Monday 28 March 2016

Stealth in Space is Possible III

In this part, we discuss active sensors and defenses against them. We'll then move onto discussing what do with the lowered detectability achieved by the various methods described so far, and how stealth affects various elements of the setting and vice versa. 

Active detection

The concept is simple: produce your own energy, send it out into space, and listen for echoes. This can be RADAR, using radio, LIDAR, using light, or various other radiations.
The USNS Howard O Lorenzen, the US Navy's latest radar ship
RADAR has been around for a long time, so you should be familiar with how it operates. In space, the biggest problem with RADAR and active detection in general is the inverse-square law. It states that energy per square meter is divided by the square of the distance. The return signal you are hoping to pick up goes through this twice.
  • Return signal: (Output * RCS * Antenna) / (157.9 * (Distance) ^4)
Return signal is measured in watts per square meter.
Distance in meters.
Output in watts is the power you put into your outgoing signal.
Antenna is the aperture of the radio receiver, measured in square meters, dependent on the frequency used and the antenna gain. 
RCS is the radar cross section, and determines how much of the radar signal is bounced off the target back towards you. It is a dimensionless coefficient.

For example, the RCS of a flat steel plate is about equal to its area. A spherical ball of steel of the same width would have an RCS of about 6% of its visible area. Radar absorbent materials can further reduce this figure, as well as shapes designed to bounce the signal into other directions.

How an object appears on radar is also quite different from how it looks visually. Corners are practically invisible, while flat surfaces look huge. The signal return is also heavily influenced by the radar frequency used. 

In space, the shortest frequencies will be used, because there is no interference from atmosphere or clutter. Due to the distances involved, a high-gain antenna will be optimal.
The high-gain antenna on the Cassini probe is its largest feature.
LIDAR uses laser light instead of radio. It has the advantage of directing its energy very effectively onto the target, with a corresponding increase in return signal strength. 

LIDAR obeys the inverse-square law, so the equation for return signal strength is very similar to that of RADAR systems. The difference is that RCS is replaced with its visual equivalent, and the Antenna factor now relates to the sensitivity of the photodetector.

It is likely that a combination of the two methods will be used to detect a target. While RADAR will likely return a weaker signal compared to LIDAR, it will penetrate through features designed to defeat LIDAR and better identify the target.

The actual detection range for RADAR and LIDAR methods is the following:
  • Detection range=(0.07958*Output*RCS*Antenna/157.9*Sensitivity)^0.17
Detection Range in meters
Output in Watts
RCS is a dimensionless coefficient
Antenna in meters
Sensitivity in watts per square meter

The sensitivity is a factor determined by the receptors used. Photodetectors are generally more sensitive than IR receptors.     

The general rule is that return signal strength drops very sharply with distance, leading to an extremely short detection range in comparison to a passive sensor. Additionally, the power output can be detected by the target craft before the return signal is strong enough, giving it time to deploy decoys or reconfigure itself in radar stealth mode.
Course correction

In some scenarios, the spaceship will have to change its trajectory after the departure burn. Whether this is to respond to enemy movements, in response to data gathered on enemy positions after departure, to evade known or likely positions of sensor platforms or defensive stations... some sort of acceleration has to be provided in a way that does not increase your detectability.

The most obvious method is using an inherently stealthy manoeuvring system, such as a cold-gas thruster.
However, propulsive performance is directly tied to the temperature of the exhaust. So, a cold-gas thruster would have very low exhaust velocity, and would require very large amounts of propellant to achieve good deltaV.

For example, a nitrogen gas thruster has an exhaust velocity of barely 700m/s.
An example of a stealthy rocket from the High Frontier board game
The ultimate cold gas thruster is the metallic hydrogen rocket, with an exhaust velocity of 17000m/s. If such a technology could be perfected, it would allow spaceships to travel across the solar system with very low thermal signatures. However, the design challenges are astounding and any inefficiency would lead to very hot exhaust.

Another approach is low-energy propulsion. 
This relies on using an efficient, high exhaust velocity but low total power engine over long periods of time, with the waste heat generated dealt with using low temperature radiators or manageable amounts of open-cycle coolant.

