Tri-Alpha article published 2 may 2011

Skipjack · Post by **Skipjack** » Tue Aug 23, 2011 1:57 am

I wasn't addressing that. I was explaining why your suggestion of a NERVA-type thermal engine was in a completely different class from the VASIMR or direct-exhaust drives, and thus not appropriate for the same mission types.

But it is not, NERVA type (fission) engines have been considered for Mars missions just like VASIMIR has been. The ISP is still more than twice that of chemical engines and you have huge structural and weight savings over VASIMIR on top of that. This is why they are still in the competition. Von Brauns Mars shuttle transport was also based on a NERVA engine.

I think that a direct exhaust fusion engine could have quite a bit more thrust than VASIMIR.
But IMHO, a high thrust to weight combined with a high Isp is what you really want, no matter what the application is and there a direct exhaust should really shine compared to a VASIMIR system.

93143 · Post by **93143** » Tue Aug 23, 2011 7:13 am

I said mission types, not targets. VASIMR with a sufficiently compact and powerful reactor could do Mars in 39 days with a continuous-thrust trajectory. NERVA could never support anything resembling that mission profile because its specific impulse is simply too low.

Also, Mars is fairly close. If you want to go to Saturn in a reasonable amount of time, you'll want Isp in the high five figures at least, or nearly two orders of magnitude greater than a solid-core NTR.

...

A burning plasma can be assumed to have a temperature in the range of at least 10 keV. Assuming 100% nozzle efficiency (and no recombination), that's a specific impulse of about 140,000/sqrt(M), where M is the mean particle mass of the plasma in daltons. Assuming a plasma containing deuterium, tritium, and helium, but somewhat diluted by protium so as to have M=1, the ideal specific impulse comes out as 140,000 seconds.

That's 0.0000014 N/W. How much power is the reactor producing, again?

...

You can improve the thrust, of course, by adding hydrogen to the plasma after it is ejected from the reactor. But the specific impulse comes down accordingly, and by the time you've gotten the thrust up to the level of an NTR, the specific impulse is terrible. (This mode, if workable, does seem to eliminate the necessity of using CSR, and it makes the idea of using a solid heat exchanger look pretty silly. So yes, your first idea was fine, but I never said it wasn't - I was only criticizing your second idea.)

There's another problem - NERVA-class reactors ran into the multiple gigawatts. A fusion reactor on a spacecraft would most likely be much less powerful and/or far heavier. So by the time you've achieved the same T/W as a NERVA-class engine, to say nothing of a Dumbo, your specific impulse is probably down below the chemical range.

If you want high Isp and high thrust at the same time, go for Orion or NSWR or something like that. MCF won't do it.

Skipjack · Post by **Skipjack** » Tue Aug 23, 2011 1:56 pm

NERVA could never support anything resembling that mission profile because its specific impulse is simply too low.

Isp is not everything!
VASIMIR has the problem that the energy supply is so heavy that it makes up for the extra weight of the fuel for a NERVA type engine. The mass of the fuel also gets smaller as the mission progresses, the mass of the powersupply does not.

93143 · Post by **93143** » Wed Aug 24, 2011 7:16 am

According to my calculations, doing a 39-day Mars transfer, with both planets at their mean distances from the Sun, takes a minimum of about 45.8 km/s impulsive, solar-orbit-to-solar-orbit, or 34.9 km/s if you can take full advantage of the Oberth effect. At 1000 s, that's a mass ratio of 35.

One way.

Good luck with that...

Skipjack · Post by **Skipjack** » Wed Aug 24, 2011 8:50 am

According to my calculations, doing a 39-day Mars transfer, with both planets at their mean distances from the Sun, takes a minimum of about 45.5 km/s impulsive, solar-orbit-to-solar-orbit, or 34.6 km/s if you can take full advantage of the Oberth effect. At 1000 s, that's a mass ratio of 34.

