KSP Interstellar Extended Practical Guide

[Voqk]

Hi all, I'm a KSP intermediate (500 or so hours) with about 80-100 hours spend with KSP Interstellar Extended. While there is a decent amount of documentation out there, I tend to learn best with specific examples of how something functions, since then I can start with a working model and see what small changes do to the overall system. As such, the documentation was a bit hard for me to work through. This guide attempts to approach this mod from a different perspective - how to build working networks and crafts for specific goals:

1. LKO Shuttle (for ferrying to my casino or hotel at 360km)
2. LKO Launch Vehicle with 120 Ton Capacity
3. Interplanetary probe
4. Nuclear Engines
5. Power Relay Setup

I'll provide a lot of comparison such as:

1. How to build an efficient power relay network for the above - the tradeoffs of different generators, transmitters, and wavelengths
2. Onboard power vs. relay power craft comparison
3. Practical implications: TWR, Cost, Fuel type selection, Transfer planning

Feel free to ctrl-f for the topic you're interested in.

There's also three Appendices:

1. Debugging Power Issues
2. Rocket Power Data
3. Beamed Power Data

To offer more information on those topics.

Note that I use a 3440x1440 ultrawide, so the screencaps are big. View on imgur and zoom in - there's tons of metrics in the screencaps. I've copied the most relevant information into this post body.

Finally - I'm only a beginner with this mod. I can now build all of the above, but I cannot say at *all* this is the most efficient way to do it. @FreeThinker may double facepalm reading this. But I hope this helps some people that learn like me get up and running with a rich, interesting mod!

1. LKO Shuttle
Target - 32 kerbals to 360km Orbit

Topics covered:

• SSTO Build
• Engine Selection

Mark 1 - Inline Power Generation

First off, I started by building this with inline power to make the shuttle completely self-contained:

Cost: $850,268
Weight: 83.34T
Heat Generation: 51.72MW
Craft File: https://kerbalx.com/Voqk/CasinoShuttle2

First stage: 3.75M Pebble Bed Reactor (for thermal power), "Phoenix" Thermal Aerospike Engine fed by two shock cones + precoolers
Orbital stage: Inline Fission Fragment Reactor and Charged Particle Genertor (for energy), two 1.25m ATTILLA thrusters for use in space.

Fairly Straightforward - you attach a thermal engine (the aerospike) to a thermal generator (Pebble Bed). Scale the pebble bed based on craft weight (start low, try to launch it, if there isn't enough thrust then scale up the reactor). Quite frankly I just opted for this size because it fits the Mk3 stack decently well. This generates tons of thrust so it's quite easy to make orbit even with my limited experience with SSTOs. You should be able to get into space with the "Phoenix" on open cycle (using atmosphere) pretty easily - and if you pitch down at about 20km altitude and build speed, you get pretty close to orbital velocity before entering space. Note: to circularize the orbit, you can't use the Aerospike natively as there is no air in space. So I switch to close cycle (does not use air) and feed it Liquid Fuel. Inefficient, but it works.

Pros: No relay network required. Self container. High ISP in space.
Cons: Somewhat range limited in atmosphere as the "Phoenix" needs some liquid fuel to run. Not really a practical concern. Bad TWR and ISP on closed-cycle "Phoenix" for circularizing orbit, so you have to start circularizing a bit sooner and the fuel burns quickly. High ISP in space but brutally low thrust. Transfers take ages.

 

Mark 2 - Beamed Thermal Power

Second, using beamed power. And this is a big topic, so there'll be further subsections.

Cost: $398,612
Weight: 61.2T
Heat Generation: 40.01MW
Craft File: https://kerbalx.com/Voqk/CasinoShuttle-Thermal2

First Stage: Two Direct Cycle Nuclear Turbojet fed by two Shock Cones with precoolers
Second Stage - 2.5m Thermal Receiver with 2.5m "Krusader" Thermal Rocket Nozzle

First stage runs purely on atmosphere. This is awesome - unlimited range in atmosphere. Plenty of thrust to fly easily. Boost up at 15 degree pitch until thrust on the thermal nozzles run out, then activate the thermal rocket engine to easily make orbit with tons of delta-v to spare.