For example, our 1GW spaceship has 294m^2 of radiators. The area/temperature detection range equation tells us that if it wishes to remain undetectable up to 10000km, then it can operate its radiators up to a maximum of 208K.

This gives us a waste heat removal capacity of 28kW. If our 'stealth' propulsion system is tailored to be more efficient at the cost of thrust, then 90% propulsive efficiency and 60% reactor efficiency is reasonable. This gives us an output of 151kW. 

Using liquid hydrogen as propellant, we can expect an exhaust velocity of 20km/s and a thrust of 15 newtons.

1GW spaceship
151kW stealth drive
20km/s exhaust velocity - 15N thrust
500 ton dry mass, 635 ton at launch, 558 tons during transit
Acceleration: Force/Mass = 0.02mm/s^2

Although it seems incredibly low, it can be operated over the course of the entire Hohmann transfer. Between Earth and Mars, it is 8.6 months. Over the course of one week, the spaceship would have deviated its trajectory by 12m/s. In a month, it is 52m/s. Over 6 months, it is 311m/s.

While it sounds small, you have to realize that a few dozens of meters per second can mean the difference between an attack and a flyby mission abort. If it detects an approaching enemy force or a dangerous area, it can hide within a volume 6.5 million km wide in a day

The third type of stealthy tactical maneuvering does not require propellants and does not generate heat. It relies on momentum transfers between spaceships in a fleet.
If two spaceships are rotating around a shared center of gravity, at a distance of 2km, to generate one gravity of force, then they can throw each other at a velocity of 100m/s. If they are separated by 20km, they can achieve 311m/s. It performs excellently as an evasive maneuver.

Departure burn

Every mission hoping to cross the vast distances between planets without being detected will rely on a Hohmann transfer, and the most important part of that trajectory is the departure burn. This is the initial acceleration performed at the home planet.

For a high-exhaust velocity, low-thrust spaceship, the total deltaV required is divided into a series of accelerations at the lowest point of the orbit. While these can be performed against the bright background of a planet, each successive maneuver multiplies the chance that you are detected. 

For very low thrust propulsion, you might have to slowly spiral out of lower orbits, into higher orbits where visibility is very high.

Maybe your propulsive power is so bright that the planet could never hide you.

In such cases, you can still achieve some level of stealth.

The primary method of identifying a spaceship is from its drive plume: the trail of hot exhaust from its rocket engine. Simple observation of the size and brightness of the exhaust provides a reading on the energy output, exhaust velocity and the thrust generated. By watching it over time, you can then determine the mass of the spaceship. Watch them until they stop accelerating, and you can calculate a trajectory, a destination, an arrival time and where to tell your telescopes to look at.

While X spaceships of Y masses accelerating at rate Z isn't much information to go on, it can be very useful when correlated with other databases

For example, the masses can point towards certain categories of military spaceship. The drive power might exclude commercial spaceships. The exhaust temperature might indicate propulsion systems from a certain manufacturer, and so on. This information can help prepare your defenses and your pre-deployment.

Equally, there are ways to distort these readings.

The easiest methods reduce your opponent's ability to identify the spaceships. These include using a different propellant, operating below full capacity and so on.

Beating trajectory predictions is harder, but also possible.

Aerobraking is one way. 
Orbital Transfer Vehicle
This is performed after the final manoeuvre. While the gigawatt drive plume of a rocket engine is extremely visible, the tenuous atmosphere at the edge of a planet is very hard to detect across interplanetary distances. An aerobrake manoeuvre can shave off dozens, if not hundreds of meters per second, and can mean the difference between going to Jupiter or swinging by to Saturn. For multi-year missions, a spacecraft can boost to a transfer trajectory, catch up with its home planet, and perform an aerobrake at the lowest point to obfuscate the rest of the mission entirely.

Another option is external boosters.
Determining thrust from an exhaust plume is easy, but external boosters can remove the option to determine mass from that reading.

For example, our 1 GW reference spaceship can be boosted out of orbit by a booster with propellant reserves and a 10 GW engine. Sensors would work out its thrust and determine that it masses 700 tons and is probably 50% engine in dry mass.
When the booster drops away, the enemy would have no idea as to what the true mass, acceleration and deltaV capacity of the spaceship is. 

