This news item is mirrored from the main ESA Web Portal.
The distinctive shape of Comet 67P/C-G. Credits: ESA/Rosetta/Navcam – CC BY-SA IGO 3.0
Two comets collided at low speed in the early Solar System to give rise to the distinctive ‘rubber duck’ shape of Comet 67P/Churyumov–Gerasimenko, say Rosetta scientists.
The origin of the comet’s double-lobed form has been a key question since Rosetta first revealed its surprising shape in July 2014.
Two leading ideas emerged: did two comets merge or did localised erosion of a single object form the ‘neck’?
Now, scientists have an unambiguous answer to the conundrum. By using high-resolution images taken between 6 August 2014 and 17 March 2015 to study the layers of material seen all over the nucleus, they have shown that the shape arose from a low-speed collision between two fully fledged, separately formed comets.
“It is clear from the images that both lobes have an outer envelope of material organised in distinct layers, and we think these extend for several hundred metres below the surface,” says Matteo Massironi, lead author from the University of Padova, Italy, and an associate scientist of the OSIRIS team.
“You can imagine the layering a bit like an onion, except in this case we are considering two separate onions of differing size that have grown independently before fusing together.”
The results of the study are reported in the journal Nature and were presented today at the European Planetary Science Congress in Nantes, France.
To reach their conclusion, Matteo and his colleagues first used images to identify over 100 terraces seen on the surface of the comet, and parallel layers of material clearly seen in exposed cliff walls and pits. A 3D shape model was then used to determine the directions in which they were sloping and to visualise how they extend into the subsurface.
It soon became clear that the features were coherently oriented all around the comet’s lobes and in some places extended to a depth of about 650 m.
“This was the first clue that the two lobes are independent, reinforced by the observation that the layers are inclined in opposite directions close to the comet’s neck,” says Matteo.
“To be sure, we also looked at the relationship between the local gravity and the orientations of the individual features all around the reconstructed comet surface.”
Broadly speaking, layers of material should form at right angles to the gravity of an object. The team used models to compute the strength and direction of the gravity at the location of each layer.
In one case, they modelled the comet as a single body with a centre of mass close to the neck. In the other, they worked with two separate comets, each with its own centre of mass.
Layers in Anubis. Credits: ESA/Rosetta/MPS for OSIRIS Team MPS/UPD/LAM/IAA/SSO/INTA/UPM/DASP/IDA
The team found that orientation of a given layer and the direction of the local gravity are closer to perpendicular in the model with two separate objects, rather than in the one with a single combined nucleus.
“This points to the layered envelopes in the comet’s head and body forming independently before the two objects merged later,” concludes Matteo. “It must have been a low-speed collision in order to preserve such ordered strata to the depths our data imply.”
“In addition, the striking structural similarities between the two lobes imply that despite their initially independent origins, they must have formed through a similar accretion process,” adds co-author Bjorn Davidsson of Uppsala University, Sweden.
Layers in Seth. Credits: ESA/Rosetta/MPS for OSIRIS Team MPS/UPD/LAM/IAA/SSO/INTA/UPM/DASP/IDA
“Layering has also been observed on the surface of other comets during previous flyby missions, suggesting that they also underwent a similar formation history.”
Finally, the team note that even though erosion is not the root cause of the comet’s double-lobed shape, it nevertheless does play an important role in the comet’s evolution today.
Local variations seen in the structure of the surface likely result from different rates of sublimation – when ice turns directly into a gas – of frozen gases embedded within the individual layers, which are not necessarily distributed evenly throughout the comet.
“How the comet got its curious shape has been a major question since we first saw it. Now, thanks to this detailed study, we can say with certainty that it is a ‘contact binary’,” says Holger Sierks, OSIRIS principal investigator at the Max Planck Institute for Solar System Research in Göttingen.
“This result adds to our growing knowledge of the comet – how it formed and its evolution,” says Rosetta project scientist Matt Taylor.
“Rosetta will continue to observe the comet for another year, to get the maximum amount of information on this celestial body and its place in the history of our Solar System.”
“The two independent and primitive envelopes of the bilobate nucleus of comet 67P/C-G,” by M. Massironi et al., is published as Advanced Online Publication on www.nature.com today.
Dr Massironi presented the study today at the European Planetary Science Congress in Nantes, France, in a dedicated press briefing.
Editor’s note: Readers may also be interested in the paper “The shape and structure of cometary nuclei as a result of low velocity accretion,” by Jutzi and Asphaug (2015); their computer simulation of an example comet nucleus collision can be seen here.
Discussion: 171 comments
This is so so wrong. The scientists have come to this conclusion based on considering two competing theories, contact binary and an eroded, single body. They have then fitted one of those theories as best they can to the evidence available.
By ignoring stretch theory, they have made the mistake of assuming that the differing angles in strata layers between the two lobes mean they are two separate bodies, favouring contact binary theory. However, the head lobe tipped by around 30° during the stretching process. That’s not a fudge. It had to tip before lifting off due to the geometry of the head with respect to the long axis and is indeed a signature of the head shearing away and stretching from the body. It completely explains the offset strata and was set out clearly in this blog post in January 2015:
https://scute1133site.wordpress.com/2015/01/22/67pchuryumov-gerasimenko-a-single-body-thats-been-stretched-part-11/
Moreover, the complexity of the strata layers within the head and body are fully explained by the clear evidence that the head lobe stretched even before it sheared from the body. This stretching deformed the strata layers in the head. There is an abundance of evidence for this as set out in Part 27 of the same blog. That part is the concluding section of a 50,000 word thesis on just this one phenomenon, the deforming head lobe.
Part 28 will address the folding onion-shaped strata in great detail and thus conclude a very robust explanation as to why the strata layers in the two lobes are different. Stretch theory predicts that they absolutely have to be different, both in their respective internal deformations and their angling to each other on a global scale. What’s more, it predicts that they should be different in the exact manner described in this newly published scientific paper.
Agree with A. Cooper.
What the text asserts does not seem to line up that well with the images. The dotted red layer lines don’t do much to show they are so different between the head and body. They do more to show the body is more of a full onion but less to show the head is as well.
The 3D shape model was given some artistic liscence in showing strata planes, although, even so, the ones around the bottom perimeter of the head look more like stretch and deformation to me. To support the text, they “should” wrap around underneath the head, like the layers of a second , smaller onion, according to accretionary theory.
The gravity vectors again show more of the body being a mostly full onion but jury is still out on the head.
I just don’t think A. Coopers ideas have been given a fair shake, why, even during the live ESA Hangout, his question was completely swept under the rug and not even properly addressed by Holger, who simply ignored the question and went on about accretionary being the only likely scenario. Sorry to rain on the paper. I am just not seeing conclusive proof by any means here, and the continued total disregard of stretch theory, ignoring it, not even addressing it at all, seems a disservice to good science.
OK… Perhaps because I am a sculptor, working in clay, putty, plaster, molten glass, metal, silicone, composites and resins for four decades now, I feel I can see a shape and with some illuminating background, such as A. Cooper’s comprehensive Stretch Theory thesis, run a 3d movie backwards and forwards in my head as to how it got that form. It just works here. How else to explain the vertical furrows in the neck like stretching partially dried Play-Do without Stretch Theory?
Well, I see less convincing arguments from this paper on accretionary or Contact Binary theory, when I run “their “movie”. And yes, II even saw the fanciful 3D computer simulation previously. Think they had to do many many sims to get one to do what they needed. But it was high speed and pretty much obliterated and melted most of the mass before reforming. There would be little to no strata to see now in that case, one would think.
I dont care who is right, I just see what I see. Bit, jist look at how even opposite camps on this blog are converging on wondering why Stretch Theory is not even being spoken about or considered.
Thank you ESA team. So glad you are letting us be a part of the discussion!
Ram, you -as an sculptor- can pause and stop at a moment when nature wouldn’t.
Have you made works thinking that nature is not going to stop working on them?
Logan,
Brilliant view! And good point. I don’t think 67P has stopped being shaped, just that there were periods of high activity, stretch and whole slabs lifting off, and quieter phases, like ever since its closest approach to Jupiter, and other unknown histories. 🙂
‘ But it was high speed and pretty much obliterated and melted most of the mass before reforming.’
Obliteration necessarily followed by mass loss, at this scale. No melting, just crushing, and breaking.
You are right on this point Ram. This new document quite probably has moved the scenario to a quieter, more sync place. Do you remember the input speed of that simulation?
Yahoo! Finally re-found NASA’s 3D Sim for accretionary binary comet theory:
Psys.org/news/2015-05-comets.html
Scroll down and play video. They ran about 100 Sims taking weeks each. So that is a lot of Sims to get this one to produce the ducky shape! I don’t wish to take anything away from their work, just saying it should not be definitive or “Unambiguous”, as Holger’s team asserts in that paper.
“Bicycle speed” impact. Whole process took 55.5 hoirs.
And Logan I should have said something like rubble-ized, rather than “melted” … although it certainly could have heated things up!
So, I just think “Stretch Theory” (enlongation theory?) as A. Cooper’s 50,000 word thesis with matchup images describes has at least an equal chance of being right!
Very limited platform. Taking your word. 55 hrs not enough for re-crystallization. Personally not discarding Cooper’s among plausible scenarios 🙂
Bicycle speed approach speed. Two objects equivalent to Coraline. 3D Model shows an apocalyptic sequence. Do you remember Philae rebooting on a planetary scale? Data input to these simulations ignore what we already know about the high levels of sinter-ing.
Ramcomet
Re bicycle speed collision, my reply went to the bottom of the thread.
(Logan – input speed was 0.375m/sec, positively sizzling!)
Andy,
Yes, good work. More like strolling speed! (Just to be clear for readers, “bicycle speed” is what the paper stated, not me.)
Again, the model took about a hundred sims taking weeks each, to get finally get this glancing (splashing?) blow and re-formation into “ducky”!
So about ninety-nine times it didn’t work. Well I wish they would give Stretch Theory just one sim chance! But that’s not their paper, is it?
We should ask someone to do a Stretch Theory animation… (Faithful to your eleven steps below, complete with spin up, slabs blowing off, head detaching with perimeter jets, neck stretching and head turning 15°, finally, spin settling down and neck hardening).
Nothing like cool “Sim” eye candy to gain some popular attention and traction, maybe even open a “door” upstairs!
Anyone know an CG animator to work with Andy, who might like some exposure on the ground floor?
For those who require a peer reviewed paper “right now”, I would like to say, the times they are a ‘changing… “The Martian” was written as a BLOG first, then became a novel, then became the highly successful movie today. Just an analogy.
Hi Ram: It’s not the same to say that an object is elongating [which conserves a plausibility of later equilibrum or return] that to say than an object is stretching [which doesn’t]. There are elongated objects out there.
Nowadays, following Cooper’s “slab” arguments closely.
_Also because of NASA’s endorsing. But that’s not the expected behavior of ‘crushed’ cometary material.
Logan,
Yes, and… regarding “crushed”, if you look at the 3D Sim to the end, over 50% of the head has been completely obliterated and reformed, while between 35% – 50% of the body has done the same!
This is completely out of sync with their using onion layers to “show” two full onions. (Which is also wrong – the head onion layers describe only a bell shape as A. Cooper described).
Most of the pertinant strata would have been utterly obliterated!!!
With all respect, they cannot have their muffins and eat them too. Using a 3D sim that undermines the onion layer vectors is contradictory.
I think the article should be retitled:
“How Rosetta’s Comet MAY Have Gotten It’s Shape”
I think we all REALLY appreciate the next post from Dr. Massaroni and Emily today. But yes, it is his “fault” he did not seriously review this blog and A Cooper’s before wrapping up the paper. (After all, why have it?)
I think it sets the stage for a fair (and CONCISE-hint hint Andy!) critique of the paper for now, then even follow his four steps, Lets work on this!
Ramcomet
Good point about the onion layers not folding round under the head lobe. The first comet photo in that link above shows a flat underside to the head lobe with a very sharp rim. The onion layers start from the rim upwards on the top side of the head. There is no sign of folding round to the underside. The rim is too sharp. However, they are there, at Hathor. They are the red lines depicted in the photo in the scientific paper. It’s just that they are end-on, torn-away strata, suggesting they were torn away from something.
In the other photo, from the link above, you can see the equivalent of the Hathor cliff sticking out roughly the same distance on the other side of the neck, i.e. on the south pole side. Its plane is in line with the Hathor plane too, showing that it is a single, flat plane all the way across. It looks as if it’s balancing on the top of the neck. This flat plane is what stretch theory says was seated on the body. Its where the onion was sliced.
One can extrapolate the beginning of Hapi, before the neck starts, across to the s pole as well. So it looks as though a flat plane existed there too for the head to sit on. This was before the neck was pulled out and perhaps expanded via Ramcomet muffin theory.
The reason the head onion layer planes don’t match to the body is due to the stretching of the head lobe before shearing. You said the rim looked stretched, not wrapping round. The best examples of this are Serqet (the ‘yellow’ triangle) and also on the head rim above the finger of Seth on the border with Babi. Here, there’s a sloping, delaminated section with a frill on its outer edge. It was clearly pulled upwards from its Seth seating while still attached- hence the three outgassing holes above it. Those three separated sections show one of the clearest and detailed matches anyway. The stretched-up section is actually the top layer of a triple delamination, like puff pastry. The middle layer below it matches the body layer. It’s in part 5 which has the actual match photos.
So the onion layers just curved round and upwards from the body like the rim of a bell before the bell sheared away from the body. That’s why the head lobe looks like a bell- it deformed to that shape prior to shearing………
Hi,
The Serquet triangular feature is the South pillar of C.Alexander gate. It has been extensively referred to in the Scute blog as the yellow triangle before it was formally named.
I would favor the stretch theory, too, have been doing so, since the beginning. I wonder why they haven’t considered it, so far.
Good science coming out of this mission to the comet. We’ve noticed these structural features since the first views of the comet and they raise philosophical questions of scale. A comet is a accretionary body with origins in the primordial solar nebula. We can see clumps and clumplets of material in the Philae surface images. We can see compositional and structural differences in fresh rock faces. We can see the size and shape of the “dino eggs” in active cliffs. We can see structural (and presumably compositional) differences in the several regions in each lobe. And of course, we see differences in each of the two lobes. So the question is , where is the line drawn between a body accreting from bits and pieces and a body made from the contact and merger of two (or more) comets?
Looking at this body, it is indeed an Escher-esque world. Ole MC was indeed before his time.
–Bill
For a long time, I thought this comet is a fragment of a much larger body, because it looked so strange how could stratification occur on a small celestial body. On the same time, we know from Philae that below a small layer of dust, the frozen rock was as solid as ice. From accretion of tiny particles, we would expect a more fluffy comet, something like snow. The fact that we see stratification and at the same time solid rocks, it means that we still have a lot to learn about how small objects are formed.
