Nuclear pulsed propulsion

[lamontagne]

I'll try to post a preliminary spreadsheet today.

Include propellant cost in this spreadsheet.  One bullet's HALEU, at 20% enrichment, would cost ~ $30,000 near-term.

Thanks, I'll try to add links to sources in the spreadsheet.

[lamontagne]

Spreadsheet with some propellant costs.

This is too early to draw any conclusion, as most vehicles are not designed, except that if the cost of the bullets for PPR is really 24-30000 $ per bullet, this is not a very likely solution.  Price needs to go way down.

Starship is a really neat solution.  It's hard to beat without significant technological advancements. 
Starship with Starship shuttles at each end and extensive ISRU on Mars would probably be even more cost competitive.

Going fast is incredibly expensive, energy wise.  The price of cargo vs people will be very different if rapid transit is a necessity.

[LMT]

Spreadsheet with some propellant costs.

Tritium breeding seems quite feasible in 3F tugs.  1 2 3

Don't ignore that. 

At $30,000 per gram, even minor excess tritium could pay for the other 3F nuclear fuels.  So here nuclear fuel cost is fairly stated as $0.  (Any system that aims to breed tritium should get the same benefit in spreadsheet, if fuel masses, neutron energies, etc., check out.)  So calc only ISRU MOX cost there, much as you would for a cycling Starship tanker, and being careful to use plausible and applicable Starship tanker operating costs overall.

[Twark_Main]

At $30,000 per gram, even minor excess tritium could pay for the other 3F nuclear fuels.  ...So calc only ISRU MOX cost

Odd how you choose to apply this clever little accounting "sleight-of-hand" trick to one, but not the other.

Care to justify this discrepancy, or is it just unabashed special pleading? 

Keep in mind investigators from the SEC should be in the office to check your books this afternoon, and they're gonna ask all the same questions... 

[InterestedEngineer]

I'm kind of confused now.

Is the idea that one can breed tritium in the PPR and pay back the $30/kg cost of the HALEU?

If that's on a spreadsheet that'd be two different items, hence the term "itemized".

I note if it is that easy to breed tritium then the market value of tritium will drop dramatically.

[Asteroza]

Video presentation to the interstellar group

Let's just take a moment to consider the person who worked this out.
So

Control drum --> spinning continuously have sine effect on neutron output -->Different rpm gives more complex output --> use Fourier analysis to create high intensity neutron pulse

Control drums are pretty much SOP for NTR and some space nuclear power designs. But I've never seen the notion of spinning them, or at different rates to create a complex modulation of the neutron output .

I could be wrong, but I think this is could be a genuinely original development in rocket engineering.

 

There are a couple of possible tweaks to the design.

My instinct is to try and keep the packages as simple as possible and put as much of the complexity on the vehicle.

So watching this video from will tell you that  the drums would need to spin at odd multiples of a baseline spin rate and have falling levels of moderation. So if the baseline is 1rpm and full moderation the 6th drum is 11rpm with 1/11 th the moderation. BTW Fourier analysis says nothing about what pattern those drums should be in. Might be an issue. Might not. 

The conceptually simple solution is to spin the drums up with a set of electric motors and a battery and they align at peak output just as it comes up to the muzzle.

My instinct is to spin them up before launch and time it so they spin down to the level that gives maximum pulse. Since this happens outside the barrel they are not being moderated at this point.

The second point is the use of HEU. The work on KiloPower showed what a monumental PITA this is IRL. In fact even HALEU is difficult unless you can get access to government uranium that's been blended down from weapons stock.

Alternatives would be to use PWR grade (about 4-5% U235) and back it with a reflector. Reflector properties were considered in SDI papers in the 80's and SNP power reactor projects in the 90's. Be and BeO were the front runners but MgO was IIRC the winner. Much cheaper and way less toxic. MgO actually has a long history in nuclear work as it's the insulator for Mineral Insulated Cable used in lots of NPP's and fire alarm systems. In powder form it's cheap, has a very high Mp and could be compressed if you wanted a high density solution.

I've never seen any details of the Orion pulse modules but AFAIK the weaponeers who worked on it had a background in implosion based bombs, which would suggest an implosion design with efficient use of uranium. If Howe's design can replace a complex explosive implosion by a design keeping most of the U on board (in the barrel and muzzle), which is in effect reused on every pulse this could be the Holy Grail of in-system space flight. Relatively cheap, fast travel within the solar system using known physics (used in a creative way).

Lastly in materials science "Sapphire" implies single crystal Alumina. If you're talking polycrystalline Alumina that's much easier to make as a coating. Multiple routes from temperatures ranging around 1000c (essentially burning AlCl3) dating from GE's work on the Nuclear Aircraft programme to lower temp methods developed for semiconductor mfg. Otherwise you're talking machining panels to fit a nozzle contour, which will be challenging.

If people can get over their aversion to fission in space this has a lot of (potentially, given no hardware has been built) very attractive features.

Nice work.

sine wave pulse makes me think...

I just had an immensely bad idea, based on an early concept I floated about a nuclear rotary ramjet that utilizes a stationary "pulse" in a ramjet compression throat.

