North Sea heli ditching: Oct 2012
From Flightglobal :
Bertling: EC225 grounding 'biggest issue' in Eurocopter history
"Eurocopter faces the "biggest issue" in its 20-year history, according to its chief executive, as the airframer attempts to deal with the fallout from two related ditchings of EC225s in the North Sea.
Speaking at a Helicopter Safety Steering Group (HSSG) meeting in November, according to minutes seen by Flightglobal, Eurocopter boss Lutz Bertling admitted the UK's oil and gas industry had "lost confidence in the EC225 helicopters and in Eurocopter".
Although the EC225 fleet used for offshore transportation in the oil and gas sector has been grounded in the UK and Norway since October, after rulings by both countries' civil aviation regulators, Bertling says Eurocopter's first priority is not to return the Super Pumas to service, but to "regain confidence in the aircraft, company and solution".
Eurocopter is still attempting to identify the root cause of the cracks in the main gearbox bevel gear shafts that forced the ditchings of helicopters operated by Bond Offshore Helicopters and CHC Scotia in May and October respectively.
Eurocopter chief technical officer Jean-Brice Dumont told the HSSG meeting the company believes the fractures were caused by "resonance" and it has embarked on a flight- and ground-test campaign to verify its hypothesis. Although speculation has focused on the potential need to replace hundreds of the gearbox shafts, Dumont says if its theory is correct, "the fix will lie in the internal operating conditions as opposed to the part itself".
However, helicopter operators and oil and gas companies may need further convincing before regaining their trust in the company. Speaking at the meeting, Richard Mintern, Bond Aviation Group chief executive, said industry should form part of the validation process.
Bertling supported this suggestion and, in addition, offered industry representatives the opportunity to visit Eurocopter's Marignane plant to see "details of the investigation, testing, general design and maintenance processes".
The minutes also highlight a disagreement between European regulator EASA and the UK's CAA on how to proceed following the second ditching. Giles Porter, representing the UK regulator, says it requested that EASA include operational limitations in its airworthiness directive, issued on 25 October, "however, they were resistant". Therefore, the CAA went ahead with its effective grounding notice.
To lift the overwater flight ban, Porter says: "CAA airworthiness experts must be satisfied that all problems have been fully resolved so that there will be no recurrence of the incident."
.
Bertling: EC225 grounding 'biggest issue' in Eurocopter history
"Eurocopter faces the "biggest issue" in its 20-year history, according to its chief executive, as the airframer attempts to deal with the fallout from two related ditchings of EC225s in the North Sea.
Speaking at a Helicopter Safety Steering Group (HSSG) meeting in November, according to minutes seen by Flightglobal, Eurocopter boss Lutz Bertling admitted the UK's oil and gas industry had "lost confidence in the EC225 helicopters and in Eurocopter".
Although the EC225 fleet used for offshore transportation in the oil and gas sector has been grounded in the UK and Norway since October, after rulings by both countries' civil aviation regulators, Bertling says Eurocopter's first priority is not to return the Super Pumas to service, but to "regain confidence in the aircraft, company and solution".
Eurocopter is still attempting to identify the root cause of the cracks in the main gearbox bevel gear shafts that forced the ditchings of helicopters operated by Bond Offshore Helicopters and CHC Scotia in May and October respectively.
Eurocopter chief technical officer Jean-Brice Dumont told the HSSG meeting the company believes the fractures were caused by "resonance" and it has embarked on a flight- and ground-test campaign to verify its hypothesis. Although speculation has focused on the potential need to replace hundreds of the gearbox shafts, Dumont says if its theory is correct, "the fix will lie in the internal operating conditions as opposed to the part itself".
However, helicopter operators and oil and gas companies may need further convincing before regaining their trust in the company. Speaking at the meeting, Richard Mintern, Bond Aviation Group chief executive, said industry should form part of the validation process.
Bertling supported this suggestion and, in addition, offered industry representatives the opportunity to visit Eurocopter's Marignane plant to see "details of the investigation, testing, general design and maintenance processes".
The minutes also highlight a disagreement between European regulator EASA and the UK's CAA on how to proceed following the second ditching. Giles Porter, representing the UK regulator, says it requested that EASA include operational limitations in its airworthiness directive, issued on 25 October, "however, they were resistant". Therefore, the CAA went ahead with its effective grounding notice.