We can speculate a lot about the processes that occurred at the formation of a comet. I cannot stop thinking about the complex processes involved in creating those layers. Also, we can speculate about how both lobes collided and what force is keeping them together. The only thing that I can say is that, if both lobes formed independently, our comet was not much larger when it moved close to Sun.
Thank you, Emily. This material is probably the most important discovery since Vega missions first came close to a comet. It gives many answers and asks many further questions.
This pair are not -by far- complete “onions”. Accretion, specially late accretion should not be necessarily one way.
“…From accretion of tiny particles, we would expect a more fluffy comet, something like snow…”.
There are “creatures” which precedes comets, maybe generations of them, by now all of them tagged as “comet-esimals”.
Does this theory apply to other comets with a somewhat similar shape, like Borelly and Halley? What about Asteroids 624 Hektor, 87 Sylvia, Eros, 25143 Itokawa, and 216 Kelopatra? So many collisions?
Yep, it looks to me the odds are very low that there are soft/low speed collisions of bodies of similar mass happening everywhere.
Not convinced by the paper.
No, Frankebe. Contact binaries and erosive necks are both separate and possible.
Comet Hartley 2 and Borrelly were also double lobed bodies. Does the “contact binary” view also apply to those two as well?
Hi,
They say : “The team found that orientation of a given layer and the direction of the local gravity are closer to perpendicular in the model with two separate objects, rather than in the one with a single combined nucleus.”
What is most important to consider is the parts facing the neck region. As you can see in the image, the flat area of Hathor crosses many layers ; Anuket doesn’t have strata so much as cracks. An assumption of two parent bodies makes the neck region hard to reconcile. In many ways, there is a continuity of features making the neck like part of the whole with no demarcation line as to divide the comet into what came from one parent body or the other. A lack of any kind of obvious border should surely give the scientists reason to hedge their brave call that they have found an “unambiguous answer to the conundrum”
Also, zones such as Hathor have strata lines with little to no connection to head lobe gravity vector and should be taken as evidence counter to the link between layers and lobe gravity vector individually.
Also, no theory has been offered to explain why accretion would lead to a layered surface for such small gravity vectors acting on “ice and dust”
The assumption that there are no other candidate mechanisms other than erosion or contact binary is false. At the very least it lacks scientific rigour in failing to mention other proffered mechanisms such as stretch as to why they are not considered.
Contact binary theory also hasn’t fully resolved the “collisional problem” as to the coincidence of a very soft collision between two similar sized bodies somewhere 4.5 billion years ago this side of an accretion disk, and travelling all that time keeping the same general shape of the two joined bodies without further collision
I have only seen two options considered for the shape of 67p: a collision of 2 bodies or sculpting. I’d like to know if the following option has been considered? Could a comet act like a very large water droplet? Water droplets can vibrate in a range of modes a first order mode is a breathing mode the second order mode forms a dumbbell shape and higher order modes form more complex shapes. 67P looks like a dumbbell shape which has frozen and become fixed. If it all melts it could start low frequency vibration again and then freeze and turn back into a snow ball as it moves away from the sun. Freezing would start on the outside and maybe that could produce onion layers.
Hi Jeremy. “…If it all melts…” How? We are having difficulties to sink heat just 10m under.
Two massive objects in space on separate trajectories can collide and just stick together and form a single glued together comet? I’d sure like to hear the detailed, step-by-step process of how exactly that happens. I’m not sure how many specialists were involved in examining this, or from how many fields, but I’m wondering if specialists from other fields that examined this same issue might come to different conclusions..
Hi Sovereign Slave. Really can not imagine how, at our proto-planetary disk. Unless 67P was already there.
The investigators may have convinced themselves it is a contact binary but they have not presented any evidence. Once again what they have done is presented a speculative interpretation. The concept of a contact binary itself is hypothetical. Nobody has demonstrated how large chunks of rock stick together. Whatever the contact speed it is a collision. The rock chunks would either bounce off each other ( the way the Rosetta lander did on the comet nucleus) or, at high enough energy, one or both would shatter. The idea that they would even remain in contact is fanciful and even if they did what process consolidates the joint. Of course the planetary accretion theory is similarly fanciful and lacking in evidence but nevertheless widely believed. This contact binary interpretation is consistent with those beliefs.
They’ve run simulations to show, that it’s physically possible.
The constraints allowing for the contact binary are reasonable for the protoplanetary disk, and worth to be considered for a collapsing dense interstellar dust cloud, the state preceeding the formation of a solar system.
Sorry Gerald, but you’re whole post is a huge reach. Would love to know how many dozens of manipulated, fanciful, hypothetical “if’s” that simulation is based on, but saying that the computer simulation shows that it’s physically possible is complete nonsense. And then saying the simulation is consistent with current creation myth dogma takes this sand castles in the sky logic to new altitudes. It still amazes me how often hypothesis, theory, assumption, modeling, interpretation, etc etc seems to morph into strong assertion then confirmed belief then accepted fact then dogma.
Only few initial hypotheses make their way to a generally accepted fact, strictly speaking to a certain confidence level.
I don’t know the meaning of “dogma” in the context of empirical sciences.
In theoretical sciences you may define a set of axioms. If they don’t contradict each other, you can define non-empty models consistent with the axioms.
The axioms define a theory.
Such a theory may or may not be applicable to a set of empirical data.
The closest you may get to “dogmas” is the convention, that data and logic are the ingredients for science.
Understood. But any discipline can be threatened by beliefs being accepted and presented as established facts and becoming dogma, defined as I use it as “a settled or established opinion, belief, or principle.” In cosmology, it’s totally taken for granted and unquestioningly asserted over and over that P67 originated from the proto-planetary disk. This is a belief that is woven into the very bedrock and fabric of cosmology (as are many others), and is now being used to support and define current observations and speculations about P67 (such as it’s a binary). But in no sense of the word has this belief been proven. And sure, there is not doubt evidence and logic to back up the belief, but evidence can cut a lot of different ways, and logic is only as good as the integrity of the interpretations of the evidence. And sure, there have to be agreed upon starting points upon which to build, but as you say, only a few hypothesis make it through the consensus process to the top, which no doubt adds tremendously to the difficulty of then ever unseating them, or questioning them once they’ve been around so many years, or making what would be considered radical assertions that would contradict them. But dogmatically asserting that we know how the universe formed billions of years ago, then interpreting and conforming today’s universe and all it’s many moving parts and activities to those assertions inevitably creates false science.
SS.
Experiments and observations are designed to test theories not confirm them.
This is a common misconception.
Experiment and observation is the very antithesis of dogma and belief.
If a theory is considered dogma or a belief system then why bother testing it?
A test in essence is questioning the dogma or belief.
Theories cannot be proven or represent “the truth”, or considered belief systems. They are approximations that can be refined or even dismissed outright as technologies for observation and experiment improve to allow greater precision.
When Apollo astronauts put mirrors on the Moon that allowed the Earth-Moon distance to be measured with greater precision, it was found the Moon was moving away from the Earth. Hence the 200 year old+ Newtonian model of the Moon being in gravitational free fall around the Earth was not strictly correct but influenced by other phenomena such a tidal forces.
The greatest critics of mainstream science are the scientists themselves.
An example of this was the recent BICEP 2 fiasco where scientists reported finding primordial gravitational waves which supported the theory of Inflation in the early history of the Universe.
The gravitational waves were observed as the polarization of light in the cosmic microwave background. This would have been a major triumph for Cosmology.
The discovery was met with large scale scepticism amongst Cosmologists themselves who considered the effects of local polarization was underestimated.
When the PLANCK data was came in this was indeed found to be the case where local polarization was caused by magnetized dust particles in our galaxy.
As a result the discovery was retracted.
What these examples do show is that mainstream science is not based on unquestionable dogma or belief systems.
@sjastro
Yes, experiments and observations should be designed to test theories. But when there are so many theories that are flat out stated as facts, and repeated over and over as facts, it should be an indication that there’s a significant problem, such as with this paper. When the authors make statements like, ““How the comet got its curious shape has been a major question since we first saw it. Now, thanks to this detailed study, we can say with certainty that it is a ‘contact binary’,” says Holger Sierks, OSIRIS principal investigator at the Max Planck Institute for Solar System Research in Göttingen,” they are removing it from the realm of theory and are stating that it is a fact. And we see this over and over with all kinds of discoveries and theories. What evidence did they find in all their research that it is not a contact binary? We don’t know. Those in the best position to gather and review the data and objectively present all the evidence for each idea haven’t done that (unless the paper is much different than this article). They’re only telling half the story, because that is now the only story since they are certain it’s a contact binary, and any evidence to the contrary has either been unsought, ignored, overlooked, or discarded. As the article says, two leading ideas emerged: did two comets merge or did localised erosion of a single object form the ‘neck’? They’ve obviously sought and outlined evidence for the first, seems time for another paper outlining how they sought and what evidence they found for the second. And as to your statement that “experiments and observations are designed to test theories not confirm them,” as this article is written at least, this does not seem to be the approach they took.
SS: Please don’t mix up cosmology (origin of the universe) with origin of the solar system, which is related to star formation and for our Sun, takes place at a late age in the history of the universe.
SS.
Let me reiterate again, theories cannot be proven and therefore cannot be presented as facts.
Experiments and observations can SUPPORT a theory but it cannot PROVE a theory.
The Moon example I provided illustrates this point.
It’s an excellent piece of timing that the perspectives of a planetary scientist on the contact theory are given here.
https://blogs.esa.int/rosetta/2015/10/12/interpreting-images-more-on-how-the-comet-got-its-shape/
I suggest you take particular note of the section.
3) Never fall in love with theories (particularly your own).
Sjastro, I’m afraid your post is simply driving home my point. The theory that P67 is a contact binary IS being stated as a fact by these scientists, which, as you point out, should not be. That’s the exact problem I’m highlighting (one of them at least). And as far as EU, I do like it but it’s got major challenges which may be insurmountable, which is fine, I have no horse in the race either way. My posts here have nothing to do with EU, they simply address what seem like rather glaring issues in the science. And I could be totally wrong on those, but it seems like you and others take questioning and skepticism of the science as personal affronts, whereas in my little world view questioning and skepticism should be the foundational working mindset of every scientist, or practitioner of whichever discipline. There should be no sacred cows in science, but I wouldn’t know that reading many of these posts.
SS,
Unless you have a very different definition of “dogma” and “belief”, I fail to see how the link I provided in my post dives home your point.
Perhaps you would like to enlighten me on how the author who explicitly mentions alternative models being investigated, or the investigation of lines of evidence that may undermine an existing theory, somehow provides examples of dogma and belief in operation.
I’m not sure why you bring EU into this, unless you think that ignoring it is an example of dogma and belief at work.
The irony is if you want to see dogma and belief at work look no further than the EU.
Skepticism has such a broad spectrum and can cover two extremes.
One is the skeptic who has an understanding of the subject matter, whose skepticism is based on logical foundations. The other extreme is the skeptic who couldn’t be bothered learning about the subject matter, let alone understand it.
In this case skepticism is little more than an expression of ignorance and prejudice.
I’m sure most individuals would have a negative view to this form of skepticism.
SS: But scientists do have a horse that they are backing. And they have to put forth their arguments ensuring they don’t make any mistakes along the way. Not silly ones and not subtle ones. For example the mathematician Cédric Villani in “Birth of a theorem” talks at some point in the book about the pressure he faces because he has to give a talk to some of the best people in his area, and he hasn’t worked out all the details of his argument to his own satisfaction.
SS: The theory that the apple falls to the ground because of gravitation is taken as a fact, and even taught to schoolchildren. The reason is that it is useful for them to learn this. Some of them may even become space scientists and send probes like Rosetta to other places.
SS: “there’s a significant problem, such as with this paper.”
Where is the significant problem? You perceive a problem, but the rest of the world does not have to share that perception. Any scientists who see a problem in this paper will be running to solve it and to argue at the next conference that their theory is better.
“Now, thanks to this detailed study, we can say with certainty that it is a ‘contact binary’.” What Sierks is saying is that he thinks he and his coauthors have solved the problem. That does not prevent someone else from trying to do better.
🙂
Even Sierks et al. themselves could come latter saying they have an ever better idea of what happened. Ideas evolve, even the good ones.
Well, if the scientists that post on this blog are any indication, I highly doubt that there will be any takers. It has not mattered one whit what the papers coming out on P67 have said, there has been an immediate support and belief by those posters that the authors are right, and whatever the authors have asserted has been vigorously defended by them against all naysayers. And as long as there is no obvious “problem” with the paper contradicting currently accepted principals of science, I doubt anyone will see a problem. But I have to wonder how many scientists actually COULD review this paper if they wanted to. I mean, to really review the paper and it’s assertions, they would have to have access to ALL the information and data that the authors have access to, not just the cherry picked information presented in the paper. Does ESA provide all that raw data to scientists who want to independently confirm the conclusions of the authors, or contradict them for that matter? Without all the data, a real peer review really isn’t possible, is it? Especially in this case where no one is in a position to gather their own data independently from that acquired by Rosetta. And who would pay a scientist to do this kind of time consuming review anyway? And what if their credentials and standing were less than the authors, would they really be taken seriously? And although much is made of the importance of providing falsification of a theory, how often is that presented as part of these type papers, or of those that have been presented so far in the P67 papers. Doesn’t seem to have made it onto the radar for the contact binary article either. Perhaps this really isn’t an aspect of being “rigorous” in one’s research. Neither can other scientists take the conclusions of these papers and replicate the results…the authors have taken proprietary information and data and drawn certain conclusions, and even if there were another scientist out there that were willing and financially able to review the raw data and even if they came to a different conclusion, there really is no scientific way for either to verify their conclusions. Nothing about it is replicatable, it would be one opinion vs another opinion based on likely interpretations. Anyway, I know no system or human endeavor is perfect, but it seems pretty apparent that there is a great deal of faith put into this much touted peer review system by scientists, and there doesn’t seem to be much consideration or regard for it’s possible weaknesses.
Forgot to include your posted quote I was responding to at the top of my long winded posted, which is, “Any scientists who see a problem in this paper will be running to solve it and to argue at the next conference that their theory is better.”
Hi SS, I completely agree with you. Who is reviewing the review process? Other than us no one is even contemplating it.
Sovereign Slave,
the raw data as well as calibrated and reduced data should be delivered to ESA’s PSA, and may be also to NASA’s PDS.
The papers released thus far are certainly not the final results.
Discussions in the science community will last for years.
All you can say by now about the released papers is, that they look good and credible.
For other missions I’ve seen, that accurate data calibration can take years, sometimes.
So, future refinements of old data wouldn’t be too unusual. Just large unexpected errors don’t look likely.
The data are usually provided with error estimates, such that you get information about the reliability of the data.
Instruments may show shadowing or warm-up effects, or oscillations due to changes in the voltage, etc.. Reduced data are then usually flagged as unreliable due to some cause.
When looking to the plots of the time series you usually see these systematic errors. Statistical errors due to noise can be calculated with standard methods.
With sufficient safety margin, results hold, despite possible future refinements.
