Amazing Fin Flutter vid!

The Rocketry Forum

Help Support The Rocketry Forum:

rstaff3

Oddroc-eteer
Joined
Jan 18, 2009
Messages
11,763
Reaction score
22
That was amazing. I wouldn't have thought the fins could flutter like that and survive.
 

cls

Well-Known Member
Joined
Jan 21, 2009
Messages
2,240
Reaction score
3
the video was discussed at length on several mailing lists. the consensus is the apparent flutter is (probably) mostly an MPEG encoding artifact; probably also a video scanning artifact (top to bottom); there was some flutter but not as bad as shown in the movie. (heck from the ground I can see my Calisto's 0.063 G10 fins fluttering on just an I161.)

it's a neat video though!!

bottom line: here is clear evidence of unsuitability of the material, for my anti-G10 campaign. :)
 

DynaSoar

Well-Known Member
Joined
Mar 15, 2004
Messages
3,022
Reaction score
0
Originally posted by cls
the video was discussed at length on several mailing lists. the consensus is the apparent flutter is (probably) mostly an MPEG encoding artifact; probably also a video scanning artifact (top to bottom); there was some flutter but not as bad as shown in the movie. (heck from the ground I can see my Calisto's 0.063 G10 fins fluttering on just an I161.)

it's a neat video though!!

bottom line: here is clear evidence of unsuitability of the material, for my anti-G10 campaign. :)
If it were a scanning artifact, there would be:
(1) no instance of 3 curve peaks in the same fin; there are,
(2) every such artifact in one fin having an artifact in the other precisely horizontally across the screen from it; there is not -- there are sharp curves that are not precisely horizontal from each other,

Nobody has:
(1) calculated the speed at which it happened (it was Mach+ according to the flight data)
(2) calculated the Mach wave overpressure converted to PSI
(3) examined rocketmaterials.com to see that 1/8" G10 can withstand 28 +/- 1 kPSI
(4) tested a fresh piece of G10 of the same type as the fins in a flexible vice (even pressure across the entire piece, just as in an overpressure situation) to determine the minimum radius it can withstand, and recorded the amount of internal microfracture created by various amounts of bend
(5) scraped the paint off the fin and looked at the microfractures under a microscope to determine if they reflect the amount of bend as determined in #4 and seen in the video.
(6) quoted the scan rate of the camera used to determine if scan artifact is even a possibility.

Whether it can occur and did occur are empirical questions, requiring data to answer. So far everyone is trying to develop answers from theory. That's rocket guesswork, not rocket science. People are apparently far more interested in showing what they "know" than in figuring out what happened. That's a shame because this is a very good opportunity to extract a large chunk of very important data.

G10 is clearly very flexible. It's also very clearly capable of taking a lot of punishment without breaking. It's so tough that adding carbon fiber laminate doesn't change its strength significantly.
 

cls

Well-Known Member
Joined
Jan 21, 2009
Messages
2,240
Reaction score
3
DynaSoar, are you on the aeropac mailing list? people did some of those things and it was pretty well determined that G10 can't flex so much (modulus too high - heck put some in a vice and try it!) so must be a visual artifact. someone with 10 years experience with MPEG encoders posted his explanation of the visual artifact and several other similar episodes he's seen.

you're right, though, more data is definitely needed!
 

bobkrech

Well-Known Member
Joined
Jan 20, 2009
Messages
8,352
Reaction score
39
DynaSoar, are you on the aeropac mailing list? people did some of those things and it was pretty well determined that G10 can't flex so much (modulus too high - heck put some in a vice and try it!) so must be a visual artifact. someone with 10 years experience with MPEG encoders posted his explanation of the visual artifact and several other similar episodes he's seen.
Cliff and Dynasoar

I'm not on the list, but I was asked for my opinion on the video based on my professional experience which is posted on the webpage referenced.

The overwhelming evidence is that it is real fin-flutter and not an MPEG artifact. I wanted to believe that it was a transonic aerodynamic fluctuation in the air density causing a mirage effect, but the magnitude of the motion rules this out.

