B Destroying asteroids shown to be very difficult

[jim mcnamara]

Charles El Mir et al, A new hybrid framework for simulating hypervelocity asteroid impacts and gravitational reaccumulation, Icarus (2018). DOI: 10.1016/j.icarus.2018.12.032

https://phys.org/news/2019-03-asteroids-stronger-harder-previously-thought.html

Modeling the destruction of asteroids in an effort to protect Earth from nasty collisions, has some new research. And a new model.

The takeaway is larger asteroids "reassemble themselves" from explosion fragments, under the asteroid's own gravity unless the explosion has far more energy than previous models indicated.

Assuming the model's results are meaningful -
In other words, we need a bigger bang. Which may mean that deflection of asteroids instead of breaking them into shards may be a more feasible approach. There is some discussion on this with Charles El Mir, the lead author, in the phys.org article.

 

[mathman]

My impression is that deflection is the only feasible method ever being considered. Destroying a large asteroid would take an enormous amount of explosive, while a lot less energy would be needed to deflect..

 

[jim mcnamara]

@mathman - your point is well taken and I think is what the phys.org article mentions as well.

 

[LURCH]

I think it is safe to say that the object may indeed reassemble itself, but not entirely. A subsurface blast would eject a significant amount of material at speeds greater than escape velocity. So, unless the blast takes place at the exact center of gravity, the newly reassembled object will be on a different course. The ejected material would be on an even greater deviation from the original (collision) course.

Also discussed in this same Forum in the thread:
Ideas to protect the Earth from possible asteroid impacts

https://www.physicsforums.com/threa...-earth-from-possible-asteroid-impacts.959077/

 

[Rive]

Same article were already presented here

Let me link the relevant wiki article
Wiki is hard to be considered a reliable source - but even there the 'blow it to pieces' approach seems to be ... missing, and the whole 'nuke it' section is actually about nuclear propelled deflection.

Can't help wondering that why scientific articles would aim against - movies?

 

[LURCH]

Thanks for the link, @Rive , that article makes a lot of the same points mentioned in the PF link in my Post#4. Nice to see that other people are thinking the same way.

The article makes some pretty flawed assumptions. For example;

For example, if there's an asteroid coming at earth, are we better off breaking it into small pieces, or nudging it to go a different direction?

These are presented as mutually exclusive scenarios, which they are not. To me the obvious question would be, “Why should we care if the object is shattered into a thousand pieces, so long as none of those pieces are headed for Earth?”

 

[Rive]

To me the obvious question would be, “Why should we care if the object is shattered into a thousand pieces, so long as none of those pieces are headed for Earth?”

The 'none' is the problem. As last resort blowing one to pieces might be better than nothing since the atmosphere has a decent shielding effect for smaller pieces so the expected damage will be most likely reduced, but it is impossible to guarantee the all the (dangerous) major pieces will miss Earth - and there is not much intervention possibility left once it become a rubble cloud without shape and handle.
But as long as it is in one piece it can be 'nudged' as a whole.

So with the 'nuke it' option the hard part is actually to give it the biggest nudge in the most effective direction without blowing it to pieces (small debris does not count).

The model mentioned in the article might be helpful in calculating the best yield/depth and other parameters.

 

[Grinkle]

So with the 'nuke it' option the hard part is actually to give it the biggest nudge in the most effective direction without blowing it to pieces (small debris does not count).

Does a nuclear explosion actually transfer very much momentum? Why would detonating a nuclear bomb adjacent to an object in space do anything except heat that object up? Why would it push the object appreciably? In an atmosphere a nuclear explosion heats up air and the resulting shock wave moves a whole lot of stuff. In a vacuum one only has the particles that are released in the nuclear explosion to do any momentum transfer - I imagine not much pushing.

In order to move the asteroid, one must use part of the asteroid mass itself for momentum transfer I think - how can it be possible to push the asteroid via explosion without ejecting part of the asteroid mass to push the rest in the other direction? Maybe you don't disagree and this is what you mean by nudge also.

 

[Rive]

It would be nice to have the linked wiki article as bottom line for any discussion about 'nuke an asteroid'.

 

[Grinkle]

@Rive Yes, sorry, should have read that first.