“a collapsing dense interstellar dust cloud, the state preceding the formation of a solar system”
Speculation too, Gerald, and consensus. No evidence. No plausible mechanism. Same with the comet formation hypothesis.
Originaljohn, there are astronomical observations of interstellar dust clouds, calculations, and simulations based on the contemporary knowledge of physics.
So we have both, observation and according theory.
You may like read more about the mechanism:
https://en.wikipedia.org/wiki/Jeans_instability
In total agreement. It seems too that drawing these kind of conclusions is akin to the elephant analogy, where the conclusions of what it is are based on feeling just one or two of the elephants body parts. There is no doubt a vast amount of information that will still be revealed from a wide variety of other tests, measurements, observations, data, and disciplines that will impact the answer to this question. Coming out with an “unambiguous answer to the conundrum” at this stage seems presumptuous, to put it mildly. But then, with the all-revered pier reviewed paper being published and the matter considered to be completely settled, as with so many other questionable conclusions, I imagine things will just continue to steamroll along to the next speculation presented as fact.
Being a work based on OSIRIS data, should include vector studios of layers and “crushing” at Ducky “shoulders” as seen at that first 3D published photo -stunning by the way-. [There is apparent evidence of force there, being body the softer object].
Eagerly awaiting for an open view of the document.
Dr. Massironi, could you please comment to the (lack of) consideration of the stretch theory of Mr. Cooper, – at least from an outsider perspective it sounds very plausible and convincing. A justification for ignoring it seems desirable, or alternatively, a substantive feedback on technical aspects of the Cooper view.
Thanks a lot, it would be really appreciated,
Michiel
Is there a peer reviewed paper relating to this ‘stretch’ theory?
Hi Matt Taylor,
All theories have to start somewhere. Before Fred Whipple published what was later to be coined “dirty snowball” theory, it was discussed at great length both formally and informally. All parts of it are still essentially hypothesis, some tests passing, others being refined by synthesis rather than repeatable observations on comets. Comets are like cats. What one comet does, activity wise, is not repeated in any predictable format by other similar comets, necessitating fairly vague knowledge of the internal processes if any knowledge at all can be certain for deep inside the nucleus.
Thus “stretch” theory is at that stage where mainstream cometary scientists have not published any papers in favour or against the concept. Like before Whipples paper, does not mean that it the ideas are not well defined, or for that matter reviewed and discussed at great length between people who greatly respect the scientific method and have found considerable evidence in its favour.
A.Cooper has progressed a blog with 28 parts which outlines a mass of, I believe, incontrovertible evidence which not only proves that stretch has occurred, but which explains surface morphology which hitherto has the cometary scientists completely confused.
Feel free to ignore stretch theory or check out the evidence for yourself. Sooner or later I know you will accept “stretch” theory as fact. Whether that takes the close up measurements taken next year and accurate surface change measurements and importantly, before and after accurate length of the comet measurements or not, it will eventually happen.
Doesn’t it suffice to look at the rims of the surfaces of the “head” part and the “body” part and see that they fit into each other like Africa fits into South America, to consider something along these lines?
Dear Hilmar,
Well, yes!
Matt
I haven’t done a peer reviewed paper on stretch theory. I have been considering it. However, in the early days I thought it would be taken up quite promptly and investigated as a possibility so there wouldn’t be a need for me to do so. The Rosetta teams could do it more thoroughly than I can. Later on, I was simply catching up on posting up the backlog of mechanisms and signatures that arise from stretch theory.
The key is the matches from the head rim to the body ‘shear line’. It’s difficult to explain them in prose, even with annotated photos- but not impossible because some people do see them. It takes much time and patience, paying attention to the finest detail. If I sat at a computer screen with someone, pointing at the matches, it would be a lot easier.
If no one takes it up then I’ll shall do a paper.
A. Cooper, please write this peer-reviewed paper.
Otherwise these discussions will be endless, and never be resolved in a satisfying way.
I think, there are several severe flaws in your reasoning.
But after a peer review process, there may be left over something valid, not necessarily stretch, but maybe something else.
Harvey and Gerald
Primarily regarding (but among other statements):
“please write this peer-reviewed paper.
Otherwise these discussions will be endless, and never be resolved in a satisfying way.” [Gerald]
And:
“Namely some *quantitative* estimates of what sort of spin-up is needed”. [Harvey]
Judging by your comments one would think that the entire Rosetta mission is on standby, awaiting the data on stretch theory that will allow them to forge ahead with an earnest, in depth analysis. What shall I do? Send a photocopy of my scrawled Woolies jumbo pad calcs so they can reproduce them on their supercomputers? Come on guys! This is very basic stuff for a space scientist. I shouldn’t be needed at all. The impetus to start investigating shouldn’t even be the calculations, it’s the principle that it could happen. That should then prompt them to do the calculations (see ref to erosion theory, below. They did it there).
As I said to Matt Taylor, I thought the idea of stretch theory would be taken up as worth investigating fairly early on in the mission. There’s a difference between accepting an idea as proof without peer review and accepting the principle in its most rudimentary form as a principle worthy of investigation by the Rosetta teams.
The blog was the way I went with it because I could get more info out that way and hoped scientists would think it may be a good idea to look into it. After all, they saw the gaping hole in the neck back in August 2014 and didn’t need prompting to start speculating on erosion theory without any proof or, of course, peer reviewed paper. And yet it was borne in mind, when looking at the data coming in, for a whole year. It was automatically accepted from the get-go as a possibility worth investigating with no initial proof- as indeed it should have been. I simply think that stretch theory should have been borne in mind in the same manner from the outset as it doesn’t violate a single law of physics.
Harvey, I have posted comments with spin-up figures before now. One was 3 hours 38 mins for loss of the Imhotep slab at a 1800m radius. Implicit in this calculation is a near-weightless head lobe centre of gravity. Any good dynamicist would see that straight away just through knowing the equations. I know that’s not your field but it is required for orbiting a comet and working out where to land on it.
Earlier in the year I posted (from memory) 8 hours or so as a rotation period for weightlessness at today’s Hatmehit radius, and, I believe, an extrapolated 4.5 hours for head lobe centre of gravity along with an informed guess that I could see the stretch scenario being plausible at 2 or 3 hours. That was based on the relevant extrapolation and weighting of root 1/r for weightlessness at lower r values in the compacted body. It was clear from those calcs that it didn’t need to be crazy rotation speeds like 0.5 or 1 hour. That was way back, possibly last year, but still I’m told there’s no spin-up data at all.
Inputs are subject to error, especially radius between head centre of gravity and body c of g when joined as one body. Also, the mass distribution between head and body if using comet c of g. And there’s the bilobed nature of the gravity field during stretch which would need modelling for exact results. Rosetta teams have more data on these three inputs than I have: via the OSIRIS shape model (for all three) and Flight dynamics data (for exact comet c of g).
As for spin-down, you must surely have realised that a very large amount of the spin-down would be brought about by the increasing radius between the two lobes during stretch. AM is dependent on the square of the radius so omega (spin rate) drops dramatically as the lobes move apart. In the highly conservative calculation below, you’ll see that the rotation period drops from 2.033 hrs at shear to 5.772 hrs at today’s radius. That leaves 6.628 hours of outgassing retro-spin needed.
We know 20 minutes of spin-up happened since 2009 (Mottola et al Sept 2014). That’s a huge amount. Marco and I have stated this at least a dozen times in the last year and still you comment as if no research has been done on it. Now the recent paper on shape model-derived torque values predicts spin-up of the same amount. So at these rates, a crude calculation would make it a bit over a century to spin down 6 hours or so, actually a bit more due to KE/omega squared issues. More realistically, 10,000 years of random outgassing torques producing a fractal graph with the current 12.4 hours as one of the unremarkable middling peaks.
And just one bout of sub-Venus-radius perihelion passages completely changes the game anyway with far greater torques- and that’s far more plausible than the fairytale kiss in the darkest recesses of the scattered disc. 67P’s perihelion distance has shifted between 2 and 3 AU many times since 1600 AD, now it’s at 1.24AU. It could easily have gone lower. As for the fairytale kiss, it really is a fairytale. It’s a virtual impossibility, which is why the recoalesced collision fragments is now the only narrative you hear. It was all addressed in Part 11, 9 months ago with detailed calculations based on real orbits.
So, for the umpteenth time, spin-up via asymmetrical outgassing is viable. As for detailed figures on torque, I suppose I could do that if I could find the time but couldn’t you? I’m not a one-man science journal. Exact torque is another paper, as is material properties. I didn’t read a detailed narrative on how the ‘bicycle speed’ model characterised 67P’s internal structure perfectly to get their paper through. No one really knows exactly what it’s made of let alone its tensile strength (but currently guessed by Thomas et al, 2015 to be very low).
For the more optimistic 2.75- to 3-hr rotation speeds, the AM conservation slowdown trick gives 8 hours for today’s stretched shape, nearly two-thirds of the way to 12.4 hrs.
The reason I haven’t published calcs, apart from wishing I could get a good r value and lobe mass distribution, is because it’s very easy to do them when you do have those two things. So I know that any Rosetta scientist will be able to knock them out during a lunch break. I obviously don’t need to hold their hand on this. So the whole concept of considering stretch theory doesn’t include some necessary revelation coming from me. It simply includes a quick glance at the basic calculations, which I don’t need to supply them, especially ones with dodgy inputs. They can see by juggling the numbers about with different inputs that it’s all viable for rotation periods between 2 and 3 hours.
But, seeing as I’m probably not going to get a good r value or mass distribution due to the published comet dimensions being so rough, I may as well paste my latest figures with the most stringent possible r value and comet extremities under significant negative g.
I used the published comet dimensions and this rotating model for measurements:
https://rosetta.jpl.nasa.gov/news/measuring-comet-67p/c-g
I took many screenshots and then used the 4.1 km base as a calibration measurement for all other measurements. I came up with a figure of 2.460km between the two lobe c of g’s today and 1.460km between them when pressed together as a single body. So the stretch would therefore be 1000m.
I then shifted a quarter of the way along the 1460m axis (of the pressed-together comet) from the body c of g to get the barycentre/comet c of g when pressed together. This assumes the head being 1/4 the mass of the comet. This gave me 1095 metres as a distance between the whole-comet c of g and the head c of g. In other words the orbital radius of head lobe c of g about the barycentre. This is needed as the axis of rotation for measurement of the circumference distance moved by the head lobe c of g and calculating the various tangential velocities for weightlessness, orbit and escape.
Some calculations are rather precise considering the inputs but that’s just to keep them tight as they sometimes get used as inputs and are then rounded in other places.
The circumference swung by the c of g at 1095m is 6881m (1095 x 2 x 3.142). I then took a tangential speed at head c of g radius that would lift the head lobe off into a low orbit. That speed is 0.94m/sec which is exactly halfway between the cusp of weightlessness, 0.780m/sec, and escape velocity, 0.780 x 1.414= 1.104 m/sec. Those two figures are specific to r=1095m, head c of g. Here is the calc for the weightless scenario:
Comet mass 1E13kg
G constant 6.674E-11Nm^2/kg^2
‘x’ means ‘multiplied by’
Using the vis viva orbital velocity equation:
v^2= GM(2/r-1/a) but r=a due to circular orbit.
v^2=6.674E-11 x 1E13 (2/1095-1/1095)
v^2=667.4 x 9.1324E-4
v^2=0.609
v= root 0.609
v= 0.780 metres per second tangential speed at a 1095-metre radius to induce weightlessness of the head lobe.
This is rough due to the bilobed gravity field but the c of g is a good enough approximation for the mass in negative g above the c of g to cancel the mass in positive g below it. In fact, it’s pessimistic because the ‘centrifugal’ forces are dependent on r^2 so mass at a higher r value carries more clout.
Escape velocity for that same radius of 1095m is 1.104 m/sec due to the root 2 rule.
So the head lobe would be ejected to orbit, but not escape, between 0.778 m/sec and 1.1 m/sec. That’s why the mid point of 0.94 m/sec was chosen as a value for ejecting the head lobe to low orbit. Meanwhile, the slabs at the extremities are well past escape v (see below).
Circumference of ‘orbit’ for seated head lobe is 1095 x 2 x 3.142 = 6881 metres. This corresponds to the still-attached head lobe rotating in a circle about the barycentre at 0.94m/sec (at its c of g), in negative g, and about to shear off.
The comet rotation period for the proposed tangential speed at r= 1095 m just before head lobe ejection is:
6881/0.94= 7320 secs or 2.033 hours.
For angular momentum conservation calculation between the pre-stretch and stretched comet, the c of g distance between head and body lobe can be used for the reduced mass version of coefficient of inertia. The reduced mass doesn’t need to be calculated because it’s the same for both versions but the r value does change and that goes from 1460m to 2460m.
The decrease in rotation period is proportional to r^2 as all AM calcs have an r^2 component multiplied by the coefficient of inertia (reduced mass in this case, which cancels).
1460m to 2460m is a 1.685 increase. 1.685 squared is 2.839. So omega (rotation period) slows by a factor of 2.839. It goes from 2.033 hours at shear to 5.772 hours at fullest stretch [today’s comet]. Today’s comet rotates at 12.4 hrs so it’s slowed by an extra 6.6 hours from 5.772 hrs (if the above 2.033 hrs is correct).
The 2.033-hr rotation period is probably getting close to the absolute maximum rotation speed needed due to the rotation radius being perhaps pessimistically low, the escape speeds reasonably high and the bilobed gravity field not being corrected for.
Varying inputs gives slightly varying but consistent outputs, all slower rotations than this example. They are up to 3 hours, or some rather more optimistic ones at 3H 30. The result is that it’s not out by an order of magnitude, requiring spin-up to 0.2 hours. It all looks plausible and most likely between 2 and 3 hours.
What makes this scenario very neat indeed is that the 0.94m /sec chosen for the head lobe to go into orbit means that Hatmehit and Imhotep had already gone through escape velocity threshold of 0.9 m/sec (using vis viva @ r=1695m for Hatmehit) and were now at 1.45 m/sec tangential speed- in highly negative g. 1.45 m/sec is derived from the 0.94 m/sec at 1095m and subjecting it to the ratio of radii, 1695/1095. So the slabs had to escape if they detached and yet the head lobe was destined never to escape. That answers your question, Harvey, as to why slabs should fly away while the head lobe conveniently remained. Its low r value and consequent higher g influence (baked into the velocity vis viva equation above) meant it needed a very high rotation rate to escape. The slabs didn’t so much.
To cap things off, the 5.772-hr rotation rate after stretch and spin-down corresponds to .558m/sec tangential speed for the head lobe c of g at 2460m from the body c of g:
Circumference=1845 x 2 x 3.142= 11,594m
Time= 5.772 x 3600 = 20,779 secs
Tangential v= 11,594/20,779= 0.558m/sec
This is very close to the 0.601 m/sec orbital speed for that r (r= 1845m from barycentre, 2460m x 0.75). The 0.601 is calculated using vis viva too, as per first calculation. That would mean that the head lobe really did go into orbit at that height and physically couldn’t get any higher, even if the neck hadn’t been helping to hold it back. That fact, by definition, ended the stretch event.