To lift the overwater flight ban, Porter says: "CAA airworthiness experts must be satisfied that all problems have been fully resolved so that there will be no recurrence of the incident."
.
Join Date: Apr 2008
Location: scotland
Posts: 29
Likes: 0
Received 0 Likes
on
0 Posts
If it could be proved beyond any doubt that a software upgrade or 14Hz mod was responsible for a harmonic / vibration change within the MGB, which subsequently led to the shaft failing, and that upgrade / mod could be reversed, would you as Pilots and Engineers be happy to operate the 225 once again?
Just putting it out there, because from an EC point of view that would surely be the easiest fix.
Just putting it out there, because from an EC point of view that would surely be the easiest fix.
Is this an impossible standard?
"CAA airworthiness experts must be satisfied that all problems have been fully resolved so that there will be no recurrence of the incident."
Terminus mos first mentioned the 14hz and FADEC V12 on 18 November in post 453...So, EC have suspected this as the potential problem since at least mid November.
So, nearly one month later, if EC goes back to V11 or even V10 FADEC software, fixes the EM Lub indications, temporarily reduces the torque, speed and TOW, and specifies 3 hour HUMS downloads, who will be the first to re instate the EC225 into service in those countries where it is not grounded and would the UK CAA accept those measures to unground the EC225?
There were some Fadec changes and EC is currently examining the effect of Version 12 Fadec software which was released in response to the 14 Hz vibration to slightly lower the Nr.
Join Date: Dec 2006
Location: UK and MALTA
Age: 61
Posts: 1,297
Likes: 0
Received 18 Likes
on
4 Posts
SAS - Sorry mate I made a bit of a Joke at your expense but it was a joke. I should not have mocked your Military Experience so please accept my apology. I am ex Army Aviation but I am one of the lucky ones that never saw any action so big respect for your Service and please accept my apology. It was really just a joke.
Best Regards
DB
Best Regards
DB
Join Date: Dec 2006
Location: UK and MALTA
Age: 61
Posts: 1,297
Likes: 0
Received 18 Likes
on
4 Posts
PITS - you are missing the point again. EASA is not responsible for Commercial Helicopter Operations Regulations - yet. It is responsible for airwoirthiness issues. It has not grounded the aircraft because, under certain operating restrictions, it is safe to fly.
UK CAA regulate Commercial Air Transportation Operations in the UK. Operators must comply with JAR-OPS 3. They have placed a restriction on operations over hostile territory (Open water areas North and South of 45 dgrees fall into this category).
EASA and CAA are in agreeemnt, they just have different legal and regulatory responsibilities and what we are suject to is the manifestation of those responsibilities.
On top of all of that the Manufacturers under Part 145, The Oil Companys under there own expositions and the Operators under the AOC approvals are also legally oblidged to conduct their own internal risk assessments and make judgements that may or may not be in line with the Regulator. However, in this case, I beleive all stakeholders are in exact agreement. We cannot tolerate hostile environment operations without the required level of safety. That is the big challenge at the moment.
I hope this makes sense.
DB
UK CAA regulate Commercial Air Transportation Operations in the UK. Operators must comply with JAR-OPS 3. They have placed a restriction on operations over hostile territory (Open water areas North and South of 45 dgrees fall into this category).
EASA and CAA are in agreeemnt, they just have different legal and regulatory responsibilities and what we are suject to is the manifestation of those responsibilities.
On top of all of that the Manufacturers under Part 145, The Oil Companys under there own expositions and the Operators under the AOC approvals are also legally oblidged to conduct their own internal risk assessments and make judgements that may or may not be in line with the Regulator. However, in this case, I beleive all stakeholders are in exact agreement. We cannot tolerate hostile environment operations without the required level of safety. That is the big challenge at the moment.
I hope this makes sense.
DB
Camper, I would say yes, because we did around 70,000 hours prior to this software version without a problem. However we did a moderate amount of flying with this software without a problem, so the issue is to be convinced that it was definitely the software change that caused the problem and not something else.
Resonance
The Flight Global posting containing a reference to a main gear box resonance reminded me that in early 1963, the CH-47A prototypes at the Test Board ( Ft Rucker ) experienced either two or three incidences of catastrophic nose gear box failure. Resulting shrapnel went into that engine intake and all in all created lots of noise and a mess, not to mention the adverse notoriety associated with the precautionary landings, one of which was on the golf course ( the old one ) adjacent to the Officers Club. Boeing determined rather quickly that the box had a resonance and the corrective action was to raise the Nr from 204 to 230. Problem solved.
