Qantas 744 Depressurisation
Evertonian
Litebulbs. I too noticed that damage & was thinking about questioning it, but after some thought about how the handle could have bent like that, I came to the conclusion that it was more than likely the door panelling that buckled into the handle. Naturally, I don't know this for fact, and all I can go by is that one photo where the damage is evident.
Join Date: Jul 2003
Location: Australia
Posts: 157
Likes: 0
Received 0 Likes
on
0 Posts
Peepeerune
My thoughts too I even suggest that there may have been a second projectile in the cabin and the green skid mark may have come from the portable bottle.
I can't see anybody would have been game enough to approach that door in flt. note the oxygen masks still in packs.
where is the small portable oxygen bottle that should be stowed next to the door at floor level?.it appears to be missing in that photo.
Maybe the crew used it ?
Join Date: Jul 2003
Location: Australia
Posts: 157
Likes: 0
Received 0 Likes
on
0 Posts
Their is no other visible damage to the bottom edge of the handle. It had been hit with enough force to deform it inwards and move it towards the open position,
Join Date: Jul 2003
Location: Australia
Posts: 157
Likes: 0
Received 0 Likes
on
0 Posts
I think it does.... the handle took the full force of the ??bottle?? and fortunately reduced its force... if it hadn't have been hit the damage around the upper door hinge point may have been more catastrophic . If the R2 door had failed then we would not have pictures to discuss all of the 'What If's'
Evertonian
Ah...Lightbulbs, while my post still stands, I misunderstood your question. That handle is actually designed that way. The top quarter (I suppose) angles into the door, but clears it of course. So, the shape of the handle has not been affected by the incident.
Last edited by Buster Hyman; 7th Aug 2008 at 05:16.
Join Date: Jul 2003
Location: Australia
Posts: 157
Likes: 0
Received 0 Likes
on
0 Posts
JAMES REASON
I have a new theme for another thread " Australian Cheddar"
By now we all know what the "Swiss cheese theory" is and how important it has been in our industry in preventing serious incidents/loss of life. Well I propose in Qantas there is another theory or phenomenon and that should be the " Aussie Cheddar" .
Cheddar has no holes and that way a catastrophic event is prevented. This theory doesn't require many slices of cheddar only one placed at random somewhere amongst all the slices of swiss cheese.
Read on and you will understand.
1. Extremely bad weather enough to cause errors in judgement, procedures and operation. But wait the golf course is soaked in water the ground is boggy and the fairways are clear. Enter one 747 which failed to stop on the runway because of the holes in the cheese but stops stuck in the bog, NO LOSS OF LIFE
2. A leaking tap/drain in the forward galley directly above the equipment bay full of avionics boxes and computers, but not a problem splash guards and drip trays will prevent damage, regular/scheduled inspections will locate damaged drip trays before anything serious happens.....NOT. So what happens when the aircraft looses all power in the middle of the Pacific or Indian Ocean??? Never fear the aircraft happens to be less than an hour away from landing and the Battery power is JUST sufficient enough to allow for safe landing NO LOSS OFF LIFE
3. A never before event in aviation HAPPENS ?? when an oxygen bottle explodes in flight on a passenger aircraft. How many slices of cheese before it got to this stage.
But wait 'Aussie Cheddar' to the rescue the bottle doesn't go through the fuel tank, the debris doesn't take out #3 engine, no one is sucked out the hole, just a couple of old bags( I didn't say old boilers). Parts of the bottle are projected into the next compartment and heavy impact on the door handle dissipates the force and prevents more serious damage to structure and possibly Door R2. The crew declare MayDay and divert to an air port that was only hours early closed due to a typhoon, fortunately this aircraft has all the latest landing instruments that will make a bad weather approach less difficult. Not need ( and didn't have all available anyway) fine weather landing. NO LOSS OF LIFE
BREAKING NEWS
Qantas have announced that they are making a take-over for the Kraft Cheese Company.
I have a new theme for another thread " Australian Cheddar"
By now we all know what the "Swiss cheese theory" is and how important it has been in our industry in preventing serious incidents/loss of life. Well I propose in Qantas there is another theory or phenomenon and that should be the " Aussie Cheddar" .
