Big Crash at Reno
Been looking for information on whether trim tabs should be mass balanced,but could not find anything conclusive; however if one `googles `SAWE`Paper3020`,by Space Electronics,there is good stuff about mass balance and the effects of `flutter`.
I have had aileron flutter on an aircraft,coincidentally an 80% scale Mustang;on completion of some testing for `G` limits and `rolling `g`pull-outs.It started at about 160kts,as I was doing a little `flypast` for the owner at an airfield at about 3-400ft.It sounded like a `Gatling` gun firing,and looking at the left wing I could not see it,the right was quite steady.Vibration was also quite severe,and I had to climb,reduce power almost to the stall to stop it.After landing, the aileron bellcrank and pushrod were found to be sheared .Also,although made to specification ,there was insufficient mass-balance on the ailerons.The wings were wooden,and stressed to +-12 g,but I`m sure if they had been metal,it may have been different. The flutter started instantly,and at a speed that I had already tested,but that had obviously weakened the internals.
So, what Lyman and M-B are probably seeing on the video is a series of snapshots of the tailplane vibrating,but the frequency may well be above the camera speed.I also agree that there is a slight distortion/alignment there as well.
If one looks at the photo on P9,#162,to me it appears that there is slightly more `up` elevator than the left one.I`m only gauging that from the shadows on the mass-balances.Now ,if one is putting UP elevator on one side ,and the tab on the opposite side is forcing Down,then there will be a difference in tension on both sets of elevator cables,and on the bolts holding both torque tubes together.Could lead to cable stretch,possibly sufficient to disengage from the pulleys and guides,or cause the bolts to `fret` or loosen.
`Voodoo` appears to have that happen,but there again does not appear to be a comprehensive report about it`s failure either.
I have had aileron flutter on an aircraft,coincidentally an 80% scale Mustang;on completion of some testing for `G` limits and `rolling `g`pull-outs.It started at about 160kts,as I was doing a little `flypast` for the owner at an airfield at about 3-400ft.It sounded like a `Gatling` gun firing,and looking at the left wing I could not see it,the right was quite steady.Vibration was also quite severe,and I had to climb,reduce power almost to the stall to stop it.After landing, the aileron bellcrank and pushrod were found to be sheared .Also,although made to specification ,there was insufficient mass-balance on the ailerons.The wings were wooden,and stressed to +-12 g,but I`m sure if they had been metal,it may have been different. The flutter started instantly,and at a speed that I had already tested,but that had obviously weakened the internals.
So, what Lyman and M-B are probably seeing on the video is a series of snapshots of the tailplane vibrating,but the frequency may well be above the camera speed.I also agree that there is a slight distortion/alignment there as well.
If one looks at the photo on P9,#162,to me it appears that there is slightly more `up` elevator than the left one.I`m only gauging that from the shadows on the mass-balances.Now ,if one is putting UP elevator on one side ,and the tab on the opposite side is forcing Down,then there will be a difference in tension on both sets of elevator cables,and on the bolts holding both torque tubes together.Could lead to cable stretch,possibly sufficient to disengage from the pulleys and guides,or cause the bolts to `fret` or loosen.
`Voodoo` appears to have that happen,but there again does not appear to be a comprehensive report about it`s failure either.
Join Date: Jun 2009
Location: ATL
Age: 67
Posts: 131
Likes: 0
Received 0 Likes
on
0 Posts
If one looks at the photo on P9,#162,to me it appears that there is slightly more `up` elevator than the left one.I`m only gauging that from the shadows on the mass-balances.Now ,if one is putting UP elevator on one side ,and the tab on the opposite side is forcing Down,then there will be a difference in tension on both sets of elevator cables,and on the bolts holding both torque tubes together.Could lead to cable stretch,possibly sufficient to disengage from the pulleys and guides,or cause the bolts to `fret` or loosen.