After joining SA and learning the numerous factors affecting the design and qualification elements in the drive train, I was always amazed that Boeing could raise the power-on Nr that much without incurring other major problems.
Although one might at first think that the overall engine response at 14 hz is an unlikely trigger, one would assume that EC would be doing ( have already done? ) extensive frequency domain testing in the qualification process for the control system change. That sort of testing ( I refer to in-flight testing ) sometimes uncovers unexpected results.
After joining SA and learning the numerous factors affecting the design and qualification elements in the drive train, I was always amazed that Boeing could raise the power-on Nr that much without incurring other major problems.
Although one might at first think that the overall engine response at 14 hz is an unlikely trigger, one would assume that EC would be doing ( have already done? ) extensive frequency domain testing in the qualification process for the control system change. That sort of testing ( I refer to in-flight testing ) sometimes uncovers unexpected results.
Last edited by JohnDixson; 7th Dec 2012 at 11:39. Reason: Additional thought
In the CH-47C....the Nr changed again for higher MAUW's. Along the way from the A Model....the Engines changed considerably as well. The Early C Models had a very serious flaw in the Power Turbines....that caused catastrophic failures and in almost all cases the loss of the aircraft when shrapnel severed the fuel lines and started massive fires.
She had her problems....but is still the best helicopter I have ever flown. Sadly, I did not get the chance to compare the big Sikorsky aircraft and think the 53E must be a treat to fly.
She had her problems....but is still the best helicopter I have ever flown. Sadly, I did not get the chance to compare the big Sikorsky aircraft and think the 53E must be a treat to fly.
Hi John
Just for clarity, the 14Hz thing is nothing directly to do with this issue. There was a possible flight control instability at 14Hz at very low density altitude - air too thick! As you know, once the loop gain of a control system exceeds unity, any disturbance will result in divergent oscillation. Thicker air = more control power = more loop gain.
In the case of the 225 this could occur with density altitude somewhere below -2000'. In order to get around the problem, the loop gain had to be reduced and this was done by reducing the Nr at low density altitude. So at normal to high density altitude, hover Nr is 103.8%. But this reduces down towards 100% as density altitude decreases. The latest version, which allows flight in density altitude down to -6000', has the Nr going as low as 98%. So operation with Nr below 100% is a relatively new thing, and there seems to be a theory that this could be the problem.
Just for clarity, the 14Hz thing is nothing directly to do with this issue. There was a possible flight control instability at 14Hz at very low density altitude - air too thick! As you know, once the loop gain of a control system exceeds unity, any disturbance will result in divergent oscillation. Thicker air = more control power = more loop gain.
In the case of the 225 this could occur with density altitude somewhere below -2000'. In order to get around the problem, the loop gain had to be reduced and this was done by reducing the Nr at low density altitude. So at normal to high density altitude, hover Nr is 103.8%. But this reduces down towards 100% as density altitude decreases. The latest version, which allows flight in density altitude down to -6000', has the Nr going as low as 98%. So operation with Nr below 100% is a relatively new thing, and there seems to be a theory that this could be the problem.
Last edited by HeliComparator; 7th Dec 2012 at 13:48.
John D:
Your post reminded me of all of the deswirl ducts that fell off of Seahawks (TFOA - arrrrgh!!!). IIRC, that clamp failure got traced to either NSV or ASV that originated in the T-700. (HMU? One of the pumps?) Been a few years, I can't remember the fix.
I do remember that on preflight we used to have some detailed looks at the deswirl duct clamp to make sure it hadn't cocked or loosened. The word on the street and in various tech bulletins was that you could detect an impending clamp failure by inspection between sorties. If it wasn't "sitting right" you'd have maintenance either replace it or confirm its soundness before launch.
As one cannot check stuff inside the gear box in pre flight ... the operator's means of mitigating (if the root cause of this is in the vibration domain) for the 225 is of course more difficult.
Your post reminded me of all of the deswirl ducts that fell off of Seahawks (TFOA - arrrrgh!!!). IIRC, that clamp failure got traced to either NSV or ASV that originated in the T-700. (HMU? One of the pumps?) Been a few years, I can't remember the fix.