Pros: Way more thrust in space using the thermal nozzle, with fine ISP and delta V. A lot easier to fly around. Way lighter too, so reaction wheels turn it fine.
Cons: None, other than needing to set up a power relay network. See Section 4.

 

Engine Selection Appendix

Direct Cycle Nuclear Turbojet
I found this is my preferred in-atmosphere engine. It generates less thrust than the TORY, but I kept having the TORY explode (more details below), so this was just easier to work with. You'll get more thrust without the precooler, but the engines may explode. Consider that these engines all function by heating up the propellent - so more heat is more thrust, but it also leads to explosions).

Pros: Easy to use. Infinite range in-atmosphere. Plenty of thrust
Cons: Not as much thrust as TORY

TORY Nuclear Ramjet Engine
Tons of thrust, but kept exploding. I tried closing the build in air intakes and feeding it cooled air (like the direct cycle), but that just made the build uglier.  I tried reducing throttle and/or output to reduce heating, but it didn't seem to help much (still blew up) - but then, what's the point? Finally, I've been told that adding more inline cooling can help with this to reduce the engine (not the propellant) temperature. I did have some luck with this, so might be worth pursuing. But I prefer the Direct Cycle's mounting, and it works well in my builds - I like to mount inline while direct cycle mounts radial, so I found direct cycle easier to work with.

Look, I know I don't like this because I can't figure it out. Not a bad engine, just a lack of knowledge. Just wanted to note that I found the direct cycle easier to work with.

Pros: Infinite range in-atmosphere. Tons of thrust
Cons: More considerations for cooling to prevent overheating to achieve that tremendous thrust. Balancing air intake cooling vs. thrust generated was more tricky for me than with direct cycle. YMMV.

"Phoenix" Thermal Aerospike Turbo Ramjet Hybrid
Works similarly to the other two in-atmosphere, but needs a trickle of liquid fuel to keep running. However, it needs to be attached to a thermal generator, so it's not an all-in-one solution

Pros: Easy to use. Great range in-atmosphere. Plenty of thrust
Cons: Needs an attached generator.

2. LKO Launch Vehicle

Target 120ton capacity

Topics covered:

• Inline vs. Beamed Power
• Sizing Engine and Power requirements
• Cost Comparison
• Fuel Comparison

Mark 1 - Stock Chemical Engines


Cost: $199,390
Weight:  793.675T
Heat Generation: 0

Purely for comparison. Not a great rocket, just threw this together as an example.
First stage: 4+5 Mastodon Engines with fuel coupling to second stage.
Second Stage: 5 Mastodon engines

Pros: Stock
Cons: Expensive, heavy.

Mark 2 - Beamed Thermal Power

Cost: $123,665
Weight: 226.284T
Heat Generation: 2.7GW
Craft File: https://kerbalx.com/Voqk/120TonLauncher

I started by throwing 120 tons of ore into tanks for something to work off of, then used a matching thermal inline receiver (mk1) + "Krusader" rocket nozzle, tried to launch it with ever increasing beamed power to size it. I found that a 3.75M receiver/rocket was overkill - I get plenty of thrust with 2.5M. Also, it seemed to want about 280MW of power, total. I didn't do this in small increments - I started with 50MW, it wasn't enough, so I sized up the generators and found that at 280MW got this to orbit cleanly, but without wasted power. This works nicely with the way the reactors scale.

The ground power is much higher - 280MW KA-band on the ground and 50MW?Near Ultraviolet (details in section 4) in orbit. As you get further from the ground based power (2KM from the KSC, offshore), the power from that declines linearly. The power from my space network increases slowly - so the TWR decreases the higher you climb (see below). However, this balance worked totally fine. The TWR was a comfortable 1.65-1 the whole flight.

Also note that the "Krusader" is much higher thrust/ISP in space - so the higher you climb, the better the engine performs - offsetting the reduction in power received.