Note that the AM conservation value of r is 2460m due to being stipulated as the ‘r’ value in the AM calculation using the reduced mass coefficient of inertia. The barycentre is ignored due to being implied in the reduced mass. The barycentre has to be used for vis viva, giving r=1845m. However, both equations are AM conservation equations (hence the name vis viva, based on Keplers 3rd law) and so I’m not quite sure why that small discrepancy is there of 0.043m/sec (0.601-0.558).
This means we’d have a scenario that seems right. The head lobe was released some time after it went into negative g and had built up some excess tangential speed to break cohesive bonds. It then sheared and simply went into a low orbit. It also just happened to drag up some neck material that caused it to orbit slightly lower than it might have. But it was all very fluid and weightless after stretch, just orbiting with this hour-glass neck that was hourglass due to being stuck between the lobes and not being at orbital speed. So it wasn’t supporting the head, just dangling between the two lobes.
It might seem obvious that the head needed to be weightless to stretch or indeed be in negative g during the stretch, and there would also be a point where it would go from negative g to positive g and be weightless by definition through the sign change.
However, in this scenario we have a spin up to a particular required rate for slabs to have escaped and the head to coincidentally be just past the cusp to break away and go into orbit, not escape. Then we have a spun-down rate that’s completely constrained by that initial required spin rate for everything to work properly and that spun down rate is dictated via AM-conservation. That final spin rate just so happens to be orbital speed for the radius we see the head lobe at today. That’s three finely tuned coincidences, all tied to one intial spin rate: escape speed threshold surpassed for slabs; orbital speed threshold surpassed for head lobe to detach but not escape and final dictated spin rate is the orbital speed required for today’s head lobe height.
So I think the head lobe orbited for some time afterwards with the neck dangling in tension between the two lobes- hence the hourglass. I think the neck probably helped with the final spin-down to 12.4 hours in some way. It would have been constantly trying to move backwards like the drifting rocks.
After spin-down from 5.772 hours to 12.4 hours, the head wanted to drop lower but simply sat down on the neck. Of course, it probably dropped a bit by compressing the neck a bit but that’s the finer detail we know must’ve happened because it really is sitting on the neck in a positive g field.
Harvey and Gerald, ESA doesn’t need the above calculations. You seem to think that Rosetta scientists had to wait for my calculations before they would be able to consider stretch as a possibility. As soon as they see outgassing…spin-up…’centrifugal’ force…any scientist can and should investigate the possibility of stretch, grab their superior inputs, crank up their supercomputers and do the calcs for themselves.
Where I may be of some help is with the matches. I have a very intimate knowledge of every last bump and dip around the head rim and shear line.
I would suggest that you might get a better audience, and chance of passing a refereeing process, with a more succinct style.
Sorry, but I totally fail to see how these calculations are relevant to ‘stretch theory’.
You are looking at a condition where loose material would be lost to space; I have no issue with that being possible. In fact I did the sum very early on, wondering if it might be true now. It might well be a way to loose chunks off the surface, but that’s pretty straightforward. If that’s all we are talking about I have no particular argument with it, but it’s not very dramatic either. I thought the idea was to stretch it to this shape.
If the comet is to stretch, plastically deform into its current shape:
What credible material is it made of, which is plastic at comet temperatures and has the right density? (Most will brittle fracture.)
What force is needed to plastically deform it, or a range of reasonable estimates?
What rotation rate does that need – far faster for sure.?
Why did whatever process that span it up then stop? Obviously the increasing moment of intertia would slow it down, but why didn’t it then carry on speeding up again as it apparently had done for quite a while?
ie, Why did the process not lead to fracture? ( I looked at that, might be possible. Unless it continues to spin up, why didn’t it.)
Why and how did a process which seems to have exactly reversed the spin up (on an object with now differing moment of inertia) occur, it seems for all such objects since all I know of have low spin rates?
I’ve always struggled with contact binary.
I still struggle just as much if not more with this.
‘ESA scientists’ will spend their time on what they judge to be most credible; if you can make a credible case, I’m sure it will get looked at; but you haven’t so far.
You will need to make that case in far, far less words to get their attention.
Hi Harvey,
I have a succinct style. pick me!
What credible material? What credible comet temperatures?
Anything from pebble sized ice/dust grains at 20 to 30 Kelvin to a liquid “mud” mixture at a more balmy 300K plus will act plastically. Low density helps. There are a fair few suggestions in various papers that the crust is hard and the interior soft or “non-rigid” as Gerald would put it. The damping of torque free precession is a HUGE clue that there is considerable plastic internal movements.
The range of rotations plausibly required is what you asked for, and ACooper supplied. NO. They are not faster than what he has calculated, and they are perfectly reasonable, and the speed ups and slow downs are perfectly reasonable at current rates of speed up and slow down.
Dear ESA. It stretched…..
Is that too many words?
Should I add – see for yourself!
That could be the whole of My paper….
I’m close to Harvey. The scenario is close to the “standard” fission / fragmentation scenarios for small asteroids with relevant tensile strength, “ridgid” model.
Rubble piles (more likely scenario for objects the size of 67P) would shed mass from the surface most distant to the axis of rotation on spin-up, and change shape, first to Maclaurin, then to Jacobi ellipsoid, with some delay relative to the liquid model, and shapes only approximately, due to friction (shear strength).
The lost mass may or may not accrete to a companion.
Surface features of the primary would go lost during this process, at least for the companion.
The second scenario might be compatible with the “onion” structures, interpreted as re-accreted mass.
Then capture/collision of the two bodies.
Both scenarios without “stretch”, since the “ridgid” scenario is brittle, and the rubble pile scenario doesn’t preserve surface topography.
Harvey
Did someone boot the goalposts over a few inches while I was gone, replying to Dr. Massironi?
For over a year you’ve been questioning whether the head could be flung from the body via spin-up. You said early on that you’d done the calculations and it didn’t stack up because it would have to be spun too fast. I made an informed guess at 2-3 hours but wanted a better head:body mass ratio.
Now I’ve done the calculations, including a reasonable spin-up speed, explanation for spin-down, AM conservation and reason for the head detaching without escaping while slabs fly off to escape at the extremities. That’s everything you asked for and more.
You now seem to be saying that the head lobe is just another fragment flying off so what’s the big deal? But if it’s dragging neck up with it, it means it stretched! That’s what we’ve been talking about for over a year- the neck stretching with the two lobes sitting passively at each end was always the definition of 67P stretching.
I’m wondering if you’ve misinterpreted the calculations above, since you say,
“It might well be a way to loose chunks off the surface, but that’s pretty straightforward. If that’s all we are talking about I have no particular argument with it, but it’s not very dramatic either. I thought the idea was to stretch it to this shape.”
Those calculations can’t really be described as dealing with chunks on the surface because it includes the entire head lobe with its c of g buried 600m below the actual surface. If this is a chunk on the surface, then it’s not conceptually very far from having a head lobe that’s half the mass, the comet splitting in two, and saying they were just chunks flying off the surface of each other. It makes no real sense. If we dance on a semantic pinhead all the way down to calling a tear through the middle of the comet the surface of the comet then fine, I can just about get my head around the topology and say yes, that’s the essence of spin-up and ‘centrifugal’ forces. And it follows that you now have a “chunk” in the form of an entire head lobe that was thrown from the “surface” of the body lobe i.e. a tear a quarter of the way into the the original comet. It’s the same principle of spin-up and centrifugal forces and it means the head rose from the body.
Chunks on the surface can only reasonably refer to quasi point masses on the actual surface. That is indeed relevant for the slabs flying off at the extremities. I think we’ve been in agreement on the physics of that, even if not agreed on the actuality.
As for the question of the neck stopping all this from happening, I’ve mentioned the low tensile stress of the neck material before in comparison with more modest spin-ups. For the neck, it’s 10-20pa (Thomas et al 2015). The highest estimate for the crust is 40pa.
I calculated the negative g force on the shear line to be 273 pa at the time of shear. Inputs: r= 1095m; T=2.033 hours; head lobe radius at shear plane 1.25km; head lobe mass 2.5E12 kg.
The 20 pa sapping of the upward tensile force by the neck is 20/273 of the whole, which is 7.5%. So the upward force that’s left for lifting the head is 92.5% of the amount used in the last comment. And due to linearity between terms, the negative g acceleration is 92.5% as well.
This hardly affects the scenario in the last comment and even with a 40pa neck resistance it would be a comparable output- the head lobe would rise to about the same place, dragging up the neck with it.
Gerald
You say ‘I’m close to Harvey’ so you might want to look at my reply to him.
Thanks for your input on models. You seemed to get pretty close to my thinking further down in the thread, in comments dated 12th Oct.
However you’re not talking here about the model in comments further down as far as I can see. Here, you mention a rigid model at the beginning and discard it and then go on to discuss a rubble pile and possible problems with that.
Further down the thread, you talked of a model that’s liquid on geological timescales, a bit like bitumen, but solid in the short term. I get the concept, it makes sense and chimes with what I see on the body of 67P.
If solid during shear, it becomes two lumps which don’t stretch apart but fly apart, reapproach each other and only then share material.
So if I put all this together, are you saying that:
1) the pebble pile should be used because it’s more likely for 67P- but it will lose surface features.
2) if we did consider the rigid model, it wouldn’t have a stretching neck but would/could share material on reapproach.
Also, (3), I’m wondering if the rigid model you mention above is, after all, the same one that’s liquid on geological timescales but you’re calling it solid because it’s at the point of shear, so it’s short term. Is that correct? //////
Hi A.Cooper,
regarding similarity to Harvey’s position, I just intended to note, that your calculations didn’t discriminate between standard fission models and a true plastic stretch.
(3) is one option.
My thoughts have been slightly different: Viscid liquids behave like a liquid, if deformed slowly. They behave like a solid, if deformed rapidly.
Most people think of glass as a solid. But over long timescales it can behave like a liquid.
Bitumen a little above room temperature is easier feasible for experiments. It breaks brittle, if bended rapidly, but deforms plastically, if bended slowly.
Water ice near 0°C can show similar properties, but a lot slower.
One of the more famous experiments:
https://en.wikipedia.org/wiki/Pitch_drop_experiment
Applied to a scenario for the comet, one might suggest a viscid liquid model which behaves like a liquid with the well-investigated equilibrium figures, but exposed to stonger forces, it may behave like a solid or a rubble pile.
Stronger forces might be result of cometary activity due to a an orbit closer to the Sun (mainly spin changes).
I guess arguing along this line isn’t quite as easily discarded.
We see ices of supervolatiles flowing on Pluto. So there is a precedence for this behaviour on an object comparable to the region where 67P may have formed, and lived for a long period of time.
That’s far from any proof for “stretch”, but might be step closer to a scenario sufficiently credible to be considered.
A paper would need to include something which appears to be absent from the ’28 parter’, which I skimmed. Maybe I missed it, its rather long!
Namely some *quantitative* estimates of what sort of spin-up is needed depending on assumed material properties, how it got spun up, and how it got spun down again. Consideration of conservation of angular momentum, energy, the implied torques, etc etc.
In particular is the process stable, why did it stop & not split in two (I looked at that a bit back & it *did* seem possible stability wise.) It needs to consider if the material properties are credible; why didn’t it just fracture.
I have always found the contact binary theory rather counter intuitive, for reasons others expressed here. I tended o favor an erosion theory, but the paper is pretty convincing. Any ‘stretch theory’ has *at least* as many problems of credibility as contact binary does I’m afraid. The latter has detailed computer models saying that, counter intuitive or not, it can happen. The former currently lacks the most basic discussion of these points.
Yet more ‘look this looks like that’ won’t forward your case.
That aspect needs to be addressed with proper statistics.
But the *mechanism* needs addressing in *far* more detail if you are to have any hope of passing a refereeing process.
Hi Harvey,
Regarding the quantitative estimates and mechanisms of spin up, spin down – we have gone through a lot of scenarios with various numbers – some on the comments of this blog, plenty via private email and some you may have missed on the Scute blog.
It necessarily has different assumptions to the rotational disruption models and the main issue is “assumed material properties” . The material properties based on the evidence on 67P are not model-able in the traditional sense. That is, a multipoint strut system in the currently available software cannot reasonably represent what would actually happen on something with a rigid outer and a soft inner, especially when even the most knowledgable scientist could not estimate how hard and thick is the outer shell, nor how soft is the inner.
Hi Gerald, Harvey,
New post perihelion data is going to overtake any work done on this. I laid out clear criteria that would settle the case well before any arguments based on the current data are half way through.
Even when stretch theory is then proved right, I know you will still say it was right to reject all the evidence which we are trying to get bona fide scientists just to notice.
Trying to even organise a peer reviewed paper with the inadequate publicly available models and *comet measurements* is like beng told to write an essay with both hands tied behind your back before the magistrate would hear your case.
Hi Marco, if I would be as convinced as you seem to be, I would simply sit down, define the model formally, write the according software, and run it, to countercheck and refine my idea, and to provide clear evidence for the community, write a paper, and submit it, e.g. to arXive.
Thousand pages of pathetic excuses cannot replace this.
Why don’t you just do it?
Hi Gerald,
I do have my reasons, but at this stage, this blog and its comment threads is what I am using to countercheck and refine my ideas. Stretch theory is only part of this.
Good Matt down here 🙂
Hi Matt,
Well, why does this blog exist? As a publicity appeasement? We dont think so, we are here for serious outside perspectives, are we not? You clearly read this from time to time, so you have seen the long time A. Cooper blog or his many posts by now.
Are ESA scientists no longer curious at the top?
I found your question of whether there is a peer reviewed paper out yet (No, of course there is only A. Cooper’s blog and numerous posts here), on Stretch Theory, curt, snooty and “ivory tower”. Hopefully I am wrong, that it was an honest question, and I will certainly have to apologize.
Otherwise:
Booooooooooooooooooooo!
If “stretch theory” is so much better than everything else scientists think, it should be an easy exercise to write an according peer-reviewed paper. Where is the problem?
If you like to be part of the game, you should acknowledge the rules.
Hi Gerald,
Do you really think we cannot possibly be right because there is no peer reviewed paper?
Any original research can be science – whether it is written by a citizen scientist on a blog or passed for publication in a science journal is immaterial. Whether there is fancy computerised models or not is immaterial.
What matters is when it is proved right (or wrong) and science is about finding the truth. To help this process stretch theory is falsifiable to a much higher degree than any other specific hypothesis with data that is certain to be gathered by Rosetta in the latter part of the mission.
Expected rates of erosion on the purported match would instantly falsify stretch. For instance the pillars defining C. Alexander gate are an integral part of the stretch signatures and cannot have eroded even a little bit since stretch occurred, so erosion there would counter the assertion that the match is not a coincidence.
