I didn't submit this news because really we barely know anything at the moment. Jason-3 is still awaiting its second burn, and without knowing anything more than "it has a broken leg" I think it's too soon to call the landing a "partial success". The second burn will be happening shortly, and they said we'd get more data about the landing in a few hours.
I've had this experience with my house design a lot.
I've been held up by a construction boom leaving the country's architects overworked - the year before last I went through three of them because they weren't getting anything done. Last year I decided not to switch again (and thus have to begin from the beginning each time) and it took my architect 11 months to even get to the stage where we need to bring the engineer in. And before that there were lots of delays in finding and buying the property to begin with. But during all that time my planning for the house has really evolved, for the better. It's still an underground steampunk cave house, but it'll be cheaper to build, more comfortable to live in, and with a better look. Even having to switch architects helped because each one helped refine my thought process with their feedback.
Example: when I started my thought process involved actually boring the house out of the ground, like a tunnel, which would have been ridiculously expensive. I moved from that to the concept of building a timber form in a naturally low place, shotcreting it, the burying that. But building big timber domes here is expensive, and shotcrete unusually expensive here too. From there I moved to the current concept of simply making a big pile of compacted ground in a low point in the shape of the house and concreting over it (with forms only needed for where the slope is too steep), and burying that. Way simpler and cheaper.
I moved from having the house physically spread out to make it like a cave, to using convoluted bends to make it actually compact (even with common walls in places) but feel like it's spread out - thus greatly reducing the amount of material. I moved from a plastic ground cloth for waterproofing to multilayered rubberized-bitumen sheeting on compacted sand over the house, as my extra time gave me time to research longevity of different membranes and I found out that that's what's used to waterproof nuclear fuel repositories over great lengths of time. I moved from the idea of a clay-based plaster on the concrete to try to give it a cave-like feel to the concept of simply pressure blasting away the inner cement and exposing the aggregate. I've refined the details on the type of concrete to use greatly, and may now even be doing a research project out of it (FRP-rebar, loose plastic and basalt fiber, basalt dust pozzolan, etc). I came to the realization that due to the lack of limitations on how thick the insulation can be I could use pumice or scoria rather than a foam-based product. Also came to the realization (after taking the time to do heat flow simulations) that while in many "umbrella earth home" designs it's uninsulated under the house, my nearness to the bedrock and potential groundwater means that I should. I've come up with dozens of new steampunk and cave stylistic features to incorporate. And on and on. None of this would have happened had there not been great delays in the process.
Composites are physically very strong, and the use of epoxy rather than polyester will improve the lifespan. Still, composites yellow in UV light (particularly if the binder isn't designed to be UV stable, but to some degree even if it is) and the binder becomes more brittle. Also while they're strong (as in resistance to breakage) they're not hard (as in Mohs hardness), so they can still get scratched up. There are some scratch resistant coatings one could use (to varying degrees of effectiveness), but if you wanted natural scratch resistance without it being brittle you'd probably want something like chromium or ceramic-matrix composites.
Well... to some extent. You can always run hotter to be able to get more efficiency - for example, that's why coal gassification is more efficient than burning solid coal, you can run it very hot through a gas turbine. Also, recovery of heat once it gets to lower temperature differentials is often not done, so some new work is in making that more affordable, which is a big boon to resources that are fundamentally lower temperature (like geothermal). And of course pairing the two, high temperature cycles with low temperature cycles.
Thermal generation tech does keep improving. But you're right, it's not easy, and there are limits.
If I remember this company's tech correctly, the wind drags it out, turning a generator as it goes. They then switch it to a low-drag mode and reel it back in. So it cycles between high power generation and low power consumption. The concept being that you'd have many of them so that you'd get continuous net generation.
The global wind resource is 72TW. You're going to have a hard time denting that. You might alter things locally - might - but not globally.
And besides that... if any evil greedy megacorp wants to move into my valley and setup something that will rob the wind of its energy... Please Do! Seriously, someone should really work on wind turbines specifically designed to act as windbreaks, in a manner that can be affordably mass produced and deployed in lieu of traditional manufactured windbreak systems.