Final note - size your receiver based on your power requirements. I seemed to need about 280MW of power, and the 2.5m inline receiver can receive 270MW, so increasing the size does *not* increase thrust. First determine power requirements, then pick the smallest receiver that can handle that power.

Some other details in the appendix.

Pros: Lower cost once relay is in place.
Cons: Cost of setting up and maintaining network

Cost Comparison

The chemical rocket costs about $200K, and the thermal only $125K. So it's far cheaper per launch.

*However* - the relay network I have costs $1.15M (yes, million) for the ground generator with a runtime of about two years. The space network is three installations costing $700K (plus launch costs). So the power network total is $3.25 million for a year. Breakeven point is 43 launches. 

But with Interstellar Fuel Switch installed (and it really needs to be), you can add more fuel (or refuel later). An IFC Radioactive Fuel Container RFC2500 set to Uranium Nitride with 3000 units onboard costs $496,500 and adds 27 years of runtime. Adding these makes the launch cost $4,939,500 total, with a runtime of 28 years. Breakeven point is 66 launches - so if you can do 66 launches in 28 years, you're just saving money past that point. *Also*, your running cost can become effectively free with ISRU processing (for example, from the Mun).

 

Fuel Comparison

The "Krusader" defaulted to LH2, and I was shocked with the low delta-v (742 m/s in this build). I then played around with switching the fuel types, and found that was the 'problem'. Here's a comparison:

Liquid Fuel:  4471 m/s, 1.67 TWR, $123,665 vessel cost
LH2:                 742 m/s, 2.45 TWR, $113, 954 vessel cost
Hydrazine:   4189 m/s, 3.16 TWR, $255,281 vessel cost
LFOx:              2812 m/s, 5.32 TWR, $114, 435 vessel cost

I believe this is due to the relative densities of the different fuels? Not sure. Regardless, in practical terms, pick the right solution for you. With hydrazine you get an excellent thrust and delta/v at a high cost. With LFOx, you get a huge amount of thrust (so you could reduce the engine size, or power requirements) and small cost with lower delta/v. Liquid fuel seemed a good balance - low thrust (but plenty, in my application) and best in class delta/v with low cost.

3. Interplanetary Probe

Target >6000 Delta V for all stock science parts

 

Topics covered:

• Electric Engine Use

Cost: $353,782
Weight: 12.613T
Heat Generation: 0 (non-retractable panels)
Craft File: https://kerbalx.com/Voqk/ScienceProbe_IonDrive

This was actually my first use of KSPIE - how to build a high delta-v probe. I wanted to use electrical power, and I wanted it to be self-container (no power network set up yet).

I think the trick here is using a cluster of ion engines - this uses nine IX-8219 "AFTER" Ion thrusters). It has a reasonable TWR of .17, so orbital transfers are simple enough. Just put it in orbit, deploy the solar panels, and go wherever you want.

4. Nuclear Engines

Target huge ship to Jool.

 

Topics covered:

• Nuclear Engine Use

Cost: $1,234, 741
Weight: 520.434T
Heat Generation: 97KW

This is a big boy - meant for long term study of the Joolean system. It consists of a command and habitation section for 24 kerbals, advanced science module, 'Explorer' craft and ISRU craft.

Explorer

Transfer stage: ~7300 m/s delta V in a transfer stage consisting of a single NERVA Solid Core Fission engine
Lander Stage with 2000m/s delta V. It attaches to the middle of the main vessel and must be docked together for interplanetary transfer.

ISRU

5280 'native' delta V - but this is meant to land, fuel up, come back to the main vessel and dump fuel in the large orange holding tanks then return to the surface and do it again. This holds 16200 units of liquid fuel (and subsequent oxidizer) - it takes three trips to fuel up the main vessel completely. Nothing special here.

Main

4623 delta V for the whole combined package. Build in three launches - main core used the four "Timberwind" engines as the main body of the second stage and launched unfueled, ISRU and Explorer also launched unfueled. Then fueled it up and was off to Jool.

There's a good amount of thrust here - I didn't have to do multiple burns to make my transfer, for example. But it is *heavy* and *expensive*.