I looked at the flight data and observed that the flutter starts and stop abruptly at M=+0.90 and that the flight had a maximum M=1.04 velocity. It not possible to do quantative calculations without measuring the mechanical properties of the fins and the acoustic energy spectrum of the motor, but we can say qualitaively the flutter is a result of a number of thing, a coupling of the eddy turbulance in the transonic flow region, the flexible construction of the fins, enhanced aerodynamic coupling due to the non-zero angle of attack resulting fronm non-parallel installation of the fins as evidenced by the second stage rotation, and the coupling of the random vibration energy of combustion instabilities in the rocket motor with the natural vibrational frequency of the fins.

I showed the video to a co-worker who is an aerospace design year with 20 year of experience in spacecraft design and testing and he said that he had observed this kind of motion in fiberglass support structures within spacecraft undergoing randon vibration testing on a shake table. In fact he spent 1 1/2years analyzing the problem and finding an engineering solution to it. I can't go into details any futher.

The natural vibration frequency is proportional to the thickness of the fin, and the amplitude of the flutter is inversely proportional to the fourth power of the thickness for a given excitaton energy. A small increase in the thickness makes a huge difference in the flutter amplitude. A 50% increase results in a factor of 5 reduction in amplitude, and a 100% increase in thickness results in an astounding 16 fold reduction in flutter amplitude.

The fins on the rocket are extremely flexible. Look at the post flight video. I have a 1' x 2' x 1/8" sheet of G-10 and I can't get it to bend like those fins. Either they are thinner, or they were substantially weakened in the flutter incident.

The stock PML QL 3K has 1/16" G-10 fins which will flex easily. For the motor combination in the flight, PML recommends using 3/32" G-10 with a layer of cloth over it. That would make the thickness ~1/8". I think the thickness is actually 3/32" rather than the 3/16" mentioned on the webpage. This would be consistent with an epoxy rich 6 oz. glass layer over the standard 1/16" fin.

A consenus opinion is irrelevant in this instance. From what I have seen on the web, the HPR community has experienced a large number of shreds that are blamed on a number of possible reasons, but very few people have actually analyzed the failure modes from an engineering standpoint. Just because a majority of the rocket community has not observed this phenomon before doesn't mean it doesn't exit.

If Peter had not done such a good job glassing the fins to the airframe, the rocket would have shredded, and without an on-board video camera, the cause of a shred at 5000 ft would be pure speculation. Fin-flutter is real and probably accounts for a majority of HPR shreds not caused by a CATO.

Bob Krech
 

Mike

Well-Known Member
Joined
Jun 3, 2011
Messages
1,330
Reaction score
0
Wow, these are the kind of discussions I like....proper rocket science!

But, could someone explain to me my the MPEG artifacts only occured during that section of flight? Were the fins suffering a small amount of flutter that was amplified by the camera?
 

lalligood

Well-Known Member
Joined
Jan 21, 2009
Messages
1,985
Reaction score
0
Ultimately, this sounds like it is a reproducable occurence/effect. So why not fly it again and make sure? :cool: :D
 

bobkrech

Well-Known Member
Joined
Jan 20, 2009
Messages
8,352
Reaction score
39
Mike

Your DVD's are encoded in MPEG-2 format. The MPEG-4 format simply compresses the audio and the video together, rather than separately as in MPEG-2. No one complains about the video quality in the DVD's, so I believe the MPEG artifact argument is hogwash. If there were an issue with the MPEG conversion algorithm as implemeted in the Aiptek DV4500 then I believe you would see it on the other moving features in the video such as the crowd and ground details and you don't. The flutter starts and stops at M=0.90 . If it were an artifact, I don't believe the visual distortions would be symmetric.

https://electronics.howstuffworks.com/question596.htm

https://www.m4if.org/

https://www.aiptek.com/Merchant2/me...&Product_Code=R-PKDV4&Category_Code=DC1#specs

To add to my previous post.

While Young's modulus in fiber glass is high, it still bends. In the vibration tests I mentioned above, the test object passed a standard swept sine wave vibration test with flying colors. It only failed in a random vibration test. The short answer for that is the damping coeffiecient in fiberglass is power dependent and is reduced as the total power is increased which eventually results in a severely underdamped situation giving rise to the large displacement amplitudes observed in the video.