(Before anyone says "just plant some trees..." I do, every year. They usually get killed by the weather, sometimes outright uprooted. I may have to start spending more time on each tree individually, encasing each one in its own individual PVC windbreak in addition to the broad rows of pallet windbreaks)
Who wants any kind of 100MW plant in their backyard? 100MW is an awful lot of power. Doesn't matter how you generate it, it's going to have to involve some combination of extreme areas, extreme pressures, and extreme temperatures - otherwise the physics doesn't work. If you want it small, it's going to have to be high pressure and/or hot. If you want it low pressure and cool, it's got to be huge. This ignores all other potential issues with each given type of power plant...
That's part of the reason that blimps are increasingly going out of favor, in favor of hybrid airships. Hybrids are based on the concept of being aircraft (generally lifting-body aircraft) that get huge lifting areas via inflation (so that they don't need any real structural strength, and thus keep the mass very low), and use helium as the inflation gas, which partially (but not completely) lifts the craft. So you have the combination of huge lifting area and much of your weight compensated for, so it takes little energy to stay aloft (which can, for example, come from solar), and unpowered landings are perfectly safe. But because (for the same payload capacity) you don't need as big of a gas bag as with a blimp, they're not as vulnerable to crosswinds and thus not as likely to break free, as well as being much easier to land without being blown into power lines or whatnot.
1) I've stood directly underneath a 1MW turbine in the middle of a large wind farm on a windy day. There wasn't much noise at all, just a light whoosh... whoosh... whoosh.... I don't know where this concept that wind turbines are "terribly noisy" comes from, but it doesn't at all match my experience.
2) You're concerned with noise about something that's hundreds of meters up? What about kite-based systems that operate at thousands of meters up, would you be worried about noise from them too?
You should see the setup they use to produce the graphene before separating out the layers with tape.
Also, may I add: it's a stupid assumption that robot operators can't innovate either. Because they do this sort of stuff all the time, inventing new techniques - using the hardware they sent - to do things that weren't expected at the time. From rovers dragging wheels to expose buried sediments while they roam, to New Horizons' doubling its planned data throughput by the realization that they could run both TWTAs at the same time with different polarization if they got the extra power by shutting down the flight computer after spin-stabilizing the craft.... the limitation isn't humans. Because humans are involved, they're the ones controlling the robots. The robots are only limited by what hardware was sent. As are humans. Send a robot with a duct tape tool and it can make things out of duct tape too that weren't planned when they launched.
It seems like every developer has a go-to technique (or four) for keeping things a little more streamlined. What are yours?
The mess is just part of coding. Deal with the mess. Accept the mess. Live with the mess. Join with the mess. Speak with the mess. Hear its answers whispered into your head. Nod assent as the mess intones its instructions. Welcome the mess in and let its tendrils throw through your veins. Understand the mess's greater plans and become its corporeal servant on this earth. Hail the mess. Hail the mess. All hail the mess. HAIL!
Really? Humans can improvise and adapt? So the human is supposed to build a mass spectrometer out of duct tape and discarded food pouches?
Modern science isn't conducted by rubbing two sticks together and seeing what happens, it requires complicated equipment. Whatever we send, that's going to be the equipment that does the science, whether it's a rover or a human behind it.
Humans can "go places and doing things" by means of us sending them and a huge amount of mass to support them, mass that could have made the robot vastly more capable of "going places" and "doing things" than the humans using up that payload mass could have. If your manned Mars mission costs 50 times more than a typical robotic mission then you're displacing 50 different robots with 50 completely different sets of capabilities sent to 50 different places on the planet. Or a dozen vastly more advanced rovers. Or hundreds of stripped down rovers.