I do remember that on preflight we used to have some detailed looks at the deswirl duct clamp to make sure it hadn't cocked or loosened. The word on the street and in various tech bulletins was that you could detect an impending clamp failure by inspection between sorties. If it wasn't "sitting right" you'd have maintenance either replace it or confirm its soundness before launch.
As one cannot check stuff inside the gear box in pre flight ... the operator's means of mitigating (if the root cause of this is in the vibration domain) for the 225 is of course more difficult.
Last edited by Lonewolf_50; 7th Dec 2012 at 14:23.
Resonance
Thanks for your note, HC. I surmised from your previous posts that they were doing something smart like that.
My personal experience with flight control excitation into the various airframe modes was that once you get beyond say 7 hz, the output rolls off significantly, mostly due to the actuators of course*. But my experience also says " never say never ", so if the gearbox shaft issue is one of resonance, it will be interesting to see how it arises.
* But the outcome can be " attention-getting ", and things can get sticky near instantaneously, thus we typically installed a cockpit " kill-switch".
You appear to have good contacts with the troops at EC, and one of the aspects that arrested my attention ( and I may have the story screwed up, so this query may be a non-starter ) is the basis for the continued flying after the first incident. On the one hand, it seemed to be perhaps novel, but then justifiable, use of the HUMS accelerometers to provide a go-no-go decision. On the other side of that procedure, I was wondering just what sort of specimen testing had been done, if any, on the lower section of the shaft. The normal questions: how many specimens, stress levels high enough to generate failure or were they low and all specimens were run-outs etc. Where I am coming from is that assigning a three hour safe life carries a message that one has a very firm handle on the fatigue strength of that piece, and on the repercussions of a failure. Then, after the second occurrence, it seemed as though there were reconsiderations as to the integrity of the remaining drive, following a shaft failure.
It's well to remember that EC will want to investigate, test and substantiate a solution, which they will then present to the authorities, on their own. OEM's learn early on not to get the qualification authorities involved in the redesign/solution process.
I'd be surprised if , by now, they are not certain of the cause, are juggling the implications of alternative solutions, and testing same. Some of the estimates that have been posted regarding get well schedules seem unreasonably pessimistic, realizing that at EC, this is Job One. It would make for a very interesting AHS paper.
My personal experience with flight control excitation into the various airframe modes was that once you get beyond say 7 hz, the output rolls off significantly, mostly due to the actuators of course*. But my experience also says " never say never ", so if the gearbox shaft issue is one of resonance, it will be interesting to see how it arises.
* But the outcome can be " attention-getting ", and things can get sticky near instantaneously, thus we typically installed a cockpit " kill-switch".
You appear to have good contacts with the troops at EC, and one of the aspects that arrested my attention ( and I may have the story screwed up, so this query may be a non-starter ) is the basis for the continued flying after the first incident. On the one hand, it seemed to be perhaps novel, but then justifiable, use of the HUMS accelerometers to provide a go-no-go decision. On the other side of that procedure, I was wondering just what sort of specimen testing had been done, if any, on the lower section of the shaft. The normal questions: how many specimens, stress levels high enough to generate failure or were they low and all specimens were run-outs etc. Where I am coming from is that assigning a three hour safe life carries a message that one has a very firm handle on the fatigue strength of that piece, and on the repercussions of a failure. Then, after the second occurrence, it seemed as though there were reconsiderations as to the integrity of the remaining drive, following a shaft failure.
It's well to remember that EC will want to investigate, test and substantiate a solution, which they will then present to the authorities, on their own. OEM's learn early on not to get the qualification authorities involved in the redesign/solution process.
I'd be surprised if , by now, they are not certain of the cause, are juggling the implications of alternative solutions, and testing same. Some of the estimates that have been posted regarding get well schedules seem unreasonably pessimistic, realizing that at EC, this is Job One. It would make for a very interesting AHS paper.
Last edited by JohnDixson; 7th Dec 2012 at 16:17. Reason: Structure and a side thought
Hi John
One of the new features of the 225 vs the L2 is the response rate of the main servos, bandwidth was over 16Hz. This is quite noticeable if, for example, you knock the cyclic by accident- the response is instant and dramatic! In response to the 14Hz problem they fitted modified servos that had reduced bandwidth, down to a little below 12Hz but of course the concept of bandwidth means that the response is only diminishing, not zero, above 12Hz. It is still sufficiently high at 14Hz to require the other solution ie reducing the Nr at low density altitude.