Bob Krech
 

powderburner

Well-Known Member
Joined
Jan 18, 2009
Messages
7,399
Reaction score
6
Originally posted by bobkrech
A small increase in the thickness makes a huge difference in the flutter amplitude. A 50% increase results in a factor of 5 reduction in amplitude, and a 100% increase in thickness results in an astounding 16 fold reduction in flutter amplitude.

******

Bob is completely correct here. Even when constructed of solid, homogeneous materials, increased thickness pays off quickly in bending stiffness. When constructed as a structural plate-skin and with intermediate supporting substructure (ribs and spars), increased thickness, bending stiffness increases with the square of effective bending depth.
 

UhClem

Well-Known Member
Joined
Feb 6, 2009
Messages
1,893
Reaction score
325
Good ol' NACA studied fin flutter long ago and published a Technical Report on the subject (#4197). The Info Central section of Rocketry Online has a good discussion of fin flutter based largly on this document.

I have been amazed at the number of people who have tried to explain away this obvious case of fin flutter as video artifacts. I am not surprised because "the data must be wrong" is something that I have heard before. What do you want, strain gauge data? I am no expert (like Bob Krech and friends) but it is obvious even to me that this is a severe case of flutter.

A better choice of reinforcing material for these fins would have been carbon fiber cloth. Carbon fiber is significant;y stiffer (about 7 times) than fiberglass.
 

bobkrech

Well-Known Member
Joined
Jan 20, 2009
Messages
8,352
Reaction score
39
Thanks for chiming in powderburner and UhClem

The report David referred to NACA TN 4197 can be downloaded from

https://naca.larc.nasa.gov/digidoc/report/tn/97/NACA-TN-4197.PDF

Aerodynamisists like to reduce problems to non-dimensional parameters which can get confusing. The algebra in the report is there in gory detail, but pictures and graphs are more illustrative.

In figure 1, the bending modes are shown pictorially. Look at them and look at the video. A pretty good comparison.

Figure 2 shows that flutter occurs in certain speed ranges. A fin may flutter over a cerain speed range, but may be well behaved both above and below it. The assumption on being well behaved above the flutter speed depends on the fin surviving the flutter when it occurs.

Figure 3 idicates the stiffness vs material, and shows that the bending modulus for a solid fin is proportional to the thickness cubed. (That coupled to th fact that the vibrational frequency is inversely proportional to the mass, and thickness leads to my statement that the maximum observed amplitude for a give acceleration is inversely proportional to thickness to the 4th power.)

Have fun reading it.

Bob Krech
 

kelltym88

Well-Known Member
Joined
Jan 18, 2009
Messages
3,910
Reaction score
18
Is it possible that if the fins were not thick enough,(I think only 1/16" inch thick), heat from the motor could have radiated to the fins and thus broke down their composition just enough to make the fins flutter? As the motor burns out the heat decrease and the fins go back to their original shape.
 

cls

Well-Known Member
Joined
Jan 21, 2009
Messages
2,240
Reaction score
3
wow, more light shed on this ... thanks guys! interesting reading.
 

hokkyokusei

Well-Known Member
Joined
Apr 16, 2012
Messages
2,908
Reaction score
3
Originally posted by cls
the video was discussed at length on several mailing lists. the consensus is the apparent flutter is (probably) mostly an MPEG encoding artifact; probably also a video scanning artifact (top to bottom); there was some flutter but not as bad as shown in the movie. (heck from the ground I can see my Calisto's 0.063 G10 fins fluttering on just an I161.)
Could you post the reasoning behind it being an MPEG encoding artifact? I have a little experience with MPEG encoders and I really don't see how an encoding artifact could produce that effect.
 

DynaSoar

Well-Known Member
Joined
Mar 15, 2004
Messages
3,022
Reaction score
0
I usually edit down long answers, but this one bears repeating in full.

LAY that science on me.

Originally posted by bobkrech
Cliff and Dynasoar

I'm not on the list, but I was asked for my opinion on the video based on my professional experience which is posted on the webpage referenced.

The overwhelming evidence is that it is real fin-flutter and not an MPEG artifact. I wanted to believe that it was a transonic aerodynamic fluctuation in the air density causing a mirage effect, but the magnitude of the motion rules this out.