This is important when the defender tries to achieve a tactical advantage by launching laser relay drones, kinetic swarms and interceptors ahead of your trajectory. To be on the safe side, your opponent will overextend their defenses. Instead, you will have the strategic advantage by launching a second wave after your opponent has expended their resources.   

A less wasteful example of this is the propulsion bus.
Several spaceships cluster together and use external propellant tanks. Only a fraction of the engines are used. Once the manoeuvres are complete, the propellant tanks are ejected and the spaceships separate. Your opponents will never know the true number or size of your spaceships. 

The final method is external propulsion.

Instead of relying on onboard reactors or rocket engines, the spaceships are driven by power generated elsewhere. For example, the fleet might be boosted on detachable laser thermal drives, riding on beams generated by boost stations. The advantages include greater efficiency, greater acceleration and a much shorter delay before they spaceships cool down enough to achieve stealth. 

In the next part, we'll discuss the uses and consequences of all the stealth methods discussed so far, and how a writer or worldbuilder can integrate them into their setting.


  1. I think you might lose your Atomic Rocket Seal of Approval over this ;)

    Aerobraking, nope, you'll see the plasma trail not to mention the insanely hot heat shield itself. The plasma trail is effectively a rocket exhaust itself. Most of the heating at interplanetary velocities is *radiative* heating from the plasma trail. Very easy to cover this avenue by simply posting "weather satellites" in orbit over the planet of opportunity.

    As for metallic hydrogen, that's going to be basically the same as a NERVA at the equivalent temperature. It's not actually sublimating like dry ice.

    Using electrodynamic tethers mmmmight work, as could other options involving magnetic fields.

    Possibly the simplest solution is to launch dumb cold pebbles with a coilgun from base or forward asteroid bases that thwack into an Orion-style pusher plate, or magnetic catcher's mitt. Or, if the strike package is small enough, simply shoot it out a big coilgun.

    1. I doubt it :)

      First of all, the ability to place a sensor satellite very close to the departure orbits of the attacking planet depends entirely on the setting and scenario the author wants to create, so I cannot take it as a given fact.

      Second, having a sensor platform over there would negate the vast majority of stealth techniques and give a massively disproportional advantage to the defenders. Also, it's like arguing against a point by changing the scenario entirely.

      Third, if the 1GW reference spaceship dipped into the atmosphere for 10 minutes to shave off 100m/s, it would only be radiating 5MW over a large area. Compare to the 300MW assumed waste heat radiated by the engine accelerating. So, negligible.

      As for metallic hydrogen, the important factor when it comes to stealth is that it produces very little waste heat. Pulsed thrusters using small pellets of metallic hydrogen can achieve extreme expansion ratios and therefore extremely hard to detect clouds of hydrogen as exhaust.

      Electrodynamic tethers are actually a terrible idea. The magnetic or electric fields required for useful acceleration in multi-ton spaceships is quite important, and that has to be provided with reactors. Those would emit waste heat.

      Cold pebbles smacking into an orion plate would release a large part of their energy as heat, so would not be stealthy. A magnetic 'scoop' which de-accelerates them to provide propulsive force is an excellent solution.

    2. How cold can this metallic hydrogen exhaust be? Are there any equations?

    3. As cold as any other fully expanded exhaust stream, so down to double digit kelvins or lower.

  2. Mr Anderson...

    It would not be cold. In fact, it would be (to quote Ruby Rhod) HOT HOT HOT! 7 000K for a chamber pressure of 100 bar, and 6 700K for 40 bar. Which would vapourise any material known to human science. The recombination energy is 216MJ/kg. About 20x better than LOX/LH2 which is why it is so amazing and SyFy.

    Matter Beam...

    How about you read this. Then look at your equation.

    1. That would be for the metastable solid metallic hydrogen engine that seems to have drawn some attention at NASA. It is not necessarily the only way to use metallic hydrogen.

