VASIMR news

10kBq Jaro · Senior Member

VASIMR news

John · Senior Member

That is good news. I'm glad to see that NASA is moving ahead with VASIMR. It will be interesting to see if we are going ahead with a nuclear thermal (NERVA-type) systems as well. I'm very undecided on this as you don't really get all that much increase in performance over chemical (hydrogen-oxygen) by going to nuclear thermal...about double the specific impulse. Yet you have all of the political issues with orbiting reactors.

On the other had we will need nuclear energy to run a VASIMR system powerful enough to be an alternative to the nuclear thermal system.

John · Senior Member

I tend to be very partial to VASIMR personally as its one the idea I can up with a few years ago independently before I found out that I was far from the first! lol. What I've been kicking around is a VASIMR manned Mars mission. If we assume a 200 metric ton spacecraft and what a thrust at peak Isp of 200 nt inorder to get an acceleration of a .001 meter/sec. That looks to me to be about 20 mega watts of thrust so we will need a nuclear power source of I'd guess at least 60 mega watts or more. This will all depend on efficiencies.

I've played around with the idea of a bi-modal system. If we run some H2 through the reactor we could reduce the mircowave heating required by the plasma thruster. In addition what if we mixed the ultra high Isp plasma thrust with ordinary super-heated H2 in order to get a higher thrust density power level. (I'm having trouble with the layout of the device. The old mixing of temperature level thing.) But perhaps the plasma thrust could heat the H2 enought to give a non-atomic hydrogen kick to the how thing. Comments?

GoogleNaut · Guru

I don't think a 'direct' co-axial bimodal system will work very well as you suggest. The reason for this is the way that VASIMR works. It uses Helicon antennas to first ionize a working fluid, usually hydrogen, but it can be almost any gas (deuterium seems to be about the best, interestingly enough...) and then superheat the plasma by using a second power stage helicon antenna with the lion's share of the rf power. Both helicons must be actively cooled because of I^2*R losses and a small amount of radiation coupling from the plasma which becomes more significant at the highest power burns.

The reason why a coaxial bimodal system wouldn't work is ironically because of the throughput--a VASIMR simply cannot process the gas fast enough. This is because as a gas becomes ionized, it becomes opaque to rf energy (this is the reason for the radio blackouts during space shuttle descents when the plasma sheath completely encloses the orbiter during reentry.) So the only way to completely ionize a gas is to process it in small enough amounts that the rf 'window' is enough to completely saturate the gas. Otherwise no matter how much power you pump into it, there will be a core of gas that will still be unionized, or only partially ionized. In a VASIMR engine, this 'unburned' propellant equates to inefficieny. With the kind of flows from a an NTR the vast majority of the gas will be unionized, so carrying a VASIMR 'afterburner' will only be dead weight. Might as well just discard the VASIMR part completely...

Another problem is cooling of the helicons. The helicons must be immersed or partially immersed in the plasma stream. In the small test units run so far, the helicons are on the outside of about a 5-10 cm quartz glass tube, completely transparent to rf energy. This accomplishes plasma segregation quite nicely. In larger units, it is likely that the helicons will probably be partially immersed in the plasma stream and will rely upon the the magnetic fields to buffer them from plasma erosion.

Further because the gas flow through a typical NTR will be much, much higher (typically about 3-5 orders of magnitude higher!) then this necessitates active cooling along the entire flow path. It won't be possible to regeneratively cool the entire flow path, so a secondary coolant loop with large heat dissipating radiators (all very heavy and complex) would be needed. Further the added active cooling would likely interfere with the helicon antennas' job of plasma heating, but I'm not sure to what degree this could be a problem. Certainly if the coolant channels were made of metal, and the helicons were outside of this cooled duct, then you will have rf induction issues--and no plasma heating at all. Clever design may alleviate these problems, but I suspect that whatever gains might be made would probably be offset by the added complexity and weight of such a system.

O.K., now a non-coaxial multimode NTR/VASIMR system might make more sense. A large NTR could provide higher thrust burns for quick planetary escape and capture, while the VASIMR cruise engine performs cruise delta-v runup and rundown and midcourse corrections. The feasability of such a combined cycle system will be apparent (or not) after a detailed engineering study--which to say the very least is a complex task. As yet, no VASIMR engineering prototype has ever been flown, so the precise scaling laws have not been determined.