I looked at the flight data and observed that the flutter starts and stop abruptly at M=+0.90 and that the flight had a maximum M=1.04 velocity. It not possible to do quantative calculations without measuring the mechanical properties of the fins and the acoustic energy spectrum of the motor, but we can say qualitaively the flutter is a result of a number of thing, a coupling of the eddy turbulance in the transonic flow region, the flexible construction of the fins, enhanced aerodynamic coupling due to the non-zero angle of attack resulting fronm non-parallel installation of the fins as evidenced by the second stage rotation, and the coupling of the random vibration energy of combustion instabilities in the rocket motor with the natural vibrational frequency of the fins.

I showed the video to a co-worker who is an aerospace design year with 20 year of experience in spacecraft design and testing and he said that he had observed this kind of motion in fiberglass support structures within spacecraft undergoing randon vibration testing on a shake table. In fact he spent 1 1/2years analyzing the problem and finding an engineering solution to it. I can't go into details any futher.

The natural vibration frequency is proportional to the thickness of the fin, and the amplitude of the flutter is inversely proportional to the fourth power of the thickness for a given excitaton energy. A small increase in the thickness makes a huge difference in the flutter amplitude. A 50% increase results in a factor of 5 reduction in amplitude, and a 100% increase in thickness results in an astounding 16 fold reduction in flutter amplitude.

The fins on the rocket are extremely flexible. Look at the post flight video. I have a 1' x 2' x 1/8" sheet of G-10 and I can't get it to bend like those fins. Either they are thinner, or they were substantially weakened in the flutter incident.

The stock PML QL 3K has 1/16" G-10 fins which will flex easily. For the motor combination in the flight, PML recommends using 3/32" G-10 with a layer of cloth over it. That would make the thickness ~1/8". I think the thickness is actually 3/32" rather than the 3/16" mentioned on the webpage. This would be consistent with an epoxy rich 6 oz. glass layer over the standard 1/16" fin.

A consenus opinion is irrelevant in this instance. From what I have seen on the web, the HPR community has experienced a large number of shreds that are blamed on a number of possible reasons, but very few people have actually analyzed the failure modes from an engineering standpoint. Just because a majority of the rocket community has not observed this phenomon before doesn't mean it doesn't exit.

If Peter had not done such a good job glassing the fins to the airframe, the rocket would have shredded, and without an on-board video camera, the cause of a shred at 5000 ft would be pure speculation. Fin-flutter is real and probably accounts for a majority of HPR shreds not caused by a CATO.

Bob Krech
The only thing I'd add is the fact that you can't test the material for handling this kind of stress by sticking it in a vice and bending it with pliers, or even using a 3 point test like Doc uses on rocketmaterials.

Although the force may have been focused more on certain parts at certain times, the entire fin was "clamped" in the overpressure of the Mach wave. To simulate this would require a "vice" that clamps the entire piece, and provides high differential pressure while still maintaining consant pressure on the entire piece. The clamp itself would have to stay tight yet flex itself. The pressure would integrate, that is, be spread out and increase across distance, rather than focus on one point and translate to another point via internal leverage (flex).
 

HeadHunter

Well-Known Member
Joined
Dec 10, 2010
Messages
248
Reaction score
0
From the AEROPAC list, I believe this was the prevailing theory. I understood the theory, but have no expertise in the field, so I must defer to the pros on this one.

Jamie,
There is an important difference between what your scan shows and what
MPEG compression does. Your scan, shows the location of the pencil when
that particular portion of the pencil was scanned.
Each portion of the pencil really was in each location shown, just not at the
same time.

The warp effect you show is entirely due to the way the image is recorded.
In this case a "line scan" method was used.
Digital still cameras record an entire frame at one instance, not line scan.
Digital movie cameras may use a "line scan" method but I have not
seen that type of sensor built in years. (To avoid this very problem)

Also, the effect you show can not explain the movement show in the clip.
We know the fin did not bend as much as shown.

MPEG will show movement that never happened.
MPEG is the likely source of the error.


SteveP



If anyone cares, here's an abbreviated description of how the error happens;

MPEG isolates portions of the image that are moving from the static background.
In order to reduce the amount of information stored or transmitted, MPEG will send
the background information once. It then sends ESTIMATED coordinates for
the portions of the image that are moving. Estimated is the key.