Hi Marco,
lack of a peer-reviewed paper is not a proof of being right.
Lack of a formal model is not a proof of being right.
And lack of a computer simulation is not a proof of being right, too.
Refusal of filling in these gaps is also not a proof of being right.
Even in the unikely case, that you would be right and you would win the jackpot, papers cannot reasonably refer to your ideas.
I know. Isn’t it fun being outside the system 🙂
This is an open blog Gerald. Peer review not necessary. Whether you consider seriously any of the ideas expressed is up to you. Others however may do so and leave you behind while you are waiting for reinforcement of your judgement from the peers..
Originaljohn, this may be right in some cases. But without the review you take a high risk to run miles into a totally wrong direction, particularly if you aren’t yet experienced in proper scientific work.
Whether people like something or not, isn’t necessarily correlated to scientific evidence.
So at some point, you may need to make a decision of whether your priority is entertainment or science.
Hi Gerald,
We are posting details of a theory here – You’re reviewing them. The miles in the wrong direction appears to be the province of peer pressure pushing that way at times helped by the peer review acceptance requirements.
On this blog we can use the rules of philosophy to look at it from the outside of the process and see its failings, as well as its successes.
Hi OriginalJohn. “…On this blog we can use the rules of philosophy”.
Is my view that you can use the rules of philosophy to define what science is going to be; but can’t use philosophy to do science. That’s meta.
Please excuse me OriginalJohn. That’s Marco’s.
I have had issues getting this posted, hence a delay.
It was a simple question which has been answered.
As for ESA scientists being curious – I am curious all over, not just at the top.
Hi Matt Taylor, Emily Baldwin,
Thanks again to Matt for engaging with this blog.
I have a question in regards to the model in this paper and others as to the calibration of measurements to the centre of gravity.
As I understand it, the models are built up from images mapped to two ellipsoids representing the two lobes. Thus very accurate 3D measurements between features on the individual lobes can be intimated from the model directly.
However, distances between features on opposite lobes depend on the distance and accuracy of the placement of the two lobes relative to each other: ie. requiring accurate measurement of the neck length.
My question is : How accurate is this neck measurement and at what points has this important measurement been calculated? I would assume measurements would have been made during close measurements in October 2014, but may have been refined again in March 2015.
What are the sources of error in this measurement and what is the possible error range?
Are there any papers written or planning to be written that determine whether the neck is completely rigid, morphs, or flexes with the torques from asymmetrical outgassing?
I ask this because this may or may not affect the research that depends on accurate placement of coordinates relative to Centre of Gravity onto the comet – eg the finding of Philae
Really going with this model from Massinori et al.
Wandering, and being remembered by the blog’s parishioners, at how could it survive dynamics of the proto-planetary disk. Also of the ineluctable general ‘integrity’ of the actual binary conjunct, as seen from angles wide open to neck’s plane.
…………
The answer could lie in the middle: It is binary, but sort of late ‘completed’ accretion into a single object [this ‘slim’ neck product of fast erosion?]. Maybe Ducky even went trough ice phase changes after the ‘contact’.
This improbable scenario introduce a new complexity: That of a binary comet going back to profundities of formative environment to get a ‘refill’. [Young stars and their ‘appropriated’ proto-planetary disk should tend to remain static, relative to their mother nebula kinetics].
Proto-planetary disks could be cyclically -‘tide’ kind- ‘shore’ washed by their mother nebula, progressively separating.
As layman, have no conscience of lots of variables and all of this could be non-sense. This is an issue too far away from actual ROSETTA hard data.
Why is bi-nary so difficult?
Accreted objects are essentially multi-nary.
Personally just find easier for them to have joined on the surface of a third than in the emptiness of space. [The ‘cementing’ material is already there].
For quite some period of time I’ve preferred a mostly erosional explanation for the duck shape, although a contact binary looked possible.
The detailed analysis of the stratigraphy and the unambiguous result came a little surprising to me, but I take it as evident, as long as there doesn’t occur some new overwhelming evidence of the contrary.
Shatter cone features (the parallel streaks) at Hathor are additional evidence for the contact binary hypothesis.
The layering as such looks reasonable. In an environment consisting of particles moving Brownian-like, with the option to adhere on contact by van-der-Waals forces, a fractal, snowflake-like nucleation starts, growing more or less concentrically.
That’s similar e.g. for growing crystals, but without lattice.
(Computer simulation of this principle is easy, just let partices move randomly, and test for contact.)
Assume some periodicity of the trajectory of the early nucleus in a not quite homogenious protoplanetary disk, and you get a stratigraphy similar to Steno’s law.
https://en.wikipedia.org/wiki/Principle_of_original_horizontality
Almost parallel trajectories in the protoplanetary disk result in low local relative velocities of the orbiting bodies.
(Otherwise accretion wouldn’t have started.)
This makes low collision velocities plausible.
Rarification after contact binary formation reduces probability for later collisions.
“Stretch” doesn’t look likely to me for several reasons, starting with the problem, that spinning up of a gravitationally bound “rubble pile” doesn’t result in a bilobic shape, but in a flattened rotational ellipsoid, a ring system, or a true two- or n-body system.
The only scenario would be a separation into a two-body system with a subsequent fusion into a contact binary.
A rigid body would simply disrupt on spin-up.
Hi Gerald,
That last line contradicts your arguments as to how torque free precession is damped. “Any non rigid model could dampen precession”
In what way is it non-rigid? Have any non-homogenous non-rigid models been tested for what happens on spin up?
If we do not have the right model of comet nucleus, the result of simulations will reflect the limitations of the model rather than what would happen in reality.
Hi Marco,
the discussion has been lasting for centuries, and lots of scenarios have been considered.
Here an overview paper about known scenarios for asteroids:
https://arxiv.org/ftp/arxiv/papers/0906/0906.4366.pdf
Another discussion about rotating “rubble piles”:
https://www.astro.umd.edu/~dcr/reprints/richardson_icarus173,349.pdf
There is a decades-long discussion about a Jacobi – dumbbell – binary transition for star formation, probably one of the closest you can get, regarding “stretch”:
https://www.astro.ulb.ac.be/~siess/pmwiki/pub/Exam_papers/Tohline.pdf
A html version with simulation movies:
https://www.phys.lsu.edu/astro/nap98/bf.final.html
A paper triggering much of the discussion (for stars) :
https://articles.adsabs.harvard.edu//full/1984PASJ…36..239H/0000239.000.html
Modelling of gravitationally bound rotating objects dates back to Newton.
https://arxiv.org/ftp/arxiv/papers/1409/1409.3858.pdf
A link to Chandrasekhar’s epochal work about rotating bodies:
https://articles.adsabs.harvard.edu//full/1963ApJ…137.1185C/0001185.000.html
A more recent overview of rotational phase transitions:
https://arxiv.org/pdf/astro-ph/9504062.pdf
Roche ellipsoids:
https://farside.ph.utexas.edu/teaching/336L/Fluidhtml/node41.html
Jacobi ellipsoids:
https://farside.ph.utexas.edu/teaching/336L/Fluidhtml/node40.html
Maclaurin spheroids:
https://farside.ph.utexas.edu/teaching/336L/Fluidhtml/node39.html
Rotating multi-body fluid systems:
https://ptp.oxfordjournals.org/content/70/6/1534.full.pdf
That’s just a small excerpt.
… regarding ridgid: The definition is a little fuzzy, and depends on the context.
In the strict sense it means only translation and rotation are allowed motions.
But real physical objects aren’t strictly ridgid. They can be deformed elastically (small force) or plastically (stronger force) to some degree. But within a considered time span they don’t flow like a liquid on small forces, and they aren’t compressible like a gas.
For damping under moderate conditions, the elastic, damped deformation is relevant; mechanical energy is transformed into heat.
The “rubble pile” model is somewhere between ridgid and liquid.
The point is 67P cannot be rigid because it has to be damping torque free precession.
However, for some reason it cannot be non-rigid in a way which allows stretch in the past or present, according to you.
You have argued that scientists believe that the surface may be rigid but the rest of the interior is porous and soft. In a model, this would behave nothing like the model of a “rubble pile” nor like the model of a rigid body, nor something in between.
You have linked to a lot of peer reviewed articles. I have looked and looked for models or even one model that shows what happens to a body that is not a rubble pile, not a liquid, not rigid, not anywhere in between those but is a complex composite of those specific to comets. I know it is exceedingly hard to model without knowing precisely when each one element is going to be a liquid, rubble, or rigid out of the multi million connected elements in a simulation.
Simulations are a poor substitute for observing what happens on a real comet in this case. I suggest you disbelieve the models and believe the comet observations.
Hi Marco, disbelieving the models would mean disbelieving physics, and disbelieving many of the best physicists and mathemeticians of the last 350 years.
All based on questionable perceptions and interpretations.
No sorry, that’s insufficient.
The least thing needed is one model simulation or calculation resulting in a body similar to that of 67P.
For this, the precise interior properties of the comet aren’t needed. Thus far, the existence of not even one solution with the “stretch” constraints has been shown.
… you may try to find arguments along the Maclaurin – Jacobi – (pear) – (dumbbell) – binary – contact binary – Maclaurin series, and try to find a solution for rubble piles.
First, there is no warrenty, that the pear and dumbbell equilibrium figures actually occur.
Second, a rubble pile of the dumbbell figure would return to the Jacoby sequence (ellipsoids) under spin-down due to its own gravity.
Assuming higher friction or adherence (ridgidity) would omit intermediate states on spin-up and disrupt directly to n-body systems (significant mass loss).
Snow-like adherent porous interior material, although ridgid in some range of parameters, would however allow for damping torque-free precession.
It would behave either like a rubble pile or like a ridgid body on a larger scale, relevant for sequences of equilibrium figures.
Heterogeneities would modify the sequences, but not fundamentally introduce new ones, since those cases have been investigated extensively by Eriguchi and Hachisu:
https://ea.c.u-tokyo.ac.jp/astro/Members/eriguchi/paper.html
You may get new shapes with rubble piles made of only few fragments. But then we are again at the contact binary solution.
To overcome this, you need to add more and more constraints. Those weill hence become a-priori less likely. But even if ignoring this, e.g. due to one-sample statistics, the necessary properties will be constraint so narrow, that they can be verified with observed properties.
Consistency is again unlikely. But if present, it would provide evidence, because of this low a-priori probability.
The problem: The latter only hypothetical right now.
Contact binaries are much better justified.
They may either be a rubble pile of just two fragments, or the result of a spin-down of a true binary.
A spin-up of the (somewhat hypothetical) dumbbell or pear figures for rubble piles would rearrange the fragments, hence show some similarity to the “stretch” scenario, but potentially matching surface features used to justify “stretch” would go lost.
… an outer crust with a rubble-pile or liquid-model interior would either be strong enough to mimic the ridgid model on spin-up, hence disrupt into a n-body system / significant mass loss, or it would be too soft to keep the bilobic structure after spin-down.
“Tower” or “needle” (extreme Jacobi ellipsoids) rubble piiles, however, can survive spin-down due to friction.
Hi Gerald,
I am suggesting an outer crust with rubble pile and liquid interior, so I guess this comment is the most pertinent to reply to. Now, when the outer crust in 67P ruptures, the internal material explosively generates gas and dust, but the internal substance appears to seal the rupture and form a new crust within seconds, minutes, or longer depending on the size of the rupture.
Interestingly there is considerable evidence of slab loss coincident with rupture and stretch, so this mainly fits in to what you are saying, rigid elements breaking off, n body system of crust components. However after rupture into two main lobes plus shrapnel, there is enough time for the internal liquid infused rubble pile to get a hardened crust itself, after having stretched into a bilobed teardrop feature. Conservation of AM does all the work in slowing the system back down to below disruption speed. The hardening neck crust has enough time to become rigid enough to hold their mutual gravity.
Note that this isn’t done backwards from engineering models that can have a stretch solution, but forwards from observing the actual properties of internal and external materials, and then patching it to those partial information heterogenous model simulations (with a thought experiment in lieu of having simulators that can model hardening liquids on vacuum exposure)
Marco, nothing wrong about your logic. But you are trying to use Aristotelean discourse in scientific, planetary issues. This is simply, ultimately incorrect.
Hi Logan,
Expand on what you mean by Aristotlean discourse. Philosophy of science and science are intertwining here. My objections to a lot of “science” is that it is not about repeatable observations, and thus the use of parsimony is fair game as being outside of science. I routinely ignore any science that at its heart relies only on parsimony against other competing theories. Truth by consensus is a way for the actual truth to hide behind a wall of burden of proof that the consensus view does not. It is very hard to look behind that wall.
Truth, for start [What’s that? (to a Scientist’s mind)]. To me Science ends at the point where it can predict [on a wining average], with a shared language, to the question: What if..?
What if today is Dec1? where is Ducky? What’s her level of activity? Is her broken onto 2, 8, hundred pieces?
That’s Science, and Scientists can answer to those questions, on a certainly wining average.
Hi Logan, I do believe in an absolute “truth” and science as a way for modelling that truth to make the predictions that they make. The closer the model gets to the essence of what it is modelling, the better the predictions will be. Checking the model’s predictive power with real data is paramount. Having data that “doesn’t contradict” a model does not strengthen the model if that model didn’t predict that data – It actually weakens it.
The Whipple model is “very weak” in the very criteria you have laid out as what science is striving towards. We need to find the truth that is hiding behind this restrictive model.
Do you realize, Marco, that we can have several models of a single process, Giving reasonable outputs, but Supported on quite heterogeneous theories?
In two words, all valid?
Do you realize that, If given a different evolution path for our history of Science, Then actual Science would rest on a different set of axioms? and have a very different set of challenges, and great enigmas?
Cheer up, Marco. Be humble and admit limits of thought. A product of nature itself, and unable to position out of the equation.
… so, what do you need? A spin-up of an overall liquid model (might be a rubble pile is possible, too) up to a pear equilibrium figure, otherwise you wouldn’t get the pear figure. Then switch to the ridgid model, spin-down. The bilobic liquid model with non-equal lobes is very unstable under spin-down and turns to the Jacobi sequence by flow of the small lobe to the large lobe. Therefore the liquid model doesn’t do the job for spin-down.
How does the switch from the liquid/rubble pile to the solid (ridgid) model to the other work?
You might assume a thickening crust just in the right instant between spin-up and spin-down.
How can this work?
Spin-up during orbital approach to the Sun, then constant spin for some period of time, together with surface sublimation and sintering, formation of a thick crust able to sustain internal pressure on spin-down. Then actual spin-down with further hardening of the crust.
The problem: The surface features during spin-up are lost during spin-up, since we needed to assume the liquid model, or at least a rubble pile model.
How can this be solved?
A crust strong enough to stay intact under its own inertial forces and gravity, but allowing an overall liquid model behavior.
Doesn’t look like fitting. But maybe another series of ifs can do somehow.
Is this a likely scenario, considering two straightforward scenarios, contact binary and erosive?
