H3

[docabn]

Could anyone explain how tritium could be used as a propellent more efficiently than the LH2 / LO2 used now? Or have I misinterpreted the science?

Thanks

 

[willy88]

Could anyone explain how tritium could be used as a propellent more efficiently than the LH2 / LO2 used now? Or have I misinterpreted the science?

Thanks

Wouldn't it be used for fusion reactors?

 

[docabn]

The question is more about how the reaction what ever kind it might be might work. are we talking about a VASIMR engine except using fusion reaction instead of plasma?

 

[jedidia]

As far as I know, HE3 would be a good FUEL (read: the substance providing the energy) for a fusion drive, while for PROPELLANT (read: the stuff being thrown out of the nozzle) normal H2 would be used.

 

[RisingFury]

As far as I know, HE3 would be a good FUEL (read: the substance providing the energy) for a fusion drive, while for PROPELLANT (read: the stuff being thrown out of the nozzle) normal H2 would be used.

Seems like a big waste of energy to me.


It would be better if you just ejected the products of the reaction - they're already at high speed

 

[jedidia]

Well, of course you would eject them too... but it wouldn't be enough mass, or would it? the concepts I read about all seemed to use hydrogen as a propellant, heating it up in the core-chamber and then channeling it through the nozzle... or not? as is obvious from my previous thread, I do not have a very deep understanding of fusion drives, so if you could help me up on it, I would greatly appreciate it. Or suggest a comprehensive read that can be followed without understanding the math...

 

[RisingFury]

What difference does it make? You only have a certain amount of power provided.

 

[Mindblast]

If you only use the reaction products as propellant mass you will have a cooling problem and also ridiculously high ISP with ultra low thrust. Depending on the mission it will be a good idea to add drive mass to raise thrust and lower ISP.

There are a lot of interesting papers about fusion drives by Robert Bussard at http://www.askmar.com/Fusion_files/ where he calculates different engine approaches for different missions from Lunar transfer to going all the way out to the OORT cloud.

 

[RisingFury]

The only reason why one should introduce more mass is to cool the reactor. That energy could go towards propulsion, but I see no reason why you should introduce more propellant just to increase thrust at the cost of Isp. You might as well build a bigger reactor to increase your thrust at no expense of Isp.

 

[jedidia]

oh my, that's a whole library... thanks. I'll be over there, reading...

 

[Mindblast]

The only reason why one should introduce more mass is to cool the reactor. That energy could go towards propulsion, but I see no reason why you should introduce more propellant just to increase thrust at the cost of Isp. You might as well build a bigger reactor to increase your thrust at no expense of Isp.

Building a bigger reactor will make your ship much heavier though so if you can increase the effective acceleration of your ship better by adding drive mass than by adding reactor size then this is probably the better way to do it. It is all a matter of the right balance. On the extreme ends you have on the one side a very small/light engine with very low ISP but high Thrust and lots of fuel/drive mass to carry along (i.e. ordinary LH2+LOX chemical thrusters) on the other side you have a huge engine with very high ISP and very low thrust but only very little drive mass to take along. So you need to find the right balance between the two extremes to match the mission requirements (deltaV and traveltime). For example you wouldn't want to use an Ion drive for a manned lunar mission and just adding more ion drives wouldn't be as effective as using chemical engines.

It all becomes appearent when you consider that the drive has a certain amount of power available that is used to increase the kinetic energy of the exhaust. Now kinetic energy is

Ekin = 0.5 * mass * v ^ 2

while the thrust is

th = mass/second * v

so while the thrust grows linear with mass and velocity of the exhaust, the energy needed grows linear with mass and squared with velocity. So if you halve the velocity you can throw in 4 times the mass for the same kinetic energy but you get 2 times the thrust.

 

[Linguofreak]

The only reason why one should introduce more mass is to cool the reactor. That energy could go towards propulsion, but I see no reason why you should introduce more propellant just to increase thrust at the cost of Isp. You might as well build a bigger reactor to increase your thrust at no expense of Isp.