When a fin makes a fast movement, the root is stationary and the tip is moving quickly.
MPEG will divided the fin into several chunks and each chunk will be given
an estimated trajectory. The chunk at the tip will be given a fast trajectory, chunk at
the root will be given no trajectory. When the MPEG file is played back the chunks
that are moving quickly can "over shoot" there actual locations.

Here's an example.
Suppose the fin tip is oscillating between -1 and +1 in the diagram below.
The first three images the camera collects, in order, are
"A" - near the left most excursion of the fin, heading back to the center
"B" - approaching the right most extreme, but not quite there.
"C" - on its way to the left after the tip made its right most excursion

If the first two images the camera collects are from "A" and "B" MPEG will generate a
trajectory for the fin tip that is large. This is because MPEG does not know that the fin
had actually been slowing as it approached +1. All MPEG knows is that the fin traveled
from "A" to "B" from one frame to the next.

MPEG will use its motion estimation and guess that the fin will be at location "D" the next time
it gets an image. When the MPEG file is being played back, it will display the fin nearing
location "D".
Once the third image is collected and the fin is shown to be at
location "C" and not "D" the motion estimator will try to quickly correct and show the
fin moving from "D" to "C". This correction will show the fin is moving twice as fast as
the toward the left, causing another overshoot in that direction.




The combination of low frame rates and MPEG compression could easily display
a movement three times as large as really happened.
 

Mike

Well-Known Member
Joined
Jun 3, 2011
Messages
1,330
Reaction score
0
Thanks for posting that Lee, certainly answered a few of my questions.
 

bobkrech

Well-Known Member
Joined
Jan 20, 2009
Messages
8,352
Reaction score
39
Dynasoar

The flow behind a shock wave is not usually steady, nor is it supersonic. In many real situations, the boundary layer flow behind a shockwave is turbulent, and the shock heating increases the sound speed so that the flow velocity is subsonic at the elevated air temperature. Additionally, while you have a large pressure/temperature jump at the shock front, the heated air has to relax back to the ambient conditions, so there is always a gradient setup behind the shock wave of constantly changing pressure, temperature and velocity conditions. There isn't any clamping by the flow, it's just the opposite. The unsteadyness of the flow induces movement of the fin, and the fluttering of the fins can and does effect the airflow. It's a coupled problem that is difficult to solve analytically, and that's the reason why wind tunnel testing and test flights are conducted on aircraft and rockets.

Recently that's also where Computational Fluid Dynamics comes into play. CFD allows you to conduct virtual flights in your computer, and an experience aeronautical engineer can use CFD to simulate the conditions where flutter or other aerodynamic effects might cause problems to a vehicle. The aerodynamic control bugs in SpaceShipOne were worked out using CFD rather than by conducting a number of wind tunnel tests because it is much faster and cheaper. The Discovery Channel did a great job showing this in it's 2 part special this month. Each test flight carefully expanded the flight envelop and collected a lot of data to compare the actual flight dynamics with the CFD predictions. A CFD model is only as good as the data input into so using CFD in the way is an iterative process. When designing a vehicle like SpaceShipOne, Rutan was developing a new concept that had not been done before and he started off with design based on almost 40 years of eperience. He got it mostly right, but the wing has a bit too much dihedral which caused the roll instability at high Mach, high altitude flight which was not fixable without a major rebuild. He did increase the tail width to increase control stability based on early test flight data and the CFD predictions.

There are some things you can learn the simple mechanical testing, and it can be useul. For example if you fins are flexible and bend easily, you can be sure that they will flutter at some velocity. The converse is not true. It your fins don't bend easily, it's not a guarantee that they won't flutter at some velocity, but it's less likely. Engineers design fins that don't flutter. If you download NACA TN 4197 you'll see the role that the mechanical tests such as those being done at rocketmaterials.org play in the design of fins that don't flutter. If you're motivated, you can get all the gory details from a college level aeronautical engineering textbook, or by surfing the net via google.

Bob Krech
 

DynaSoar

Well-Known Member
Joined
Mar 15, 2004
Messages
3,022
Reaction score
0
Originally posted by HeadHunter
From the AEROPAC list, I believe this was the prevailing theory. I understood the theory, but have no expertise in the field, so I must defer to the pros on this one.