Humans just simply cannot compare, gram for gram. And gram for gram is what matters when delivering payload to the surface of Mars costs upwards of $100 per gram. Humans really only buy you latency. And who gives a rat's arse about latency when budgetary constraints limit how often you can fund something like that? If you can get the money for such a "supermission" once every 20 years, what does it really matter if the data comes back after one year or three?
"Artificial gravity" is actually a well understood technology - centripetal/centrifugal force. We still need prototype testing though before we try it out, particularly if we want to do it with tethers rather than rigid structures (our experience with tethers in space has been less than stellar). It also imposes minimum size constraints on the diameter of the craft, as you don't want people exposed to too much tidal forces between their head and their feet.
Regardless, while living in space isn't "good for a person", the sort of timespans being talked about aren't fatal. The effects of the radiation exposure are more concerning. And the easiest way around that is huge amounts of mass being launched. And the most realistic way to do that at this point in time is probably to make launch costs lower (that's probably easier than dragging large enough asteroids into orbit and building spacecraft into them or breaking them into bits and cementing them onto a spacecraft). Low launch costs also have a tremendous number of side benefits. So in terms of "keeping humans safe and healthy during long voyages", the focus should be on reducing launch costs. Dramatically.
Thankfully physics doesn't stand in the way - our costs to orbit are several orders of magnitude higher than the energy costs, and even taking the rocket equation and "normal" chemical propellants as a given (not necessarily a fundamental assumption) it's "simply" a matter of engineering to get the costs down**. But said engineering takes serious funding. The sort of funding that we generally prefer to put into moon voyages, space stations, and the latest trend, manned Mars missions.;)
** Assuming a 20:1 mass ratio and $1/kg propellants, both of which are actually rather unfavorable assumptions, the propellant-cost-to-orbit is a mere $20/kg. Current launch costs are 100 to 1000 times that. So there's a massive amount of room for improvement.
We've already detonated over 2000 atomic bombs right here on Earth. The miniscule fraction of one bomb that might possibly return is pretty much irrelevant.
The tiny stuff could be nuked easily enough, but the really big stuff would just create a lot of really big (and now somewhat radioactive) rubble to carpet-bomb whatever place gets the impact.
This is a common myth. The reality is that simulations show that nuclear weapons can readily both deflect large asteroids without destroying them, or alternatively destroy them into bits too small to pose a threat and with too much momentum to reform into a large impactor. And even if that wasn't the case, there's also significant dispute among experts to the popularly repeated concept that a bunch of small pieces are worse than one big piece, as smaller pieces come in at varying time and thus spread out the heat load, ejecta load, etc, experience more burnup, produce much less powerful tsunamis that don't "echo" around the Earth as much, etc.
There's really no other option that has the sort of combination of A) near term technology, B) little lead time to deploy, C) minimal lead time required for effective deflection, D) low odds of failure, and E) capability of deflecting very large objects with small payloads - nuclear detonation, whether via standoff deflection or explosive disassembly, in is better than all or almost all competitors in every single category (kinetic impactors are slightly better in some categories but provide orders of magnitude less deflection capability for a given payload size)
Oh, another possibility for rule relaxation: they could allow electrical weapons, but only very low voltage (no limits on current). So it'd be quite possible to protect sensitive electronics (it takes voltage to get past electrical insulation), but robots will still be able to burn through each other with electricity.
Corrosives, if allowed, would probably have to have a requirement that they can't, in reaction with common metals and plastics, outgas anything dangerous. Flammables (and robots in general) should probably have a requirement about a minimum time they would take to burn up if everything flammable was ignited at once - you don't want anything too violent going on. Any uncertainties should come down to the opinion of a judge. No chemical should be allowed that reacts violently with water (or whatever fire extinguishing system they use).
Hmm, water hazards in the battle arena... that'd be neat.