The plunging, or first flutter mode is when the tips go up and down and the node, the part that doesn't move, is the centerline, as shown in the first part of the video. The counterweights and elevator deflection can drive this;
This would be different than the 2nd mode;
Hard to tell, but this excited the 3rd mode;
God knows what this is, turning a solid to a fluid maybe, and back again'
All of these are different than simple control surface flutter that sycamore experienced with the aileron, and up till now, everyone thought the trim tab was doing. Since the HS went into 1st mode flutter, if the video can be believed, then this could also be explained by an elevator getting loose from the torque tube, the fuselage snapping back from ti's twist, and getting things going. Bending the tab spanwise in this mode would explain why the inner part of the tab separated leaving the outer section attached. If the tab rod failed, the whole tab would flutter and rip off in one piece.
If the right elevator torque tube attachment failed, and the pilot was pulling 30 pounds of so, the left elevator would be over deflected and the right elevator up load would decrease causing the left roll without a pitch change. It would take a few 1st mode oscillations for the tab to fail, which would happen sooner if it was stiffened after Voodoo, and would happen very fast from the maybe 50 to a 100 hertz. Don't know if he camera can capture 50 hertz, so we might be looking at an illusion. But the theory ties everything in so far; the roll, the fuselage wrinkles, the tab breaking in half, the g occuring after the roll correction... .
Found the hinge moment report. The elevator bulge is standard for the D model
and there's still a slope change with Mach number on hinge moments.
http://naca.central.cranfield.ac.uk/...ca-tn-1302.pdf
Join Date: Jul 2009
Location: Not far from a big Lake
Age: 81
Posts: 1,454
Likes: 0
Received 0 Likes
on
0 Posts
Nice job with the flutter videos Mr. Cub. 
You have posted almost all my favorite ones and posted them faster than myself, although Fred Haise's tail flutter video posted earlier on this thread is my all time favorite.
You have posted almost all my favorite ones and posted them faster than myself, although Fred Haise's tail flutter video posted earlier on this thread is my all time favorite.
CC/Anyone, any idea about the tailplane section; was it the earlier,or lower profile.or perhaps an H-model(p51H),which I`m assuming would be somewhat different. Do you know if there are any later reports which would deal with the later models,either Mustang,Bearcat,Jug..?
Join Date: Jun 2009
Location: ATL
Age: 67
Posts: 131
Likes: 0
Received 0 Likes
on
0 Posts
Do you know if there are any later reports which would deal with the later models,either Mustang,Bearcat,Jug..?
NASA Technical Reports Server
The sampling rate of the video is 30 frames per second. Thinking the HS isn't showing flutter, but the left HS deflected down could be a simple overload which would arise if the right elevator let go and the left elevator was over deflected as a result. This would still put all kinds of stress on the tab and the results would be the same.
The elevator was the weak point in early dive testing, and with the q's at Reno, the elevator load cycles during the race, the loads increasing on the right elevator due to the tab on the left one, and the higher horsepower, the elevator just gave up in the high g turn.
Join Date: Aug 2011
Location: Grassy Valley
Posts: 2,074
Likes: 0
Received 0 Likes
on
0 Posts
The wrinkling of the skin is consistent with the stress on the fuselage, whether the image is part of the failure sequence or not, imo. Downforce and displacement reduces the "area" of the substrate, and the skin, unable to "shrink", wrinkles instead.
I have consistently been impressed with the life cycle of the skin on these birds. It is flexible, and the frame is definitely flexible. This suggests fatigue, and added drag in the circuit. I think there are areas of the airframe that would perform far better wrapped in FRP. It has greater elasticity, and the weight penalty might payoff in better times.
Control would be better, the contour and drag would be more consistent, and airflow would be more predictable.
In the turns, (all left, of course), the drag of the left Trim Tab on the elevator (with the PITCH UP) would be an ad hoc spoiler, and bleeding some energy with an already "deployed" fence, is an elegant use of otherwise wasted drag. It would be interesting to know if these issues resulted in some additional strengthening in this control.