...
The combination of low frame rates and MPEG compression could easily display
a movement three times as large as really happened.
Precisely the same argument could be presented with equal validity, stating that the actual flutter was greater than what was observed. The presumed error works both ways. Given sufficient sample (overcorrected plus undercorrected plus non-corrected) the video would approximate the real in variance. As a result, the period of flutter shown may be inaccurate, but not the variance.

Furthermore, the error would not
(1) result in complex curvature in a frame unless it were already present, and
(2) (as presented in the above argument) never result in excessive travel in the same direction on two consecutive frames.
 

DynaSoar

Well-Known Member
Joined
Mar 15, 2004
Messages
3,022
Reaction score
0
Originally posted by bobkrech
Dynasoar

The flow behind a shock wave is not usually steady, nor is it supersonic. In many real situations, the boundary layer flow behind a shockwave is turbulent, and the shock heating increases the sound speed so that the flow velocity is subsonic at the elevated air temperature. Additionally, while you have a large pressure/temperature jump at the shock front, the heated air has to relax back to the ambient conditions, so there is always a gradient setup behind the shock wave of constantly changing pressure, temperature and velocity conditions. There isn't any clamping by the flow, it's just the opposite. The unsteadyness of the flow induces movement of the fin, and the fluttering of the fins can and does effect the airflow. It's a coupled problem that is difficult to solve analytically, and that's the reason why wind tunnel testing and test flights are conducted on aircraft and rockets.
I agree that there'd be turbulence in the flow, with great differences in pressure over fairly short distances, but I still think the differences would all be relatively clustered within the greater ambient pressure of the flow.

There are some things you can learn the simple mechanical testing, and it can be useul. For example if you fins are flexible and bend easily, you can be sure that they will flutter at some velocity. The converse is not true. It your fins don't bend easily, it's not a guarantee that they won't flutter at some velocity, but it's less likely. Engineers design fins that don't flutter. If you download NACA TN 4197 you'll see the role that the mechanical tests such as those being done at rocketmaterials.org play in the design of fins that don't flutter. If you're motivated, you can get all the gory details from a college level aeronautical engineering textbook, or by surfing the net via google.

Bob Krech
I'm sure the flutter could be predicted fairly accurately from three point testing, but the failure/survivability couldn't. Something kept the fins from snapping an a bend radius well beyond what three point testing would predict. An integrated force would not result in focus of the force and localized failure such as would occur in a three point test.

No, I've not studied CFD, but I have studied statistical mechanics at Santa fe Institute, taught by folks from Los Alamos. Conceptually, they're close.
 

SwingWing

Well-Known Member
Joined
Jan 18, 2009
Messages
1,871
Reaction score
157
How much would air friction heating of the fins come into play at these speeds?
 

bobkrech

Well-Known Member
Joined
Jan 20, 2009
Messages
8,352
Reaction score
39
SwingWing

Aerodynamic heating is not an issue for the transonic flight regime. If you look at the references below you will see that in the transonic velocity region, 0.6 < M < 1.1 heating isn't a problem.

https://aerodyn.org/Atm-flight/sregime.html

https://aerodyn.org/Atm-flight/tlimit.html

Below are some typical skin temperature observed on aircraft at flight altitudes and velocities.

Vehicle M T (°K) T(°F) Shock Temp(°F)
Convair B-58 2 420 297 475
North American XB-70 3 550 531 1101
North American X-15 6 900 1161 3941

https://aerodyn.org/Atm-flight/table-skin-t.html

If you calculate the values for M=1.3 (1450 fps) at 1300 ft, you will see the shock temperature is ~230 F (assuming a sea level temperature of 59 F) which isn't anything to worry about.

Many useful on-line calculators for the shock temperature and other aerodynamic properties can be found at the webpages below. Most are graphical so you get to see what's happening.

https://www.grc.nasa.gov/WWW/K-12/airplane/isentrop.html

https://www.grc.nasa.gov/WWW/K-12/airplane/sound.html

https://www.grc.nasa.gov/WWW/K-12/airplane/atmosi.html

https://www.grc.nasa.gov/WWW/K-12/airplane/machang.html

https://www.grc.nasa.gov/WWW/K-12/airplane/normal.html

https://www.grc.nasa.gov/WWW/K-12/airplane/shock.html

https://www.grc.nasa.gov/WWW/K-12/airplane/oblique.html

Have fun. This is a pretty good website.