 

General Nuclear Use Appendix

I was having trouble make nuclear engines work for me. Their relatively low thrust (NERVA, for example) made timing transfer windows a pain in the ass, and capture burns (to Moho, for example) a bit scary as it took so long to decelerate for capture. Note that making a burn over 5 minutes for transfer is pretty much impossible, you very easily end up burning into kerbin's atmosphere as your manuever node at the start of the burn may be radially in to make sure you're resulting trajectory is correct.

If you're having this issue, plan to make two burns - one to get to a higher orbit with a period of about 4-8 hours, then a second burn to transfer. The reason I say a 4-8 hour period is that if you burn further - such as burning to the edge of SOI on your first burn, and the remainder on a second burn, your orbital period may be too high to time your transfer effectively - you may be waiting twelve days to get back to your maneuver node and find yourself in a poorer transfer window. I found this to be just not fun, so I just used more engines on this craft to provide sufficient thrust and eliminate that necessity.

 

4. Power Relay Network

Target 50GW continuous coverage in Kerbin system (LKO out to minmus) and 200GW of ground power for heavy launch vehicle.

Topics covered:

• Wavelength Selection
• Orbital Height Selection
• Orbital Placement
• Ground based Wavelength and Placement
• Generator Comparison

Orbital Wavelength Selection / Orbital Height and Placement

First off - check the wiki for ideal orbital heights for continuous coverage of a body. https://wiki.kerbalspaceprogram.com/wiki/Tutorial:Setting_up_a_CommNet_system. 600km orbital height gives continuous *ground* coverage. Push that out a bit for LKO coverage (for example, a shuttle at 80km elevation will not have good coverage with a relay network at 600km. Shoot higher - 800km or 1Mm for the relays, for example. This is the first thing to determine - what orbit height does your relay need to be at to give coverage to the orbit heights your shuttle/receivers will be at.

Now that we have that baseline, consider the application. The larger the wavelength (Microwave is big, X Ray is small), the better efficiency you get, but the shorter distance it will be effective. For example, a Ka-Band microwave in LKO is a very big wavelength - 8.5mm vs. Near Hard X-ray 100pm. Beams spread out as the travel - so if you start at 8.5mm in LKO but then go to minmus, that beam may become tens of meters (or more - I didn't calculate this) wide. Think about a flashlight. If you point it at something close, it's a very small circle. If you point it at something a few meters away, the area that's light will be much larger, but much less bright. Same concept. This is what's meant by Spot Size - if you look in the Power Receiver interface on a shuttle you're trying to run on thermal power, you'll see how much that beam has spread out. You can combat this two ways:

1. Use a smaller wavelength. If I want to send power a long way, the beam will spread out a lot before getting to the target. So I need a smaller wavelength. X-Ray, for example, can provide power over many millions of miles, while microwave only provides good power over a few KM. NOTE: the tradeoff is efficiency. Near Hard X-Ray is about 32% efficient - so if you have 100MW generator and send out Near Hard X-Ray, you'll only have a maximum of 32MW in your network. Ka-band microwave is 94% efficient, but only provides power over tens of kilometers. This is really important - beaming power a long way requires a smaller wavelength, which will be less efficient. So, you design your network based on this.
2. Larger transmitters. A 5M transmitter will generate a larger spot size than a 20M one. I'll show an example below.

That's a bit dense. There's specific data in 'More data in 'Appendix - Beamed Power Data', but time for some practical examples.

Again, I want coverage throughout Kerbin, so I placed my relays halfway between Minmus and Kerbin. Thus, the *worst* power is in LKO or at Minmus, but it gets no worse than that. So that's the minimum power I get anywhere in the system. Here's some examples. Note that the shuttle/receiver is at an orbital height of about 50 Million meters - 45 million meters from the LKO generators. 