On the subject of the 3hrs, as I understand it the crack develops perhaps 20 hrs before failure, but is only detectable by HUMS within the last 4 -6 hrs or so when the shaft starts to lose rigidity. The crack progression can of course be determined by the beach marks as discussed in the AAIB bulletin. As far as i know the 3 hrs was practically tested only by the 2 events and so it does not seem to have huge integrity. This of course is one reason why the operators and oil companies decided not to fly crew change flights using this procedure, and CAA followed suite with the ban on public transport over hostile terrain. However if you are operating over landable terrain it is less of a big deal - if the worst happens, you just land and no big deal. So I think it is right that the procedure is there to support certain types of operation - just not N Sea crew change!
Edited to add-just re-read your question and I see I misinterpreted it - you are asking about the procedure after the first event. Again I think it was only based on the observed propagation rate of the event. Not very robust, but presumably adequate to persuade EASA to allow it.
I don't think anyone actually used it - in Bristow we had none of the relatively few shafts affected (those with the new-profile hole) and we waited to take delivery of our new EC225s until they had re-worked shafts without the cavity.
One of the new features of the 225 vs the L2 is the response rate of the main servos, bandwidth was over 16Hz. This is quite noticeable if, for example, you knock the cyclic by accident- the response is instant and dramatic! In response to the 14Hz problem they fitted modified servos that had reduced bandwidth, down to a little below 12Hz but of course the concept of bandwidth means that the response is only diminishing, not zero, above 12Hz. It is still sufficiently high at 14Hz to require the other solution ie reducing the Nr at low density altitude.
On the subject of the 3hrs, as I understand it the crack develops perhaps 20 hrs before failure, but is only detectable by HUMS within the last 4 -6 hrs or so when the shaft starts to lose rigidity. The crack progression can of course be determined by the beach marks as discussed in the AAIB bulletin. As far as i know the 3 hrs was practically tested only by the 2 events and so it does not seem to have huge integrity. This of course is one reason why the operators and oil companies decided not to fly crew change flights using this procedure, and CAA followed suite with the ban on public transport over hostile terrain. However if you are operating over landable terrain it is less of a big deal - if the worst happens, you just land and no big deal. So I think it is right that the procedure is there to support certain types of operation - just not N Sea crew change!
Edited to add-just re-read your question and I see I misinterpreted it - you are asking about the procedure after the first event. Again I think it was only based on the observed propagation rate of the event. Not very robust, but presumably adequate to persuade EASA to allow it.
I don't think anyone actually used it - in Bristow we had none of the relatively few shafts affected (those with the new-profile hole) and we waited to take delivery of our new EC225s until they had re-worked shafts without the cavity.
Last edited by HeliComparator; 7th Dec 2012 at 16:38.
Resonance
Should have been more precise, HC. Sorry. SA, for the non-fly-by-wire controls typically ratchets down the hydraulic feed to the AFCS actuators to 1500 psi or less, and they are the reason for decreased response above about 7 hz ( and of course the typical filtering around the basic aircraft modes ). Do the Eurocopter machines get 14 hz out of their Autopilot actuators? In any case, it would be interesting if any of those inputs make that MR shaft "ring".
Hi John
Maybe we are slightly at cross-purposes here. I am talking about the main hydraulic servo controls in the head. These run at 175bar (~2600 psi) and the instability arises from having the transmission (on which the servos are mounted) flexibly attached to the airframe, whilst the pilot's flight controls are rigidly attached to the airframe.
The 225 has two different types of AFCS actuators in series for each channel (for redundancy) - hydraulic servos in the Autopilot Hydraulics - I think very similar in concept to those fitted to the S61 - running at about half the pressure of the main servos, plus electric SEMAs which are digitally controlled fast electric screw jacks with built-in positioning feedback loops. I am not sure what the bandwidth of these is, but I suppose there is not much point in having main servos that are significantly faster than the AFCS series actuators.