Cheddar has no holes and that way a catastrophic event is prevented. This theory doesn't require many slices of cheddar only one placed at random somewhere amongst all the slices of swiss cheese.
Read on and you will understand.
1. Extremely bad weather enough to cause errors in judgement, procedures and operation. But wait the golf course is soaked in water the ground is boggy and the fairways are clear. Enter one 747 which failed to stop on the runway because of the holes in the cheese but stops stuck in the bog, NO LOSS OF LIFE
2. A leaking tap/drain in the forward galley directly above the equipment bay full of avionics boxes and computers, but not a problem splash guards and drip trays will prevent damage, regular/scheduled inspections will locate damaged drip trays before anything serious happens.....NOT. So what happens when the aircraft looses all power in the middle of the Pacific or Indian Ocean??? Never fear the aircraft happens to be less than an hour away from landing and the Battery power is JUST sufficient enough to allow for safe landing NO LOSS OFF LIFE
3. A never before event in aviation HAPPENS ?? when an oxygen bottle explodes in flight on a passenger aircraft. How many slices of cheese before it got to this stage.
But wait 'Aussie Cheddar' to the rescue the bottle doesn't go through the fuel tank, the debris doesn't take out #3 engine, no one is sucked out the hole, just a couple of old bags( I didn't say old boilers). Parts of the bottle are projected into the next compartment and heavy impact on the door handle dissipates the force and prevents more serious damage to structure and possibly Door R2. The crew declare MayDay and divert to an air port that was only hours early closed due to a typhoon, fortunately this aircraft has all the latest landing instruments that will make a bad weather approach less difficult. Not need ( and didn't have all available anyway) fine weather landing. NO LOSS OF LIFE
BREAKING NEWS
Qantas have announced that they are making a take-over for the Kraft Cheese Company.
The Reverend
Join Date: Oct 1999
Location: Sydney,NSW,Australia
Posts: 2,020
Likes: 0
Received 0 Likes
on
0 Posts
Nobody has commented so far on the outward bulge of the top right hand R2 door hinge panel on the exterior view on post #745. That panel is supposed to be flush with the fuselage skin.
Evertonian
Hmm...still not seeing 80 degrees. The handle is designed with an inward 'bend' at the tip & that's all I see. I am seeing an armed door (if down is armed...I can't recall) in the Open position though...if comparing that KLM door to the QF one. The handle looks ok to me.
Anyway, the people that count will know what they are looking at.
Anyway, the people that count will know what they are looking at.
Join Date: Jul 2002
Location: 40N, 80W
Posts: 233
Likes: 0
Received 0 Likes
on
0 Posts
I am just a private pilot with no special knowledge of oxygen systems. My trade is (was) Physics.
This is a fascinating thread generated by many smart and knowledgably people.
But I am curious about some reactions.
Lack of curiosity
For instance, several comments have been along the lines of “an explosion of an oxygen cylinder in flight has never happened before, so how could it possibly have happened now“?
Yet when presented with a case of an oxygen cylinder exploding in a vehicle on the ground, complete with a photo in glorious Technicolor (from geemul in Post #832 on 31st July 2008), nobody seems to have expressed curiosity about how that could have happened.
The vehicle was said to be unoccupied and no one was hurt. Nobody was doing anything to the cylinder. There was no turbulence. No G-forces. Maybe the temperature in the vehicle was exceptionally high (less likely since the explosion occurred at night).
Question: what was determined to be the cause of the explosion? Did the insurance company investigate, and pay out on the loss? Where were the fractures in the cylinder? Presumably the fragments were available to re-construct the cylinder. Was any defect found in the cylinder? Even more interesting, was no defect found in the cylinder (other that its failure to hold pressure). Why is nobody interested enough to ask?
A perfect cylinder
One aspect which has not been mentioned yet (maybe I missed it) is the fact that for a perfect cylinder of uniform thickness under internal pressure, the circumferential tensile stress in the cylinder wall is about twice the tensile stress in the cylinder wall parallel to the cylinder axis. So a “good” cylinder should fail by a crack developing parallel to the cylinder axis. This seems to have happened to the failed cylinder shown by Manchaca (Post #832 on July 31st 2008), part of that photo is reproduced below.