      The ultimate one would be the way more difficult liquid metallic hydrogen, which requires the highest pressure and is the most hypothetical. Liquid metallic hydrogen is a superfluid, which is why it has such an astounding speed of sound.
      The flow speed of a fluid under pressure is limited by its speed of sound (which why hotter is better, as it increases speed of sound). So with liquid metallic hydrogen, it allows for a cold "gas" thruster with theoretical Isp on par with fusion rockets (though not the same mass ratios). Even better, it may absorb considerable heat by expanding, for regenerative cooling.
      That is, if you can get past the slight engineering challenge of keeping the stuff at pressures found in Jupiter's core, of course.
      There may be easier ways by using those high-pressure solid metallic hydrogen pellets. Solid metallic hydrogen is more studied, and require slightly less mind-blowing pressures.
      All those (including the NASA proposal) are dependent on the still poorly known physics of metallic hydrogen, though.

      For aerobraking, this is not about a full re-entry, but skimming the upper atmosphere to alter velocity by only a few hundreds of m/s. This would be considerably less energetic, and may be missed by interplanetary sensors.
      On the other extreme, you have today's probes using aerobraking with their solar panels to shave off a few m/s here and there.

    2. And as with all exotic monopropellants, things get rather exciting when the fuel tank takes a hit... although it now brings up the prospect of submarine battles in Jupiter's oceans!

      Regarding the aerobraking... why bother? A gravity assist is 100% inert and can confer enormous amounts of delta-V. If you're already stealthy coming in (assuming the amateur astronomers on Planet Hick haven't spotted you yet)then you're good to go.

    3. what is the speed of sound (or even better, other performances?) in liquid metallic hydrogen? I'm having trouble with opening the link that troy sent

  3. Hi there ! Just passing by to say that I love your contributions to "hardscience" SF, and as I plan to write such kind of novels soon (not in English though, not my mother language), I try to understand all these science things and make my own universe with it... I'm already a lover of ProjectRho website and the few others that earned their Atomic Rockets approval !

    Let's be honest your website content is the most difficult for me to handle... cuz' I don't work or study physics, hehe.

    1. Thank you for your interest. You are the audience that I most wish to help.

      If you need further explanations of something, just mention it in the comments.

      Don't worry about the lack of new posts lately, it's just the exam period.

    2. Just a quick thought, of especially great importance for helping those with less experience of the maths, it might be good if you noted in any equations you quote what units are to be used. I guess you're generally using SI units but that might not be obvious to everyone, and when you quote constants numerically rather than as the expressions which go to form them you're further locking the equation towards a specfic choice of units. It might help people if you quoted the units you're using. "Laser power (Watts)", "Range (metres)"... because if someone used megawatts and astronomical units without altering constants accordingly (an unpleasant task for those not keen on maths) their results would be very wrong.

    3. Hi!

      I'll update the post tomorrow as you suggested.

  4. As much as I respect Atomic Rockets, they get a few things wrong about drive thrust.
    The spectral bands will tell you what propellants are being emitted. However, they provide no information about the proportions if there are multiple propellants. They also say nothing about the density not energy of the exhaust.
    Doppler shifts in the bands will tell you the linear component of the exhaust velocity, either toward or away from the observer, but this is NOT the actual exhaust velocity. There is absolutely no information about the lateral velocity components. Furthermore, I am uncertain as to how precise this information will be. Will the doppler shift of the measurable component be accurate within 1 km/s? 1 m/s? 1 mm/s? Even if you could acquire an interferomtetric estimation of the exhaust velocity from different observation posts, it is questionable if the precision would be sufficient for useful information.
    next, intensity should be able to provide some information about the kinetic energy of the exhaust, but this would be average temperature only. It would be rather easy to defeat this by "burning" different propellants at different temperatures, because there would be no means of determining which propellants were "burning" at which temperatures, or the relative proportions of each.
    Nor would it likely be possible to acquire useful information regarding the size of the plume. There are physical limitations to the resolution possible with any given telescopic array, no matter how technologically advanced. An array the size of the arecebo observatory, would have a resolution of about 1 pixel per 300 m * 300 m. If you want a resolution on the order of 1 m^2, you would need an array baseline spanning 1 million km. Even if you had excellent resolution, this would only give you the size of the plume... not it's density, and therefore not it's mass. It certainly does not provide information on mass flow.
    Finally, AR oversimplified the data requirements for measuring thrust. Yes, thrust is determined by the mass flow of the exhaust and the exhaust velocity... but mass dispersion is also important. You need to know the dispersion angles of the exhaust and the dispersion curve to accurately calculate how efficient that exhaust is in applying thrust in the vector of motion. Without this information, you are not going to be able to calculate the mass of the vessel.
    All of the information you DO get assumes that there is going to be an emission of photons in the direction of the observatories, and that in itself is not a given, especially with ion exhaust.