You still have a limit on thrust/weight ratio when only using exhaust for propellant. If you need a higher acceleration than the maximum you can get with that setup, you'll need to add propellant. It will give you better accelleration both by reducing ISP and increasing thrust, and by cooling the thing better, which allows you to add power, which reduces the amount of ISP you need to sacrifice to get a given amount of thrust.

Also, if you're using a Deuterium-Tritium fusion cycle, you'll need to dump alot of extra propellant in to capture the neutrons from the reaction (which carry most of the energy from the reaction and can't be magnetically confined on their own, since they have no charge). If you don't, they'll just go flying off into space (wasting energy, since they'll be going out in every direction) or hitting parts of your spacecraft, which will have all kinds of negative effects, such as embrittling the structural materials, requiring extra radiation shielding for the crew, etc. If you dump a good bit of propellant in, the neutrons will hit atoms in the propellant stream, transferring their energy, and also being carried off with the propellant stream as it is guided by the magnetic nozzle.

Of course, for any fusion engine design where you're dumping propellant in, you'll need to figure out how to do so without quenching the reaction.

 

[RisingFury]

I am well aware of the added mass the bigger reactor would bring and even if the the mass penalty for the reactor would be twice as great as the original reactor and the thrust would be doubled, the reactor does not comprise the entire mass of the ship. The power to mass ratio of the ship in general would increase, resulting in a higher acceleration.

Think of it in terms of Dawn, a probe heading to Cares and Vesta. It has 3 ion engines - the same design as used on Deep Space 1 - and has a total of 425 kg of Xenon propellant. The total mass of the spacecraft at launch was around 1250 kg. Now... only one of the ion thrusters can be fired at the same time, because of insufficient power... but despite that fact, there are 3 on board. That means that even though 1 extra thruster could have been used as a backup, 3 were put on board regardless of the mass penalty. It's safe to assume that the future vessels will not consist of an engine and fuel and that most of the weight will be provided by the payload.

Dawn is able to achieve around 10 km/s Delta-V with just over 400 kg of propellant. That's about as much as a rocket needs to get into LEO, with tons and tons of propellant.


@ Mindblast:

For one, I can imagine that a vessel with a fusion drive will not be used on an Earth-Moon trip, but rather a trip around the solar system, or as part of a Buzzard ramjet.

Therefore it is safe to assume that a ridiculously high Isp is desirable, as the engine will undoubtedly operate in the range of weeks, months or more.


Which brings me to your equations...

While you certainly have the ability to read equations off Wikipedia, you don't seem to have any understanding of them.

These things we know:
The mass of the ship: m
The mass flow rate through the reactor: M
The power produced by the reactor: P (I'm going to assume for now that power is *not* a function of mass flow and that 100% of the power gets used for propulsion).
I will be using v as exhaust velocity.

Here is the acceleration produced by the engine:

F = M * v
m * a = M * v
a = M * v / m

Here is the exhaust velocity, from the power provided by the engine:

A = 0.5 * m * v^2
P = 0.5 * M * v^2
v^2 = 2 * P / M

Now, if I take the equation for the acceleration and square it
a^2 = M^2 * v^2 / m^2

And if I plug the second equation into the first one:

a^2 = M^2 * 2 * P / (M * m^2)

And if I simplify:

a^2 = 2 * M * P / m^2

And square root:

a = sqrt(2 * M * P) / m

Now... as you can see from this equation, doubling the mass flow rate will NOT increase your acceleration by two times. It will increase it by square root of 2 =~ 1.41. You'll need to increase it by 4 times to double your acceleration, which quite clearly goes against your assumptions.

And if we work out the power provided by a mass flow of fuel:
I will be using q as energy provided by fusing a certain amount of mass (q = W / m)

Keep in mind however, that this is the mass flow rate after the fusion - after we've extracted energy. There will be a slight mass deficit due to fusion, but I'm gonna neglect that for now, so I don't need to introduce another variable. The error is not great and the equation that follows is a very good approximation.

Therefore:

P = q * M

And if we plug that into the equation for acceleration:

a = Sqrt(2 * M * M * q) / m
a = Sqrt(2 * q) * M / m

And as you can see, acceleration increases linearly with the mass flow through the reactor, however, this equation will only apply if the mass flow is actually being burned. If you introduce mass after burning has been finished, the first equation will apply.