The circumferential stress would be somewhat lower towards the cylinder ends, but once the crack starts it could tend to continue in a perfect cylinder in the same direction and end by splitting the neck containing the valve (even though its considerably thicker than the walls). Which is what seems to have happened to the cylinder in the photo above. In which case the valve might be released from the cylinder.
Cylinder failure in any other mode (such as the top half of the cylinder separating from the bottom half) seems to me to imply the presence of a significant defect or external influence. The cylinder necks are considerably thicker than the walls and look to be hard to “knock off”. The valve found on the aircraft was reported to be still attached to part of the cylinder, so the valve did not “pop out of the neck”.
Don’t believe everything you read
Posts #948 &951 discuss a cylinder “containing 115 cubic feet of oxygen @ 1,850 psig at 70ºF”. A cylinder of the proportions shown in photographs of typical installations with an internal volume of 115 cubic feet would have to be approximately 4 ft in diameter and 9 ft long internal dimensions. If this 115 number comes from a published specification, either the specification is wrong or it is not relevant to this thread. Possibly the volume quoted may be for the gas content at atmospheric pressure, or maybe a decimal point is missing. In contrast, the specifications for “3HT seamless steel cylinders for aircraft use” in Post #852 is for “seamless steel cylinders with a water capacity (nominal) of not over 150 pounds”, or about 2.4 cubic feet - nothing like 115 cubic feet.
Adiabatic expansion
The expansion calculation by Mech-prentice in Post #942 looks right. Modifying it along the lines suggested by Machaca in Post #948 for expansion to 12 psi in the cargo hold instead of 4 psi outside the aircraft gives an expansion factor of about 35.8 instead of 78.6. This gas would be very very cold. If I have done my sums correctly (Mech-prentice check?), the adiabatic cooling of oxygen starting at 1,850 psi and room temperature expanding to 12 psia would cool it enough to liquefy some of it at near -170ºC. As the oxygen expands it does work on the surrounding air, adiabatically heating it (a little), but the oxygen itself gets cold in the process. In the case of the cylinder in the car, the oxygen is also does work on the vehicle (as you can see), and this energy comes from that stored in the pressurized oxygen, which consequently cools. As the oxygen warms up to ambient temperature it would expand by another factor of about 4 - 5.
Ignition process
A once heard a speaker at a safety meeting explain how some oxygen accidents occur when grease is present in tubing between the cylinder and the regulator. The operator typically attaches a regulator to a cylinder and then abruptly opens the cylinder valve. The air in the tubing between the cylinder valve and the regulator is thus suddenly adiabatically compressed and heated, and any hydrocarbon ignites, as in the cylinder of a diesel engine. The oxygen itself is cold - it’s the heated air in front of it which initially ignites the grease. Of course, once ignited, the oxygen burns the remaining grease together with the metal tubing, regulator, etc. So the valve should always be cracked open slowly in case grease is present. If there were no adiabatic heating, oxygen at room temperature would not ignite the grease, even under pressure.
Oxygen lances
Hot steel will burn in oxygen. I understand that the standard method of cutting thick (e.g. 1 inch and thicker) steel plates is by an oxygen lance. The cut is started by an oxy-acetylene torch. Once the steel is hot the acetylene is turned off and the cut is continued using pure oxygen - the burning steel provides all the required heat. This is analogous to what happens in grease-initiated accidents - the grease is ignited, after which metal burning in oxygen provides all the required energy.
A completely different case
Below are stills from the Youtube video mentioned by FlexibleResponse in Post #805.
http://www.youtube.com/watch?v=9lw_fhNAIQc
The voice commentary says that the inside of the top of the cylinder (Photo left below), the cylinder threads (next photo below), and the valve threads (last two photos below), were all melted . This suggests chemical energy to me, involving burning inside the cylinder before it exploded. This cylinder burst consistent with being heated inside near the valve, not by a split parallel to the axis as described above. Quite different from the aircraft incident.
Final note - I have no idea why the cylinder burst in the aircraft.
This is a fascinating thread generated by many smart and knowledgably people.
But I am curious about some reactions.
Lack of curiosity
For instance, several comments have been along the lines of “an explosion of an oxygen cylinder in flight has never happened before, so how could it possibly have happened now“?
