Nuclear Safe Trajectories?

John Creighton · 2005-07-06 13:36:43

I am going to start this thread to show my ignorance in orbital mechanics. It has been argued the launching a hot nuclear reactor is unsafe in the event the reactor fails. Well do we need to launch it directly at earth? Is there a way that the rocket can spiral into earth so that if it fails it will never hit earth?

Think about either returning from earth or retuning from the moon. Perhaps from the moon the rocket could fly passed earth moon L1 closer to the moon then just as it was passing L1 do a corrective maneuver to transfer from a high moon orbit to a high earth orbit. For going from mars to earth perhaps the rocket could approach earth sun L2 but just short of L2 on the side closer to the orbit of Mars. It could then do some kind of holding maneuver. Then a chemical, solar ion or solar thermal stage detaches to transfer the crew or cargo the rest of the way.

Fledi · 2005-07-06 14:19:36

A NTR could go all the way from LEO to Lunar or Mars orbit, you aren't much better off if the reactor fails in a higher orbit, as long as you can't predict which way it will come off course. If it goes intert in LEO for example it must have an option to go into a higher orbit or it will enter the atmosphere years later (that means we will have a few years of time to figure out what to do if it strands in LEO).

But if the engine keeps thrusting in some random direction you can't tell where it's going to go anyways. But you could probably self destruct it in this case or something.
The whole reason for using nukes is to have a high thrust, high ISP engine for the whole orbit to orbit part of the journey, it doesn't seem to make much sense to bother with extra chemical stages along the way.

For nuclear electric propulsion Earth-Moon travel will take longer because of the low thrust (at current tech), so that's only a good option for places that are much farther away, like Mars.

John Creighton · 2005-07-06 15:05:52

A NTR could go all the way from LEO to Lunar or Mars orbit, you aren't much better off if the reactor fails in a higher orbit, as long as you can't predict which way it will come off course. If it goes intert in LEO for example it must have an option to go into a higher orbit or it will enter the atmosphere years later (that means we will have a few years of time to figure out what to do if it strands in LEO).
But if the engine keeps thrusting in some random direction you can't tell where it's going to go anyways. But you could probably self destruct it in this case or something.
The whole reason for using nukes is to have a high thrust, high ISP engine for the whole orbit to orbit part of the journey, it doesn't seem to make much sense to bother with extra chemical stages along the way.
For nuclear electric propulsion Earth-Moon travel will take longer because of the low thrust (at current tech), so that's only a good option for places that are much farther away, like Mars.

Well, for cargo Nuclear Electric might make since. Nuclear thermal strikes me as a minimal benefit over chemical and sometimes I wonder if it is worth it.

BWhite · 2005-07-06 15:24:02

How would a cis-lunar nuclear tug fit into these scenarios?

Starting from L1 a tug could accelerate towards the Moon, take a gravity assist fly-by and head towards Earth for a Terra assist fly-by then on to Mars or wherever. Cut the (pushed) payload loose from the tug somewhere between the Moon and Earth and let the payload travel ballistically thereafter. With full power burns between L1 and Luna and Luna and Earth, that payload could easily have substantial velocity before being cut loose.

Maybe then spin the tug around (for braking engine burns) and/or aero-brake to remain in a high Earth orbit returning to L1 or LEO as needed using some fancy math and doing gravity assists as needed to minimize fuel consumed on the way back to LEO or L1.

This way a nuke tug could throw Mars bound payloads every week or so, with the transit time and net payload a function of the calendar (relative orbital positions of Earth & Mars).

= = =

Unfortunately, the risk of miscalculation and dumping a hot reactor into the atmosphere was something I had not considered.

Edited By BWhite on 1120685309

John Creighton · 2005-07-06 15:39:35

Yeah, low earth orbit I think is too close to earth in the event of a failure or a miscalculation.

Fledi · 2005-07-06 16:01:27

And this scheme also does not change the fact that you can only launch payloads to Mars every 26 months without lots of extra delta v.

Aerobraking a nuke doesn't seem to be a good idea to me either, at least not in the Earth's atmosphere.

Nuclear thermal strikes me as a minimal benefit over chemical and sometimes I wonder if it is worth it.

It depends on the size of your spacecraft, for smaller ones definitely not, for larger it has the advantage of requiring far less fuel than chemical.

BWhite · 2005-07-06 16:10:57

And this scheme also does not change the fact that you can only launch payloads to Mars every 26 months without lots of extra delta v.
Aerobraking a nuke doesn't seem to be a good idea to me either, at least not in the Earth's atmosphere.