Still both VASIMR and multimodal NTR are exciting and promising technologies. And I have confidence that their development and adaptation will pave the road to humanity's colonization of the solar system.

-- Edited by GoogleNaut at 21:26, 2006-02-11

-- Edited by GoogleNaut at 21:28, 2006-02-11

10kBq Jaro · Senior Member

John wrote:

you don't really get all that much increase in performance over chemical (hydrogen-oxygen) by going to nuclear thermal...about double the specific impulse.

I think that the particular application of NTRs is at least as important as comparisons with chemical rockets.

In some cases, it may be very advantageous to use an NTR even if its performance is equal to or slightly inferior to a chemical rocket.

For instance, ISRU using NTRs is typically much less demanding of infrastructure (i.e. water electrolysis or methane manufacture and long-term storage) than rockets requiring cryo tankage for H2 & O2, or CH4 & O2.

Analogously, its odd that some folks can readily understand the tremendous potential of Orion-type propulsion, despite their horrendously poor fuel efficiency, but fail to credit the obvious ISRU advantages of low-Isp NTRs: Only a small fraction of each Orion blast is utilised for thrust, while a very high fraction of energy released in H2-O2 combustion contributes to chemical rocket thrust. But who cares, if it can't get the job done, due to other factors (low total energy content, in the case of interplanetary transfer application).

John wrote:

you have all of the political issues with orbiting reactors.

For trans-lunar shuttle service, yes, because presumably the reactor would be run several dozen times before being retired (and subsequently released to decay in a very long-lived orbit). That type of service would certainly create a large fission product inventory, plus a small amount of long-lived transuranics.

But the same issues also arise with a high-power VASIMR using a reactor & conversion system for electricity generation.

In other words, if the "political issues with orbiting reactors" are deemed too challenging for NTRs, then the same will also be true for NEPs. In that case, you might as well forget about nuke propulsion anywhere but in deep space.

But is that a reasonable assumption, given that we're safely operating dozens of nuke subs & ships right here on Earth ? ...I think not !

Anyway, how many of these nuke lunar shuttles are we likely to operate concurrently during the first half of this century ? ....two at the most ? ...give me a break !!!

John · Senior Member

I also said:

It will be interesting to see if we are going ahead with a nuclear thermal (NERVA-type) systems as well. I'm very undecided on this

So I'm not anti-nuclear thermal. As far a things nuclear reators, impluse units, etc. in space. That's ultimately going to be necessary. What I was addressing was specifically would it be wise to go nuclear in the near term for the limited advantage of nuclear thermal and for near term missions? I still remain undecided. One other issue that bothers me about nuclear thermal systems is the very large quantities of LH2 that would need to be maintained through out the duration of a Mars mission for example. If you go to water for example your specific impulse is cut from ~900 to ~300 that is the order of magnitude of good hypergolics like UDMH and NO4. CO2 would reduce it further to ~200. The other issue is that the reactor and shielding weigh considerably more than a chemical engine with similar thrust. But you are right that if you got the CO2 on Mars to use to propel your lander/exploration vehicle or for the return trip that would change things a lot.

It seems to me that we could go forward with nuclear thermal systems if the political decision is made to do so. They don't require any fundamental advances in technology.

Getting back to VASIMR it might even be possible for mission no further out than Mars and in particular for cargo transport from Earth orbit to Moon orbit to use solar power. For one thing solar energy is availiable in this region as compared with the outer planets. Second the nuclear electric systems require large radiators to remove the waste heat. One could consider a craft with similar sized solar panels and not have the weight of a reactor/turbine/generator. I woud expect that the solar panels are lighter per unit of area than the radiators. Beyond Mars we must have nuclear.

-- Edited by John at 14:17, 2006-02-12

10kBq Jaro · Senior Member

John wrote:

Getting back to VASIMR it might even be possible for mission no further out than Mars and in particular for cargo transport from Earth orbit to Moon orbit to use solar power.

AGREED !

.....the other side of the coin is of course that crewed ships need quick transit, ergo nuke power.