Yet when presented with a case of an oxygen cylinder exploding in a vehicle on the ground, complete with a photo in glorious Technicolor (from geemul in Post #832 on 31st July 2008), nobody seems to have expressed curiosity about how that could have happened.
The vehicle was said to be unoccupied and no one was hurt. Nobody was doing anything to the cylinder. There was no turbulence. No G-forces. Maybe the temperature in the vehicle was exceptionally high (less likely since the explosion occurred at night).
Question: what was determined to be the cause of the explosion? Did the insurance company investigate, and pay out on the loss? Where were the fractures in the cylinder? Presumably the fragments were available to re-construct the cylinder. Was any defect found in the cylinder? Even more interesting, was no defect found in the cylinder (other that its failure to hold pressure). Why is nobody interested enough to ask?
A perfect cylinder
One aspect which has not been mentioned yet (maybe I missed it) is the fact that for a perfect cylinder of uniform thickness under internal pressure, the circumferential tensile stress in the cylinder wall is about twice the tensile stress in the cylinder wall parallel to the cylinder axis. So a “good” cylinder should fail by a crack developing parallel to the cylinder axis. This seems to have happened to the failed cylinder shown by Manchaca (Post #832 on July 31st 2008), part of that photo is reproduced below.
The circumferential stress would be somewhat lower towards the cylinder ends, but once the crack starts it could tend to continue in a perfect cylinder in the same direction and end by splitting the neck containing the valve (even though its considerably thicker than the walls). Which is what seems to have happened to the cylinder in the photo above. In which case the valve might be released from the cylinder.
Cylinder failure in any other mode (such as the top half of the cylinder separating from the bottom half) seems to me to imply the presence of a significant defect or external influence. The cylinder necks are considerably thicker than the walls and look to be hard to “knock off”. The valve found on the aircraft was reported to be still attached to part of the cylinder, so the valve did not “pop out of the neck”.
Don’t believe everything you read
Posts #948 &951 discuss a cylinder “containing 115 cubic feet of oxygen @ 1,850 psig at 70ºF”. A cylinder of the proportions shown in photographs of typical installations with an internal volume of 115 cubic feet would have to be approximately 4 ft in diameter and 9 ft long internal dimensions. If this 115 number comes from a published specification, either the specification is wrong or it is not relevant to this thread. Possibly the volume quoted may be for the gas content at atmospheric pressure, or maybe a decimal point is missing. In contrast, the specifications for “3HT seamless steel cylinders for aircraft use” in Post #852 is for “seamless steel cylinders with a water capacity (nominal) of not over 150 pounds”, or about 2.4 cubic feet - nothing like 115 cubic feet.
Adiabatic expansion
The expansion calculation by Mech-prentice in Post #942 looks right. Modifying it along the lines suggested by Machaca in Post #948 for expansion to 12 psi in the cargo hold instead of 4 psi outside the aircraft gives an expansion factor of about 35.8 instead of 78.6. This gas would be very very cold. If I have done my sums correctly (Mech-prentice check?), the adiabatic cooling of oxygen starting at 1,850 psi and room temperature expanding to 12 psia would cool it enough to liquefy some of it at near -170ºC. As the oxygen expands it does work on the surrounding air, adiabatically heating it (a little), but the oxygen itself gets cold in the process. In the case of the cylinder in the car, the oxygen is also does work on the vehicle (as you can see), and this energy comes from that stored in the pressurized oxygen, which consequently cools. As the oxygen warms up to ambient temperature it would expand by another factor of about 4 - 5.
Ignition process
A once heard a speaker at a safety meeting explain how some oxygen accidents occur when grease is present in tubing between the cylinder and the regulator. The operator typically attaches a regulator to a cylinder and then abruptly opens the cylinder valve. The air in the tubing between the cylinder valve and the regulator is thus suddenly adiabatically compressed and heated, and any hydrocarbon ignites, as in the cylinder of a diesel engine. The oxygen itself is cold - it’s the heated air in front of it which initially ignites the grease. Of course, once ignited, the oxygen burns the remaining grease together with the metal tubing, regulator, etc. So the valve should always be cracked open slowly in case grease is present. If there were no adiabatic heating, oxygen at room temperature would not ignite the grease, even under pressure.