Aren't delta V and travel time functions of each other?

In other words, you can go anytime, it just might take a long long time to get there? My understanding is that package goods can be launched towards Mars whenever (every two weeks for example).

Arrival times at Mars will vary, perhaps greatly and deceleration can be a problem meaning multiple aerobrake passes.

I thought the 26 month windows were for Hohmann trajectories with human crews. The free return Mars to Earth trajectory opening in 2014 supposedly takes years without needing ANY fuel.

Edited By BWhite on 1120687882

Fledi · 2005-07-06 17:00:40

Does that trajetory include gravity assist?
If that's the case it still depends on special planetary constellations to work.
The Grand Tours of the Voyager probes were such special trajectories for example.
I'm not sure how often gravity assist for a Mars trajectory in the Earth-Moon system is possible, maybe even as often as every month or so (?), but going to Mars at different trajectories than Hohmann requires more delta v, regardless where it comes from.

GCNRevenger · 2005-07-06 17:05:02

You do need a minimum Delta-V of escape velocity to get to Moon/Mars from LEO Bill, and probobly a little more then that to make sure you get there. For fairly low-energy trajectories to Mars like MarsDirect and NASA DRM, I bet that the minimum delta isn't much lower then the one already planned.

For the Moon, chemical engines are just fine, because you don't need a huge ship to get there, eventually you want a reuseable lander (leverage Lunar LOX, no radiation shield), and a small nuclear tug makes limited sense. For MARS however, it is possible to get there with chemical engines, but probobly not without breaking the bank... each sortie would need far more fuel, such that a vehicle like NASA DRM would need three rather then two HLLV launches and an additional docking. Launching six rather then four HLLVs in time to make the launch window would be hard too.

I have no problem with putting reactors into Earth orbit and activating them there, so long as their orbit doesn't decay for a few decades (preferably a century or two) then there is no problem. The same with using nuclear rockets to leave orbit, since even if they do fail, they can't lower their orbit. The idea of an NTR rocket "running wild" and deorbiting itself is silly and won't happen since its a turbopump rocket like any other.

The problem with nuclear is trying to bring the reactor BACK to Earth, any trajectory that would in the event of vehicle or guidence failure that could cause a recently used reactor to enter Earth's atmosphere is a serious risk, one that is not generally justifiable.

GCNRevenger · 2005-07-06 17:08:33

Does that trajetory include gravity assist?
If that's the case it still depends on special planetary constellations to work.
The Grand Tours of the Voyager probes were such special trajectories for example.
I'm not sure how often gravity assist for a Mars trajectory in the Earth-Moon system is possible, maybe even as often as every month or so (?), but going to Mars at different trajectories than Hohmann requires more delta v, regardless where it comes from.

Gravity assists to get to Mars increase the travel time, either coming or going, and aren't acceptable.

John Creighton · 2005-07-06 17:35:57

I have no problem with putting reactors into Earth orbit and activating them there, so long as their orbit doesn't decay for a few decades (preferably a century or two) then there is no problem. The same with using nuclear rockets to leave orbit, since even if they do fail, they can't lower their orbit. The idea of an NTR rocket "running wild" and deorbiting itself is silly and won't happen since its a turbopump rocket like any other.
The problem with nuclear is trying to bring the reactor BACK to Earth, any trajectory that would in the event of vehicle or guidence failure that could cause a recently used reactor to enter Earth's atmosphere is a serious risk, one that is not generally justifiable.

So what I was looking for was a trajectory that wouldn’t cause the vehicle to enter earths atmosphere in the event of a failure. To do this you don’t try to hit the target with one burn rather you aim short of your target, find your new trajectory then aim short of your target again. So what we are looking for is something faster then a spiral but slower then a Hoffmann transfer. Thus you aim short enough that any reasonably likely failure can not cause you to renter earths atmosphere. Say we undershoot earth by 10% from mars, would there be any possible way we would hit it in the event of a failure. Can we really miss that badly?

GCNRevenger · 2005-07-06 18:28:20

You can't really make only half a burn to get back to Earth, it would increase trip time and decrease safety too much (fall into the Sun?). Besides, the very last burn to enter Earth orbit will be just as dangerous.

I say that with today's material technology, that we not even bother risking bringing the reactor back to Earth at all. There are basically two and a half types of NTR fission rockets: NERVA type, Timberwind type, and (my favorite) Gas-Core type.