Dont forget that the fuel will be consumed over the course of the mission. You are also totally neglecting thrust in your calculation (or the lack of it for VASIMIR). The acceleration using VASIMIR will be very weak and so it will take a long time to get to the 34 km/s.
Also, the concept for the Mars transport using NERVA type engines is not mine, nor is it new. It was around since at least von Braun, but probably earlier than that.
Anyway, lets see what a mass ratio VASIMIR will be able to do. The energy supply problem still has not been solved and from all we know for such a short duration mission, it would have to be quite massive and huge too. From my POV VASIMIR is a bit overhyped and oversold.
I think that Sloughs ELF thruster and other concepts look a lot better.

93143 · Post by **93143** » Wed Aug 24, 2011 8:25 pm

My apologies; I forgot to decrease the integration time step in the trajectory optimizer I wrote for this occasion after I finished testing it. The numbers in my earlier post have been replaced with converged values. The conclusion remains the same.

Skipjack wrote:
At 1000 s, that's a mass ratio of 34.
Dont forget that the fuel will be consumed over the course of the mission.

What part of "mass ratio" did you not understand?

You are also totally neglecting thrust in your calculation (or the lack of it for VASIMIR). The acceleration using VASIMIR will be very weak and so it will take a long time to get to the 34 km/s.

Did you seriously think I was high-school enough to just take the distance between orbits and divide by the desired trip time? The Oberth-effect number is total delta-V for starting from L2, swinging by Earth, and propulsively arriving in LMO. Solar gravity is of course taken into account.

And yes, these numbers are for impulsive burns and thus completely inapplicable to high-Isp, low-thrust drives like VASIMR or ELF, which would most likely be used for continuous-thrust trajectories. They're optimistic even for NTR, since the T/W is probably too low to take full advantage of the Oberth effect during such massive burns...

[You might be able to shed several km/s with aerocapture, but not anywhere near the whole 20+ km/s arrival delta-V, and even letting the MTV hit the Martian atmosphere at 12 km/s still requires it to have a mass ratio of almost 15, which is extremely dubious even for a hydrolox MTV, even discounting the weight of the necessary heat shield. With nuclear engines and straight hydrogen tankage, I'd call it impossible, and Kirk Sorensen's stage design spreadsheet heartily concurs... you could stage, of course, but that makes the required initial mass even larger, and since you're throwing away most of your MTV, it no longer qualifies as the same class of mission... also keep in mind that this is a one-way trip; to get back without a very large LH2 depot in LMO, you need to square the mass ratio... more than square, actually; Earth aerocapture probably can't handle as much delta-V, because the orbital speed is higher.]

Also, the concept for the Mars transport using NERVA type engines is not mine, nor is it new. It was around since at least von Braun, but probably earlier than that.

Who are you arguing with? I never said NERVA-type engines couldn't be used for Mars. I said they couldn't be used for ultra-fast Mars transit on the same timetable as a high-Isp thruster like VASIMR or a fusion direct-ejection drive, because the Isp is too low and the required mass ratio gets ridiculous (also remember that for a return trip without refueling, the delta-V budget doubles, so the mass ratio gets squared).

For comparison, 90 km/s at an Isp of 30,000 s is a mass ratio of 1.36. 180 km/s is a mass ratio of 1.84. 35 km/s (which won't do 39 days on a continuous-thrust trajectory, but hey) is a mass ratio of just 1.13.

Anyway, lets see what a mass ratio VASIMIR will be able to do. The energy supply problem still has not been solved and from all we know for such a short duration mission, it would have to be quite massive and huge too. From my POV VASIMIR is a bit overhyped and oversold.
I think that Sloughs ELF thruster and other concepts look a lot better.

You don't listen.

93143 wrote:I was only criticizing your second idea.

93143 wrote:VASIMR with a sufficiently compact and powerful reactor

I'm not arguing for VASIMR specifically or against direct-ejection. Neither am I claiming that currently-available power sources are adequate for the described mission. Please don't try to commandeer the argument.

[Digression: If your power source is thermal, VASIMR definitely has issues (though perhaps not insuperable ones; note Figure 5 in particular). With direct-conversion Polywell it (perhaps along with other electric thrusters in its class) seems to fill a niche that DFP can't efficiently cover.]

