you can move a single flight control, one time, to its full deflection, for one axis of airplane rotation only (pitch, roll or yaw), in smooth air, without risk of damage to the airplane.

Even though the accident discussed above is a part 25 airplane, VA is applicable to part 23, CAR 3, and LSA airplanes. Also, even though experimental airplanes may not have a published VA , they will still have some maximum maneuvering speed associated with the maximum structural design loads. Therefore, the pilot should be aware of what speed this is, and adhere to the guidance herein. The regulations governing the design strength requirements for airplane structure require adequate strength for full control deflection (below VA). However, they do not require the manufacturer to make the airplane strong enough to withstand full control input followed by a full control input in the opposite direction, even below VA. Neither do they require the manufacturer to design the airplane for more than one simultaneous full control input such as full ailerons with full elevator and/or rudder.

http://rgl.faa.gov/Regulatory_and_Guidance_Library/rgSAIB.nsf/dc7bd4f27e5f107486257221005f069d/3c00e5aa64a2827e8625781c00744393/$FILE/CE-11-17.pdf

 

[Henning]

I'm no marine engineer obviously, but what constitutes "operating not as designed?" Most ships are generally not designed to run in to hard objects such as icebergs and other ships, and should be operated as such. I get that. But tipping over from a full hard rudder? Loading limits were grossly exceeded? Other ships don't seem to do that.

Loading of cargo and passengers within the design stability parameters set forth in the stability book developed by the Naval Architect. One can also add to that operating in a negligent condition.