Hi Gerald,
This last series of posts at least addresses the question of justifying ignoring stretch as a possibility. Holger should have answered the question in such a way that stretch scenarios are considered impossible for comets due to no simulations coming up with a stretch solution. If you are a pseudonym for Holger, I apologise because you kind of said that at an earlier point of the blog but I wouldn’t have known it came from the authority of a PI in the Rosetta mission.
However, you pretty much answered my question in the last line. It essentially boils down to parsimony – two straightforward scenarios easy to model and contrast, versus a complex set of conditions seemingly having to be engineered just right to have happened.
That complex set of conditions really does appear to be the case from morphology evidence on the comet, if anybody other than A.Cooper could be bothered to test surface features with thought experiments rather than assuming everything on the comet is eroding primordial ice and dust, no matter how it is shaped and the relationship with other features.
Gerald
You’ve pretty much come to the model scenario as constrained by the evidence in the stretch blog.
Before moving on to showing how close you are to how I think the stretch event played out, I should assuage your concerns regarding the two caveats you have.
The first caveat appears to be based on an assumption that pre-shear spin-up and post-shear spin-down necessarily have to be over a short period. It sounds as if you are saying your scenario is happening or perhaps even needs to happen over a particular perihelion passage with spin-up on the way in and spin-down on the way out.
The second caveat is to do with what you refer to as ‘ifs’, a good term for any reasonable worry to do with engineering a ‘just-so’ scenario. The ifs are in relation to your concern that the crust needs to be hard enough to retain the matches during spin-up and yet allow stretch to happen.
These two caveats can be resolved by bearing in mind the following 9 general parameters that I have always considered applicable to the stretch event. These aren’t ifs, just basic, reasonable assumptions based on the evidence. After I’ve laid them out, I shall go through the stretch event chronology, as I see it played out, in 11 stages. All the while, I bear in mind your model and point out the junctures where your caveats are now addressed.
List of 9 general parameters:
1) Firstly, both spin-up and spin-down could happen over many orbits and probably did.
2) The shear event itself may not have happened at perihelion, because it only needed to reach a cusp rotation period at which the head lobe could no longer stay adhered to the body. That would most likely happen at perihelion, yes, but would be quite likely on approach or even on retreat depending on surface morphology presented to radiative heating and also stored heat.
3) Thirdly, post-shear rotation rate could be maintained for a very long time, allowing the neck to harden. This would be an unremarkable panning-out of the post-strech scenario if we assume point 1 is in play i.e. much longer overall time frames. So it wouldn’t be an ‘if’. You touch on the idea of spin-up being maintained for long enough for sintering and hardening: “then constant spin for some period of time, together with surface sublimation and sintering, formation of a thick crust able to sustain internal pressure on spin-down. Then actual spin-down with further hardening of the crust”. I agree, especially if you’re saying that the constant spin is over a long period of time, not necessarily over one retreat from perihelion (although, who knows, maybe 6 months is enough to harden up the neck, though I doubt it). Also, just a quick word on “constant spin” for other readers: this wouldn’t be the spin-up speed before shear. It would slow dramatically directly after shear and stretch (see my workings above for 2.033 hrs down to 5.779 hrs). One might think that this slowdown means there’s already a problem with imminent neck collapse but if the head was flung away and ended up in orbit at or near the current height, it was weightless with respect to any compression on the neck, i.e. weightless at a 5.779-hour, 1000-metre orbit. My calculations above suggest this is quite likely. So “constant spin” means constant post-stretch spin rate. I’m pretty sure you meant a slowed post-stretch spin too, Gerald, even if you don’t agree on 5.779 hours. Correct me if you meant something else by “constant spin”.
4) If the head lobe was in orbit after stretch and the post stretch spin-down is thought to have happened over a very long period, many orbits, it means that there was plenty of time for sintering and neck hardening.
5) On shearing, almost all tensile forces got transferred to the growing neck and the two lobes, especially their crusts (and crucially, their matches) were completely relieved of stress. This means the matches were preserved from that point onwards, both during stretch and spin-down and probably for centuries or millennia thereafter. Being crust, spent of volatiles, there is no scope for suface sublimation to erode the matches. Even subsurface gases appear to emerge from the neck next to the matches, going transversely under the crust towards the line of least resistance. This way, the matches perched right on the edge of the crust remain pristine.
6) This is related to (5). You say, “The surface features during spin-up are lost during spin-up, since we needed to assume the liquid model, or at least a rubble pile model.” However, it doesn’t matter what happens to the surface features during spin-up. They could deform as much as they liked, even along the soon-to-be shear line/head rim. That’s because they were still attached. They could deform and stretch as much as they liked up to the very last second because it was only at the second of shearing that the matches were generated….and since all tensile stress on the crust and the shear line were instantaneously relieved by stress transfer to the neck as per (5), the matches were instantaneously generated and preserved at once (in practice it seems to have been an unzipping so this process would have occurred over a short period but would be instantaneous for any two matching points being generated as they tore apart along the zip). There are clear signs of deformation-before-shear of the joined head and body in the form of the bell-shaped frill around the back of the head, especially the triangular “yellow match” that’s triangular because it was stretched.
7) The crust had some capability to deform as evidenced by the V shapes on both head and body (body is a diamond of two V’s). Also, there’s the fact that the head, including its crust deformed far more than the body showing that it was actually quite pliable. This is especially noticeable at the head rim where highly localised areas were bent round at the base as it ‘herniated’ out of the body and into a bell shape (see below). Also, the onion layers in the crust were capable of sliding over each other as is visible on the head lobe. This capacity to deform might be akin to the Earth’s mantle deforming over long periods, essentially solid but deforming gradually. This supports the idea that pre-shear stretch happened over many orbits. I’m also now wondering if the onion layers were actually caused by the stretching, causing concentric strata to differentiate and slide over each other. If that’s the case then the crust was actually deforming by definition of the stretching.
8) Despite the crust deforming in the long term i.e. in sympathy with the stretching core beneath (stretching into the diamond shape), it was able to allow the core to deform faster and slide beneath it. This is evidenced by the ‘red triangle recoil’ in Part 26 (signature 6) where the crust slid back 200 metres from its original core position.
9) Although (7) and (8) were at play, deformation and sliding, the crust couldn’t deform and slide in all places under all stretch vector forces. Hence the flayed area next to Landing Site A (to the right looking down). It cracked, like stretching chocolate-covered Turkish delight. Also some slabs were probably loosened via this mechanism before being ejected by further spin-up. It’s probably worth imagining chocolate-covered Turkish delight for the stretching core/crust. But without the elastic properties of jelly for the core, which was more plastic.
Now we can revisit your scenario and romp through it unencumbered by any objections. Here are the 11 spin-up scenario points:
1) Spin-up was gradual over many orbits hence gradual deformation of the crust accommodating the core stretching beneath it. Deformation lines appeared in sympathy with the V-shaped aspect of the increasingly obvious diamond shape. On the body, this is evidenced by the so-called ‘red’ triangle of craters and two lines in the flayed area, parallel to its border. The perimeter of the diamond is a stretch vector in itself. Some crust was flayed off too. Neither deformation nor flaying matters because the matches weren’t generated yet.
2) The head lobe (possibly a proto lump) towards the front end of the stretching diamond, started stretching in an accelerated manner due to the fact that it had a head start being a lump at one extremity (higher r= higher ‘centrifugal’ force). It herniated into a bell-shape atop the diamond and towards one end.
3) This herniation lead to a sort of pear shape: a diamond with an obvious bell-shape at one end. Is this what you mean by a pear? We know it was bell-shaped because today’s head lobe retains that shape and the bottom curve of the bell around its back rim.
*It is for this reason that the surface layers in the scientific paper for this post (Massironi et al 2015) appear to be normal to the gravity vector- they were curved up from the body with the bulging head.*
That’s why the gravity vector relationship breaks down for the underside of the head, which has no ancient crust but does have the visible side-on fracture planes of the onion layers. So the head is in fact half an onion, sliced from the body.
4) The spin-up cusp was reached (spin rate of between 2 and 3 hours) and the head sheared. It was more complicated than a single shear but for the purposes of your modelling it’s good.
5) On shear, the tensile forces were instantaneously transferred to the neck and the matches were therefore preserved for the long term in the crust of the shear line (on the body) and head rim.
6) The head was ejected 1000 metres over around 90 to 150 minutes into orbit at that height, dragging neck material up with it. The dragging-up signature is preserved in the four zigzagging but notionally vertical lines at Anuket.
7) Around half the spin-down or more was achieved during the stretch event itself due to angular momentum conservation. Immediately post stretch the rotation rate was between 5.8 and 8 hours.
8) The head swung 15° anticlockwise on an eccentric axis during the last third of the stretch (clear 15° offsets, torsion in the 4 lines).
9) The rotation plane precessed anticlockwise about the longest axis by about 20° in response to the head rotation and slab loss.
10) Spin-down from the quoted 5.779 hours (or possibly 8 hours) to today’s 12.4 hours happened over many centuries. This meant that the head lobe could stay almost in orbit, i.e. almost weightless for centuries before slowly seating itself down on the neck. This allowed the neck to harden at its leisure. This satisfies your switching to the solid model from the liquid model. In fact the orbiting head acts as a stand-in for the solid model while neck-hardening takes place. It was only the neck that needed to harden. That’s because only the neck needed to take the compressive forces after spin-down.
11) The spin-down to 12.4 hours means the head might never have experienced the full g acceleration of the gravity field since the stretch event, meaning it has to this day carried on in its partial stand-in role to help the neck resist the compression. In other words, the moderate spin rate is still acting as a proxy for the solid model.
This isn’t the whole detailed story of the stretch as I see it from the visible evidence but it does describe the key dynamical and morphological points in the stretch event and, I believe, all the points necessary to make your latest model (and model switching in mid stretch) work with your caveats resolved and no ifs.
Great…
Now I’m craving Turkish Delight, and the shops aren’t even open yet ;-(
Hi Marco,
the theoretical existence of an engineered “stretch” model is still to be shown.
A long way to show 67P actually being consistent with such a model.
…And funny, but I’m not Holger Sierks.
Hi A.Cooper,
here a link to the pear-shaped equilibrium figure, as investigated by G.H.Darwin:
https://rsta.royalsocietypublishing.org/content/roypta/200/321-330/251.full.pdf
A.Cooper,
… In order to get closer to a plausible scenario, without too much of engineering voodoo:
Assume the material of the comet behaving like a (vortex-free) liquid in geological time-scales, and short-term like a solid.
Examples on Earth are rock salt or bitumen. That’s probably closer to the physical conditions on the comet than those of Earth mantle material.
Assume a spin-up within geological time-scales e.g. by YORP up to a liquid-model (asymmetric) pear shape.
Then a disruption (fission) to an (asymmetric) binary of a solid model due to either rapid spin-up by cometary activity, tidal forces by close encounter to Jupiter, or collision. Dumbbell sequence is omitted, since timescales have been too short to get an equilibrium.
The rubble pile model probably would’t work, since it would predict equatorial mass shedding,
see e.g. section 5.3 in this overview paper about asteroid surface geophysics:
https://arxiv.org/pdf/1503.01931.pdf
Let the binary fuse to a contact binary (two-“pebble” rubble pile) after spin down, e.g. again due to cometary activity.
This sequence might come into reach for simulations, and it wouldn’t contradict a contact binary scenario.
Consistence with observation is another question.
Hi Gerald,
I’m not sure what you mean by engineering voodoo. All we are doing is reconstructing ie. Reverse engineering using the evidence and what we know about physics directly. If we knew 100% it stretched we would have to explain it regardless of not having a model solution for it. If it can be engineered, it can happen in nature. I think engineering is a very good way to look at it. If it cannot be engineered using known physics, then it can’t happen, but if it can, then we can look at the reasonableness of what intelligence or luck is required.
Hi Marco. “If it can be engineered, it can happen in nature”.
??
Aliens needed at some point on the middle of the process 😉
ROSETTAs can be engineered, middling human work. If you take nature in the wider of definitions, Then yes, ROSETTAs can happen in nature.
Hi Marco,
in case the other post has been swallowed, here the short version:
Why haven’t the presumed chemicals of the presumed liquid interior been observed during the outbursts?
Logan
Hi Marco. “If it can be engineered, it can happen in nature”.
??
Aliens needed at some point on the middle of the process
You laugh, but if we find something that is engineered, we have to postulate something like that. I really am not ruling t out……
Oh, I think, the primary formation led to a rotational ellipsoid, which got stretched later on, probably by the comet entering a strong gravitational field! The comet was almost torn apart, when it flew on. Remove the “neck” (some of the innards that kept the parts sticking together, look at how the “head” part is sunk in, feeding some material to the “neck”) and put together the “head” and the “body” to convice yourself :-).
Perhaps it would be useful to put your math skills to work and determine what the amount of torque is at the neck that’s being caused by the rotation of the comet, because whatever is gluing the (supposedly) two separate comets together has to be strong enough to continually overcome that force. And the underground surface area of the fusion is very large, what exactly was the process that maintained that much complete contact between the two (supposedly) comets long enough to fuse them together, and how exactly did the fusing happen? I mean, there are not flat surfaces on either (supposedly) comet, so the two (supposedly) melded surfaces were no doubt not flat either, yet here they are seamlessly melded together. And aren’t just about all free traveling bodies in space spinning to some degree (like P67 is now)? How do two spinning, round, uneven bodies get glued together? And why are both (supposedly) comets so similar in appearance and composition, and why do the opposite cliff tops of the neck line up so perfectly? I guess I should just accept that yet again, what the comet obviously looks like it’s not.
Capta problems again, retry.
“Perhaps it would be useful to put your math skills to work and determine what the amount of torque is at the neck that’s being caused by the rotation of the comet, because whatever is gluing the (supposedly) two separate comets together has to be strong enough to continually overcome that force”
Maths skills required are limited, because the answer is zero.
Torques cause angular *accelerations*. For the body to continue in constant rotation requires no torque at all.
It’s just the rotating nag version of Newton’s first law of motion!
Torques *are* required to spin it up or spin it down, exactly the sort of quantitative discussion missing entirely from the ‘stretch theory’.
Hi Harvey,
If he had said torque at the neck due to the increase in rotation of the comet, he may have a point. Asymmetrical outgassing causing accelerating rotation which has been measured would have some sort of torsional effect at the neck.
I think your implication that we have not justified the level of torque required to spin it up to partial fracture is unfair.
But that is *not* what was said.
The statement is crysal clear.
“torque is at the neck that’s being caused by the rotation of the comet,”
The is no such torque.
Its says nothing about *acceleration*.
It demonstates a basic failure to understand mechanics.
Hi Harvey,
Ok. But I am interested now.
Perhaps it would be useful to put your math skills to work and determine what the amount of torque is at the neck that’s being caused by the change in rotation of the comet.
I am also not impressed by SS’s poor understanding of mechanics, but I am politely leaving open the possibility that it was a typo and he just left out a couple of pertinent words.
