EatVacuum

I'm just thinking that complete sections with oxidizer and (whichever) fuel we go with would be more reusable than giving them a tank with fuel only, especially if the oxidizer is in the cargo bay section. That was why I was also thinking of providing liquid fuel plus oxidizer versions even if I do figure out the hydrogen resource thing - more reusability. Also - Design conversation in the forum is good because it gives us useful feedback from interested players. But I'm thinking we should have a contact channel other than going through the forum, both in case it goes down again and to allow us to exchange our WIP files at some point. I'll need parts to be able to validate the configs, and you'll need the vice versa. And if we end up having to get someone else to import your parts into Unity, they will as well. Speaking of, I did spend time trying to get a sample part into Unity and there is some learning curve, I haven't gotten it right yet, I guess it's more tutorial watching for me.

Now that the forum is back, I have a few thoughts about the parts list and earlier comments; I think this list is good, but a couple of thoughts. Some people will want this as a kit to fly a realistic Skylon, but it would be a bonus if the parts were also useful for building other craft. You're making a start on that by going with one of the existing part diameters rather than going with the exact 64% or 50% resizing. I think we should continue this a bit further - I'd like to keep the fuel and oxidizer in the same tank/part, but that said, it doesn't change where you cut the parts. How much fuel/oxidizer goes into a part is controlled by the config file, not the actual volume of the part. We also only need to reach 40% or so of real world orbital velocity to get into Kerbin orbit, so having a bit less fuel than we could won't likely be a problem and should even leave a reserve for powered landings. Another thought, related to the previous - if you cut the nose cones etc. at points where the diameters match existing diameters then it would make the parts more reusable, for instance maybe parts like the nose core can be cut off where it reaches the 1.25m or 2.5m diameter point, even if it is a bit off from the logical divisions based on function. Same thing for the orbital maneouvering system, RCS and so on. I was going to suggest splitting the engine into two rather than three, but on second thought, doing it does allow someone to use the intake and engine without having to stick to the "banana" shaped overall shape, so go with that. It would also make it easier for people to model the Lapcat, since the Scimitar engine doesn't seem to have the banana curve. With the forum down, this suggestion may have arrived too late, but if not, give it some consideration. As always though, this is largely your project, so the modelling choices are best left to the modeller. Yep, looked at the video, that and Figure 11 "Suborbital deployment..." in the Skylon user manual definitely confirm the positioning of the pitch/yaw thrusters right at the bow and stern. That leaves the roll control up for question, but making separate attachable RCS for the wings and including roll thrust in the bow and stern would allow for both options. I'll look at that for sure. Depending on when .22 comes out, we may not have to do that. If it should happen that different fuel technologies are part of the R&D tree (and I think that would be a realistic option to allow for developing more efficient engines), then we may have more options than just going with generic liquid fuel or making a custom hydrogen (Krydrogen?) resource. If not, then I should provide both options - hydrogen-based fuel tanks and generic fuel. Now a last question - what colour should we go with? The black paint scheme shown in the RE videos is appealing, but should we consider going with a paint scheme that better goes the existing common colors? I expect that once/if Skylon becomes a reality, it would sport different colors for the companies that fly it, just like airliners, since it looks like they are expecting to sell them commercially. So we don't have to be tied to what Reaction Engines is using. Any thoughts?

To quote what someone earlier said about my speculation on where to put the thrusters "That sounds suspiciously like an opinion...". I haven't come across anything stating that the fuselage is a lifting body, but if it is, we can do it. On the other hand, I did come across something that indicates that Skylon has a pretty high takeoff speed in this forum; "With a takeoff speed of M0,5 (over 250KIAS) and rocket engine like thrust, it should climb... almost like a rocket". If this is correct, at Mach 0.5 (actually 330 KIAS, knots indicated airspeed, which is about 350mph) Skylon would be going more than twice as fast as a typical airliner, or almost twice as fast as a typically combat loaded F16 at takeoff. That will go a long way to making up for the small wings. But I haven't found anything official that confirms the forum quote is correct. So if anyone can prove either of these statements, we'd appreciate you providing the source. In the meantime, in case we do need to go down that road... It's a simple change to the config file, I can probably make any KSP part into a lifting body. As proof I offer the amazing lift-cubes; all the lift of a delta wing, but much more space efficient. I hope to market them to any Kerbal nation that uses aircraft carriers or is plagued by narrow hangar doors... And in case anyone is doubting the lift is provided by the modified QBE probe cores, you will note in the SPH shot below the Centre of Lift is clearly on the cubes. The flaps are needed for flight control and the 1x1m structural panels are only there because I needed something to attach flaps to, they wouldn't go directly on the cubes.

