sevenperforce

No way an Ant can sport the 4700 m/s that a minimal Dawn stage has.

The Spark has more thrust than I need, but I've been using it to burn off the Juno. Maybe it's worth the mass loss to add a decoupler. **looks** Yep, the mass of a decoupler is less than the mass difference between a Spark and an Ant. So I may swap them out and see if I can ditch one of those donut tanks entirely. Also you can save more mass on your current vehicle by attaching the cubic strut to the lower stage, attaching the prop decoupler to it, and then rotating/translating it up to the top.

To be clear, you can certainly do it -- it's just that control surfaces are less massive and more effective. If your projectile is hypersonic then traditional control surfaces don't work quite as well. You can use grid fins but they are draggy. The preferred guidance method is always going to be vectoring the main engine, either through gimbal or some other design. The Sprint ABM, mentioned upthread by fellow physics god @p1t1o, used inert fluid injection inside the main engine nozzle to vector thrust at 100-gee accelerations. It was very successful. The one place where an ABM (anti-ballistic missile) might be able to utilize cold-gas thrusters would be in unpowered terminal guidance. If your missile is designed to boost to hypersonic velocities and then coast to impact/detonation, then it's going to need to maneuver. Traditional control surfaces are draggy at hypersonic velocities and require extremely heavy/strong actuators; grid fins are even draggier. So using cold-gas thrusters would actually be a good idea. This is what SpaceX does with its fairing entry, actually. Its fairings hit the atmosphere at over 2.3 km/s or about Mach 7. They have no control surfaces to maneuver, so they use nitrogen cold-gas to control the orientation of the entire fairing as a lifting body. If you want to use this for an ABM with a terminal coast phase, you may want to look at base bleed projectiles. A major component of drag on artillery shells is the low-pressure region which forms behind the base of the shell. Base-bleed projectiles add a small, weak rocket/powder cartridge just behind the base which produces a flow of warm gas. By itself it wouldn't produce much thrust, but increasing the pressure in that region reduces base drag and extends range. An ABM with a terminal coast phase could benefit by using cold-gas thrusters embedded in the tail of the vehicle (to reduce tail drag); it could also use these thrusters differentially to produce torque and guide the warhead to its target. Might be something worth looking into. If you do, you'll want to start by looking at the lift-to-drag ratios of cylinders and cones at hypersonic velocities to get an idea of how much force it would take. Fortunately NASA has studied this in detail. They could be used both inside and outside the atmosphere because they fire through the center of mass. Accordingly, they act like the divert thrusters mentioned above. Divert thrusters are actually less efficient than gimbal steering, but they work the same way at any velocity so it's not an issue. I have noticed that folks here are very helpful. I think KSP is just complicated enough that it prompts people to get better at explaining rocket science, which makes them good at, well, explaining rocket science.

Mother of god, that was impressive. I've never used anything from Breaking Ground but maybe it is time.

Hat trick: if you flip a 1.25-m circular intake and slap it on the back of a fairing, then slide it up almost all the way into the fairing, an engine mounted inside the fairing can fire through it without injury. It also completely kills the tail drag of the fairing, which is the important thing. I hit 680 m/s on the Juno at 8 km before I fired up the Spark, blew off the Juno, and pulled up under power using props in the fairing to get my apoapse to 50+ km. Then I used the decoupler to blow off the fairing and burned what remained of the Spark to get just shy of a circular orbit.

Good idea using the Twitches -- they have better SL isp. On my end, I just managed to go interstellar on 1.812 tonnes using a Dawn on top of a Spark on top of a Juno. It took a little cleverness with a drag-killing intake, though. It has 800 m/s to spare past interstellar velocity but other than just reducing some of the initial propellant load I'm not sure what else I could do.

No maximum payload. If you want a bigger payload just bring a bigger rocket.

You are correct, my bad!

Rail launch is not great for human cargo. As a rule, an airbreathing lifting-body SSTO is a poor choice for crew to begin with because there's no good abort mode when you're pointed face-first into what is nearly a re-entry environment. The original design uses high-energy-density fuel, with liquid methane and aluminum gellant. If you're swapping in nukes you no longer need high energy density. What I'd do, if the government gave me a bunch of nukes to roll in, would be to replace the LOX tanks with water tanks and feed the nuclear-thermal engine with a mixture of water and regular liquid methane in varying mixture ratio so I could get high thrust (with water) at the beginning and high specific impulse (with methane) at the end, varying as needed. You'd need multiple injectors because of temperature issues but otherwise it works very well. And yes, much bigger. More like the shuttle layout -- maybe with the crew in the nose having canards and being an independent re-entry vehicle.

Yeah, that was the impression I got It seems like it could put a lot of torque on the joint/hinge, but I suppose if you can make it work under tension then that's a positive anyway. I wonder how to get larger spread.