GoogleNaut · Guru

I would just add that it might be possible to utilize some 'lower quality' propellants by going to nuclear power sources. What I mean is that VASIMR has an amazingly unique capability to tune its rf energy stream to a wide range of frequencies. I don't have the paper in front of me, but the basic idea is to tune the the first helicon to the ion-cyclotron freqency for a particular atomic arrangement. These frequencies are well known, and relationships exist to easily calculte the necessary freqency for just about any magnetic field strength desired. For the laboratory VASIMR device using hydrogen, the first Helicon was tuned for electron-cyclotron resonance with a 1 Tesla (10,000 gauss) magnetic field, and had a peak rf output in the 10-12 MHz range. This ionzies the propellant stream but doesn't 'heat' the plasma very much. The second helicon was tuned to achieve ion-cyclotron resonance with most of the power going into this stage. If memory serves this was around 6 MHz, and ate up most of the rf power. This achieves most of the plasma heating. Peak heating can apparently be close to 100 million Kelvins, but the plasma get's awfully thin and erosion of engine materials begins to be a problem...

Anyways, it is possible to retune and engine by changing the magnetic field strength and the rf frequency so that something like Oxygen can be used. Why use oxygen? Well, about 60% of lunar regolith is oxygen by weight. An orbiting lunar supply depot, catching mass from the moon using a mass driver on the surface, would process lunar regoltih to seperate out metals like iron, aluminum, and titanium. Eventully oxygen will be a waste product of this 'smelting' operation. Oxygen can be compressed, liquified and stored in large cryogenic tanks. So a nuclear powered vessel using tunable VASIMR engines comes by, picks up its cargo of metals and tanks up on liquid oxygen. What kind of performance can be had with oxygen? Well, if at the high end hydrogen can produce about 15,000 to 30,000 seconds of specific impulse, then because hydrogen has a molecular weight of about one and oxygen has a weight of about 16, then oxygen should give 1/16 the engine performance, so say between 1000 seconds to 2000 seconds of specific impulse. What this means is that our spacecraft will have to carry about 2.8 times (at the 1000 second specific impulse level) the propellant weight of oxygen as it would need if hydrogen were used to achieve the same delta-v with the same payload (and dry vehicle mass.)

What this means is that if the vehicle needed 50 tons of LH2 to achieve a particular mission goal, it will need atleast 140 tons of LO2 to achieve the same result. Since LO2 is much denser than LH2, even with the increased mass loading, the propellant tanks will still be much smaller. This improves the carrying efficiency of the tanks by quite a bit.

The possibility of creating an NTR system using liquid oxygen as a propellant is not impossible, but would present many unique challanges. By definition, once the oxygen was evaporated and the gas heated even a little bit, the result will be a hot oxidizing flow. The Russians are the only ones with much experience in dealing with hot oxidizing flows in their RD-170 and RD-180 engines. A nuclear thermal engine in this regime is techinically formidable. It would probably require substantial use of thorium oxide based ceramics for core mechanical and structural fittings. Also the reactor fuel could not be clad with anything that would oxidize, which seems to necessitate some kind of bare oxide fuel element technology--which I don't think has ever been done before (certainly not with hot oxidizing flows.) Also as Jaro has pointed out to me on several occasions, oxide fuel elements do not have great thermal conductivities, so this necessitates very small fuel structures. Perhaps fused fuel/oxide ceramic fibers spun into something like concentric Coleman lantern mantels. As gas permiates the fibers heat is picked up, but whether enough is picked up remains to be seen. It is entirely probable that the fibers would just melt, and spray and soak into the fuel mats downstream, clogging and decreasing coolant flow while increasing back pressure. I can't see how this could end in anything other than an explosion. So hot oxygen flows through a LO2 NTR system just seems a little too exotic.

But I think VASIMR could do it.

John · Senior Member

GoogleNaut, 10k or anyone else,

What is the efficiency of the VASIMR in converting electrical energy into thrust? This is going to be critical in making this concept work.

GoogleNaut · Guru

The power conversion efficiency going from electricity to rf is about 90%, and the rf coupling efficiency I think is about 60-70% and if a Brayton Cycle gas turbine is used for electricity generation then this can be nearly 50%, so conservatively let's say 40%. So reactor power to thrust power efficiency will be about: 0.4*0.9*0.65=0.234 or about 23.4% overall. This sounds dismal, but this is fairly good. Going from a conventional (more well understood) Brayton Cycle, heterogenous solid core reactor to a homogenous MHD Vapor Core Reactor system could improve the reactor power to electrical power conversion to nearly 50%. If a combined cycle is used (combined Brayton gas turbine cycle with MHD) this could be nearly 70%, but this entails a lot of foundation work. Nobody has ever built a vapor core reactor with the powers needed to for a manned spacecraft propelled by say 100MWe of VASIMR engines. Nobody has even flown an engineering testbed of the VASIMR engine, and it doesn't look too good in the NASA budget lately either.