 

[jbriggs444]

The model mentioned in the article might be helpful in calculating the best yield/depth and other parameters.

how can it be possible to push the asteroid via explosion without ejecting part of the asteroid mass to push the rest in the other direction? Maybe you don't disagree and this is what you mean by nudge also.

It seems clear that Rive is contemplating a sub-surface detonation which does eject part of the asteroid mass to obtain effective thrust.

I'd be thinking in terms of a well-aimed armor piercing warhead with a delay fuse timed to achieve a burst at the desired depth. The guys who design missiles are pretty good at that stuff.

 

[PAllen]

Actually, it seems to me not critical that any asteroid mass be ejected. Given an isotropic explosion at some point on the asteroid surface, nearly half the momentum ejected by the explosion goes to deflecting the asteroid. The tendency of the asteroid to coalesce if initially blown apart is only to the good for the purposes of deflection. With current technology, I don’t see any better deflection approach than a properly placed nuke.

 

[Tom.G]

Given an isotropic explosion at some point on the asteroid surface, nearly half the momentum ejected by the explosion goes to deflecting the asteroid.

Hmm... seems to me that the mass is in the plasma of the fireball, which would be the gross mass of the bomb and casing. Doesn't seem like that would be very effective in moving a megaton rock.

Here is a slightly informative link. https://www.labroots.com/trending/space/12897/here-s-what-d-happen-detonated-nuclear-bomb-space

Try a Google search for nuclear bomb explosion, or for nuclear bomb explosion in space

 

[Rive]

it seems to me not critical that any asteroid mass be ejected.

When available energy is to be converted to impulse, it is about the classic E= 1/2mv² equation. With digging out the impulse it is p=2E/v, which is actually inversely proportional to the speed for any given energy.
There is energy loss and many other considerations, but at the end this is the basis of the deflection.

With abundant energy and low mass, you get a flashy explosion but just low impulse at the end: to get more impulse you need low speed and just the more mass the better.
Of course the speed still should be above the escape velocity of the asteroid in question to get effect.

So the key point of the 'nuke it' approach is to determine the optimal yield and depth of detonation for the material constitution of a specific asteroid to avoid breaking it up yet providing the most mass barely escaping it: this would provide the maximal thrust. (The direction of the thrust to get maximal effect is a different matter.)

That's why I dare to say that the linked model seems to be spot on, yet at the form it was presented it is just a capital misfire.

 

[PAllen]

When available energy is to be converted to impulse, it is about the classic E= 1/2mv² equation. With digging out the impulse it is p=2E/v, which is actually inversely proportional to the speed for any given energy.
There is energy loss and many other considerations, but at the end this is the basis of the deflection.

With abundant energy and low mass, you get a flashy explosion but just low impulse at the end: to get more impulse you need low speed and just the more mass the better.
Of course the speed still should be above the escape velocity of the asteroid in question to get effect.

So the key point of the 'nuke it' approach is to determine the optimal yield and depth of detonation for the material constitution of a specific asteroid to avoid breaking it up yet providing the most mass barely escaping it: this would provide the maximal thrust. (The direction of the thrust to get maximal effect is a different matter.)

That's why I dare to say that the linked model seems to be spot on, yet at the form it was presented it is just a capital misfire.

That's of course true, but impulse is still proportional to energy given that plasma velocity will not vary much as bomb yield increases. While it is not as efficient as optimal burial, a large enough energy produces a large impulse. It is also technologically easier.

 

[LURCH]

It also appears to me that the researchers were pretty responsible in their wording, saying that their new model shows some “possible” differences from previous models, and the old conclusions are “called into question,” while the writers of the article make it sound like the old models have now been proved wrong. Always a good idea in these situations to remember that all we have now are two or more models that do not agree. We have no observations to support one over the other. The new model includes factors that the old model did not, and that is a good thing, but there is still a lot of room for factors that no one has even guessed at yet.

 

[mfb]

That's of course true, but impulse is still proportional to energy given that plasma velocity will not vary much as bomb yield increases. While it is not as efficient as optimal burial, a large enough energy produces a large impulse. It is also technologically easier.