Grid fins are draggier than traditional control surfaces so they are no good for boost but they are suitable for something falling if you don't mind it slowing down. They have high control authority without stall risk and with minimal hydraulic requirements. And yes, I know the N-1 used them. Still didn't make it a good idea.

Seems like a LOT of torque on those hinges. With Falcon 9 the force is distributed between the hinge and the piston.

You need 1.8-2.1 km/s to circularize from GTO to GEO and then you need 1.4-1.8 km/s to spiral from GEO to EML-1. Lots of essentially zero-energy transfers from EML-1 to EML-2 and NRHO. So around 3.55 km/s. Six months of burn time for the HALO (plus time for the transfer to NRHO), 20 months (and 5 tonnes xenon) for a 30-tonne landing module.

If you induce rotation via RCS or control surfaces, the angular momentum will resist changes in pitch. This is how projectiles and certain rockets are spin-stabilized. If the center of pressure or center of mass changes during rotation, it will cause precession of the axis of rotation, like how a gyroscope acts weird. The axis of rotation will still pass through the center of mass. If you have control surfaces, of course, the atmosphere will damp the rotation rather rapidly. Think of it as a feedback chain. You fire the RCS, which produces a perpendicular force displaced from the center of mass, which induces rotation around the center of mass. The rotation changes the angle of attack. The new angle of attack produces induced body lift. Induced lift at the center of mass changes the overall direction while induced lift distant from the center of mass attempts to push the cylinder back to a zero AoA. That push is what pushes back against your RCS. If you keep firing your RCS, the AoA will increase until the resistive force of that induced lift is equal to the thrust of your RCS and it will then hold that AoA, causing a continuous turn. Think of it like turning the helm of a sailing ship. The harder you turn the wheel, the more the rudder will angle and the more flow of water will push back against the rudder. The ship will turn as long as you keep applying force (the more force, the more the rudder is angled and the faster it turns) then start going straight again once you release the wheel. Using RCS for aiming, then? In atmosphere you can use RCS for translational/diverting thrust but it is more mass-efficient to use guide fins to steer. It's happening in zero gee free fall with no atmosphere. No airstream means no change in the angle of attack because there's nothing to attack. Puffing the RCS begins rotation as before, but without atmosphere the rotation continues at a constant rate until a puff of opposite RCS stops it. The kill vehicle is hovering on a thruster so it is not moving. Each of the RCS thrusters used to change position are pointed through the center of mass, so they nudge it in any direction desired without inducing any rotation. Because it's hovering, there's virtually no friction, and so even a very small puff of RCS will produce sufficient force to move it; the question is how quickly you want it to move around. You need more powerful RCS to make quicker movements back and forth.

Just made orbit on 3.469 tonnes, 55 parts, 7 stages, 40 Seps. Two fairings and a minimum-mass final stage did the trick. Could probably shave it down to 38 seps but will probably end it here. Incidentally this is the first KSP vid I've ever made.

Using ion engines to launch a craft from LEO to TLI is a bridge too far, the final stage of the LV would be preferred, or the upper stage of another LV, separately launched and docked in LEO. I discussed this once with @tater before in the context of delivering landing modules. The time for the PPE to spiral out from LEO to the moon using ions while carrying substantial mass is something on the order of a year of continuous burn time. Assumptions: 7 km/s from LEO to EML-1 using low-thrust spirals PPE propulsion: 4x AEPS Hall thrusters at 14 kW consumption, 2600 s isp, 600 mN each Propellant: 5000 kg xenon System dry mass (estimate): 3 tonnes HALO mass: 3.4 tonnes Maths: Total solar-electric dV: 14,710 m/s (good so far) Propellant consumption at max thrust: 94.13 mg/s Propellant usage to reach 7 km/s: 2.738 tonnes Burn time: 2.71e7 sec or 336.66 days Not undoable -- Hall effect thrusters have built up 8000-hour burn times in testing -- but challenging. For larger modules it is prohibitive. A notional 30-tonne multistage Artemis crew lander would require 10.4 tonnes of xenon and would take 42 months of continuous thrust, or around 3.5 years.

Hot damn, that's nice.