Bob Krech
 

Chuck Rudy

Well-Known Member
Joined
Jan 20, 2009
Messages
829
Reaction score
0
I decided to bite and look at the video. First goround I noticed something which made me shake my head. The base root of the fin, where it's attached to the rocket, on the right fin moved around the rocket. So along with fin flutter is fin rotational realignment, I just made that up so don't do a search on it ;-). I'm gonna guess this rocket had a vibration at transsonic, affecting the video input, and didn't move into mach immediatley. Vibration makes for odd effects at times.

And I, of all people, complain about the compromise of DVD. Digital averages data to change the fps using wondeful algorithms, but this average makes for unrealistic freeze frames with shadows of preceeding frames averaged into the next frame. Thereby allowing 29fps to be shown at 24fps to save memory and allowing more minutes on DVDs. Most don't even notice it, but it's terribly obvious if you spend way too much time watching high speed video.

In short, repeat the flight with faster motor and review the video, the same effect may be visible with this camera at transsonic, busting into mach will answer a lot of questions, as will another camera. This is looking to be a bargain basement camera, which is not a bad idea in case it's sacrificed.

When we see more video we won't be discussing in a vacuum. To make too much out of one piece of video is wreckless.
 

UhClem

Well-Known Member
Joined
Feb 6, 2009
Messages
1,893
Reaction score
325
I find the idea that MPEG compression could have amplified the actual fin movement to be interesting. Interesting ienough that I started reading up on MPEG compression.

The first think I discovered is that there are several frame types. The first is the "intra" or I frame which only has spatial compression. This is the reference for succeeding frames.

Then there is the predicted or p frame. This is the frame type that includes modeling of motion. But this modeling is done on a finer scale than the entire picture. The algorithm breaks the picture up into 16X16 pixel blocks and runs a prediction algorithm on it.

Then there is the bidirectional frame which uses information from both preceeding and following frames.

Because of the high roll rate in this particular video, prediction algorithms are useless because the features move outside of the analysis blocks.

Unless the MPEG algorithm mixes both intra and p frame algorithms (which is possible, I just haven't seen anything to say that it does) in a single frame, the temporal compression could not amplify the fin movement.

As a test to see if any temporal compression method was being used in the blocks around the fins I looked for adjacent images where there was some ground detail around the fin in one frame and not in the next. It turns out that the smoke trail shadows provide this. If temporal comression was being used then some of the smoke trail shadow would follow the fin. It didn't.

I would really like to be able to look at each frame and be able to identify which frame type it is. I suspect that because of the high roll rate no predicted or bidirectional frames were used. But I haven't been able to locate a reasonably priced tool that shows this information.

I did look at on board video from a rocket that does exhibit p-frame anomalies. The rocket was roll stabilized (mostly) so the roll rate was fairly low. Roads picked up ghosts.

https://www.deltavrocketry.com/videos.htm
 

Stymye

Well-Known Member
Joined
Jan 21, 2009
Messages
7,599
Reaction score
6
fin flutter enhanced by a video anomaly.
 

dave carver

....what hump?
Joined
Jan 18, 2009
Messages
4,973
Reaction score
5
Location
Idaho
Dave Thidens's "MOOSE" a 5.5x 80 some inch rocket, had large Nike like fins out of 1\8 G-10 . When flown on an Ellis Mtn. L at about 500 ft. it started to go unstable shredding it's self to peices.
Post flight analisys showed cresent shaped scrapeing from bottom to top where the fin tore loose as it bent radicaly back and forth. As the Ellis Mtn. motor is a long burn motor "MOOSE"
spent too long at max Q, ergo, shred.
 

Terrier

Well-Known Member
Joined
Jan 18, 2004
Messages
51
Reaction score
0
Just wondering if the other two fins on the opposite side from the suspected fins and camera were affected. Could it be just turbulance from the camera fairing on these two fins? I know the video doesn't show the other two fins and the post flight video doesn't address these fins. What about them?
 

Latest posts

Top