Oxygen lances
Hot steel will burn in oxygen. I understand that the standard method of cutting thick (e.g. 1 inch and thicker) steel plates is by an oxygen lance. The cut is started by an oxy-acetylene torch. Once the steel is hot the acetylene is turned off and the cut is continued using pure oxygen - the burning steel provides all the required heat. This is analogous to what happens in grease-initiated accidents - the grease is ignited, after which metal burning in oxygen provides all the required energy.
A completely different case
Below are stills from the Youtube video mentioned by FlexibleResponse in Post #805.
http://www.youtube.com/watch?v=9lw_fhNAIQc
The voice commentary says that the inside of the top of the cylinder (Photo left below), the cylinder threads (next photo below), and the valve threads (last two photos below), were all melted . This suggests chemical energy to me, involving burning inside the cylinder before it exploded. This cylinder burst consistent with being heated inside near the valve, not by a split parallel to the axis as described above. Quite different from the aircraft incident.
Final note - I have no idea why the cylinder burst in the aircraft.
Last edited by PickyPerkins; 7th Aug 2008 at 18:20.
Join Date: Jul 2007
Location: My Stringy Brane
Posts: 377
Likes: 0
Received 0 Likes
on
0 Posts
PickyPerkins
Don't believe everything you read too quickly 
Yes, the stated volume of 115 cubic feet is the gas content (industry standard), not internal volume (2.4 cubic feet of O2 wouldn't provide much relief!).
True. The method/specs for hydrostatic testing are also included in Post #852. Cylinder walls are constantly flexing due to changes in pressure differential. When lifespan limit (24 years) is reached or testing indicates excessive "bulging," the cylinder is scrapped. This rupture is considered an "ideal failure" from an over-pressure burst test:
If QF30's suspect cylinder turns out to be the culprit due to a manufacturing flaw determined from the fragment(s), then I expect we'll see more stringent post-manufacturing testing & inspection. Otherwise, we may see the current 3-year test frequency shortened significantly.
I'd love to know the life & service history of the green meanie!
Posts #948 &951 discuss a cylinder “containing 115 cubic feet of oxygen @ 1,850 psig at 70ºF”. A cylinder of the proportions shown in photographs of typical installations with an internal volume of 115 cubic feet would have to be approximately 4 ft in diameter and 9 ft long internal dimensions. If this 115 number comes from a published specification, either the specification is wrong or it is not relevant to this thread. Possibly the volume quoted may be for the gas content at atmospheric pressure, or maybe a decimal point is missing. In contrast, the specifications for “3HT seamless steel cylinders for aircraft use” in Post #852 is for “seamless steel cylinders with a water capacity (nominal) of not over 150 pounds”, or about 2.4 cubic feet - nothing like 115 cubic feet.
Yes, the stated volume of 115 cubic feet is the gas content (industry standard), not internal volume (2.4 cubic feet of O2 wouldn't provide much relief!).
So a “good” cylinder should fail by a crack developing parallel to the cylinder axis.
If QF30's suspect cylinder turns out to be the culprit due to a manufacturing flaw determined from the fragment(s), then I expect we'll see more stringent post-manufacturing testing & inspection. Otherwise, we may see the current 3-year test frequency shortened significantly.
I'd love to know the life & service history of the green meanie!
Last edited by Machaca; 8th Aug 2008 at 00:40. Reason: corrected test freq
Join Date: Jul 2007
Location: My Stringy Brane
Posts: 377
Likes: 0
Received 0 Likes
on
0 Posts
From CASA's April 2006 AD on Inspection, Test and Retirement of Compressed Gas Cylinders:
NSEU -- Does anything besides a complete de-pressurisation & re-pressurisation constitute a cycle?
Permanent volumetric expansion must not exceed
10% of total volumetric expansion at test pressure or a permanent increase in
volume of more than 1/5000 of its original volume.
10% of total volumetric expansion at test pressure or a permanent increase in
volume of more than 1/5000 of its original volume.
In accordance with whichever of the following occurs first:
(a) the manufacturer's specification,
(b) for 3HT cylinders:
(i) 4 380 pressurisations (cycles), or
(ii) 24 years from date of manufacture
(a) the manufacturer's specification,
(b) for 3HT cylinders:
(i) 4 380 pressurisations (cycles), or
(ii) 24 years from date of manufacture