The difficulty is that while tilting a rocket seems conceptually similar to changing its angle in flight, they are fundamentally different problems. A stationary, standing cylinder is "tilted" by a perpendicular force applied above the center of mass. The force must be large enough to rotate the cylinder and lift one edge of the rocket off the ground against the force of gravity, using the opposite edge as a fulcrum. Applying perpendicular force at or below the center of mass, in contrast, will only cause the rocket to slide laterally. The tipping point at which gravity takes over and brings the cylinder smashing to the ground is determined by the point at which the rotation of the cylinder moves the center of mass past the fulcrum point. Gravity is the primary actor being considered here. Changing the orientation of a cylinder in flight also involves a force applied perpendicularly, but everything else is different. There is no interaction between the cylinder and a stationary surface, so you are no longer pushing against gravity -- in fact, gravity is completely irrelevant to orientation (apart from incidental things like the force of gravity determining the density and pressure of air, etc.). The fulcrum is now the center of mass of the cylinder. Normal force applied at the fulcrum produces lateral displacement (like the diverting thrusters in the examples shown); normal force anywhere else rotates the entire cylinder around the center of mass. The concept of "tilting" the rocket is nonexistent. During boost phase, a rocket is intended to fly at a zero angle of attack whenever possible. A nonzero angle of attack results in lift-induced drag which slows the rocket, which you don't want. The only time the angle of attack deviates from zero is when the rocket is trying to change direction, either using vectored thrust, guide fins, or RCS. To change direction, you want to use the air stream to "turn" the rocket in flight; you do this by using control surfaces (or vectored thrust or RCS) to produce a nonzero angle of attack, which produces a lift force which turns the rocket. A rocket needs to keep its center of mass as far forward as possible and its center of pressure as far back as possible in order to remain aerodynamically stable (think of a shuttlecock as the extreme example of mass-forward and pressure-rearward). So it usually make sense to put control surfaces at the back. What type of project are you doing?

The issue will be not acceleration, but forward airspeed. The body of the rocket acts like a gigantic lifting body. Trying to push it out of alignment is going to produce differential lift which forces it back into the lowest air-resistance orientation relative to the airstream. Just after liftoff, the forward airspeed is negligible so the required reaction force from the side thruster is essentially the same as it would be at a standstill. In fact, it may be lower, because it does not need to produce tipover (tipover requires the center of mass to be moved beyond the base). But as it accelerates, the forward airspeed increases, and the required force will increase.

Helpful terminology... Jet. Any engine which uses the expansion of a working fluid from a nozzle to directly produce thrust via Newton's Third Law. Combustion Pulsejet. A jet engine which uses a simple inlet to mix air with fuel to create repeated explosions for pulses of thrust. Combustion Turbojet. A jet engine which uses an axial turbocompressor to compress and mix air with fuel to burn and create continuous thrust. Turboprop. A turbojet engine which gears its axial turbocompressor to a propeller to create secondary thrust at low-subsonic airspeeds. Bypass Turbofan. A turbojet engine which gears its axial turbocompressor to a large ducted fan to create secondary thrust at high subsonic airspeeds. Combustion Ramjet. A jet engine which uses shock cone intake geometry to trade inlet air's speed for compression to combust at subsonic speeds. Supersonic Combustion Ramjet or scramjet. A ramjet which uses a wider shock cone intake geometry to allow compressed combustion at supersonic speeds. Ejector jet or ducted rocket. A rocket with an air inlet and duct. Atmospheric air is entrained by the exhaust's low static pressure, then compressed and heated by its high dynamic pressure to expand against the duct as additional working fluid. Turborocket. An ejector jet which uses an axial turbocompressor to compress air at the inlet before mixing with the rocket exhaust. Ramrocket. An ejector jet with shock cone intake geometry to compress the air to subsonic flow speed before mixing with the rocket exhaust. Scramrocket. An ejector jet with wider shock cone intake geometry that compresses the air, but not so far that it reaches subsonic flow speed. Rocket-Based Combined-Cycle (RBCC) Engine. Any engine which operates in different modes (ejector jet, ramjet, turborocket, etc.) at different speeds and altitudes. Also, I love this design: http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.526.7376&rep=rep1&type=pdf Uses jelled aluminum in liquid methane to create a high-density, high-energy propellant. Sled-launched RBCC SSTO that operates as an ejector jet, a ramjet, a scramjet, a scramrocket, and a pure rocket.

You are describing what is colloquially described as a scramrocket or more properly a supersonic-bypass ejector rocket. Minor correction: a scramjet constricts airflow, but not enough to cause a normal shock which renders the flow subsonic. The combustion takes place within the supersonic airflow -- hence "Supersonic Combusting Ramjet" or SCRamjet. This is not the image I was looking for before, but I love this one:

Reading this and wondering if it will be possible to bring an asteroid to the KSC and then return it to LKO. In before Bradley Winstance does this in an SSTO that lifts off with Junos and lands on Duna......

Got it down to 2.794 tonnes with three stages. Got about 400 m/s to spare on the ion stage after reaching Kerbol escape but not sure if that's enough margin to swap out the Wheesley for a couple of Junos.

I wonder if they would ever go "full-expendable" on the upper stage to deliver a particularly hearty monolith to LEO then "refuel" the empty to recover it. I wonder how much performance that would yield.

Bravo! I am just over 3 tonnes using a Wheesley first stage.