Craft File: https://kerbalx.com/Voqk/85MW_Gen_ShortInfrared_Launcher

Spot Size Examples

1. Bad: I have two 85GW generators in HKO at an orbital height of 5.286Mm - million meters , both with 85GW generators. One uses Near Infrared (1.05um wavelength) and one using Short Infrared (700nm wavelength, larger).  You can see the effect of spot size here - Near Infrared has a wavelength of 1.05um - really small! But 45 million meters from it, my spotsize is 12.75 meters because the beam has spread out. My receiver is 2.5 meters, so only 1/5th the size of the power beam - thus I only get about 17.92% efficiency on that beam. The Short Infrared (wavelength 2.2um, larger) has a spotsize of 26.73m and efficiency of 8.55%.
2. OK: I have three generators at an orbital height of 20Mm - again, about halfway between Minmus and Kerbin. So my vessel is about 25Mm from these generators - closer than the ones in HKO. One is broadcasting Near Ultraviolet - a wavelength of 400nm - and a 3.75m transmitter. I see a network efficiency of 58.45% here, not great. This is because the spot size is 3.9m - bigger than my receiver (2.5m). This can be fixed two ways - I can use a smaller wavelength OR a bigger transmitter. Either will reduce my spot size, thus making my receiver able to use more of the power. Using a smaller wavelength will be less efficient (covered earlier), using a bigger transmitter is just going to be a heavier/harder launch. So...
3. Best: My generator at 20Mm using Red Visible (700nm wavelength, so a larger wavelength than the Near Ultraviolet in 'OK'). On it's own, this would mean it was less efficient - again, larger wavelength = bigger spot size. However, I'm using a massive 'Inline Wrapped Phase Array) transmitter which is 50M big once deployed. As such, my spot size from this larger wavelength is actually smaller - 512mm - for a 99.17% network efficiency. So, I get the benefit of a larger wavelength being higher efficiency (Red Visible is 66% efficient, vs Near Ultraviolet is 56%). So - I get the best of both world. A 10% more efficient wavelength, still with a really tiny spot size. You'll even note that my spot size is 512mm - way tinier than it needs to be. So I could change this Generator/Transmitter combination to an even larger wavelength like something in the infrared spectrum (76% efficiency) and still have a tiny spot size.

More data in 'Appendix - Beamed Power Data'

Finally, check out my 120 ton launcher on the ground at the KSC, to see the ground based transmitter:

Here, I'm receiving two beams: one from the 85GW generator orbiting Kerbin at 20Mm (the same one from above) and one from my ground based generator. You can see I still get decent power from my 85GW generator - the spot size is only 719mm, add some atmospheric absorption and you see a 66.835% efficiency. Not bad!

Ground based Wavelength Selection and Placement

Craft File: https://kerbalx.com/Voqk/200GWLandGen-KaBand_Launcher

Finally, to get my 120 ton launcher off the ground, a moderate efficiency from an orbital network isn't going to do it. A lot of that power is absorbed by the atmosphere, for one thing. For another, I just need a lot more power to get into space than to move around in space. So I build a ground based generator producing 200GW. In this case, I don't need the power to go very far - it's only meant to get me to space (70km). So I can choose a really large wavelength (again, because I'm not worried about the beam spreading out over long distances). I compared the various microwave bands (X, Ka, W, D) and found Ka to be a good balance. It gave a high enough efficiency to get the rocket off the ground, but spot size was still workable when high in the Kerbin atmosphere.

Other details in the 'Appendix - Rocket Power Data'.

Generator Comparison and Cost

I haven't played around with this as much, to be honest. I didn't look at early game generators, for one. I did compare the Stellerator to the Dusty Plasma Fission Fragment. @FreeThinkercan probably add more here if he wants, I'll update if so and do a further comparison to assist in early game stages.

The tradeoffs were:

Stellerator
Pros: Earlier in the tech tree. No actinide production = no need to reprocess fuel.
Cons: Heavier. More Expensive for given power. Shorter lifespan (I was only seeing 167 days of usage)

Dusty Plasma Fission Fragment
Pros: Lighter, cheaper, longer lifespan (1.8 years). No actinide production = no need to reprocess fuel.
Cons: None that I know of.

This is not a fair comparison *at all* - I'm comparing maxxed tech tree versions of two generators, one is later in the tech tree and one isn't. I only bring this up to note that when picking a generator, consider both the power, the cost, the weight (if launching to space) and the lifetime.