I have the picture that with trim defeated on the right, it is somewhat like driving an automobile with only one front wheel power assisted.
I have consistently been impressed with the life cycle of the skin on these birds. It is flexible, and the frame is definitely flexible. This suggests fatigue, and added drag in the circuit. I think there are areas of the airframe that would perform far better wrapped in FRP. It has greater elasticity, and the weight penalty might payoff in better times.
Control would be better, the contour and drag would be more consistent, and airflow would be more predictable.
In the turns, (all left, of course), the drag of the left Trim Tab on the elevator (with the PITCH UP) would be an ad hoc spoiler, and bleeding some energy with an already "deployed" fence, is an elegant use of otherwise wasted drag. It would be interesting to know if these issues resulted in some additional strengthening in this control.
I have the picture that with trim defeated on the right, it is somewhat like driving an automobile with only one front wheel power assisted.
Join Date: Jun 2009
Location: ATL
Age: 67
Posts: 131
Likes: 0
Received 0 Likes
on
0 Posts
The wrinkling of the skin is consistent with the stress on the fuselage, whether the image is part of the failure sequence or not, imo.
Heard talk of the tailwheel be forced down against hydraulic pressure in the actuator. The maintenance manual shows an uplock and uplocks wouldn't be designed to take torsional loads from the fuselage excessively twisting.
There's also what appears to be denting on the turtle deck aboce the, '1', which would indicate the turtle deck in this area was stressed beyond the elasticity of the aluminum from the fuselage overly twisting.
http://www.avweb.com/newspics/reno-c...rien_large.jpg
Join Date: Jul 2009
Location: Not far from a big Lake
Age: 81
Posts: 1,454
Likes: 0
Received 0 Likes
on
0 Posts
Notice how the remainder of the trim tab appears indistinct in the photograph that CC has just posted the link to. I wouldn't be surprised if it if fluttering at a high rate and causing that visual effect. Of course anything fluttering at that point in time was not relevant to the final outcome.
Join Date: Jul 2009
Location: Not far from a big Lake
Age: 81
Posts: 1,454
Likes: 0
Received 0 Likes
on
0 Posts
We have found two versions of the telemetry data informally relayed to us. One version indicates that the engine lost power due to fuel interruption from the high g level. Depending on the geometry of the fuel system in the Ghost, it may be possible to estimate peak g required to cause a fuel interruption.
Interruption of fuel can occur due to pump inlet conditions or pump outlet/engine inlet conditions.
Was the fuel pump on the Ghost a centrifugal pump or a positive displacement pump?
Was it tank mounted or engine mounted?
Is the fuel source a wing tank or some other tank?
Is there more than one pump involved?
Interruption of fuel can occur due to pump inlet conditions or pump outlet/engine inlet conditions.
Was the fuel pump on the Ghost a centrifugal pump or a positive displacement pump?
Was it tank mounted or engine mounted?
Is the fuel source a wing tank or some other tank?
Is there more than one pump involved?
Join Date: Sep 2011
Location: Nevada
Age: 57
Posts: 18
Likes: 0
Received 0 Likes
on
0 Posts
I also notice in the slo-mo vid that the TT virtually vanishes for an instant before its inboard mount breaks away and then trails the elevator before departing the A/C. Please excuse my ignorance, but are the TTs cable operated, and if so how do they operate with one actuating tab above and one below the elevator surface? Does that mean when one is being pulled on the other is being pushed by the cables (in normal operation)? Also that dent over the number one, is evident in the build photos from 2009, I think its just a characteristic of the modification done to GGs spine.
...2009 Build photos
WarbirdAeroPress.com
....2009 Run-up/ck flt
WarbirdAeroPress.com
I am certainly not questioning the maintainers or the maintenance on this airplane but.......3 of what appear to be 1/4 bolts that attach the inboard elevator hub to its bellcrank? I dont know but seems mighty fragile if one were to consider the torsional stresses placed by pinning one of the TTs stationary. I personally am seeing the logic in the thought that a disconnect between the two elevators may have been the initial failure. Can somebody explain the twitch of GG in the video after passing Rare Bear (approx 7 min into the vid with GG above and in front of RB) (possibly rounding home or #1 pylon)?