 

[Linguofreak]

I am well aware of the added mass the bigger reactor would bring and even if the the mass penalty for the reactor would be twice as great as the original reactor and the thrust would be doubled, the reactor does not comprise the entire mass of the ship. The power to mass ratio of the ship in general would increase, resulting in a higher acceleration.

But it will never exceed the thrust to weight ratio of the engine itself. Which, in this case, will be very low.

Think of it in terms of Dawn, a probe heading to Cares and Vesta. It has 3 ion engines - the same design as used on Deep Space 1 - and has a total of 425 kg of Xenon propellant. The total mass of the spacecraft at launch was around 1250 kg. Now... only one of the ion thrusters can be fired at the same time, because of insufficient power... but despite that fact, there are 3 on board. That means that even though 1 extra thruster could have been used as a backup, 3 were put on board regardless of the mass penalty. It's safe to assume that the future vessels will not consist of an engine and fuel and that most of the weight will be provided by the payload.

Dawn is able to achieve around 10 km/s Delta-V with just over 400 kg of propellant. That's about as much as a rocket needs to get into LEO, with tons and tons of propellant.

But it is provided at very, very low thrust. The thrust to weight ratio of the whole engine is not that great, let alone a spacecraft for which a reasonably small fraction of the mass is engine mass.

@ Mindblast:

For one, I can imagine that a vessel with a fusion drive will not be used on an Earth-Moon trip, but rather a trip around the solar system, or as part of a Buzzard ramjet.

Vultures don't need fusion drives. (Bussard).

 

[RisingFury]

If the mass of the ship is m and the mass of one reactor is m/10, the thrust provided by the reactor is F, then the original thrust to mass ratio is F/m, but the new one is 2 * F / (11/10) * m = 20/11 * F/m, therefore, the thrust to mass ratio of the ship overall has almost doubled.

 

[Linguofreak]

If the mass of the ship is m and the mass of one reactor is m/10, the thrust provided by the reactor is F, then the original thrust to mass ratio is F/m, but the new one is 2 * F / (11/10) * m = 20/11 * F/m, therefore, the thrust to mass ratio of the ship overall has almost doubled.

Yes, but if your acceleration is only 1 km/s per week, doubling it to 2 km/s per week may not be very helpful.

 

[RisingFury]

The ion engine powering Deep Space 1 took 4 days to increase velocity by some 30 m/s and yet we're pretty keen on developing such technology.


On a trip around the solar system, 1 km/s per week would be quite favorable.

 

[Mindblast]

If the mass of the ship is m and the mass of one reactor is m/10, the thrust provided by the reactor is F, then the original thrust to mass ratio is F/m, but the new one is 2 * F / (11/10) * m = 20/11 * F/m, therefore, the thrust to mass ratio of the ship overall has almost doubled.

Allright then. Lets take that example and go from there..
Assume our ship has a total mass of 1000 tons so the Reactor would be 100 tons. Also assume we can use a D-He3 Reaction because it creates a low amount of neutrons and we want charged particles to be able to direct them out the rear somehow. Then assume the reactor has a maximum power of 10GW.

The D-He3 Reaction is:
D + He3 -> He4 + p + 18.3 MeV
18.3 MeV =~ 2.932e-12 J
For 1 mol D and 1 mol He3 the reaction would yield ~1.76568 TJ (!).
So to burn at the maximum power rating of our reactor of 10GW we could burn 0.005663535 mol D and He3 per second. That would be 0.01708g of He3 and 0.0114g of D per second. The reaction products would weigh 0.02267g (He4) + 0.0057g (p) = 0.0284g.

So we use 10GW power to accelerate 0.0284g of reaction mass per second. Assuming we can magically convert all the energy to kinetic energy that gives:

Ekin = 0.5 m * v^2

v = sqrt(Ekin * 2 / m)
= sqrt(10GJ * 2 / 0.0000284kg)
= 26537244 m/s

ISP = 2705121s

That gives a thrust of
F = 0.0000284 kg/s * 26537244 m/s
= 753.7 N

So that would give our ship an acceleration of
a = F / mship = 753.7 / 1000000kg = 0.0007537 m/s^2 not too much to work with .
Even if we add 10 more engines we would then have F = 753.7 * 11 = 8290.7 N and mship = 1000000kg + 10 * 100000kg = 2000000 kg so the acceleration would then be 0.004145 m/s^2.