Anyway, you were talking about NERVA-type NTRs, and seemed to be claiming that they could do the same class of fast transit as the higher-Isp drives, which is very much not true. A modern "fast" Mars transit using NTR is six months.

...

I also mentioned the fact that going further than Mars is even more demanding, and that NTR is even more incapable of ultra-fast transit for targets such as Saturn.

No, NERVA is not remotely in the same class as VASIMR or ELF or any such high-Isp nuclear/electric drive. Proposing NERVA-type solid-wall propellant heating as a drop-in substitute for reactor plasma ejection, as you did, throws away the Isp advantage of a fusion drive, which is the only thing that can justify the immense mass of a fusion reactor.

Skipjack · Post by **Skipjack** » Thu Aug 25, 2011 1:14 am

Did you seriously think I was high-school enough to just take the distance between orbits and divide by the desired trip time? The Oberth-effect number is total delta-V for starting from L2, swinging by Earth, and propulsively arriving in LMO.

I apologize

For comparison, 90 km/s at an Isp of 30,000 s is a mass ratio of 1.36. 180 km/s is a mass ratio of 1.84. 35 km/s (which won't do 39 days on a continuous-thrust trajectory, but hey) is a mass ratio of just 1.13.

Ok, are we still talking about VASIMIR here? Last time I checked it had an Isp of about 3000, not 30,000. Maybe this is where my (or your) confusion comes from?

93143 · Post by **93143** » Thu Aug 25, 2011 1:19 am

IIRC, VASIMR was supposed to operate optimally at 5000 s with argon, or 30,000 s with hydrogen. You can go lower, of course, if you need more thrust; hence the name...

I agree that if it maxed out at 3000 s, a 39-day Mars transit would be a lot more dubious regardless of the power source...

Skipjack · Post by **Skipjack** » Thu Aug 25, 2011 1:44 am

http://en.wikipedia.org/wiki/Variable_S ... sma_Rocket

Ok, I am completely confused now. Guess it is too late in the day for me.
I read up on it again on Wikipedia, since I was not sure about my numbers anymore:
http://en.wikipedia.org/wiki/Variable_S ... sma_Rocket

There it says that the optimum achievable Isp for VASIMIR would be 5,000 secs. That is more than my 3,000 but still no 30,000...
However and this is where my theory seems to come in again, the thrust to weight ratio at 5.000 Isp is so low that you would need 23 days just to get from Earth to a Lunar orbit (14 days at 3,000 Isp and more thrust).
So, I really dont know anymore what to think.
You may be right, I may be wrong, we may both be wrong...

93143 · Post by **93143** » Thu Aug 25, 2011 2:47 am

5000 s is with argon. The higher number is for hydrogen, and seems to correspond roughly to what you'd get by scaling with the square root of the atomic weight, which is approximately correct if temperature is the limiting factor.

The Earth-Moon transit is for a relatively near-term system; the ultra-fast Mars transit requires a more advanced power source. They had a more conservative design that took something like 80 days to get to Mars, I think.

It is also true that spiraling out from LEO takes a significant chunk of the time in a low-thrust high-speed Mars transit, which is why I like L-point staging.

Skipjack · Post by **Skipjack** » Thu Aug 25, 2011 6:50 am

5000 s is with argon. The higher number is for hydrogen, and seems to correspond roughly to what you'd get by scaling with the square root of the atomic weight, which is approximately correct if temperature is the limiting factor.

With hydrogen you would have even less thrust and would need even more spiraling to get out of earths gravity well.

The Earth-Moon transit is for a relatively near-term system; the ultra-fast Mars transit requires a more advanced power source. They had a more conservative design that took something like 80 days to get to Mars, I think.

Hmm, I would assume that more input power would only result in a higher Isp, which still leaves the issue of having to spiral out of LEO, which would take the bulk of the time, I suppose.