If you are rotating control drums at high-ish speeds, why not something resembling a rotating detonation wave engine? You spin a balanced outer cylinder carrying a pair of control drums, and squirt high pressure water into the RDE chamber. Control drum actuation and positioning relative to the "pulse" can vary. Chamber setup to get a slotted uranium drum spinning probably makes this difficult though...

[LMT]

I'm kind of confused now.

Is the idea that one can breed tritium in the PPR and pay back the $30/kg cost of the HALEU?

If that's on a spreadsheet that'd be two different items, hence the term "itemized".

I note if it is that easy to breed tritium then the market value of tritium will drop dramatically.

3F could breed tritium beyond self-sufficiency, to zero out nuclear fuel cost.  I think that hasn't been attempted with PPR.

Price wouldn't "drop dramatically"; most of the tritium must go back into the drive, you know.

A fusion drive that can pay its own way -- where else have we seen such a concept?

[LMT]

An example of a fusion drive concept with breeding:

Ball 2019 uses a water blanket to breed deuterium, increasing a notional D-D drive's specific energy by up to 65%, and increasing Isp by 29% or more.

Image:  "A cartoon fusion-powered spacecraft with neutron pusher/blanket..."  Ball 2019.  Blanket angular extent in red.

Refs.

Ball, J., 2019. Maximizing specific energy by breeding deuterium. Nuclear Fusion, 59(10), p.106043.

[Twark_Main]

I'm kind of confused now.

Is the idea that one can breed tritium in the PPR and pay back the $30/kg cost of the HALEU?

3F could breed tritium beyond self-sufficiency, to zero out nuclear fuel cost.

For the third time (after your outburst got nuked), you don't "zero out" any other (arbitrarily chosen) cost buckets. That's not how accounting works, that's how shady underhanded accounting works. 

You need to account for everything separately. You can't just ignore one cost and one income by assuming that the two dollar amount will always be equal. You risk either falling short, or (arguably worse) leaving money "on the table" by not fully optimizing your revenue-generating process — reasoning that if it covers "its" cost, why bother doing more?

Fortunately the stakes in this miscalculation are pretty low. The risk isn't that you'll deceive investors (which would be tragic), but that you'll deceive yourself about the financial landscape, and waste a bunch of your own time pursuing a path that's less-than-opt........

...you know, on second thought...   carry on LMT!   

[lamontagne]

I've added a page summing up the values of the engine.  There are a few discrepancies, likely typos or the results of multiple iterations.

I find the burn up fraction to be quite low, well bellow 0.1%.  This seems a bit wasteful of nuclear fuel, although I guess this is consistent with the experience from smaller nuclear weapons?

In general, for the nuclear industry, the costs of the fuel might be significantly reduced is the isotopes were laser separated rather than centrifugally separated, so perhaps the present value of Haleu is not representative of what it would be for such a vehicle.  I guess we need more references.

Clearly, there is no way this could compete with Winterberg's compressed fission proposal, if it was workable, as Winterberg proposed 10% burn up fractions.

My next step is looking at the cooling requirements for removing 127 MW from a barrel the size of small desk.  Which really was my original intent :-)

[LMT]

I'm kind of confused now.

Is the idea that one can breed tritium in the PPR and pay back the $30/kg cost of the HALEU?

3F could breed tritium beyond self-sufficiency, to zero out nuclear fuel cost.

For the third time (after your outburst got nuked), you don't "zero out" any other (arbitrarily chosen) cost buckets. That's not how accounting works, that's how shady underhanded accounting works. 

You need to account for everything separately. You can't just ignore one cost and one income by assuming that the two dollar amount will always be equal. You risk either falling short, or (arguably worse) leaving money "on the table" by not fully optimizing your revenue-generating process — reasoning that if it covers "its" cost, why bother doing more?

Fortunately the stakes in this miscalculation are pretty low. The risk isn't that you'll deceive investors (which would be tragic), but that you'll deceive yourself about the financial landscape, and waste a bunch of your own time pursuing a path that's less-than-opt........

...you know, on second thought...   carry on LMT!   

Who noticed hybrid fusion's ready production of excess tritium, at $30,000 per gram?  It's an extraordinary gift, when you think about it.

Pulsed drives that don't breed fuel would be at a terrible economic disadvantage, whatever the Isp.  Drive proponents might give the challenge some creative thought.

[LMT]

Clearly, there is no way this could compete with Winterberg's compressed fission proposal...

Leveraging Thermosphere Oxygen

Beyond D-T and breeding advantage, ISRU propellant also gives a cost advantage in many cases.  Oxygen comes first because it's the easiest to harvest:  a VLEO LOX ISRU system can precede even Martian ISRU, itself preceding the need for nuclear propulsion.  So it's fair to zero out your "cost to put propellant in LEO", for LOX. 

-  For PPR, that's ~ 89% of the cost (water).

-  For 3F / MOX, that's ~ 78% of the cost (methalox).

Cases for other ISRU propellants could be considered separately.

-

The target is delivering a payload of 100 tonnes to Mars...

200 t is SpaceX baseline.