Still, a combined cycle VCR/MHD generator concept with 100-200 MWe of VASIMR propulsion should just about open the doors to the solar system--atleast out to about the asteroid belt. My money is on VASIMR as a good NEP concept.

-- Edited by GoogleNaut at 02:29, 2006-02-15

10kBq Jaro · Senior Member

In nuclear electric propulsion (NEP) the key factor is always power per unit mass, not efficiency.
We've got oodles of thermal power from the reactor, but trying to make the conversion process to electricity efficient inevitably leads to very massive systems (particularly the waste heat radiators), which we want to avoid, even if it means lower efficiency (think of Orion again).
By going to a relatively inefficient (22% in INSPI's design) but light-weight energy conversion, in this case using MHD in combination with the VCR, the radiator is far smaller & lighter, because it operates at over 2000K (2600K inlet, 1550K outlet).
If this were a powerplant on earth, where we don't much care about mass, the VCR-MHD would be just the topping cycle, after which you could have a gas turbine and then a steam turbine, for an overall efficiency of nearly 70% as Googlenaut says.
But its not how you build spaceship propulsion.

GoogleNaut · Guru

Agreed. And of course there is always the engineering rule: K.I.S.S. (Keep It Simple, Stupid!) which basicly means that as a systems complexity increases, so does the insipient failure rate. So keep the power conversion simple, light weight and make it as robust as possible (parallel paths for redundancy.) It's not rocket science...err, well, yes it is...

Philipum · Veteran Member

This is a reply to GoogleNaut's post three days ago.

I don't quite undestand why the specific impulse for oxygen should be 1/16 the one for hydrogen. As I undestand it, the specific impulse is the momentum per unit weight of the propellant. I agree that oxygen weights 16 times more than hydrogen; however, if one assumes that an oxygen ion would be accelerated up to the same energy than an hydrogen ion, then, because P=sqrt(2*m*E), the oxygen ion would carry four times as much momentum than the hydrogen ion. In this case, then, using oxygen would give 4/16=1/4 as much specific impulse as using hydrogen. But I don't know if the VASIMR accelerator would be able to accelerate oxygen to the same energy than hydrogen.

GoogleNaut · Guru

Philipum,

[My Bad!]

You are correct. Since K=1/2*m*v^2, then all other things being equal: (i.e., the thermalization of kinetic energy of a population of particles will tend to make the average energy per ion the same,) then if mass is increased by 16, the velocity is infact only decreased by a factor of 4. So you are correct, oxygen should achieve about 1/4 the specific impulse of hydrogen run in an equivalent engine.

Ergo, for hydrogen with specific impulse at the high end of 15,000 to 30,000 seconds, oxygen should achieve about 3000 to 6000 seconds, which is still pretty good.

I like the idea of using something which will eventually exist in quantity and will essentially be a waste product. Also, the flexibility of reaction masses available to VASIMR's should greatly increase their safety for deep space missions to the Asteroids, because conceivably anything that can be vaporized could be processed by VASIMR, it's just a matter of selecting the right magnetic field strength and the right electron/ion cyclotron frequencies to excite.

One problem though is radiant energy coupling from the plasma to the inside wall of the engine (or the gas tube, if a fuse quartz duct is used.) This could become very important if plasma temperatures and densities achieve high enough values that Bremstraulung radiation becomes the predominant radiating mechanism.

Thanks again for the correction!

Ty Moore
"GoogleNaut"

John · Senior Member

Thanks for the analysis of the VASIMR/nuclear electric efficiency. It looks like about 25% with 60% being radiated heat. When compared with the VASIMR/solar one of the issues is which weighs more: solar panels to absorb sunlight to make electricity at say the orbit of Mars or radiators to eliminate waste heat from the Brayton cycle system.