Let's take a really big bomb then. The Tsar Bomba, a yield of 50 MT with a mass of 27 tonnes. Too heavy to be launched on an Earth escape trajectory by any existing rocket. Let it explode on the surface, half of its mass hits the asteroid, half of it escapes using half of its energy. For simplification let's assume this half moves away with the same velocity for all the material, I'll do the same for the asteroid material so we ignore a similar prefactor in both cases. You shoot away 13.5 tonnes at a speed of sqrt(2E/m)=3900 km/s, a bit over 1% the speed of light. Total momentum transfer: sqrt(2Em) = 5.3*10¹⁰ kg m/s.

What happens if we can bury a 10 MT bomb ~10 meter below the surface and shoot away the material above it? Let's use 1000 tonnes here. Let's assume again that half the yield is used to shoot away this mass. We shoot the material out at a speed of 290 km/s and get a total momentum transfer of 29*10¹⁰ kg m/s. Despite using a much smaller weapon we got 5.5 times the effect.

What happens if we can shoot away 1 million tonnes with the 10 MT bomb (might need some sort of drilling to reach ~100 m depth)? You get a velocity of 9 km/s and a momentum transfer of 914*10¹⁰ kg m/s, we gain another factor 30 over the already improved result.

 

[sophiecentaur]

Why 'Nuke it'? (Apart from the obvious popular appeal.)
The thread has already established that we would need to land on and penetrate the asteroid in order to make best use of the 'nuke' then why not use a conventional rocket? Less energy available of course but energy is not momentum, as we have recognised and what's needed is a Match of source against load, to use an electrical analogy.
The Momentum change which is what we are actually after, could be directed in the optimum direction.

 

[Rive]

Why 'Nuke it'?

Apart from the obvious reason (nice big blast, at a convenient distance) it's about our actual limits regarding rocket technology. The available mass would be only a few hundred kg (based on the deep space vehicles produced so far) on the asteroid, and in that mass it's hard to pack anything else what would make any difference on a rock what's big enough to worth (or: better) deflected.

So with a nuke it would be already feasible with our current technology.

 

[mfb]

Why 'Nuke it'? (Apart from the obvious popular appeal.)
The thread has already established that we would need to land on and penetrate the asteroid in order to make best use of the 'nuke' then why not use a conventional rocket? Less energy available of course but energy is not momentum, as we have recognised and what's needed is a Match of source against load, to use an electrical analogy.
The Momentum change which is what we are actually after, could be directed in the optimum direction.

A 27 tonne rocket shooting "all" its mass away at 4.5 km/s (better use hydrolox) produces a momentum change by 1.2*10⁸ kg m/s. That is a factor 400 worse than the Tsar Bomba surface explosion, a factor 2000 worse than the shallow explosion of the 10 MT bomb, and a factor 70,000 worse than the deep explosion.

Why nuke it? Because it is better if you need a significant push.

 

[rizheng1]

Hmm... this is an interesting topic. I think that the hypervelocity impact study is quite relevant to some of our ways of defending against an asteroid, but not necessarily to nukes.

In regards to nuke vs asteroid, I think that if we are to deploy an explosive device, we will likely emphasize on range. We want it to meet the asteroid as farm away from Earth as possible, so that it only has to make a small change in the asteroid's trajectory in order to make it completely miss Earth. This means that a heavier payload may not be better if we cannot get it to a detonation point as far as a lighter payload. That being said, the missile aimed at an asteroid will probably be nuclear not because it makes a bigger boom, but because the missile needs to be propelled by a nuclear source. Since even the largest impact we can make onto asteroids is pretty small, we will need to have a detonation point somewhere beyond at least Mars orbit. Preferably Saturn or Jupiter orbit. Mars orbit along takes 6 months to get to, and the only fuel source that lasts that long is really nuclear.
In this method, the hypervelocity impact study wouldn't really matter, because the explosion will not be big enough to create a system where the asteroid and the impactor are shattered to a degree in which the particles' gravity end up very uniform in strength and are also spherical. (implying that the pieces are uniform in mass, shape, and therefore composition)

The way we might defend ourselves from asteroids which does involve these hypervelocity impacts, and a personal favorite of mine, is to nudge objects from the asteroid belt into their paths. This method is actually pretty half-baked as if we are getting hit by an asteroid, it probably comes from the asteroid belt and this method solves none of that. But, for the sake of having a hypervelocity impact, this is one of the more plausible ones.