From the above launcher cost comparison:

The relay network $1.15M (yes, million) for the ground generator with a runtime of about two years. The basic space network is three installations costing $700K (plus launch costs). So the power network total is $3.25 million for a year.

But with Interstellar Fuel Switch installed (and it really needs to be), you can add more fuel at launch (or refuel later). An IFC Radioactive Fuel Container RFC2500 set to Uranium Nitride with 3000 units onboard costs $496,500 and adds 27 years of runtime. Adding these makes the space network cost $4,939,500 total, with a runtime of 28 years. *Also*, your running cost can become effectively free with ISRU processing (for example, from the Mun). No need for fuel reprocessing with Dusty Plasma, it doesn't generate actinides.

 

Appendix - Potential Issues:

My TWR is too low with thermal power!!

Open the thermal receiver, open the power receiver interface. Check these:

1. Are you connected to the generators? Make sure that you enabled the transmitters on them if you don't see them - it doesn't default to being on in the VAB, so you might have launched a generator, have power onboard, but forgot to turn on the transmitter.
2. Do you see the generator, but have very low power in 'Available Power'? This is a combination of wavelength, spot size and your generator's efficiency 
   1. Make sure you selected an appropriate wavelength (see section 4). You can tell this by looking at 'spot size'. You want this to be smaller than the receiver size - so, for a 2.5M receiver, you want this to be less than 2.5M. If it isn't, see section 4 on wavelength selection. Also note that spot size is based on the size of the transmitter - a 20M transmitter will have a much smaller spot size than a 5M Phased Array Transceiver. See Section 4 - I outline this a bit more there.
   2. Check the 'Facing' metric. If your receiver is not facing your transmitter, you're going to get bad power reception. On my 120 ton launcher tests, I'd have 98% facing my ground generator at launch, trailing off to 20-30% facing in space. At that point, I'd have 80-90% facing my space based network, so I'd have good coverage throughout. If you're on the ground and getting <80% facing at launch, you need to move your ground generator. 
   3. If the network power is far less than what you know the generator is capable of (the generator produces 9GW, but your available power is 1MW) you need to check the generator. Do you have enough heat dissipation? Use the Thermal Helper in the VAB to check. Is your Convertor sized appropriately? If you scaled up your generator to 5M, but left the convertor at 2.5m, you aren't converting all of that power to useable energy. They need to be scaled appropriately.
   4. Make sure your transmitter can handle the power the generator produces. A Phased Array Transceiver only transmits 5MW, while an inline wrapped phase array transmits 1GW.
   5. Make sure your transmitter is sized appropriately. The bigger the transmitter size, the smaller the spot size - a 20M transmitter will have a much smaller spot size than a 5M Phased Array Transceiver. See Section 4 - I outline this a bit more there.
   6. Check your fuel type (outlined in section 2) - for example, Liquid Fuel gives a lower TWR than LH2 (but more delta V). You might be able to fix this by changing your fuel
   7. Size up your receiver and generator. A 2.5M thermal generator and 2.5M "Krusader" can put 120 tons in orbit with 27GW of power. If you have that much power in your network, but your receiver is tiny, it can't convert all of it. Check 'Max Power Capacity' on the receiver in the VAB.

 

Appendix - Rocket Power Data:

Here's a lot more screencaps that I used to pull the above information.

First off, using the same generator and transmitter, comparing D-band Microwave to Ka-Band Microwave at 10Km and once target Apoapsis of 80km is reached (not at the apoapsis but once the speed to reach it is achieved, while still in atmosphere). Also comparing 2.5m receiver to 3.75. All using liquid fuel.

Takeaways:

1. Atmospheric/water absorption can kill you. D Band is bad for that.
2. Bigger receivers are only necessary if you need to receive more power. 2.5M can receive 270MW, so a 3.75 receiver is unnecessary here.