It is also true that spiraling out from LEO takes a significant chunk of the time in a low-thrust high-speed Mars transit, which is why I like L-point staging

Ahh, ok. That is of course a quite different story. You would still have to get the whole thing to the Lagrange point first of course. But then you may be able to use a NERVA type engine for doing that and then stage

There, NERVA engine used in your mission type as well.
Everybody is happy

93143 · Post by **93143** » Thu Aug 25, 2011 7:28 am

Skipjack wrote:With hydrogen you would have even less thrust and would need even more spiraling to get out of earths gravity well.

I suppose so. But the advantage of VAriable Specific Impulse is that you can trade it for more thrust...

I would assume that more input power would only result in a higher Isp

No. VASIMR, as I understand it, is designed as a constant-power system. The power ratio between the ionization and heating stages is what determines the balance between thrust and Isp, and it can be varied in-flight.

Adding more power means using a bigger VASIMR. More thrust at the same Isp. And it's linear; a VASIMR running on a 200 MWe gas-core nuclear plant will generate exactly (neglecting higher-order scaling effects in the engine) 20 times as much thrust as one running on a 10 MWe solar array at the same Isp, using the same propellant. And the available Isp range should be pretty much the same for both.

(As a thought experiment, you can notionally replace the 200 MWe VASIMR engine (or three 67 MWe engines) with 20 10 MWe engines. Same result, except that larger engines are likely to weigh less than an equivalent cluster of small ones.)

You would still have to get the whole thing to the Lagrange point first of course.

Yes, but ideally you only have to do it once per ship. Then, after the return to L2, you can refurbish, resupply, refuel (the moon may have enough water that it isn't a catastrophic waste to use it for this), send a new crew up in a couple of Orions via good old hydrolox (or send them faster with a nuclear thermal lunar shuttle), and go to Mars again.

Mars windows don't happen every month, so there's plenty of time to resupply (I actually favour electric tugs for this too, at least for stuff that has to come from Earth), but there's no point having the crew wait around in space for weeks for their ship to claw its way up out of Earth's gravity well, not to mention that returning to LEO would be similarly time-consuming...

...well, there might be a point. I'm sure I haven't thought this through as carefully as most of the people whose job it is to think things like this through, but L-point staging does seem to be popular, so...

Skipjack · Post by **Skipjack** » Thu Aug 25, 2011 7:35 pm

a new crew up in a couple of Orions via good old hydrolox

Called the MPCV now and I would rather send Dragons

93143 · Post by **93143** » Thu Aug 25, 2011 8:26 pm

The Authorization Act calls it MPCV. Lockheed Martin, as far as I can tell, still calls it Orion. A lot of other people do too, since it's faster.

It is a little annoying that you have to be so specific nowadays when talking about that ridiculous outlier on the T/W vs. Isp curve that we could have built back in the '60s and didn't...

But yes, Dragon is an option (though a bit cramped for the same number of people). You could probably launch it on a Falcon Heavy...

...I just now noticed that the claimed TLI performance of the Falcon Heavy matches the claimed LEO performance of the projected upgraded/stretched Falcon 9...

Munchausen · Post by **Munchausen** » Thu Feb 02, 2012 11:34 am

Tr-Alpha hiring

http://jobview.monster.com/Scientific-S ... 12087.aspx

Scientific Software Engineer
About the Job

Summary:

Applications are invited for a full-time position in the Data Acquisition and Processing (DAP) group at Tri Alpha Energy (TAE). The group primarily aids with the data acquisition, analysis, processing, and/or visualization needs in support of both the experiment and simulation groups at TAE. Utilizing your broad computational and scientific background, you will understand the needs of the physicist, translate them into requirements, develop prototype applications, iterate with physicist and then deliver quality scientific applications. You will therefore be involved in the entire software life-cycle: planning, development, deployment, and maintenance.

Requirements:

Advanced degree (MS or PhD) in computer science, plasma physics, or related field. Minimum one year experience with at least one compiled and one scripting language with a strong preference towards C++ and python, respectively. Experienced with Linux.