Just some interesting info:
Tech INFO - Aircorps Aviation
...2009 Build photos
WarbirdAeroPress.com
....2009 Run-up/ck flt
WarbirdAeroPress.com
I am certainly not questioning the maintainers or the maintenance on this airplane but.......3 of what appear to be 1/4 bolts that attach the inboard elevator hub to its bellcrank? I dont know but seems mighty fragile if one were to consider the torsional stresses placed by pinning one of the TTs stationary. I personally am seeing the logic in the thought that a disconnect between the two elevators may have been the initial failure. Can somebody explain the twitch of GG in the video after passing Rare Bear (approx 7 min into the vid with GG above and in front of RB) (possibly rounding home or #1 pylon)?
Just some interesting info:
Tech INFO - Aircorps Aviation
Join Date: Jun 2009
Location: ATL
Age: 67
Posts: 131
Likes: 0
Received 0 Likes
on
0 Posts
Is there more than one pump involved?
but are the TTs cable operated
Also that dent over the number one, is evident in the build photos from 2009, I think its just a characteristic of the modification done to GGs spine.
Can somebody explain the twitch of GG in the video after passing Rare Bear
Join Date: Aug 2011
Location: Grassy Valley
Posts: 2,074
Likes: 0
Received 0 Likes
on
0 Posts
"If you can fly in the mountains, you can fly anywhere".......
Reno/Tahoe is an....interesting environment, for the pilot, year round.
Stead is in a "bowl", a well stirred one.
I lean toward elevator failure, as well. The right. The left elevator has stress attenuation (at the join/bellcrank) via the tab, the right takes more impact. The difference in load between the two is a chronic source of strain on the mounting cups at the bellcrank. I think the right elevator may have sheared, scrubbed some of the throughbolts that join it to the bellcrank. When it relaxed out of the airstream, the left elevator went stress critical, v/v the Trim Tab. Depending on the severity of the rotational failure of the Right elevator in its mounting, the LHS took on the load we see in Clipped Cub's image. The degree of torsional stress displacement of the tail would have unrigged the Tail feathers out of alignment with the wings and forward fuselage.
Permitted an extrapolation, this unrig makes the aft fuselage an emphatic right rudder, with sufficient input to recover the left roll, and roll the a/c back right. The fuselage recoils, and the a/c 'settles' into its ballistic path upward. It also recovers the Pitch Up, as the aft would be drooped, and deflected right.
This instant unrig/rerig suggests to me the loud noise heard at the time. The a/c at this speed is a drum, a skinned barrel, no different.
Reno/Tahoe is an....interesting environment, for the pilot, year round.
Stead is in a "bowl", a well stirred one.
I lean toward elevator failure, as well. The right. The left elevator has stress attenuation (at the join/bellcrank) via the tab, the right takes more impact. The difference in load between the two is a chronic source of strain on the mounting cups at the bellcrank. I think the right elevator may have sheared, scrubbed some of the throughbolts that join it to the bellcrank. When it relaxed out of the airstream, the left elevator went stress critical, v/v the Trim Tab. Depending on the severity of the rotational failure of the Right elevator in its mounting, the LHS took on the load we see in Clipped Cub's image. The degree of torsional stress displacement of the tail would have unrigged the Tail feathers out of alignment with the wings and forward fuselage.
Permitted an extrapolation, this unrig makes the aft fuselage an emphatic right rudder, with sufficient input to recover the left roll, and roll the a/c back right. The fuselage recoils, and the a/c 'settles' into its ballistic path upward. It also recovers the Pitch Up, as the aft would be drooped, and deflected right.
This instant unrig/rerig suggests to me the loud noise heard at the time. The a/c at this speed is a drum, a skinned barrel, no different.