 

[mfb]

@rizheng1: Most near Earth asteroids fly around in the inner Solar System, their distance to Earth changes all the time and their orbit intersects Earth's orbit. You don't need to go far away to get there.

nudge objects from the asteroid belt into their paths

It is challenging already to reach an object with a rocket. With an asteroid, that is mainly dead mass? Much more difficult.

 

[Rive]

...to nudge objects from the asteroid belt into their paths.

We are not even able to tell for sure if an asteroid will hit Earth or not: we can give a probability only, and for this we have a relatively big target. But then we nudge a small asteroid to hit another small asteroid with 100% chance and give it a known change in trajectory?

 

[sophiecentaur]

A 27 tonne rocket shooting "all" its mass away at 4.5 km/s (better use hydrolox) produces a momentum change by 1.2*10⁸ kg m/s. That is a factor 400 worse than the Tsar Bomba surface explosion, a factor 2000 worse than the shallow explosion of the 10 MT bomb, and a factor 70,000 worse than the deep explosion.

Why nuke it? Because it is better if you need a significant push.

I know the numbers could be important but how important would it be if you could choose a very early intervention? Your figures imply a no-brainer solution to a last minute deflection problem.
As rizheng says:

We want it to meet the asteroid as far away from Earth as possible, so that it only has to make a small change in the asteroid's trajectory in order to make it completely miss Earth.

A one minute of arc deflection a year beforehand would give massive amplification of course direction. But I appreciate that the orbit prediction error goes up, the further in advance you take action so that could mean many more interventions with more asteroids would be necessary - just to be on the safe side. There are actually a lot of relevant factors when deciding on a suitable solution for the collision problem and different generations could find different optimum solutions as technology progresses.

 

[Rive]

I think the most relevant factors are the time of noticing the danger: the size of the asteroid and our capability to deliver mass to the asteroid.

For the size of the asteroid, we have a bottom line: below that any intervention is just does not worth it. And that size is actually pretty big already.
About our capability to send a device, a few hundred kg delivered in limited time is already a big deal.

Please correct me if I'm wrong, but right now on scientific/engineering basis nothing else seems feasible but a nuke with a surface (!) explosion, and even for that we are a bit short, and the progress does not seem promising at all.

 

[sophiecentaur]

I think the most relevant factors are the time of noticing the danger:

On top of that there are prediction errors that multiply with time. Looking too far ahead would produce an embarrassing number of possible candidates. I could imagine a number of craft, placed in suitable orbits, that could rendezvous with threatening objects when they've been identified. That could save a lot of time and even save on overall cost.
What we actually need is a 'serious' collision that would bring the problem to the public's attention and that would justify vast amounts of global spending. But the 'event' would need to be pretty big; more visible than the present evidence of climate change, which is not yet serious enough to make people sit up and take notice of that (probably) more relevant problem. Most humans don't think the same way that most PF members think, unfortunately.
[Edit: a large new crater on the Moon (our side) would be an ideal advert for a very expensive defence system for Earth. I guess the downside could be that, for a very visible crater formation, a significant amount of debris could reach Earth and do some low level damage here.]

 

[mfb]

I know the numbers could be important but how important would it be if you could choose a very early intervention?

It doesn't matter. If the asteroid is so small and you have so much time that a tiny impactor is sufficient to deflect it: Sure, do that. That's not the point. The question is how to deal with larger asteroids efficiently.

A one minute of arc deflection a year beforehand would give massive amplification of course direction.

Unless the asteroid will make close fly-bys afterwards you only get an amplification in the direction of the orbit - the orbit stays a closed ellipse, changing its shape but not the orbital period doesn't lead to a deviation that accumulates over time.
sin(1 arcminute)*1 AU = 43,000 km. Enough if you know the orbit very precisely already.

 

[sophiecentaur]

To put the asteroid into an entirely different orbit, wouldn't that (a least sometimes) involve more than one burn?

deviation that accumulates over time

I realize that a deflection will not cause a helical path (you didn't mean that?) but what can you do to ensure that it will always be out of harm's way? You are perhaps implying that there is a minimum energy requirement to get a permanent miss trajectory? That makes sense.

 

[mfb]

Every burn will change its orbit. With a single short burn the new orbit will intersect the old orbit at the point of the burn.