The 225 autopilot is extremely precise but smooth. I think the digital control loops are not a digital replication of an analogue control loop as they were in the 332L2. But even so, it smacks of a high gain control loop to me which, unless bandwidth is high and hence phase shift is low, would result in instability. However there is no hint of instability in the 225's control loops and this may be in part due to the high bandwidth of the various control components thus minimal phase shift.
Waffling a bit there, but the relevant point is that I don't think the problem is any hint of the 14Hz oscillation, rather it is the steps taken to avoid the oscillation that might be having unintended consequences.
Maybe we are slightly at cross-purposes here. I am talking about the main hydraulic servo controls in the head. These run at 175bar (~2600 psi) and the instability arises from having the transmission (on which the servos are mounted) flexibly attached to the airframe, whilst the pilot's flight controls are rigidly attached to the airframe.
The 225 has two different types of AFCS actuators in series for each channel (for redundancy) - hydraulic servos in the Autopilot Hydraulics - I think very similar in concept to those fitted to the S61 - running at about half the pressure of the main servos, plus electric SEMAs which are digitally controlled fast electric screw jacks with built-in positioning feedback loops. I am not sure what the bandwidth of these is, but I suppose there is not much point in having main servos that are significantly faster than the AFCS series actuators.
The 225 autopilot is extremely precise but smooth. I think the digital control loops are not a digital replication of an analogue control loop as they were in the 332L2. But even so, it smacks of a high gain control loop to me which, unless bandwidth is high and hence phase shift is low, would result in instability. However there is no hint of instability in the 225's control loops and this may be in part due to the high bandwidth of the various control components thus minimal phase shift.
Waffling a bit there, but the relevant point is that I don't think the problem is any hint of the 14Hz oscillation, rather it is the steps taken to avoid the oscillation that might be having unintended consequences.
Servos
HC, I think your description is what I had envisioned, up to the difference in servo mounting, which I knew at one time but completely forgot until you mentioned it. Thanks for clearing it up. All of these considerations will be history soon, as the industry transitions to FBW and all of the mechanical mixing, boost actuators, Autopilot/AFCS actuators and related claptrap vanish. I think I can hear the Maintenance and the Safety Community saying a loud " Adios ".
Eeek - I have got used to no rods/ cables / bits of string between me and the engines, not sure I am ready for same between me and the swash plate! Is there currently a civil certification standard for FBW in helicopters?
Join Date: Sep 2007
Location: USA
Posts: 601
Likes: 0
Received 0 Likes
on
0 Posts
Riff Raff- Any thoughts on the seemingly random start/finish points between individual shafts of the weld in relation to the 2 large holes in the lower part of the shaft?
Sorry for the tardy response, but I only browse this thread once every couple of days.
As for the two larger, diametrically opposed holes in the conical section of the shaft below the weld joint, I could not say for certain what their purpose is. The only thing I could speculate is that they are windows to allow the spray from a lube jet to deposit some oil onto the inner wall of the spinning bevel gear shaft. The oil deposited on the inner shaft wall would work its way up to the spline connecting the bevel shaft to the sun gear of the first epicyclic stage.
If you look at the nice cross section drawing someone posted a while back, you'll note that there is also an o-ring seal just above the spline and a carefully located vent hole thru the sun gear wall just inboard/below the seal. The purpose for this arrangement is so that the spline joint operates submerged in lube oil, and that the oil flowing from one end of the spline to the other flushes out any microscopic metallic debris generated by rubbing between the spline teeth. This arrangement is common practice with highly loaded spline joints in aircraft gearboxes, and it is done to minimize fretting wear in the spline tooth flanks.
Sorry for the long winded explanation. Hope you found it useful.
riff_raff
Join Date: Feb 2009
Location: A nice place
Posts: 160
Likes: 0
Received 0 Likes
on
0 Posts
riff_raf
Thanks for that explanation of the 2 large holes, when I think back I do remember an oil jet adjacent to them.
My main query was with the start/finish point of the weld. On some shaft pictures it’s directly above one of the holes, on others it’s not. From a stress strain point of view would it not be better to have this point directly above one of the large holes or exactly 90 degrees to it?
Thanks for that explanation of the 2 large holes, when I think back I do remember an oil jet adjacent to them.
My main query was with the start/finish point of the weld. On some shaft pictures it’s directly above one of the holes, on others it’s not. From a stress strain point of view would it not be better to have this point directly above one of the large holes or exactly 90 degrees to it?