Join Date: Sep 2011
Location: Nevada
Age: 57
Posts: 18
Likes: 0
Received 0 Likes
on
0 Posts
Wow I had the same though regarding Voodoo.....I wonder if there was much of a roll in his pitch up and accent. Regarding the stock function of the TTs so for ND trimming the left would be pulled and the right would be pushed into the airstream.
Join Date: Aug 2011
Location: Grassy Valley
Posts: 2,074
Likes: 0
Received 0 Likes
on
0 Posts
Don't put too much into the parked image of VooDoo, the right elevator is not in alignment with the left, this does not mean what you see is what presented at the moment of failure. It would be interesting to read the record of VooDoo.
I don't see the Trim Tab as failing first, it is stressed by position, not load. It is comfortable in the configuration at failure, whilst the Elevators are not.
IOW, the tab is not experiencing any untoward loading. For it to fail, it needs a critical stress point to be exceeded. The failure of the right ELEVATOR explains this critical load nicely. I think Clipped Cub is on to this?
I don't see the Trim Tab as failing first, it is stressed by position, not load. It is comfortable in the configuration at failure, whilst the Elevators are not.
IOW, the tab is not experiencing any untoward loading. For it to fail, it needs a critical stress point to be exceeded. The failure of the right ELEVATOR explains this critical load nicely. I think Clipped Cub is on to this?
Join Date: Sep 2011
Location: Nevada
Age: 57
Posts: 18
Likes: 0
Received 0 Likes
on
0 Posts
Thanks CC for your response and explanations.
EDIT: I now see I had the stock operation of the elevator trim tabs reversed, my bad..... Also, upon review of some cockpit photos I believe GGs trim were possibly electric, both pitch and roll.....switches on the left side of the cockpit.
EDIT: I now see I had the stock operation of the elevator trim tabs reversed, my bad..... Also, upon review of some cockpit photos I believe GGs trim were possibly electric, both pitch and roll.....switches on the left side of the cockpit.
Last edited by xmh53wrench; 14th Oct 2011 at 04:09.
Join Date: Jul 2009
Location: Not far from a big Lake
Age: 81
Posts: 1,454
Likes: 0
Received 0 Likes
on
0 Posts
CC
I don't think so. As the float gets heavier in the high g environment, the fuel it is displacing should get heavier also. The two should balance each other exactly out.
Stock P-51 fuel pressure appears to be in the 12-16 psi range with 19 psi max.
This is not such a high pressure that it could not be over-ridden at very high g.
The Stock P-51 used what appears to be a centrifugal fuel pump on the inboard aft corners of the wing fuel tanks (identified as the booster pump I believe.)
Sample calculations:
For 40 inches of height change (this is a SWAG* number), the pressure differential in a column of gasoline is about 1.04 psi.
At 15 g, the pressure differential in that same 40 inch column of gasoline is 15.6 psi. These numbers are the same orders of magnitude as the fuel pressure and indicate that interruption of fuel delivery may be possible at high g.
The Ghosts fuel arrangement was not stock, and the above calculations may not be relevant to the Ghost.
*For the uninitiated, SWAG = Scientific Wild Ass Guess.
Has a carb with a float. Float could have been forced down under excessive g creating flooding?
Stock P-51 fuel pressure appears to be in the 12-16 psi range with 19 psi max.
This is not such a high pressure that it could not be over-ridden at very high g.
The Stock P-51 used what appears to be a centrifugal fuel pump on the inboard aft corners of the wing fuel tanks (identified as the booster pump I believe.)
Sample calculations:
For 40 inches of height change (this is a SWAG* number), the pressure differential in a column of gasoline is about 1.04 psi.
At 15 g, the pressure differential in that same 40 inch column of gasoline is 15.6 psi. These numbers are the same orders of magnitude as the fuel pressure and indicate that interruption of fuel delivery may be possible at high g.