The more varied the weapons, the more difficult it'll be for a robot to defend against all possibilities. And some weapons may prove most effective in combination. Maybe a weapon that cuts small holes in a foe's armour might not be good on its own, and maybe a water gun might not be great on its own, but punch a hole on a robot and then get the robot wet, and well, your results may be quite different;) I'd think that the most effective combinations would be like that - one to get you past the armor (spinning blades, hammers, vises, etc), and then one to damage the insides through the holes (fire, water, corrosives, etc). And wouldn't that be more satisfying to the audience, a multi-pronged attack ending with a robot frying its foe from the inside?:)
Letting drones compete would be a nice twist. I'm not sure how one could fairly handle letting drones compete with non-drones, though - seems like an either-or situation. Unless... hmm...
Other: Uneven floors and obstacles would be good, and could be varied, even with their own weather - a boulder-strewn badlands, a rainy forest, a post-apocalyptic hellscape, an icy barrens in a blizzard, etc. Maybe get rid of some of their old constraints, such as no fire-based or fluid weapons (although their arena would need be able to be able to handle it). Fire would not just be aesthetically pleasing but open up brand new optimization constraints on defensive techniques, and fluids like adhesives, solvents/corrosives or lubricants could also lead to rather interesting battles. Ranged weapons would be nice, within limits of course. Maybe simply replace the prohibition on ranged weapons to "no weapons that make use of an explosive charge" so that you don't just get a bunch of people mounting handguns to RC cars... maybe also set a maximum projectile velocity and/or kinetic energy too. Hmm, you know, if you allowed some degree of ranged weapons and had a realistic ceiling height, you might be able to fairly compete drones with non-drones. Otherwise you can still use weapons on a tether, but that's a lot harder.
The concern is and always has been stepping over the red "no nukes allowed" line. Once both sides of a conflict start playing with nukes, even if it starts out with small, tactical, targeted nukes, the other side will too, and whichever side is losing will be tempted to scale up, and ultimately you're opening up the risk of escalation into full blown strategic nuclear exchange. It's been the key reason why tactical nukes have been largely avoided for so long. The consequences of nuclear war ever breaking out between major powers are just so atrocious that one doesn't want to play lightly with anything that could make it easier to happen. Even when one side isn't a major power, we all know that these regional conflicts have a habit of escalating, and that different sides have a habit of misestimating how much of a line they're stepping over from the perspective of their rivals.
That said, the US may forced into this whether they want to or not, given that Russia's been developing - and has started deploying - tactical nuclear delivery systems. They've really been waving around their "nuclear card" a lot lately - my favorite was when they "accidentally" let a news camera capture a picture of design plans for a submarine-based cobalt bomb doomsday device among papers an officer was carrying.
Nuclear war gaming is a really morbid topic... the whole "if we do X, then they're going to do Y, then we'll have to do Z" thing, because the casualty numbers are so absurd... "If we do X here, then their attack will only kill between 6,3 and 7,5 million people, but our counter will kill between 23 and 26 million people, so that works out well to our favor..." The fact that even a "win" is really a devastating loss to the victor is what led to the concept of MAD.
Even on the battlefield it leads to weird situations. For example, part of the reason that neutron bombs were developed was the realization that should Soviet forces (which stressed a "deep battle" doctrine involving huge numbers of rapidly advancing tanks) flood into western Europe, the west could use nuclear weapons against their forces to try to stop them, but tanks tend to have a habit of surviving nuclear blasts unless they're near the epicenter. The radiation load might be fatal to the crew, but that could take days or more, and meanwhile the Soviet "zombie crews" could have taken control of a large chunk of Europe before they become too sick to continue. With neutron bombs, Soviets would have to respond by spreading their tanks out more, which greatly reduces their ability to be defended and supplied. It's possible to make tanks resistant to neutrons by incorporating neutron absorbers, such as boron, or moderators like hydrogen... but ironically the depleted uranium sometimes used in tank armour these days could actually enhance the yield of the radiation by undergoing fast fission.
Oh, and it's worth pointing out that dial-a-yield nuclear bombs are often effectively neutron bombs at their lower yield settings.
Slashdot Mirror
Jason-3 circularization burn underway.