If we add 1000tons of additional drive mass to the ship instead and add 100 kg of drive mass per second to the engine we get
v = sqrt(10GJ * 2 / 100.0000284kg)
= 14142 m/s

F = 14142 m/s * 100.0000284kg/s
= 1414200 N

Then we get an acceleration of a = 1414200 / 2000000 = 0.7 m/s^2 sustainable for 10000 seconds or 2.78 hours. It will give us a total deltaV of
dv = v * ln((shipmass + drivemass) / shipmass) = 141419 * ln(2) = 9802 m/s which would be nice for a quick trip to the moon in less than 24 hours.

Instead with the 11 engines arrangement above we would already need 5-6 days to get to the 2000 m/s or so minimum deltav needed for a Hohmann transfer orbit to the moon. (2000 m/s / 0.004145 m / s^2 = 482509s)
Probably followed by a couple of weeks for orbit circularisation in lunar orbit..

So you can see it is not always the best thing to just add more engines.

---------- Post added at 09:24 AM ---------- Previous post was at 09:17 AM ----------

The ion engine powering Deep Space 1 took 4 days to increase velocity by some 30 m/s and yet we're pretty keen on developing such technology.


On a trip around the solar system, 1 km/s per week would be quite favorable.

It all depends on the mission.. for a deep space probe it is absolutely the right thing to go high ISP low thrust but for manned missions to the Moon or Mars or maybe even Titan you want fast transfer times, so high thrust low isp.

---------- Post added at 10:39 AM ---------- Previous post was at 09:24 AM ----------

Oh it seems i have overlooked your post RisingFury.. sorry for that..

@ Mindblast:

For one, I can imagine that a vessel with a fusion drive will not be used on an Earth-Moon trip, but rather a trip around the solar system, or as part of a Buzzard ramjet.

Therefore it is safe to assume that a ridiculously high Isp is desirable, as the engine will undoubtedly operate in the range of weeks, months or more.

Even if you run it for months it will be desirable to add some drive mass to reach a point where you spend all your propellant mass while constantly accelerating/decelerating throughout the trip and still optimize traveltime with payloadfraction.

And apart from that, why not use it for moon trips too ? As i calculated in my last post you can get a nice payload fraction of about 45% for a fast moon transfer. Chemical thrusters can't do that.

Which brings me to your equations...

While you certainly have the ability to read equations off Wikipedia, you don't seem to have any understanding of them.

No need to get offensive..

Now... as you can see from this equation, doubling the mass flow rate will NOT increase your acceleration by two times. It will increase it by square root of 2 =~ 1.41. You'll need to increase it by 4 times to double your acceleration, which quite clearly goes against your assumptions.

No in fact it exactly corresponds to my assumptions:

So if you halve the velocity you can throw in 4 times the mass for the same kinetic energy but you get 2 times the thrust.

a = Sqrt(2 * M * M * q) / m
a = Sqrt(2 * q) * M / m

And as you can see, acceleration increases linearly with the mass flow through the reactor, however, this equation will only apply if the mass flow is actually being burned. If you introduce mass after burning has been finished, the first equation will apply.

Thats all fine if you can build light weight fusion reactors with powerlevels in the TeraWatt range to get enough thrust while only using the fusion products as drive mass. I guess this will not be within our reach for quite a while though.

 

[RisingFury]

Who do you think would spend 10 billion dollars building a reactor in space to go to the Moon? Do we build ion engines to send probes to the Moon? No, we build them to send them off on years long voyages.


Also, I don't think anyone would spend that much money to get 10 km/s Delta-V and burn for 3 days. Besides, even with this "high thrust", to get to the Moon, you'd still have to slowly spiral out into a higher orbit... kinda like how they get unmanned probes to the Moon. So it's clear that for ejects, you'd still need propulsion that's in the range of 10 times more powerful.