Definitely, but there's a couple of obvious ways to do RCS, depending on the geometry of your space craft. For a relatively short shape like a sphere or a relatively stubby cylinder like an Apollo service module, you can just stick a bunch radially around the middle/waist, usually four works well because of our tendency to think up/down. left/right, forwards/backwards - it's how we define space and how our controls tend to work, from joysticks on up. Four RCS allows them to work in balanced pairs for most of the combinations of rotation and translation, except for forwards/backwards and roll where all four can be used. Something long and thin like Skylon has a higher moment of inertia (don't flame me if I'm using the wrong term, university physics was a long time ago) for yaw and pitch. So it makes sense to have sets at each end to give you more leverage to rotate the thing. Given that the Skylon is proportionately much longer and thinner than a shuttle, it's pretty much going to have to have RCS at the ends. On the other hand, those wings are stubby, so you'd think that if the Space Shuttle didn't need wing thrusters, then Skylon wouldn't either. But there's a couple of heavy engines going on the ends of those wings, which is going to increase the roll inertia, which wasn't the case for the shuttle. So I'd think that Skylon might need them after all. That's my opinion, but it's based on some consideration and analysis, your mileage may vary. It's likely that the wing RCS are going to be implemented as separate components, based on Cpt. Kipard's parts list earlier in this thread, so I guess you will have a choice in any event.

The pictures and explanation are great, that helps a lot. You just hit another big concern, those wings are tiny compared to the size. I haven't figured out exactly how lift works, but getting this thing off the ground on the existing runway might be a challenge. To add to it, KSP rocket fuel tanks are heavy foor their size, this thing (and a lot of the new rockets) are using hydrogen which is very light compared to kerosene and the other historical fuels. Compare the relative size of the Skylon's hydrogen and oxygen tanks here (expand the picture next to the "Material Construction" title) with that of the Saturn Five where the oxygen tanks are double the size of the RP-1 (Kerosene) fuel tanks, here. Admittedly part of that huge difference is because so much of the oxygen needed (about 250 tons) is used in the air-breathing stage, but you can look at other hydrogen-burning rockets to see. Note to self - in the .cfg files put in a lot less fuel than usual for the equivalent size, and also, don't use the usual 11/9 ratio. I am intrigued by the OMS system - from the description it doesn't sound like just a typical RCS system (which is generally a single propellant gas under pressure or a hypergolic mixture), sounds more like a lower thrust conventional rocket motor. The blurb describing it, found on the Skylon link above says; "Whilst in orbit the main propellant tanks are vented and allowed to warm to ambient conditions. Propulsion and attitude control are provided by the Orbital Manoeuvering System (OMS) or Reaction Control System (RCS). This uses a common LH2/LO2 propellant storage which is heavily insulated and cryogenically cooled. This system can remain operational on orbit up to 7 days. The RCS employs gaseous propellants supplied by the Gaseous Propellant Supply System (GPSS). The GPSS also supplies reactants to the fuel cells and the auxiliary power turbines." The second sentence makes it sound like the OMS and RCS are the same system, but then the third and fifth sentences imply they are separate. From the point of simulating it in KSP, I think we need to treat them as separate. It's also interesting that the main tanks are allowed to vent, it may be that only the auxillary fuel tank, which is very small, will have fuel for deorbit, unless it's cold enough up there that they still have some usable hydrogen and oxygen after it comes up to ambient temperature (and presumably vaporizes). That also implies a completely unpowered landing, which is pretty hard to do in KSP. Powered space plane landings are challenging enough. I would have thought 50% would be the scale, based on Kerbals being 1m tall, here's good ol' Bill standing next to a 1.25m tank and Jeb's Big stick (a measuring tool someone made). The thinner lines are .1m, the wider one second from the top is the 1m mark. Bill comes up to the 8th 0.1m line, but there's about .2m below the stick. So definitely 1m tall, but I guess most people aren't 2m (6'6") tall which 50% scale would make Bill if we rescaled him. But on the other hand 64% only makes him 1.56m (5'1"). I guess 64% is good, and if it seems to be the standard, it works for me. Once you have built the models we could always use scale and rescale in the configs to make 3.75M version, or even smaller. And related to your ponderings on where to put the thrusters, there is a convention described in the Wikipedia article on RCS; "Location of thrusters on spaceplanes The suborbital X-15 and a companion training aero-spacecraft, the NF-104 AST, both intended to travel to an altitude that rendered their aerodynamic control surfaces unusable, established a convention for locations for thrusters on winged vehicles not intended to dock in space; that is, those that only have attitude control thrusters. Those for pitch and yaw are located in the nose, forward of the cockpit, and replace a standard radar system. Those for roll are located at the wingtips. The X-20, which would have gone into orbit, continued this pattern." If we find out otherwise, we can correct, but I don't think we will go far wrong is you go with this as a basis. So, I am thinking the fuselage itself should break down into six components, from the bow back - small nose cone w/rcs, big forward fuel tank, cargo bay, big aft fuel tank (including RCS fuel), a thin aft tail section w/RCS, a small OMS rocket. There'd also be two more wingtip RCS units. The more I look at this, the more I realize how huge a job modelling this is going to be, what with animating the cargo bay doors, retractable under carriage etc. I guess I do have the easy part