Obviously this isn't for deep space out to Jupiter and beyond but one of the most interesting areas for space travel would be the Earth-Moon system, Mars and the inner asteroids. Current solar panels are 14% efficient (the ISS panels). This site claims near term advantages could increase this to 50%.

http://www.thespacereview.com/article/214/1

As say a radiator temperture of 1000 K would release about 56 Kilowatts per sq. meter per side or 112 KW per sq meter. Solar energy at Mars would be about 10 KW per sq. meter so at 50% one would have 5 KW per Sq. meter. So unless the weight of the reactor/turbine/generator is more that 22 times the weight of the solar panels the advantage would be with the nuclear system. Any thoughs on this?

GoogleNaut · Guru

Solar Insolation Constant (the amount of energy flux striking a unit area) at Earth Orbit (1 A.U.) is about 1373 Watts/m^2 according to the CRC Handbook of Chemistry and Physics ( 14-2.) Moving out to Mars, which averages about 1.52AU from the sun, then the solar constant at Mars should be about 1/(1.52^2)*1373=594 Watts/m^2

All other things being equal, a solar panel at Mars orbit will generate only about 594/1373=43% the power that the same panel will generate at Earth Orbit.

It gets much worse the farther you move out. By the time one reaches the orbit of Jupiter at 5.2 Au from the sun, the solar constant is down to about:
1373/(5.2^2)=50.8 Watts/m^2 or about 3.7% of solar flux at Earth Orbit. Thus to achieve the mission power output that is possible at Earth Orbit, would require solar panels with almost 30 times the area (and weight) of the same panels at Earth orbit. And at Jupiter, with its powerful belts of charged particles circulating around in the magnetosphere, the panels would quickly get fried so they won't last very long. This is why nuclear power will be so important for really deep space missions, because solar power becomes very difficult and ultimately impractical much beyond the orbit of Mars.

-- Edited by GoogleNaut at 19:53, 2006-02-17

John · Senior Member

Oops! I don't know where I got number. However, even if one didn't have a copy of The Handbook of Chemistry and Physics one could have gotten closer. I seem to remember that the Sum has a suface temperture of 10000 deg F and a diamter of 864,000 miles. If convert from F to K you get (10000+460)*5/9 = 5811 deg K. The Stefan-Boltzman law states that

Power/m^2=sigma * T^4 where sigma is a constant equal to 5.6705 * 10^-8 W m^-2 K^-4.

so 5.6705*10^-8*5811^4=6.38545^10^7 watts/m^2 at the suns surface.

Using the inverse square law to scale this (432,000 miles/93,000,000 miles)^2 =

2.1577* 10^-5. So multiplying we get 1378 watts/m^2. That's pretty close for approximate numbers from memory other than the Stafan-Boltzman constant from my calculator. I must have been tired!

Anyway using your numbers 1373 * .43 = 593 watts/m^2. So dividing the 112,000 / 593 =189. So the nuclear system would need to be 189 times heavier to have equal performance. So the advantage is clearly with nuclear.

john fraz · Veteran Member

Googlenaut (feb 12)

> Anyways, it is possible to retune and engine by changing the magnetic field strength and the rf frequency so that something like Oxygen can be used. Why use oxygen? Well, about 60% of lunar regolith is oxygen by weight

You all have also touched on other fuels here. What I'd like to toss into the mix is dual-mode NTRs. In these, O2 is injected into the throat of the engine bell, so the O2 combusts with the supersonic, superheated exhaust. This eliminates the concerns of the hot oxidizing O2 on the insides of the engine.

Also there's the possibility of using CO or other things like CH4 as the propellant; isp goes down while thrust is higher than for a comparable H2 fueled engine, but it's available and not needed for other things. It's also combustible with O2, so it can also use the O2 "afterburner".

Another point of such Dual mode NTRs is the possibility of making them tri-modal: "afterburning" with LOX, and then run them as fission power plants, to power ISRU plants, on-site exploration while landed, or to power an NEW engine. This capabillity gives an NTR all the capabilities of the VASIMR. The only questions (and which I doubt we can answer here) is the size, cost and complexity increase of an NTR to encompass the power-hungry NEP rocket. It might not be worth it.

Any way, NTR or nuclear fission in general has all the capabilities of VASIMR, with none of the great technological hurdles needed for VASIMR -and none has the advantages of Orion, which is admittedly not efficient, but has its benefits over all others.