 

Height	Receiver	Wavelength	Facing	Spotsize	Network Efficiency	Available Power	Thrust
10km	3.75	D	48.127%	144mm	80.23%	216.6MW	4134kN
Ap	3.75	D	35.491%	3.2m	2.98% (*absorption)	8.057MW	71kN
10km	3.75	Ka	70.124%	719mm	92.22%	251.2MW	3856kN
Ap	3.75	Ka	31.658%	6.87m	27.75%	74.92MW	121kN
10km	2.5	Ka	49.510%	528mm	92.45%	249.6MW	3673kN
Ap	2.5	Ka	36.405%	5.48m	31.07%	83.89MW	448Kn

 

D Band at 10KM, 3.75m Receiver:
Facing: 48.127%
Spotsize: 144mm
Network Efficiency: 80.23%
Available Power: 216.6MW
Thrust: 4134kN


D Band at Ap, 3.75m Receiver:
Facing: 35.491%
Spotsize: 3.2m
Network Efficiency: 2.98% (*note the high water/atmospheric absorption here)
Available Power: 8.057MW
Thrust: 71kN


Ka Band at 10KM, 3.75m Receiver:
Facing: 70.124%
Spotsize: 719mm
Network Efficiency: 92.22%
Available Power: 251.2MW
Thrust: 3856kN

Ka Band at Ap, 3.75m Receiver:
Facing: 31.658%
Spotsize: 6.87m
Network Efficiency: 27.75%
Available Power: 74.92MW
Thrust: 121kN


Ka Band at 10Km, 2.5M receiver:
Facing: 49.510%
Spotsize: 528mm
Network Efficiency: 92.45%
Available Power: 249.6MW
Thrust: 3673kN


Ka Band at Ap, 2.5M receiver:
Facing: 36.405%
Spotsize: 5.48m
Network Efficiency: 31.07%
Available Power: 83.89MW
Thrust: 448kN

 

 

Appendix - Beamed Power Data:

This information shows the impact of wavelength selection on spot size and network efficiency for relays set up at 686km orbit height and 6,876km orbit height. Generators all produce 85GW natively, all using Phased Array Transceiver (Sphere). Receiver shuttle is orbiting at 86km or 72km (forgot to standardize but it doesn't skew the results too much.

Takeaways:

1. Pick an appropriate wavelength based on your relay positioning and craft orbit heights. This illustrates that.
2. You can see that if I set up a close to kerbin / HKO relay network for LKO use, I get a huge efficiency with Near infrared - but practically, Long/Short/Near all work as they cap out the available power (42.11MW) due to receiver size, etc.
3. For a higher orbit relay to provide deeper space coverage, Red Visible and Near infrared are both fine options.

 

Orbit	Type	Wavelength	Spotsize	Transmit Power	Available Power	Network Efficiency
686km	Ka	8.5655mm	160.37m	67.11MW	1.041MW	1.55%
686km	Far Infrared	33um	6.06m	61.4MW	25.18MW	41.01%
686km	Long Infrared	11um	1.96m	57.83MW	42.11MW	99.92%
686km	Short Infrared	2.2um	378mm	54.26MW	42.11MW	99.89%
686km	Near Infrared	1.05um	204mm	50,69MW	42.11MW	99.97%
5286km	Red Visible	700nm	1.31m	47.12MW	42.11MW	99.46%
5286km	Short Infrared	2.2um	4.11m	54.26MW	31.08MW	57.29%
5286km	Near Infrared	1.05um	1.96m	50.69MW	42.11MW	99.44%

686km Transmitters


6876km Transmitters

Edited January 21, 2022 by Voqk

[ssd21345]

Noice guide!

Quote

I believe this is due to the relative densities of the different fuels? Not sure. Regardless, in practical terms, pick the right solution for you. With hydrazine you get an excellent thrust and delta/v at a high cost. With LFOx, you get a huge amount of thrust (so you could reduce the engine size, or power requirements) and small cost with lower delta/v. Liquid fuel seemed a good balance - low thrust (but plenty, in my application) and best in class delta/v with low cost.