The reference baseline is a 4 month (120 days) transfer by the SpaceX Starship...

~ 6 months with 200 t.

[lamontagne]

Two views of the possible arrangement for the PPR fission engine.

The bullet is the red element lurking in the center of the model.

The large (2m) yellow thingy is the magnetic nozzle.

The nozzle is tentatively mounted on six green shock absorbers, that have an arbitrary 15 cm of travel.  They operate at 1 hz, like the engine.  the six drums also rotate at 1 hz, or 60 rpm.

Now, to develop the radiators for 127 MW of cooling!

[LMT]

six green shock absorbers

MMO has a greater blast, to 120 t TNT equivalent (5e11 J vs. PPR's 3e9 J), and MMO doesn't use shock absorbers.

Regenerative shock absorbers can recover some energy, as with a greater, superconducting magnetic mirror, but they're likely an optional comfort and not required for the crew, even at MMO scale.

[lamontagne]

six green shock absorbers

MMO has a greater blast, to 120 t TNT equivalent (5e11 J vs. PPR's 3e9 J), and MMO doesn't use shock absorbers.

Regenerative shock absorbers can recover some energy, as with a greater, superconducting magnetic mirror, but they're likely an optional comfort and not required for the crew, even at MMO scale.

Interesting reference, thanks.  The pulse unit has a 10% burn up fraction, compared to a fractional burn-up fraction for PPR, so that explains most of the difference. 

Note the nozzle is 20m in diameter and masses 100 tonnes, so these is a vast difference in scale between the two vehicles.

The joined paragraph, extracted from the paper implies, to me, a mechanism to transmit the energy of the nozzle to the ship.  So I'll keep my design for the time being, but eventually I'll be interested to check the shock limits you also sent. 

[LMT]

Note the nozzle is 20m in diameter...

Jerk (g/s) scales with pulse energy, not nozzle diameter.

[lamontagne]

A question on Nozzle/pulse interaction.  How long is the propulsive phase in a pulsed propulsion drive?

The reaction of the PPR produces a plasma pulse.  This plasma is leaving at the exhaust velocity after having interacted with the magnetic field of the nozzle.  So it won't be there very long. At best, as far as I can understand, it will be in the nozzle for Ve/d=t.  So for Ve= 50 000 m/s, and d=1 for a 2m diameter nozzle t=0,00002 seconds.    It follows the actual thrust will be much higher than the average thrust, something like 100 kN/0,0002 = 5e9 N.  And for a 500 tons vehicle, the acceleration will then be about 5e9 N /500 000 kg = 10 000 m/s, or 1000g.

That is quite a shock.  The shock absorber system is used to spread out that acceleration over time and bring the shock down to a somewhat more gentle continuous thrust.

There is also the issue that for a nozzle, in one direction it is pushing against a 500 tons ship, while in the other direction, as it bounces back, it is not pushing against anything.  Hence the two step system used in Orion.

This should be true for any pulsed propulsion vehicle.  but it's very dependent on how much time the blast interacts with the nozzle, so I wonder if something might be wrong with my original equation of speed and distance. 
1- Does the interaction last more time than I expect? 
2- Is this a good reason to build really big nozzles?

I've joined some interesting information gathered by Scott Lowther in his Aerospace Project Review publication on the subject of Orion.  A bargain at 8.00$ an issue, if you've never come across these. https://www.aerospaceprojectsreview.com/npp.htm

note : Oups, changed my nozzle size by mistake.  new table.

[LMT]

How long is the propulsive phase in a pulsed propulsion drive?

You're looking for acceleration and jerk (g/s), varying as plasma interacts briefly with the magnetic mirror.  You might ballpark MMO example evolution from the time progression of density contours, Fig. 11.

[lamontagne]

How long is the propulsive phase in a pulsed propulsion drive?

You're looking for acceleration and jerk (g/s), varying as plasma interacts briefly with the magnetic mirror.  You might ballpark MMO example evolution from the time progression of density contours, Fig. 11.

Fig 11, image 5 is after 120 microseconds and seems mostly clear of plasma. So about 0,00012 seconds. Very roughly 3 times longer for a 10m radius nozzle than for my calculation of a 1m radius nozzle, at 0,00004s. The MMO has twice the exhaust velocity.  If I apply my simplistic model 10m/100 000 m/s = 0.0001 which is fairly close.  So 1000g +-20% for the PPR.

[LMT]

How long is the propulsive phase in a pulsed propulsion drive?

You're looking for acceleration and jerk (g/s), varying as plasma interacts briefly with the magnetic mirror.  You might ballpark MMO example evolution from the time progression of density contours, Fig. 11.

Fig 11, image 5 is after 120 microseconds and seems mostly clear of plasma. So about 0,00012 seconds. Very roughly 3 times longer for a 10m radius nozzle than for my calculation of a 1m radius nozzle, at 0,00004s. The MMO has twice the exhaust velocity.  If I apply my simplistic model 10m/100 000 m/s = 0.0001 which is fairly close.  So 1000g +-20% for the PPR.

The rate of change of acceleration is jerk, which determines the need for a shock absorber.  You scale that plot to total MMO thrust, 1,870 kN.