The latter has been tested, has only modest performance, but is restartable. Its the easiest option.

Timberwind was theorized and played with a little, but never tested much. It has a little better performance as NERVA type, would have been lighter, but can only fire once. Only good for expendable booster roles.

Either of the above would be great for Mars exploration, and NASA-DRM calls for such an engine for its expendable TMI rocket.

Gas-Core type engines are special... rather then using a solid reactor core, the Uranium itself is intentionally melted down, and becomes so hot that it even becomes a gas... perhaps even a plasma. This type is "down the road" a ways because of the technical challenges, and would be pretty big most likely, but its performance is perhaps an order of magnetude beyond the other two... the trick is, you can just flush the Uranium out the back of the engine when you are done firing, so no "hot" reactor need return to Earth onboard. Definatly a mid/far future option.

John Creighton · 2005-07-06 19:25:43

You can't really make only half a burn to get back to Earth, it would increase trip time and decrease safety too much (fall into the Sun?).

It would increase trip time which would be expectable to cargo and may or may not be acceptable to people. For cargo I think it is a reasonable option because electric propulsion is pretty slow even for cargo. Of course I am not sure why we would need to bring large quantities of cargo back to earth from mars.

As for safty why would it be less safe? By making burns later in flight it could be safer because some of the error in the initially burn will be corrected for by the later burn. And if you undershoot earth your ship won’t fall into the sun it will end up in some relatively stable (100’s of years?) orbit between earth and mars.

Besides, the very last burn to enter Earth orbit will be just as dangerous.

No the last burn will be safer because an error in the last burn of a multiple burn approach will cause a smaller error in the final destination then an error in an initial burn of a double burn approach (Hoffmann).

I would argue that a multiple burn approach is safer then a double burn approach because in the Hoffmann trajectory all the last burn can do is circularize and if the rocket arrives to early or too lately it is too bad. It is simply too late to correct for this kind of error.

The reason for the safety boost is the sooner the errors are corrected the less fuel that will be needed to correct the errors, however by correcting the errors early, the accuracy of the corrections needs to be that much greater. The initial burn of the Hoffmann trajectory is were the errors are introduced so it may be too early to correct them, while in the final burn it is too late.

In a multiple burn approach the initial undershoot is not an error it is deliberate. And each burn is timed so that the ship arrives at earths orbit the same time as earth does. So I don’t believe that the multiple burn approach requires any more fuel then a Hoffmann trajectory. However, if the timing is off then more energetic trajectories will be required and thus more fuel will be needed. However if the timing is off on the injection burn of a Hoffmann approach it is too late to correct by the time of the circularizing burn.

What I notice about a Hoffmann trajectory is each burn is tangential to the velocity. So I assume that if you have multiple burns which are also tangential to the velocity of the space ship then no more fuel will be needed then a Hoffmann trajectory.

Fledi · 2005-07-06 19:31:33

The idea of an NTR rocket "running wild" and deorbiting itself is silly and won't happen since its a turbopump rocket like any other.

Following scenario: guidance computer malfunction during a burn, communication is lost, through bad luck the spacecraft orients itself in a way that the engine thrusts against its velocity vector and doesn't shut down (computer malfunction).

Or the computer simply goes crazy, like HAL

By the way, sidenote: make guidance systems resistant against terrorist hacker attacks.

Gravity assists to get to Mars increase the travel time, either coming or going, and aren't acceptable.

Yeah I agree, something like a rotating tether is a much better option for launching cargo capsules than all this gravity assist hocus-pocus.

GCNRevenger · 2005-07-06 19:38:44

My worry isn't so much the accuracy of your burn, its that the rocket engine won't reliably handle multiple restarts.

There is the issue of gravitational losses to the Sun that only burning part of the way to your destination too, and could bite into your fuel bill a bit.

Saving your fuel for a burn later on ensures the total trip time is longer, because you aren't spending as much time at a higher speed even for identical total Delta-V.

And finally, we don't have a way to store liquid hydrogen for more then a month or two, maybe three, without a complex & heavy cyrogenic condenser. You won't have any propellant to feed that engine late in the transit.

Fledi · 2005-07-06 19:47:41

No the last burn will be safer because an error in the last burn of a multiple burn approach will cause a smaller error in the final destination then an error in an initial burn of a double burn approach (Hoffmann).

Correction burns are done with about every space probe mission after a rough insertion into the interplanetary trajectory.
It's just that the first burn has by far the largest delta v.