The point is, despite your protestation, there is torque acting to spin it up (Mottola et al 2014) and this may be causing the ephemeral cracks in the neck. The damped rocking may even be ratcheting the neck longer and thinner as it does this. It is likely to be damped because otherwise precession would ensue, and it is likely to lengthen the neck very slightly due to angular momentum exchange between the lobes in the process.
A little speculative perhaps, but there are cracks, there is torque, there is spin up, there is damping and conservation of AM is a physical law. I am sure the mission scientists are calculating the effect of torque on the neck, observing the occurrence of cracks, keeping track of the measurements of the neck and measuring the rotation period. I await with baited breath for the results we should expect sometime next year….
Marco. I don’t have accurate values for the angular acceleration. One could roughly approximate the moment of inertia easily enough. But even over many months the reported changes in rotational period that I’ve seen have been very small. The implied angular accelerations is thus really very small indeed; so despite the high moment of inertia I think the torques will be pretty small, but I’ve not run the numbers.
Hi Harvey,
The numbers might not matter anyway. Numbers are a poor substitute to seeing the effects with accurate before and after measurements off the neck.
Marco, it’s at least obvious, that the mean acceleration by torque is only a tiny fraction of the gravity, and for almost (except very close to the axis of rotation) all the surface only a tiny fraction of the centrifugal force.
The delta-v per day due to spin changes is certainly well below one centimeter.
A valid argument might go along the lines of a very virulent outburst, meaning a velocity change of just 1mm/s could cause damage, if it just happens sufficiently fast.
A remote option might be some yet to discover mechanical resonance.
Due to long-term spin changes you may get a shift of the balance involving gravity and centrifugal forces.
The residual forces may be the best candidates for (indirect) effects caused by spin changes.
@Harvey,
When you say, “It demonstrates a basic failure to understand mechanics,” you’re right, I find this the case every time I take my car in for repairs and listen to one tell me what’s wrong. But seriously, point taken, I’m sure us peanut gallery folk post all kinds of crazy stuff, and me more than most no doubt. I’m out of my depth with the technical science, and as such generally try to keep my posts more along the lines of what are more my strengths, which are critiquing systems of thought and logic. For any discipline or system of thought to be matured, both inductive and deductive reasonings must eventually merge and support each other. As Sylvia Wassertheil-Smoller says, “In science there is a constant interplay between inductive inference (based on observations) and deductive inference (based on theory), until we get closer and closer to the ‘truth,’ which we can only approach but not ascertain with complete certainty.” My strong sense is that many in cosmology often seem to have too strong a belief in the accurateness of the picture on the box of the cosmic jigsaw puzzle, based on a pretty limited number of puzzle pieces. I think the article of this post is a case in point. Perhaps if the authors had said that there is evidence that could support P67 being a binary, and laid out that evidence without the unequivocal conclusion, it would have left the question open enough to invite further investigation and debate. Instead, they have proclaimed that they have ascertained the “truth,” which is always a rather precarious position to take.
I don’t think that they seriously claim to know “the” truth in a strict sense. It’s always just up to some confidence level.
To call something a “discovery” the usual convention is at least a 5-sigma confidence, meaning a risk of about one in three millions, at most, to be wrong.
I’m not even sure, whether the paper claims a 5-sigma confidence at all.
Headlines tend to simplify.
Sovereign Slave, some of your questions have been answered by the simulation of the fusion a while ago.
The fusing comet(esimals) have been sufficiently soft to be compressed to some degree by the collision, but not destroyed. The porosity allowed for some compression, and plastic damping of the collision, including spin differences.
As “glue” the low gravity plus some friction and adherence or sintering, are sufficient.
Similar composition (although I’m not sure whether that’s really the case) is explained by the lack of gravitational differentiation. It would have needed some liquid medium plus sufficient gravity.
Planets are sufficiently massive, and initially liquid in most of their interior. So heavy materials like iron can sink to the center, and light material like water or gasses swim to the top. As a result the initially homogeniously mixed dust is partially sorted, leading to rather different and separated kinds of rocks, liquids, gasses.
A comet, on the other hand, consists mostly of well-mixed primordial dust and ice grains of the protoplanetary (or protostellar) disk.
Although different zones of this disk may (or probably) have been of a different composition, but that’s locally less relevant.
Due to similar composition and environment, surface processes are similar and lead to similar surface features.
But I think, there is more detail to be investigated regarding this question.
Yes, “looking like” is only one ingredient of the whole story.
Well, Gerald, thanks for penning a serious response to my oftentimes smart-as* postings, but at the risk of carrying my tradition a bit further, I’ll do my best to interpret your above post:
So, what you are saying is that IF the comets are soft enough (first, what is the definition of “soft,” are there any real world measurements or solid data indicating exactly how soft they’d have to be, or to what depth, and area size, and would they both have to be the exact same softness?); and IF they’re compressed to some degree (how much exactly, to what depth, over how big an area, and wouldn’t they have to compress in absolute exact balance (center of gravity to center of gravity, so to speak) to prevent the creation of spin and the subsequent spinning off from each other?); IF the porousity allowed for some compression and plastic damping of the collision, including spin (how much, has this been tested to scale? Also, you mistakenly presented this as an assertion or fact, but it is actually a supposition; also, if I’m reading you right, you seem to have slipped in that these factors would be sufficient to overcome spin, which is even more speculative and unlikely); IF the gravity is just right; IF the dampening is just right; IF low gravity plus some friction and adherence actually are sufficient as glue (again, supposition stated as fact); IF there is a lack of gravitational differentiation; IF there was a liquid medium and sufficient gravity; IF comets consist mostly of well-mixed primordial dust and ice grains of the protoplanetary (or protostellar) disk (another assumption, belief actually, stated as fact). Anyway, these are just a fraction of the many speculative IF’s that no doubt had to be fed and re-fed into the computer over and over again to get it to behave and spit out the desired results so that it can be claimed that, “Ta-da! We have a winner, a computer generated model that totally proves that two comets can get glued together.” Sure, modeling can be useful, but this just smacks of unlikelihood on par with proving that sus scrofa domestica have volar tendencies.
Well, Sovereign Slave, sequencing 99.9% confident IFs is different from sequening 0.001% confident IFs.
The IFs they’ve sequenced still multiply up to a reasonable value, such that the simulation can be used as a demonstration, that the scenario is physically reasonable and possible.
Gerald, in law there is an important distinction between a simulation presenting evidence vs a simulation being presented AS evidence. It is acknowledged that presenting a simulation AS evidence is a very dicey proposition filled with all kinds of potential errors, deceptive variables, and unreliability, and is treated accordingly. Might be something to consider.
In a word, simulating that something might happen is not evidence that it did.
Sovereign Slave, nothing else that it’s physically possible, did I say.
Other simulations with different model assumptions may return a similar result.
But currently there are no such other published simulations.
So we have actual simulations, and the hypothesis, that there might exist other simulations at some point in the future.
About hypothetial future simulations we can only say, that they currently neither prove the existence nor the absence of other options.
Have been missing 3 body scenario since April.
https://blogs.esa.int/rosetta/2015/10/01/rosettas-first-peek-at-the-comets-south-pole/#comment-553222
Should be noted that dominant layering not necessarily correspond to accretion. [Impact wave propagation could dominate, by example].
At Imhotep dominant layering is that of flow deposition.
Also, in that Bill’s Escher -esque spirit, layers could belong to smaller objects within.
Given 1/ the considerable distances (relative to their size) which have presumably always separated comets from each other, 2/ their necessarily heterogeneous trajectories and 3/ their relatively high velocities with respect to their rest-frame and to each other, it would seem, from the purely mathematical viewpoint of probability, that the eventuality of two comets anywhere in the solar system ever coming into extremely gentle contact with each other at all (let alone then, somehow, managing to stick together…) is vanishingly small. A second reason for scepticism resides in a second level of (im)probability, namely that it is precisely this extremely rare object (if such objects even exist at all…) which was somehow chosen, by an infinitely improbable stroke of luck, for Man’s first-ever true rendez-vous with a comet in the shape of the Rosetta mission… It looks to me, in that case, like the winning billion-dollar ticket in a cosmic lottery.
For this reason alone, in addition to the various other arguments put forward by other contributors, I certainly don’t buy the ‘contact binary’ idea either. It is indeed worrying and dispiriting to see that such an improbable idea is being rushed into print in the shape of a peer-reviewed article whereby it will henceforth be considered as being ‘cast in stone’.
Thomas: I don’t think any one is proposing that the two cometesimals had different origins.
Neither am I, Kamal, quite the contrary! Where did you get that impression? Personally, I ‘ve stated several times that I believe 76P is exactly what it looks like; a large misshapen lump of rock (made from one piece).
THOMAS, I agree with Kamal. None of your 3 assumptions is plausible.
The comet must have formed. Otherwise it wouldn’t exist. The poor differentiation, and high porosity exclude fragmentation of a large (planet-like) parent body as a source.
Hence the cometesimals of whatever size must have travelled on intersecting, and moderate trajectories with respect to each other.
Otherwise we wouldn’t find nothing but fine dust as a result of ongoing virulent collisions, or as a result of lack of intersecting orbits.
The comets formed early in the history of the solar system. The conditions of today aren’t the same as in those early days.
In the beginnings of the protostellar disk, all the mass of the Sun (and more) was distributed over the young system. It rarified over time, as some of the material fell into the forming young Sun, some material was ejected from the solar system, some accreted to planets, asteroids, comets, etc.
Gerald, your statement that “comets formed early in the history of the solar system” is a HUGE “assumption”, required by the “dirty snowball” theory (and vice versa). Nothing in the multitude of notoriously “unexpected” images and data which have been acquired of 67P supports this assumption, so the rest of your argument simply falls to pieces.
In addition, even supposing that your basic assumption is correct, what is the mathematical probability of Rosetta having chanced upon one of these alleged “binary contacts”? Unless you are also claiming that there is a very high percentage of such objects produced by the conditions you describe? In which case, there should also be a pretty high percentage of “trinary contacts” and even higher multiples.
And you still need to explain how the cometessimals not only didn’t bounce apart on impacting each other, however “gently”, but actually stuck lastingly and seamlessly together. If the standard model truly requires to defend the “contact binary” hypothesis in spite of its extreme improbability, then this is further clear proof of just how severely challenged it has become.
Detecting asteroid binaries is difficult, but nevertheless:
A list of true binary asteroids:
https://www.johnstonsarchive.net/astro/asteroidmoons.html
The suspected contact binaries (asteroids and comets) :
https://www.johnstonsarchive.net/astro/contactbinast.html
Ternaries occur, but are much less frequently uniquely identified:
https://en.wikipedia.org/wiki/(153591)_2001_SN263
Besides one asteroid with a ring, I’m not aware of n-ary asteroids, with n larger than 3.
https://www.johnstonsarchive.net/astro/astmoons/am-10199.html
Fission by YORP is probably the widest-accepted mechanism for binary asteroid systems:
https://www.helsinki.fi/acm2014/pdf-material/Day-3/Session-2/Room-4/JACOBSON-D0FA.pdf
https://www.jpl.nasa.gov/news/news.php?feature=4459
“In the near-Earth population, about 16 percent of asteroids that are about 655 feet (200 meters) or larger are a binary (the primary asteroid with a smaller asteroid moon orbiting it) or even triple systems (two moons).”
The percentage does’t refer directly to contact binaries, but it provides a rough idea, that several percent of the asteroids and comets should be expected to be contact binaries.
https://www.astro.umd.edu/~dcr/reprints/walsh_icarus220,514.pdf, page 527, right column, (iv):
“The observed fraction of asteroids with bifurcated mass istributions, sometimes refered to as ‘‘contact-binaries’’, is
estimated to be around 10%.”
Gerald, obviously no-one is disputing the existence of binaries if you mean members of a same family of bodies which are orbiting either each other or a common centre of gravity: the most obvious example is stars, *over half* of which are estimated, on the basis of observations, to prefer such a configuration, according to Wikipedia. This proves nothing, however, about CONTACT binaries and in particular about the reality of the hypothetical less-than-half-walking-pace “docking” mechanism which seems to be required for 67P to be one of these rare, exotic, species…
And we would still like some more detailed explanation about how, once this hypothetical docking procedure has been completed and minimal contact has been established between two presumably tiny areas of each comet, the head and the body then somehow acquire a very thick neck (as in the case of 67P, and even more so in the case of comets such as Hartley 2), which displays perfectly seamless junctions with both lobes. It reminds me of how we make a snowman: once the body has been shaped and the round head has been placed on top of it, to make sure the head doesn’t topple off, we have to add in additional bits of packed snow all around between the neck and the body, which we pat firm and smooth with our hands. This is presumably not how 67P’s or Hartley 2’s necks were formed, yet I can’t think of any other possible mechanism. And the question of the vanishingly small mathematical probability of Rosetta just happening to visit one of these improbable bodies still needs to be addressed.
Hi THOMAS,
Docking mechanism 🙂
I like that. They may as well say it. It requires alien intelligence for it to be viable.
THOMAS, you might have failed to notice, that the observed fraction of asteroid binaries is about 10%.
Your snowman example isn’t necessarily too far off. Just Earth’s gravity is absent, and the involved objects are much larger, such that their inertia plays a major role.
Although in literature a fission/mass shedding scenario of a rubble pile by spin-up due to YORP seems to be favored for asteroid contact binaries. As I understand it, this includes comets whenever close enough to the Sun to allow for YORP.
Gerald
“Although in literature a fission/mass shedding scenario of a rubble pile by spin-up due to YORP seems to be favored for asteroid contact binaries. As I understand it, this includes comets whenever close enough to the Sun to allow for YORP.”
Does this mean you do not have an objection to spin-up to fragmentation? It would then be simply a question of the solid/liquid stretch models not panning out for you (or for us as you would say!). Is my thinking correct on that?
Gerald
Hmm, I didn’t quite word that right. You are clearly entertaining the idea of spin-up to fragmentation for comets in order to arrive at the spin-up required in the models. That’s because the models run and show fragmentation.
But I had assumed you were just temporarily allowing the high spin-up as a given in order to start the model runs- but that it was something you didn’t really subscribe to.
It seems, from what you say, that you’ve seen a reference to YORP spinning up comets to fragmentation speeds and therefore do subscribe to the underlying idea of spin-up to fragmentation rather than just entertaining it so as to crank up the model run. That’s the essence of my question above.
Hi A.Cooper,
I don’t know the actual history of 67P. But two things can be done, at least to some degree: Checking whether an idea is physically possible (by model runs or analytical calculations), and comparing an idea with actual observations.
Spin-changes by several causes are certainly physically possible. Common causes are collisions, and YORP and cometary activity are generally accepted. For moons gravitational effects are also accepted as causes for spin changes. For YORP one needs to pay attention with the size of the object, the distance to the Sun, and the period of time to be considered.
Cometary ativity can lead to rapid spin-changes.
There are plenty of models running spin-up until fragmentation, and beyond.