I didn't submit this news because really we barely know anything at the moment. Jason-3 is still awaiting its second burn, and without knowing anything more than "it has a broken leg" I think it's too soon to call the landing a "partial success". The second burn will be happening shortly, and they said we'd get more data about the landing in a few hours.
Be patient, grasshopper.
I've had this experience with my house design a lot.
I've been held up by a construction boom leaving the country's architects overworked - the year before last I went through three of them because they weren't getting anything done. Last year I decided not to switch again (and thus have to begin from the beginning each time) and it took my architect 11 months to even get to the stage where we need to bring the engineer in. And before that there were lots of delays in finding and buying the property to begin with. But during all that time my planning for the house has really evolved, for the better. It's still an underground steampunk cave house, but it'll be cheaper to build, more comfortable to live in, and with a better look. Even having to switch architects helped because each one helped refine my thought process with their feedback.
Example: when I started my thought process involved actually boring the house out of the ground, like a tunnel, which would have been ridiculously expensive. I moved from that to the concept of building a timber form in a naturally low place, shotcreting it, the burying that. But building big timber domes here is expensive, and shotcrete unusually expensive here too. From there I moved to the current concept of simply making a big pile of compacted ground in a low point in the shape of the house and concreting over it (with forms only needed for where the slope is too steep), and burying that. Way simpler and cheaper.
I moved from having the house physically spread out to make it like a cave, to using convoluted bends to make it actually compact (even with common walls in places) but feel like it's spread out - thus greatly reducing the amount of material. I moved from a plastic ground cloth for waterproofing to multilayered rubberized-bitumen sheeting on compacted sand over the house, as my extra time gave me time to research longevity of different membranes and I found out that that's what's used to waterproof nuclear fuel repositories over great lengths of time. I moved from the idea of a clay-based plaster on the concrete to try to give it a cave-like feel to the concept of simply pressure blasting away the inner cement and exposing the aggregate. I've refined the details on the type of concrete to use greatly, and may now even be doing a research project out of it (FRP-rebar, loose plastic and basalt fiber, basalt dust pozzolan, etc). I came to the realization that due to the lack of limitations on how thick the insulation can be I could use pumice or scoria rather than a foam-based product. Also came to the realization (after taking the time to do heat flow simulations) that while in many "umbrella earth home" designs it's uninsulated under the house, my nearness to the bedrock and potential groundwater means that I should. I've come up with dozens of new steampunk and cave stylistic features to incorporate. And on and on. None of this would have happened had there not been great delays in the process.
Composites are physically very strong, and the use of epoxy rather than polyester will improve the lifespan. Still, composites yellow in UV light (particularly if the binder isn't designed to be UV stable, but to some degree even if it is) and the binder becomes more brittle. Also while they're strong (as in resistance to breakage) they're not hard (as in Mohs hardness), so they can still get scratched up. There are some scratch resistant coatings one could use (to varying degrees of effectiveness), but if you wanted natural scratch resistance without it being brittle you'd probably want something like chromium or ceramic-matrix composites.
Thank you for repeating my point.
Wow, someone's got an axe to grind. Use my whetstone, my good man.
Well... to some extent. You can always run hotter to be able to get more efficiency - for example, that's why coal gassification is more efficient than burning solid coal, you can run it very hot through a gas turbine. Also, recovery of heat once it gets to lower temperature differentials is often not done, so some new work is in making that more affordable, which is a big boon to resources that are fundamentally lower temperature (like geothermal). And of course pairing the two, high temperature cycles with low temperature cycles.
Thermal generation tech does keep improving. But you're right, it's not easy, and there are limits.
Sounds a lot easier than for a large scale wind turbine, the RPM is surely much higher - even if you need to gear it up somewhat.
If I remember this company's tech correctly, the wind drags it out, turning a generator as it goes. They then switch it to a low-drag mode and reel it back in. So it cycles between high power generation and low power consumption. The concept being that you'd have many of them so that you'd get continuous net generation.
The global wind resource is 72TW. You're going to have a hard time denting that. You might alter things locally - might - but not globally.