To get the object away from Earth permanently you need a much larger trajectory change (~1000 times larger) than a simple impact avoidance maneuver needs. Moving it by a few thousand kilometers vs. moving it by a few million kilometers.

 

[jrmichler]

Project Orion, 1958 through 1963, investigated the use of atomic bombs to move large spaceships. The Wikipedia article has some good discussion on momentum transfer: https://en.wikipedia.org/wiki/Project_Orion_(nuclear_propulsion).

Good reading, especially if you are interested in uses for machine guns that shoot atomic bomb bullets.

 

[sophiecentaur]

To get the object away from Earth permanently you need a much larger trajectory change (~1000 times larger) than a simple impact avoidance maneuver needs. Moving it by a few thousand kilometers vs. moving it by a few million kilometers.

Is there any point in considering complete removal? That would imply the need to hoover up a vast number of those things. I should have thought that there would only be a point in ad hoc deflection, as and when it's identified as a problem. Rather than go for expensive total clearance, is it not best to plan for an orbiting fleet of ships? One thing that experience has taught us is that these space borne systems can be made ultra reliable for many decades of use, at least.

 

[mfb]

I should have thought that there would only be a point in ad hoc deflection, as and when it's identified as a problem.

That's the idea.
You asked about some other procedure.

 

[sophiecentaur]

That's the idea.
You asked about some other procedure.

I don't think so; which bit gave that impression? All I did was to point out that the further in advance you apply the deflection, the greater the effect. When you get down to it, all that's necessary is to miss by , say, 20 thousand km.

 

[mfb]

What gave me the impression was you asking about it:

but what can you do to ensure that it will always be out of harm's way? You are perhaps implying that there is a minimum energy requirement to get a permanent miss trajectory?

 

[sophiecentaur]

What gave me the impression was you asking about it:

I see. It wasn't a particular interest of mine but I was taking up your point

Every burn will change its orbit. With a single short burn the new orbit will intersect the old orbit at the point of the burn.

Which I took to mean that the short solution was effectively useless - just putting things off till later (next time round, even). You were quoting a bare fact and it wasn't clear what your opinion was about the consequences of that fact.
But experience tells us that it's a pretty low probability event which can (in the future) be dealt with ad hoc and that would reduce the probability of disaster even further.

 

[mfb]

In general Earth is a small target. If you can avoid a predicted impact it is likely that the object won't be a threat for hundreds of years or longer.

 

[TeethWhitener]

Possibly of relevance--some experiments scheduled for 2-4 years from now:
https://en.wikipedia.org/wiki/AIDA_(mission)

 

[LURCH]

To get the object away from Earth permanently you need a much larger trajectory change (~1000 times larger) than a simple impact avoidance maneuver needs. Moving it by a few thousand kilometers vs. moving it by a few million kilometers

I think this may not be true in all instances. I’m hoping somebody will check my reasoning here, but if the deflection is away from the ecliptic, then only two changes, each of equal magnitude, would be required. I’ll describe the plan and see if it meets with reason:

Let me start by observing that earth-crossing asteroids generally have elliptical orbits with eccentricities around 2.5-2.75 or so, and orbital periods of around 2.5-3 years. The only part of this orbit we need to address is the two points at which that ellipse crosses the line of 1AU distance from the Sun.

If a potential impactor is spotted, a slight deflection at an early enough moment will turn a “possible hit” into a “near miss.” On the object’s next approach, probably about three years later, it will miss by a slightly wider margin, and so on for each successive orbit.

If the initial acceleration is at 90° to the ecliptic plane, then the orbit will become progressively more inclined as time goes by. Once the orbit is inclined to a high enough degree, the two points at which the orbit passes inside of the 1AU line will both occur well outside of the ecliptic, at which time a second acceleration, equal and opposite to the first, will stop the continuous change, making the new, highly inclined orbit permanent (more or less), right?

 

[jbriggs444]

If the initial acceleration is at 90° to the ecliptic plane, then the orbit will become progressively more inclined as time goes by.

What effect would be responsible for increasing the tilt of an elliptical orbit over time?

If such an effect existed, one would expect all of the planets to be orbiting at 90 degrees relative to Jupiter.