The Ghosts fuel arrangement was not stock, and the above calculations may not be relevant to the Ghost.
*For the uninitiated, SWAG = Scientific Wild Ass Guess.
Last edited by Machinbird; 14th Oct 2011 at 04:42. Reason: Add Picture
Wrench, if you go to those Tech Notes,and the `flight controls`,you can see the trim layout.Rudder and elevators used the same trim gearboxes.Also in the `tailwheel` section it shows the layout of the elevator/tailwheel /rudder steering mechanism,and locks.
Join Date: Jun 2008
Location: Nouvion
Posts: 42
Likes: 0
Received 0 Likes
on
0 Posts
Just some thoughts.
How do you know the GG was running a float carburettor ? Merlins were fitted with normal carburettors, pressure carburettors and injection carburettors. With all the modifications made to this a/c and its engine, I would not take for granted that anything was standard on this engine.
Re the engine cut(if it had a float carb)......if it was due to a rich cut, in all likelyhood you would have seen black smoke from the exhausts due to the over rich mixture....don't see that on any video or still. If it was running a stock fuel system , 12-16psi doesn't seem a lot of pressure to lift fuel to the carb inlet, especially with the reported extremely high G level experienced. So the engine cut could more likely be due to a brief interuption of fuel supply. Again I would be very sceptical that this a/c ran a stock pump/fuel system...maybe it does. With the amount of HP this engine was making , it would require much higher fuel flows and pressures surely ?
What makes you so sure the RH elevator or its torque tube failed ? If the RH tab was indeed fixed in the streamlined position , then the LH elevator/tab combination was imparting a force through the LH elevator structure to the RH elevator via the torque tubes and bellcrank. You are forgetting Newtons 3rd law of action and reaction. The forces are equal throughout the system but will be in opposite directions. When it fails, it will be at the weakest point, thats for the NTSB to find out !
Re the tail wheel extension..........it seems it was designed to extend under large G which the pilot was required to pull in the event of tailwheel extension failure. Not supprising it extended here in this case. It has also been mentioned that the doors also came off and this was an indication of structural damage to the tail due to flutter/twist/or the tab failure. The a/c was doing close to 500mph, I don't know of any a/c with a gear limiting speed in excess of 500mph ??? If you extend any gear at that speed, the doors are surely gonna come off !!
How do you know the GG was running a float carburettor ? Merlins were fitted with normal carburettors, pressure carburettors and injection carburettors. With all the modifications made to this a/c and its engine, I would not take for granted that anything was standard on this engine.
Re the engine cut(if it had a float carb)......if it was due to a rich cut, in all likelyhood you would have seen black smoke from the exhausts due to the over rich mixture....don't see that on any video or still. If it was running a stock fuel system , 12-16psi doesn't seem a lot of pressure to lift fuel to the carb inlet, especially with the reported extremely high G level experienced. So the engine cut could more likely be due to a brief interuption of fuel supply. Again I would be very sceptical that this a/c ran a stock pump/fuel system...maybe it does. With the amount of HP this engine was making , it would require much higher fuel flows and pressures surely ?
What makes you so sure the RH elevator or its torque tube failed ? If the RH tab was indeed fixed in the streamlined position , then the LH elevator/tab combination was imparting a force through the LH elevator structure to the RH elevator via the torque tubes and bellcrank. You are forgetting Newtons 3rd law of action and reaction. The forces are equal throughout the system but will be in opposite directions. When it fails, it will be at the weakest point, thats for the NTSB to find out !
Re the tail wheel extension..........it seems it was designed to extend under large G which the pilot was required to pull in the event of tailwheel extension failure. Not supprising it extended here in this case. It has also been mentioned that the doors also came off and this was an indication of structural damage to the tail due to flutter/twist/or the tab failure. The a/c was doing close to 500mph, I don't know of any a/c with a gear limiting speed in excess of 500mph ??? If you extend any gear at that speed, the doors are surely gonna come off !!