What I notice about a Hoffmann trajectory is each burn is tangential to the velocity. So I assume that if you have multiple burns which are also tangential to the velocity of the space ship then no more fuel will be needed then a Hoffmann trajectory.

The Hohmann trajectory is an ellipse with the perihel (small axis) having the same size as the inner planet's orbit and aphel same size as outer planet's orbit.

If you make major midcourse engine firings you will change the shape of the ellipse into something more fuel consuming to reach the same aphel, but at a different place and time, so you will miss the planet if you originally aimed for it before the burn.
Orbital insertion will not become any easier by these burns since you will have to be very close (by interplanetary standards) tho the target planet when you do the orbital insertion burn anyway.

John Creighton · 2005-07-06 20:22:24

The Hohmann trajectory is an ellipse with the perihel (small axis) having the same size as the inner planet's orbit and aphel same size as outer planet's orbit.
If you make major midcourse engine firings you will change the shape of the ellipse into something more fuel consuming to reach the same aphel, but at a different place and time, so you will miss the planet if you originally aimed for it before the burn.
Orbital insertion will not become any easier by these burns since you will have to be very close (by interplanetary standards) tho the target planet when you do the orbital insertion burn anyway.

I’m not sure what your saying here. Surely no corrections will put the ship on a parabolic or hyperbolic correction. There is no where near enough fuel to do that. I think we could measure the wasted delta V as the change in velocity which is exerted perpendicular to the velocity of the space craft. At least maybe this would give a rough measure. I am not sure though if it would be necessary to do much velocity correction in a direction perpendicular to the direction of the space ship. By making subsequent burns start late or earlier we should be able to effect the time the ship arrives at earths orbit while keeping all burns tangential to the trajectory of the ship.

The objective of the burns really isn’t to make it much easier, rather to make sure that enough fuel isn’t given to the rocket to put it on a collision course to earth with a hot reactor. What I am having trouble doing is anticipating the errors. Where is the rocket going to fail? Are the sensors going to fail? Can we make them redundant enough to feel safe? Is the navigation not going to be accurate enough? How many satellites do we need to deduce the position? Will the pups shut off to late? Can we build enough redundancy in the valves to shut them off? Can we use multiple pumps so that if only one pump stays on a little too long it is not going to make enough difference? I am told there is a danger of a nuclear rocket colliding with earth. Well can we successfully mitigate this risk. If not why, why not?

John Creighton · 2005-07-06 20:27:58

My worry isn't so much the accuracy of your burn, its that the rocket engine won't reliably handle multiple restarts.

A legitimate concern.

There is the issue of gravitational losses to the Sun that only burning part of the way to your destination too, and could bite into your fuel bill a bit.

I’ll have to think about how this would work.

Saving your fuel for a burn later on ensures the total trip time is longer, because you aren't spending as much time at a higher speed even for identical total Delta-V.
And finally, we don't have a way to store liquid hydrogen for more then a month or two, maybe three, without a complex & heavy cyrogenic condenser. You won't have any propellant to feed that engine late in the transit.

That is a very good point. This is one thing I don’t have a grasp on. That is how much difference does nuclear thermal make to the mass fraction. Does it make enough difference to justify the cryogenic condenser. I understand for the moon nuclear thermal is of little benefit but since the trip is longer to mars it gives a better performance on the mass fraction to make it worth while. But are these performance gains enough to justify a condenser.

Fledi · 2005-07-07 05:21:58

Surely no corrections will put the ship on a parabolic or hyperbolic correction

Relative to what? To the sun, no. But the ship is already hyperbolic relative to the destination planet when it's flying between the planets.
That's why you can't avoid doing a high delta v braking when you get close to the planet to come down from this hyperbolic to elliptical or circular orbit anyway.
What the small correction burns of space probes influense is at what point exactly the probe will reach the destination planet's orbit.

John Creighton · 2005-07-07 10:59:52

Could you come into earth sun L2 gently then spiral into the earth. And how does this work if you are coming from mars into earth on the L2 side are you going the direction around earth with the moon or the direction against the moon. I know at certain velocities and positions near L2 orbital predictions exhibit some chaos. Is there a velocity angle and position you can approach L2 that easy to capture the ship in earth orbit and the orbit is predictable enough that you know you won’t hit the earth or the moon despite making relatively large errors.