 

[LURCH]

What effect would be responsible for increasing the tilt of an elliptical orbit over time?

If such an effect existed, one would expect all of the planets to be orbiting at 90 degrees relative to Jupiter.

Thank you @jbriggs444, that’s exactly the sort of double-checking I was hoping to get. If the tilt were started with a push, would it not continue until another force stopped it? Or would it require continuous thrust to keep going.? My impression is that, once the motion is begun, it will continue until some force makes it cease.

 

[mfb]

If a potential impactor is spotted, a slight deflection at an early enough moment will turn a “possible hit” into a “near miss.” On the object’s next approach, probably about three years later, it will miss by a slightly wider margin, and so on for each successive orbit.

Why would the closest approach follow such a pattern? Most likely it will jump around wildly.

Thank you @jbriggs444, that’s exactly the sort of double-checking I was hoping to get. If the tilt were started with a push, would it not continue until another force stopped it? Or would it require continuous thrust to keep going.? My impression is that, once the motion is begun, it will continue until some force makes it cease.

Orbits are closed (neglecting three-body interactions and general relativity). A simple push just changes the orientation of the orbit a bit but then it orbits stably in this new orientation.

 

[sophiecentaur]

What effect would be responsible for increasing the tilt of an elliptical orbit over time?

I was wondering about that too. The orbital period would change a bit but that's all I can imagine happening. If the orbital plane were tilted away from that of the Earth's and then the eccentricity were increased, you could permanently remove the object from danger. But I think that (as with most useful manoeuvres) would require two 'burns'.

 

[jbriggs444]

If the tilt were started with a push, would it not continue until another force stopped it?

The effect of a force that stops is a change in the orbit -- to a new stable elliptical orbit.

 

[jbriggs444]

But I think that (as with most useful manoeuvres) would require two 'burns'.

A single burn is adequate. A pair of distinct elliptical orbits about the same primary can intersect in at most two points. With a single burn you can tilt one orbit so that its position near the one intersection point rises up out of the plane and its position near the other intersection point is depressed down out of the plane.

I believe that it is even easier than that. A random burn applied at a random time will, with probability 1, result in orbits that do not intersect.

This is in theory -- treating the two orbits as separate two-body solutions. In practice, the solar system is a many body system. There is no such thing as a stable orbit and almost everything you can do is temporary. Also, the probability 1 thing applies for point-like satellites. Real planets and asteroids have non-zero cross-sections.

 

[sophiecentaur]

With a single burn you can tilt one orbit so that its position near the one intersection point rises up out of the plane and its position near the other intersection point is depressed down out of the plane.

I can almost picture that in my mind. The ellipse is narrower than the circle where the planes coincide?

 

[mfb]

Or wider. Or just not aligned with Earth. Here is a sketch - while that object doesn't intersect Earth's orbit you can see how it would intersect an orbit somewhere between Earth and Mars - but only once. The other intersection with the plane of the planets is much farther away from the Sun.

Another 3D sketch

A different 2 D sketch but the color code indicates where the asteroid is "above"/"below" the plane of the planets.

 

[sophiecentaur]

The other intersection with the plane of the planets is much farther away from the Sun.

You will have to help me with this one. Doesn't the major axis of the ellipse pass through the Sun - so why isn't there symmetry about that axis? Is it the effect of other bodies?

 

[jbriggs444]

You will have to help me with this one. Doesn't the major axis of the ellipse pass through the Sun - so why isn't there symmetry about that axis? Is it the effect of other bodies?

There is no reason for the line that is the intersection of the orbital planes to align with the major axis of either ellipse.

 

[sophiecentaur]

Of course. . . . Tilt. Durrrr

 

[LURCH]

@mrb; thanks for the images, they make the concept much more clear. Also thanks @jbriggs444 for pointing out that changing the inclination takes only one acceleration, and not two. I now remember learning that, many years ago, and having difficulty accepting it. It was in relation to artificial satellites, and I couldn’t see how a vehicle in space could move to a new orbit and be expected to stop without executing a burn once it reached that new orbit. Now it seems obvious. Live and learn, I guess.

I believe this sort of inclination change used to be called an “orbital plane adjustment”, or something similar, and it was quite problematic. I’m off to search for what those problems were, because I can’t remember.