And besides that... if any evil greedy megacorp wants to move into my valley and setup something that will rob the wind of its energy... Please Do! Seriously, someone should really work on wind turbines specifically designed to act as windbreaks, in a manner that can be affordably mass produced and deployed in lieu of traditional manufactured windbreak systems.
(Before anyone says "just plant some trees..." I do, every year. They usually get killed by the weather, sometimes outright uprooted. I may have to start spending more time on each tree individually, encasing each one in its own individual PVC windbreak in addition to the broad rows of pallet windbreaks)
Who wants any kind of 100MW plant in their backyard? 100MW is an awful lot of power. Doesn't matter how you generate it, it's going to have to involve some combination of extreme areas, extreme pressures, and extreme temperatures - otherwise the physics doesn't work. If you want it small, it's going to have to be high pressure and/or hot. If you want it low pressure and cool, it's got to be huge. This ignores all other potential issues with each given type of power plant...
That's part of the reason that blimps are increasingly going out of favor, in favor of hybrid airships. Hybrids are based on the concept of being aircraft (generally lifting-body aircraft) that get huge lifting areas via inflation (so that they don't need any real structural strength, and thus keep the mass very low), and use helium as the inflation gas, which partially (but not completely) lifts the craft. So you have the combination of huge lifting area and much of your weight compensated for, so it takes little energy to stay aloft (which can, for example, come from solar), and unpowered landings are perfectly safe. But because (for the same payload capacity) you don't need as big of a gas bag as with a blimp, they're not as vulnerable to crosswinds and thus not as likely to break free, as well as being much easier to land without being blown into power lines or whatnot.
1) I've stood directly underneath a 1MW turbine in the middle of a large wind farm on a windy day. There wasn't much noise at all, just a light whoosh... whoosh... whoosh.... I don't know where this concept that wind turbines are "terribly noisy" comes from, but it doesn't at all match my experience.
2) You're concerned with noise about something that's hundreds of meters up? What about kite-based systems that operate at thousands of meters up, would you be worried about noise from them too?
You should see the setup they use to produce the graphene before separating out the layers with tape.
Also, may I add: it's a stupid assumption that robot operators can't innovate either. Because they do this sort of stuff all the time, inventing new techniques - using the hardware they sent - to do things that weren't expected at the time. From rovers dragging wheels to expose buried sediments while they roam, to New Horizons' doubling its planned data throughput by the realization that they could run both TWTAs at the same time with different polarization if they got the extra power by shutting down the flight computer after spin-stabilizing the craft.... the limitation isn't humans. Because humans are involved, they're the ones controlling the robots. The robots are only limited by what hardware was sent. As are humans. Send a robot with a duct tape tool and it can make things out of duct tape too that weren't planned when they launched.
Come now, do you really expect me to do coordinate substitution in my head while strapped to a centrifuge?
It seems like every developer has a go-to technique (or four) for keeping things a little more streamlined. What are yours?
The mess is just part of coding. Deal with the mess. Accept the mess. Live with the mess. Join with the mess. Speak with the mess. Hear its answers whispered into your head. Nod assent as the mess intones its instructions. Welcome the mess in and let its tendrils throw through your veins. Understand the mess's greater plans and become its corporeal servant on this earth. Hail the mess. Hail the mess. All hail the mess. HAIL!
Really? Humans can improvise and adapt? So the human is supposed to build a mass spectrometer out of duct tape and discarded food pouches?
Modern science isn't conducted by rubbing two sticks together and seeing what happens, it requires complicated equipment. Whatever we send, that's going to be the equipment that does the science, whether it's a rover or a human behind it.
Humans can "go places and doing things" by means of us sending them and a huge amount of mass to support them, mass that could have made the robot vastly more capable of "going places" and "doing things" than the humans using up that payload mass could have. If your manned Mars mission costs 50 times more than a typical robotic mission then you're displacing 50 different robots with 50 completely different sets of capabilities sent to 50 different places on the planet. Or a dozen vastly more advanced rovers. Or hundreds of stripped down rovers.