So... assuming 1000 tons of propellant on a ship that has such a low acceleration would mean a delta-V of roughly 18 400 km/s. That's a horribly high number. I'm pretty sure getting to Saturn with this high Isp would be quicker then by getting there at "high thrust". Even optimizing the thrust to increase velocity all the way there and then turn around and start thrusting to kill that velocity... this kind of travel, which is the fastest, would still need an insanely high Isp - far higher then what you proposed.

 

[Mindblast]

Who do you think would spend 10 billion dollars building a reactor in space to go to the Moon? Do we build ion engines to send probes to the Moon? No, we build them to send them off on years long voyages.

A probe mission is usually a one way trip so using an expensive drive on a probe would only be worth it for longer missions. An earth/moon transporter is not a use-once-throw-away ship so if there is enough demand for earth/moon transfers then yes for 10 billion dollars the thing would be built.
If you just look at the estimated constellation program costs an earth/moon transfer ship that gets 900tons of payload (or maybe just 500-700 accounting for ship mass) to the moon within 24 hours with just 1000tons of drive mass (H2 or even Water would do) for just 10 billion dollars would be something they would be VERY happy to build.

Also, I don't think anyone would spend that much money to get 10 km/s Delta-V and burn for 3 days. Besides, even with this "high thrust", to get to the Moon, you'd still have to slowly spiral out into a higher orbit... kinda like how they get unmanned probes to the Moon. So it's clear that for ejects, you'd still need propulsion that's in the range of 10 times more powerful.

The burntime is not 3 days but 2.78 hours for the proposal in my post above. You don't need to spiral out with that acceleration..

But oh well if thats not working take 400kg/s massflow. Then you have only ~5000m/s total deltav but 1.4 m/s^2 acceleration so that would be ~41 minutes total burntime, thats roughly a half orbit in LEO so definately no spiraling there.

So... assuming 1000 tons of propellant on a ship that has such a low acceleration would mean a delta-V of roughly 18 400 km/s. That's a horribly high number. I'm pretty sure getting to Saturn with this high Isp would be quicker then by getting there at "high thrust". Even optimizing the thrust to increase velocity all the way there and then turn around and start thrusting to kill that velocity... this kind of travel, which is the fastest, would still need an insanely high Isp - far higher then what you proposed.

Ok then lets take the Saturn example. Saturn is how far, 8-9 AU ? For simplicity lets take 10AU and calculate how long it would take us to go half the way with constant acceleration for the 3 different setups proposed:

5 AU are roughly 750e9 m

Given the acceleration we can calculate the time it takes to fly the leg as

750e9 m = 0.5 a * t ^ 2

t = sqrt(750e9 * 2 / a)

For simplicity i will not use the rocket equation and take the ships mass as constant which will shift the results a bit in favor of the high isp low thrust setups because they don't use much fuel on the way compared to the high thrust setup.

So lets see.. the first setup was the 1000tons ship with a 100tons reactor.
a = 0.0007537 m / s^2

t = sqrt(750e9 * 2 / 0.0007537) = 44611453s =~ 516days
fuel massflow for this design is 0.00002848kg/s so we use
0.00002848kg/s * 44611453s = 1270.5 kg of fuel on the way out to 5AU


The next setup was 2000tons ship with 1100tons for 11 reactors:
a = 0.004145 m/s^2

t = sqrt(750e9 * 2 / 0.0007537) = 19023191s =~ 220days
fuel massflow here is 0.00031328kg/s so we use
0.00031328kg/s * 19023191s = 5959.6kg to go the 5AU


Finally we have the setup with 2000tons with 100ton reactor and 1000ton drive mass:
i figured out an optimal mass flow of 51.7g/s for this mission so lets see..

exhaust velocity
v = sqrt(10GW * 2 / 0.0517kg/s) = 621970 m/s

thrust
F = v * 0.0517kg/s = 32156N

acceleration
a = 32156N / 2000000kg = 0.0161 m/s^2

t = sqrt(750e9m * 2 / 0.0161 m/s^2) = 9652342s =~ 112days
at the fuel massflow of 0.0517kg/s we use
0.0517kg/s * 9652342s = 499026kg of fuel to go the 5AU so half of the fuel on board.. as already noted if we had used the rocket equation here things would even look better for this setup as the acceleration would double towards the end of the mission.