    1. Thank you for commenting on this issue. I was unable to gather any data on this subject, so your insight is useful.
      I have a gut feeling that multi-spectral observation of an exhaust plume can defeat the exhaust 'salting' you mentioned. Two elements at the same temperature emit at different wavelengths, so observing the plume's emissions on a per-wavelength level might determine the energy of each propellant used.

    2. The multi-spectral observation you mention (with doppler) will allow you to determine the average velocities of each of the different gases, and it is even possible that you might detect doppler "echoes", or a thicker-than-usual spectral band, which will tell you that there are different exhaust velocities. However, there is still no way to determine the proportions of the different gases, nor the proportions of the same gas being emitted at different at different velocities. Thus, an accurate measuremnt of thrust would not be possible.
      Actually, there is another clue that this would not work: no space agency uses this technique to measure the thrust of engines during testing. If it worked, it would be simple and inexpensive. There would be no need of complex test stands. It would especially make measurement of ion engine exhaust much easier, since the current method requires putting multiple sensors at various locations within the exhaust stream.

      On a separate matter, in my analysis of resolution above, I forgot to mention that the resolution of 1 pixel per linear 300 m ias valid for a range of less than 1 AU. We are actually dealing with angular resolution being translated into distance, so each time the distance doubles the linear resolution is cut in half, and the area resolution stripped to one-quarter. This is also dependent upon the wavelength(s) being detected. My figures assume visible spectrum light. Resolution for IR or radio-wave is much lower (if you have UV or cosmic radiation, the resolution is higher).

    3. On exhaust: that is a point I will have to look into further.

      On resolution: I believe your argument can be translated as every single hard detect requiring active sensors, such as low wavelength LIDAR, at interplanetary ranges.

    4. On resolution: active sensors will be alomst useless at interplanetary ranges, since return signals decrease with distance at d^-4. This means you would need a multi-GW power source for every transmitter. Also, this really would not help you with resolution. In fact, it essentially cuts the resolution in half because the distance is now doubled.
      Instead, there are two options:
      First, use an AGEIS type system to link all the detectors in your fleet to one another, creating a baseline as large as the distance as your most remote communicating ships. The problem here is that size limitations reduce you to slow scans, reducing probability of detection... but at least you will have remarkable resolution for whatever you DO detect. There are more limitations though, that tend to get really complex.
      The second option is to install multiple large recievers on any astronomical body you control, and build other (huge) dedicated observation platforms and place them in planetary or solar orbits. The downside here is the astronomical cost, and the relative ease of locating and destroying the observatories.
      IOW, my argument can be translated as 'every hard detect will require multiple arrays with communicating components spread at least the distance between the Earth and the moon.'

      Something to keep in mind: the resolution will depend upon the size of the baseline (the distance between two communicating components within a single array); but the actual sensitivity is dependent upon the actual total 'apperture' area of the collected components. If you have two 35m telescopes, one on earth and the other on the moon, you will have the resolution of a 384 000 km aperture telescope; but it would have the sensitivity of a 50m telescope.

    5. Considering that modern telescopes are sensitive to 10^-19 watt emissions using only 5m mirrors, I guess that synthetic apertures will be more than enough to detect any emitter in the Solar system.

      In most settings, travel times between planets means that you can even use a small group of satellites sweeping across their orbits to artificially increase your aperture, like in Synthetic Aperture Radar.

  5. Stealth in space cannot exist unless you can achieve invisibility. Stars in space do not twinkle and every spot in the night sky contains stars.