Humans just simply cannot compare, gram for gram. And gram for gram is what matters when delivering payload to the surface of Mars costs upwards of $100 per gram. Humans really only buy you latency. And who gives a rat's arse about latency when budgetary constraints limit how often you can fund something like that? If you can get the money for such a "supermission" once every 20 years, what does it really matter if the data comes back after one year or three?
I nanodisagree with you.
"Artificial gravity" is actually a well understood technology - centripetal/centrifugal force. We still need prototype testing though before we try it out, particularly if we want to do it with tethers rather than rigid structures (our experience with tethers in space has been less than stellar). It also imposes minimum size constraints on the diameter of the craft, as you don't want people exposed to too much tidal forces between their head and their feet.
Regardless, while living in space isn't "good for a person", the sort of timespans being talked about aren't fatal. The effects of the radiation exposure are more concerning. And the easiest way around that is huge amounts of mass being launched. And the most realistic way to do that at this point in time is probably to make launch costs lower (that's probably easier than dragging large enough asteroids into orbit and building spacecraft into them or breaking them into bits and cementing them onto a spacecraft). Low launch costs also have a tremendous number of side benefits. So in terms of "keeping humans safe and healthy during long voyages", the focus should be on reducing launch costs. Dramatically.
Thankfully physics doesn't stand in the way - our costs to orbit are several orders of magnitude higher than the energy costs, and even taking the rocket equation and "normal" chemical propellants as a given (not necessarily a fundamental assumption) it's "simply" a matter of engineering to get the costs down**. But said engineering takes serious funding. The sort of funding that we generally prefer to put into moon voyages, space stations, and the latest trend, manned Mars missions. ;)
** Assuming a 20:1 mass ratio and $1/kg propellants, both of which are actually rather unfavorable assumptions, the propellant-cost-to-orbit is a mere $20/kg. Current launch costs are 100 to 1000 times that. So there's a massive amount of room for improvement.
We've already detonated over 2000 atomic bombs right here on Earth. The miniscule fraction of one bomb that might possibly return is pretty much irrelevant.
This is a common myth. The reality is that simulations show that nuclear weapons can readily both deflect large asteroids without destroying them, or alternatively destroy them into bits too small to pose a threat and with too much momentum to reform into a large impactor. And even if that wasn't the case, there's also significant dispute among experts to the popularly repeated concept that a bunch of small pieces are worse than one big piece, as smaller pieces come in at varying time and thus spread out the heat load, ejecta load, etc, experience more burnup, produce much less powerful tsunamis that don't "echo" around the Earth as much, etc.
There's really no other option that has the sort of combination of A) near term technology, B) little lead time to deploy, C) minimal lead time required for effective deflection, D) low odds of failure, and E) capability of deflecting very large objects with small payloads - nuclear detonation, whether via standoff deflection or explosive disassembly, in is better than all or almost all competitors in every single category (kinetic impactors are slightly better in some categories but provide orders of magnitude less deflection capability for a given payload size)
Oh, another possibility for rule relaxation: they could allow electrical weapons, but only very low voltage (no limits on current). So it'd be quite possible to protect sensitive electronics (it takes voltage to get past electrical insulation), but robots will still be able to burn through each other with electricity.
Corrosives, if allowed, would probably have to have a requirement that they can't, in reaction with common metals and plastics, outgas anything dangerous. Flammables (and robots in general) should probably have a requirement about a minimum time they would take to burn up if everything flammable was ignited at once - you don't want anything too violent going on. Any uncertainties should come down to the opinion of a judge. No chemical should be allowed that reacts violently with water (or whatever fire extinguishing system they use).
Hmm, water hazards in the battle arena... that'd be neat.