The theoretical difficulty with stretch is mostly – as far as I can see – that the model runs show a different behaviour, either mass shedding near the equator for rubble piles, with possible re-accretion to a companion, implying initial surface features being lost, or catastrophic disruption for solids with relevant tensile strength.
Liquid models form typical equilibrium figures, with several alternatives, mostly depending on the properties of the liquid.
An idea about spin-up needs to pass a theroretical criterion (physically possible) to be reasonably considered.
And it needs to provide observational evidence, which should be at least mostly compatible with generally accepted methods, or else the new method needs to be proven valid.
I see deficits in both criteria for stretch. I think, both can be improved significantly. Although I’d expect, that refinement of the methods will show, that a “stretch” scenario doesn’t fit 67P. But I wouldn’t rule out entirely, that some fragmentation could have taken place during the long history of the comet.
A criteria would first be needed to find out whether the strata are a result of early accretion directly from dust of the protoplanetary disk, or whether a secondary accretion has taken place after a fragmentation.
Another approach would be a check for a stretch-like scenario within the model-run of the collisional contact binary scenario.
There are certainly still degrees of freedom to play with.
Generally, the overall simplicity of a scenario is correlated to its likelihood.
Hi Gerald,
We are trying to keep it simple and to the point. The models don’t have a stretch solution. You are interpreting this as evidence that stretch is “virtually” impossible. I don’t.
This is because we can calculate “possibility” by calculated forces and materials that would result in stretch. We can have a look for circumstantial evidence that fits the required material properties. We can also note that the required material properties are dynamic in a way that no computerised model has yet tried to mimick.
Try “Comparing an idea with actual observations”
As one option in your own advice. Is the comet morphing or eroding? We don’t even know yet, but the shape model is taking erosion into account (as evidenced by the smoothing of small features) but is not taking morphing into account (as evidenced by the correct CONSERT triangulation mapping to the wrong spot by about 50m on the “latest” shape model) ..
Hi Marco,
please don’t try to exploit obvious measurement errors as evidence. That’s a clear dead-end.
One of your idols, Alfred Wegener, used distance measurements which were 1 km off as evidence for continental drift. Don’t repeat this error. It reduces credibility. Reviewers just discard this kind of pseudo-arguments as noise.
… if you can actually *calculate* the properties needed for stretch, it shouldn’t be too difficult to feed it into a “computerized” model.
This would be a big progress for stretch.
If the results would fit at least roughly with observations (including laws of physics), I’m sure it would have a chance to end up in the “canon”.
The ifs still to fulfill.
Hi Gerald, regarding your quote “One of your idols, Alfred Wegener, used distance measurements which were 1 km off as evidence for continental drift”
It makes me much happier about using this kind of argument knowing that Wegener did.
Thomas: “comets formed early in the history of the solar system” is a HUGE “assumption.
The earlier we go the less we know so we have to make assumptions. Gerald makes the standard assumptions about the model that the scientists working in the field are agreed upon.
Since you do not have a model of your own, you make any assumptions which help you to win an argument, and no one can check whether it is consistent with your world view. Fair enough, you have no stake in this mission and are just having fun.
Marco, Andy and others have a point of view, and they are making a case for it. One can agree or disagree with their point of view, but one cannot deny that they have a stake in their ideas.
Thomas: “vanishingly small mathematical probability of Rosetta just happening to visit one of these improbable bodies”.
I am afraid you are ignoring the fact that planning and probability do not interact well together. They did not toss a coin and say let us go to 67p.
Speculatively, there could be a ‘shake it off’ 🙂 of whatever non strongly attached to the objects.
More than an ‘arrow and aim-table scenario’ imagining more like a ‘CRUSH, Crush, crush’ one. Until differential kinetic energies dissipated. Lots of mass loss.
Would be interesting if someone did a simulation in “chocolate” of the erosion theory and showed how the outcome in terms of surface-perpendicular vectors looks different from 67p.
Hi Kamal,
Just for comedic relief, chocolate has come up in conversation and elsewhere. I proposed microgravity experiments with molten chocolate to represent how ephemeral liquids may behave on the surface of a comet. A colleague has coined “malteser theory” to describe how comets end up with a hard shell and soft porous interior. Then there is features on the comet that look like pieces of toblerone. Just looking at this comet makes me crave chocolate now .
Marco: The collision mechanism appears to favour a one-shot process in the early history of the comet. Then the question is what effects would subsequent erosion have? The erosion mechanism favours a gradual process. Then we have to have an explanation for the surface-perpendicular vectors (hence the chocolate experiment). The stretch mechanism may even work with a punctuated process, every time 67p passes Jupiter closely. I do not know what the implication would be on the surface vectors.
Ramcomet
I dug a bit deeper because I always had grave doubts about using this paper as proof for 67P being a contact binary. The paper is at Science Mag, paywalled, but as usual, the supplementary pages tell us what we need to know:
https://m.sciencemag.org/content/348/6241/1355.figures-only
First of all, with the Psys article, I looked at the video and the stills and ascertained that the relative v looked like about 0.5m/sec. This was based on the 2.8 hours between the first two frames, 5km relative travel and little slowdown- a big red flag, because they must be near escape velocity if there’s no slowing, i.e. very slow. And if “around 1km” diameter escape v is tiny. 5km relative travel was established by taking screen shots and roughly measuring the distance travelled by each sphere and adding the two. Measuring tool was the average width of the two spheres equaling the assumed 1km. But all done very roughly. 2.8 hrs is 10,080 seconds so the relative speed is roughly 5000/10,800= 0.49m/sec.
I didn’t really have do include the above calculations because I found the exact speed in the supplementary pages and it’s even a little less. But this served as a good check for corroboration, that I haven’t misinterpreted something.
So on finding this supplementary page linked above, I found the exact relative speed. It’s stated as “1.5v esc” meaning 1.5 times escape velocity. Lower down it says that escape velocity is 0.25m/sec. So the collision speed is 1.5 x 0.25= 0.375 m/sec.
Which is decidedly not “bicycle speeds”.
So who started promulgating the “bicycle speed” analogy. Everyone now thinks that the 67P lobes can crash into each other at, what, 5m/sec (12mph)? Or 10m/sec (road race speed)? Ordinary comfortable bicycle speed I would say is 10mph or 4.5m/sec. That’s 12 times faster than the stated 0.375m/sec.
Walking speed of 3 mph is 1.34 m/sec. So the impact speed is 3.5 times less than the speed at which you stroll to the shops.
If you extrapolate to 67P at 1.5 times escape velocity, the collision speed would be 1.5 x 0.85= 1.275 m/sec or walking speed. So we’re right back to where we started with the assumed 1 metre per second needed plus a little extra for attenuation that might come about via collision friction. So whilst the paper is useful in telling us how a near-escape-speed collision might play out, it’s been coopted and exaggerated for the purposes of enhancing 67P’s contact binary credentials. Those credentials have always been shaky, ever since the ‘collisional problem’ was unveiled at AGU14. It doesn’t inform the original debate over how you can get 10m/sec collisions to stick, or 5m/sec collisions or 2m/sec.
We’re back to colliding at angles of 0.15° at 900AU, or, if it’s 1.5 v esc at 1.275 m/sec, it’s 0.18°. That’s also assuming the two bodies are a) exactly alongside each other on their 3000 billion km orbit and b) doing *exactly* the same speed, because if you increase the absolute speed of one of them you exceed the maximum allowable speed of 1.275m/sec which is entirely due to the sideways approach vector @ 0.18° and 400m/sec. This is clearly a ludicrous scenario to contemplate, which is why they have finally ditched it and gone for the reaccretion scenario.
This paper does nothing to answer the questions on the dynamics necessary for the fairytale kiss at 900AU. It certainly shouldn’t be used to back up CB theory for 67P.
When this first came out in May, it was Marco who sent me the link. I used the 55hrs and whole travel distance that time to work out that it was at most 1m/sec. I signed off saying if by “bicycle speeds” you mean a French onion seller with a wooden leg and a dodgy derailleur, then yes, it’s true. Now I find it’s one third of that already miniscule speed.
Acknowledged, Cooper. 0.375 km/s. Crawling baby speed.
Also, there is something that doesn’t match when contemplating miserable scape velocity of Ducky. This is no more VDW forces territory. Neither fluffy-dancing along- comet-esimals territory. These are km big, water ice ‘cocooned’ objects. Their own scars tells that [as you say] kisses are very unlikely.
As we go up in the scale ‘other rules apply’.
This morning still believed in ‘big’ binaries. For this two to be ‘mated’ something much bigger did the marriage. Otherwise, they’ve always been a single object.
logan
“when contemplating miserable scape velocity of Ducky…As we go up in the scale ‘other rules apply’.”
Exactly. In my reply to Dr. Massironi, I mentioned only half the problem with using this modelled impact to extrapolate to 67P. I made no mention of the specific kinetic energy problem. That is, the KE per unit mass or per kg.
The squaring of velocity for KE is the ‘other rule that applies’.
Scaling up the model with respect to escape velocity parity gives 1m/sec divided by 0.25m/sec = 4 times the the escape velocity for 67P. This in turn means 1.5m/sec impact velocity for 1.5x escape velocity, again in parity with the model. So far so good, that’s what I described above and in my reply to Dr. Massironi.
However, by scaling up the impact velocity by 4 times they scale up the impact energy by 16 times. Whatever the mass of the two bodies, the KE per kg is timesed by 16 if the velocity is timesed by 4. There are just as many separate 1kg blocks of comet as before but each one carries 16x more kinetic energy…and that means that each one of those kg blocks has to dissipate its own energy, 16x more energy, on impact. That has to go in heat (friction). This implies 16x more heat energy input and the disruptive smashing, cracking, grinding and swirling of matrix material to generate that heat. This is a highly unlike scenario, especially as it took 100 runs to achieve it at the lower, modelled speeds.
The more likely scenario is that the two bodies would grind past each other, dissipating about 1/16th of their KE, per the model, and resume their multi-billion-year odyssey as separate bodies.
They would still retain somewhere near to 15/16ths of their original KE energy. Translated to velocity loss, they’d be continuing with 96.8% of their original velocity, or 1.45 m/sec. That’s a 5cm/sec velocity loss in the collision and still a very healthy 0.45m/sec above escape velocity.
Simply put, the two bodies would pass right through each other and escape at almost exactly the same speed. Each one would be accompanied by a cloud of escaping debris that was ground off.
And that’s for the strictly extrapolated 4x escape velocity scenario of 1.5m/sec which is walking speed. For the quoted “bicycle speeds” (let’s say sedate, shopping trip bicycle speeds of 10mph or 4.5m/sec), the velocity disparity between model and real life is 12x more and the KE disparity is a whopping 144x more.
This is all based on the model’s heavy glancing blow. You may dissipate a little more of that 16x more KE with a head on, but I’m pretty sure you’d end up with a small cloud of orbiting debris with an escaping ellipsoid shell containing 90% of the material. If a head-on was viable, it would probably have been culled from those 100 model runs.
The modelling paper looks fine to me. But it should be taken off the list of supporting evidence for 67P being a contact binary resulting from a collision at bicycle speeds.
Whoops, I forgot to use the cosmic approach speed, which is the slower one, outside the gravity well, before the well accelerates it in to 0.375m/sec. That’s the speed that counts because it’s the one at which the two bodies are approaching over millions of km before the hit. The well speed-up is only in the last few tens or hundreds of km.
If the impact velocity is 1.5 times escape velocity, the cosmic approach speed is 1.118 times escape velocity. So we really are down to where we were before the “bicycle speeds” paper appeared to change everything: for an escape velocity of 1m/sec at 67P and an impact of 1.5 times escape velocity (1.5 m/sec) the cosmic radiant speed is 1.118m/sec.
It is a given that an impact speed of 1m/sec would keep the impactors inside the gravity well either orbiting each other or stuck together. That’s because they can’t escape at above 1m/sec if they impacted at 1m/sec. That’s why before, we were talking about impacts of 1m/sec were the only ones that would safely keep the impactors together and that a little surplus speed just might get absorbed. The surplus in this model is 0.118 metres per second or 11.8 centimetres per second, just shy of five inches per second.
So, whereas before it seemed we had gained 5 or 10 metres per second over and above the known lower limit of 1metre per second, we have actually gained 0.118 metres per second or around 1-2% of what is claimed (if you assume bicycle speeds to be 5-10 m/sec).
Hi A.Cooper,
a simple question may add some insight:
What’s the relative collision velocity of (dust) grains in Saturnian rings?
Hi Gerald,
Are you suggesting that comet contact binaries impacted in the sub Roche zone of a planet?
How would they get into that situation, and how would they get out of it without a catastrophic impact?
I am doubting that any realistic accretion ring scenario can herd them in such a convenient way (I’m not quite sure of the answer to the question of what the average impact would be, but I doubt it could scale up to km size bodies anyway)
Unless there is a docking mechanism at play where perhaps electric fields exactly counter mutual gravitation and absorb some of their kinetic energy in a longer distance plastic collision…..
Last paragraph tongue in cheek by the way. EU proponents should take it up and run with it.
Hi Marco,
I see an analogy between planetary ring systems and the protoplanetary disk in terms of relative velocities.
Whether grains of such a ring system accrete to larger bodies depends on the density of the ring, and on the distance relative to the Roche limit.
Most of the planetary disk has been outside the Roche limit of the Sun, hence accretion to larger objects.
Only outer parts of the Saturnian ring system are outside the Roche limit, but those parts are too dilute to accrete to moons.
The same reason moons accrete from rings outside the Roche limit of planets, comets and planets accrete outside the Roche limit of the Sun.
Relative velocities are sufficiently small to allow gravity for accretion. This applies to dust grains as well as to cometesimals with contact binaries as a possible result.
A.Cooper: “The paper is at Science Mag, paywalled…”.
Not for me, at least:
https://m.sciencemag.org/content/348/6241/1355.full.pdf
Comets which grew bigger at super-volatile accretion zone have better chances of continuing accretion at volatile accretion zone.
Water ices imprint a different, harder mechanics to accretion process.
This is cometary fiction
Dear all,
Following up on my previous comment regarding further discussion from Dr Massironi about how the comet got its shape, please see the link below.
Because Dr Massironi provided a lot of really interesting insight into geological interpretation of planetary mages in general, we decided to make a separate blog post out of it, which we hope you find just as interesting and useful reading as we did. He also provided a few comments about the stretch hypothesis.
https://blogs.esa.int/rosetta/2015/10/12/interpreting-images-more-on-how-the-comet-got-its-shape/
Best wishes,
Emily
If not already done, I’d suggest to run model simulations consistent with the observed porosity of 67P/C-G.
Dense packages of spheres predict a compressive strength significantly different from packages just sufficiently dense to withstand its own gravity.
The result should show significantly different elastic and plastic properties.
Those properties result in significantly different probabilities for contact binary formation.
Porosity should allow for an inference of a close to one-dimensional subspace of the two-dimensional friction/tensile strength parameter space.
🙂