Fledi · 2005-07-09 11:46:04

Hmm aiming for a L point is probably an option but you have to come close to its coordinates in space and time (I don't know how close) and match speed there.

RobS · 2005-07-10 00:44:22

If you are approaching the Earth from Mars and you are on a Hohmann trajectory, you are moving 7,000 miles per hour (almost 11,000 km/hr) faster around the than Earth and than the (L2 point). So you'd need a delta-v of 7000 mph to "stop at the L2 point. If you then fell to Earth from L2, you'd be moving about 25,000 mph when you passed Earth and if you wanted to go into low Earth orbit you'd need another delta-v of about 7,000 mph. Total delta-v: 14,000 mph.

But if instead you roared by L2 at 7,000 mph and approached Earth, a single delta-v of about 8,000 mph would suffice to put you in a low Earth orbit. In other words, firing your engines deep in Earth's gravity well saves a LOT of fuel and delta-v.

-- RobS

Fledi · 2005-07-10 05:09:31

Yes you're right RobS, it's a lot of extra delta v, it depends on how effective your nuclear drive is and how safe you want to make the return if going to an L point is worth it.
But then you also could make a compromise, go in closer than L2 but not quite low Earth orbit and make the second burn there.

Of course with chemical engines there is the option of aerobraking all the way down.

RobS · 2005-07-10 20:37:47

Why not fire the nuclear engines in the moon's gravity well, perhaps over the far side, then once your trajectory enters the Earth's gravity field, fire them again at, say, 100,000 kilometers altitude and put the nuclear engines into a 1000 km by 100,000 km orbit. Or if there are crew on board, make the first firing over the far side of the moon and the second firing near the Earth-moon L1; it would be relatively easy for the crew to dock to a vehicle there. Or you could eject the crew in a capsule there for the flight to Earth and put the nuke into a holding pattern around L1 (or Earth-moon L2, which is stable).

-- RobS

GCNRevenger · 2005-07-11 17:12:04

Sounds like the best option would be to park the nuclear stage in Lunar orbit as far as safety goes, and use whatever Earth/Moon transit system to get back home. Refueling the thing might be problematic though, unless it had such a high Isp that it didn't need so much propellant or a decent supply of H2 around/on the Moon. "Conventional" solid-core engines like NERVA or Timerberwind will never reach this level of performance.

New Mars Forums

Announcement

#1 2005-07-06 13:36:43

Re: Nuclear Safe Trajectories?

#2 2005-07-06 14:19:36

Re: Nuclear Safe Trajectories?

#3 2005-07-06 15:05:52

Re: Nuclear Safe Trajectories?

#4 2005-07-06 15:24:02

Re: Nuclear Safe Trajectories?

#5 2005-07-06 15:39:35

Re: Nuclear Safe Trajectories?

#6 2005-07-06 16:01:27

Re: Nuclear Safe Trajectories?

#7 2005-07-06 16:10:57

Re: Nuclear Safe Trajectories?

#8 2005-07-06 17:00:40

Re: Nuclear Safe Trajectories?

#9 2005-07-06 17:05:02

Re: Nuclear Safe Trajectories?

#10 2005-07-06 17:08:33

Re: Nuclear Safe Trajectories?

#11 2005-07-06 17:35:57

Re: Nuclear Safe Trajectories?

#12 2005-07-06 18:28:20

Re: Nuclear Safe Trajectories?

#13 2005-07-06 19:25:43

Re: Nuclear Safe Trajectories?

#14 2005-07-06 19:31:33

Re: Nuclear Safe Trajectories?

#15 2005-07-06 19:38:44

Re: Nuclear Safe Trajectories?

#16 2005-07-06 19:47:41

Re: Nuclear Safe Trajectories?

#17 2005-07-06 20:22:24

Re: Nuclear Safe Trajectories?

#18 2005-07-06 20:27:58

Re: Nuclear Safe Trajectories?

#19 2005-07-07 05:21:58

Re: Nuclear Safe Trajectories?

#20 2005-07-07 10:59:52

Re: Nuclear Safe Trajectories?

#21 2005-07-09 11:46:04

Re: Nuclear Safe Trajectories?

#22 2005-07-10 00:44:22

Re: Nuclear Safe Trajectories?

#23 2005-07-10 05:09:31

Re: Nuclear Safe Trajectories?

#24 2005-07-10 20:37:47

Re: Nuclear Safe Trajectories?

#25 2005-07-11 17:12:04

Re: Nuclear Safe Trajectories?

Board footer