The more varied the weapons, the more difficult it'll be for a robot to defend against all possibilities. And some weapons may prove most effective in combination. Maybe a weapon that cuts small holes in a foe's armour might not be good on its own, and maybe a water gun might not be great on its own, but punch a hole on a robot and then get the robot wet, and well, your results may be quite different ;) I'd think that the most effective combinations would be like that - one to get you past the armor (spinning blades, hammers, vises, etc), and then one to damage the insides through the holes (fire, water, corrosives, etc). And wouldn't that be more satisfying to the audience, a multi-pronged attack ending with a robot frying its foe from the inside? :)
Letting drones compete would be a nice twist. I'm not sure how one could fairly handle letting drones compete with non-drones, though - seems like an either-or situation. Unless... hmm...
Other: Uneven floors and obstacles would be good, and could be varied, even with their own weather - a boulder-strewn badlands, a rainy forest, a post-apocalyptic hellscape, an icy barrens in a blizzard, etc. Maybe get rid of some of their old constraints, such as no fire-based or fluid weapons (although their arena would need be able to be able to handle it). Fire would not just be aesthetically pleasing but open up brand new optimization constraints on defensive techniques, and fluids like adhesives, solvents/corrosives or lubricants could also lead to rather interesting battles. Ranged weapons would be nice, within limits of course. Maybe simply replace the prohibition on ranged weapons to "no weapons that make use of an explosive charge" so that you don't just get a bunch of people mounting handguns to RC cars... maybe also set a maximum projectile velocity and/or kinetic energy too. Hmm, you know, if you allowed some degree of ranged weapons and had a realistic ceiling height, you might be able to fairly compete drones with non-drones. Otherwise you can still use weapons on a tether, but that's a lot harder.
Second. Why Battlebots decided to emulate WWF is beyond me.
The concern is and always has been stepping over the red "no nukes allowed" line. Once both sides of a conflict start playing with nukes, even if it starts out with small, tactical, targeted nukes, the other side will too, and whichever side is losing will be tempted to scale up, and ultimately you're opening up the risk of escalation into full blown strategic nuclear exchange. It's been the key reason why tactical nukes have been largely avoided for so long. The consequences of nuclear war ever breaking out between major powers are just so atrocious that one doesn't want to play lightly with anything that could make it easier to happen. Even when one side isn't a major power, we all know that these regional conflicts have a habit of escalating, and that different sides have a habit of misestimating how much of a line they're stepping over from the perspective of their rivals.
That said, the US may forced into this whether they want to or not, given that Russia's been developing - and has started deploying - tactical nuclear delivery systems. They've really been waving around their "nuclear card" a lot lately - my favorite was when they "accidentally" let a news camera capture a picture of design plans for a submarine-based cobalt bomb doomsday device among papers an officer was carrying.
Nuclear war gaming is a really morbid topic... the whole "if we do X, then they're going to do Y, then we'll have to do Z" thing, because the casualty numbers are so absurd... "If we do X here, then their attack will only kill between 6,3 and 7,5 million people, but our counter will kill between 23 and 26 million people, so that works out well to our favor..." The fact that even a "win" is really a devastating loss to the victor is what led to the concept of MAD.
Even on the battlefield it leads to weird situations. For example, part of the reason that neutron bombs were developed was the realization that should Soviet forces (which stressed a "deep battle" doctrine involving huge numbers of rapidly advancing tanks) flood into western Europe, the west could use nuclear weapons against their forces to try to stop them, but tanks tend to have a habit of surviving nuclear blasts unless they're near the epicenter. The radiation load might be fatal to the crew, but that could take days or more, and meanwhile the Soviet "zombie crews" could have taken control of a large chunk of Europe before they become too sick to continue. With neutron bombs, Soviets would have to respond by spreading their tanks out more, which greatly reduces their ability to be defended and supplied. It's possible to make tanks resistant to neutrons by incorporating neutron absorbers, such as boron, or moderators like hydrogen... but ironically the depleted uranium sometimes used in tank armour these days could actually enhance the yield of the radiation by undergoing fast fission.
Oh, and it's worth pointing out that dial-a-yield nuclear bombs are often effectively neutron bombs at their lower yield settings.