Yes, hydrazine is a high volumetric energy density liquid fuel. Think of OP version of liquid fuel/RP-1

Quote

I was having trouble make nuclear engines work for me. Their relatively low thrust (NERVA, for example) made timing transfer windows a pain in the ass, and capture burns (to Moho, for example) a bit scary as it took so long to decelerate for capture. Note that making a burn over 5 minutes for transfer is pretty much impossible, you very easily end up burning into kerbin's atmosphere as your manuever node at the start of the burn may be radially in to make sure you're resulting trajectory is correct.

Freethinker maintained persistent thrust since kspie warrant this kind of burn. They're low thrust but still much higher than ion engine, so it is kinda middle ground for isp and thurst. Yes, similar to stock ion engine, you have to start burning at high orbit and burning step by step. Also check out Brachistochrone Trajectories if you have high isp engine:

Spoiler

 

 

Edited January 21, 2022 by ssd21345

[Voqk]

4 hours ago, ssd21345 said:

Noice guide!

Yes, hydrazine is a high volumetric energy density liquid fuel. Think of OP version of liquid fuel/RP-1

Freethinker maintained persistent thrust since kspie warrant this kind of burn. They're low thrust but still much higher than ion engine, so it is kinda middle ground for isp and thurst. Yes, similar to stock ion engine, you have to start burning at high orbit and burning step by step. Also check out Brachistochrone Trajectories if you have high isp engine:

  Hide contents

 

 

Looked into Brachistochrone - this would work, although I wonder about the maths for a two-burn hohman transfer for more delta v efficiency. Will see if I can figure that out and make a new post.

Edited January 21, 2022 by Voqk

[kurgut]

On 1/21/2022 at 10:39 AM, Voqk said:

Looked into Brachistochrone - this would work, although I wonder about the maths for a two-burn hohman transfer for more delta v efficiency. Will see if I can figure that out and make a new post.

Hi,

Thanks for those insights and tips you gathered and neatly shared here ! That was very pleasent to read and go through.

I apreciate seing the processes and toughts you went trough, and this is really what makes KSPIE unique in that sense, experimenting !
This is really a good guide someone should link to beginners, as the tought process is easy to follow, and there's plenty of good tips along the way  

I am myself quite experienced in KSPIE, not a pro, not a noob, but this mod needs this type of content shared for the community, please continue !

 

A quick note about TWR/transfer burns/vessels design :

In general, you could go with a much, much lower Vacuum TWR, to improve your efficiency, and thus increase your available payload fraction on your spacecrafts. eg. 0.17 TWR for you Ion probe is huge!
You could halve this, at least. Same on the Jool mothership, could be a 2 to 3 times lower. But I understand the blocking point for you here was those low twr interplanetary transfers, and maneuvers (insertions burns, etc), since they require long burn time, and are complicated, at first glance.

Yes it's very ineffective, and if you lower your TWR even more as I advice you, It becomes impossible to eject in one pass... 

That's where the science of maneuver splitting comes and shine ! 

Basically, start ~20 days before your ejection burn. Plan it. An then subdivide into ~200 to ~300 m/s burn at periapsis, until you are close to eject (keep your final Ap bellow the Mun's SOI, or she's gonna mess with you), then for your last split burn, adjust its value, so actual ejection burn your planned comes back at the right place. Sorry this is a mess of an explanation. BUT, a very good KSP player made a super concise, clear and detailled guide on the subject, that I highly recommend you watching !

In my opinion, this is a fondamental technique to have in his pocket for KSPIE, as low twr/highly efficient are your bread and soup for quite some time, until more late game tech.. And once you start to understand, it becomes even fun to play with and plan !

May the Gods of ISP and Fission be with you  

Peace

On 1/21/2022 at 6:09 AM, ssd21345 said:

you have to start burning at high orbit and burning step by step

You actually lose a subtantial amount of Deltav and efficiency by doing this, ejecting from high orbit. Most efficient way is to burn as close as you can from the planet you eject from, so just above the atmosphere for Kerbin, to fully take advantage of the Oberth effect. And then split your burn into multiple passes of range going to 100 m/s to 300 m/s depending on your twr/isp/irl time ^^
Check just above where I linked a very clear and informative video about those precomputed Low-TWR interplanetary transfer burns  

cheers

Edited May 11, 2023 by kurgut