I'm trying to find some on-line source for the papers which have been published about the dual-mode "afterburning" NTRs...

GoogleNaut · Guru

Dual mode and Tri mode NTR engines are on the drawing board. I am especially keen on the Pratt and Whitney "Triton" engine which is a multi-mode engine with LOX afterburning. This gives it very good thrust to weight ratios and ought to work pretty well in the Earth-Moon and Earth-Mars regime. The engine also has the capability of generating electricity via a helium/xenon Brayton cycle gas-turbine unit which circulates gas through seperate channels within the core--while the engine runs in normal mode, electric power is generated. The Brayton unit also provides cooling to the engine during shutdown transients and provides decay heat removal later on...also the engine can be 'run' in a thermal only mode which will still generate 60-100 kwe from the Brayton unit.

Pratt and Whitney has a white paper on the Triton, but I have yet to purchase it. I'll dig for it tomorrow and see if I can find the source (I think its available at the AIAA for about $25)

John Fraz:

The O2 is actually injected just downstream of the throat in an afterburning NTR. The reason for this is that the flow must achieve supersonic speed within the nozzle before injection is done. The flow is subsonic within the chamber (the plenum between the exit of the reactor core and the throat;) the flow is transonic in the throat; and the flow is supersonic downstream of the throat. Injecting O2 within the throat will allow pressure instabilities to propogate back into the core--the engine will choke itself, and will 'chug' until the whole thing blows apart. Injecting O2 downstream within the supersonic flow allows supersonic combustion to occur, downstream of the nozzle throat. As long as the flow is supersonic where O2 afterburning occurs, and as long as the rate-of-change of O2 addition (kg/s^2) is reasonably small, then no sonic disturbances can propogate back to the throat (and hence the reactor.)

If O2 is added too quickly, then the danger exists that a shockwave will form within the nozzle bell which can cause the flow to stagnate (almost stop) within the nozzle. If the flow stagnates, then a reflected pressure wave can back propogate up to the nozzle throat and beyond--immediate and total destruction of the engine will result!

I'm not sure what specific impulse is possible with O2 afterburning, but I'd imagine that something better than 500 seconds isn't unreasonable. Whereas, with pure hydrogen, the engine may achieve 900-1000 seconds of Isp. And the thrust will generally be pretty high: 75,000 - 100,000 Newtons with hydrogen, and perhaps something like 125,000 to 175,000 Newtons with afterburning.

VASIMR is an electric rocket, with specific impulse in the high range of 15,000 to 30,000 seconds; 1000-3000 seconds in the low range; and thrusts anywhere from 10-100 Newtons in the High Range; 1000-2000 N in the Low Range. VASIMR is an excellent choice for deep space missions requiring fast interplanetary transits, or really high delta-v's. The performance gives more flexibility with mission abort scenarios--which can improve mission safety significantly. NTR with afterburning is excellent for reasonably fast interplanetary transfer; landing and ascent engines for a ISRU (Insitu Resource Utilization) capable lander; cogeneration of electric power and heat for surface operations and habitat life support.

And, again Jaro, you are right about the possibility of using insitu resources for propellant manufacture on site. If Mars has as much water as many think, then an afterburning NTR lander using liquid methane and oxygen is a definate possibility ought to work pretty well.

I'm not sure how much 'coking' could be a problem for NTR. If methane is used within an NTR thermal disassociation could deposite significant amounts of carbon within cooling channels of the motor. Depending on how fast this accumulates, there could be definate motor life issue here. Perhaps with a basic mission scenario of down; a couple of surface hops, then a long burn either back to mother ship in orbit, or a long jump directly back to Earth--coking within the motor might not be as important because the lifecycle is so short anyway...

John · Senior Member

Not too long ago it was announced that the Ad Astra Rocket Co would collaborate with NASA on the development of the VASIMR engine. I found this website that might be of some interest for the company.

http://www.adastrarocket.com/AdAstraRocket.html

The also have a notional four month VASIMR powered Mars mission based on a 12 mega watt system. With a 30 day spiral out of LEO and an 85 day transist to Mars.

"Spiral through another day..."

http://www.adastrarocket.com/missions.html

This seems like very good news that development is underway. If we could just get some of these block-headed politicians...er I mean national leaders to back this thing...