You are right, they are different engines, a lot of shared technology but different. I'm not sure why the Sabre is curved the way it is, you'd think the airflow inside the engine would work better if it was symmetrical/straight. In any event, the geometry is going to make it challenging for you. The wings and engines do appear to be below the centre of mass, but I'm wondering if the downward tilt of the thrusters actually lines them up with the CoM. If they do, I think you are right that the Skylon will always be flying a bit nose down when it is in space. On another note, I'm a bit worried about the positioning of the wings and canard - I tried building a smaller but proportionate space plane using mostly stock parts, and with a canard that far forward, it dragged the centre of lift forward of the CoM, and that makes for an unstable plane. Of course, the stock model was only a rough approximation of the real thing but it is something I'll worry about until we can test fly. I was thinking about scale of this thing - if you build the model true to real world scale, it will be enormous, the hull will be more than 5m radius, i.e. bigger than the Nova Punch rocket parts. On the other hand the Kerblanauts are less than half the size of a human, so maybe we should consider making the scale to be proportionate to the size of the Kerbals not to conform to real world measurements? I don't want to end up with something like the Buran and Canadarm in the Robotic Arms pack - they seem to be made to real world scale and the result is the Canadarm is taller than most of my rockets. Right now I am frankly simulating Sabre by using an air intake to collect air and convert it to oxidizer. My engine is always burning oxidizer but I'll quote Aerospaceweb.org to justify this and then I'll explain the actual process; "The purpose of both the jet engine and the rocket engine is to combust a mixture of fuel and oxidizer. This combustion process generates a high-pressure exhaust that creates thrust to push a vehicle forward. The fundamental difference between the two types of engines, however, is where the oxidizer comes from." So in reality, a rocket can be thought of as a jet engine that gets it's oxidizer in liquid form, or alternately, a jet engine is just a rocket engine with a hole in the front that gets its oxidizer for free. I'm using an air intake to produce and feed oxidizer to a rocket engine. How it is done is as follows; 1) The air intake, while open takes in air and converts it to oxidizer. This is purely a game mechanic, air is oxidizer, or to be accurate air is 20% oxygen and oxygen is by definition an oxidizer. The rate is such that one intake can feed one engine up to about 22km, after which it falls rapidly behind. At that altitude you should close the intake to reduce drag since the oxidizer collected is too little to be useful. 2) The fuel feed mechanism of KSP delivers the oxidizer to where it can be used, either to the rocket motor to make thrust or to a fuel tank. 3) The rocket engine is always burning oxidizer, initially the intake provides enough but eventually it has to start using the oxidizer stored in the rocket fuel tanks. I've modified the engine's atmosphere curve to be equivalent to the Sabre's - a jet-like Isp of 3240 up to about 22km and a rocket-like Isp of 420 from 23km up, with a brief transition between the two Isps in between. A purist might point out that I'm really simulating a LACE-type engine (Liquid Air Cycle Engine), but the net effect is the same as if the air was actually being consumed directly. Sabre super cools the air to allow it to function at such height and speed, but it doesn't actually cool the air down to liquid like LACE does. If you compare the Isp and altitude numbers to those quoted for Sabre, you'll see I've cut them back by about 10%. Kerbin's atmosphere height and orbital velocities are smaller than Earth's, using the full values makes it more effective than I think it should be. I'd like to have a true jet/rocket hybrid, I've been looking at Careo's HydraEngineController which is in the B9 pack. Using it would give me a true dual mode engine, but I haven't found any licensing info for his stuff in B9 Aerospace. Until I do, I won't rip off his work, hopefully he'll be willing to let us use it in return for giving him credit.

Cpt. Kipard, I responded to your last post in the "[REQUEST] Skylon & Saber Engine - Reaction Engines Ltd" (sorry, I haven't figured out how to put in the URL to link the title to that thread), if you're good with it, I will join you here. Whether or not you agree, I do have a couple of thoughts; I'm still barely into understanding how Unity works, but I believe that the actual orientation of the models does not control how thrust, lift or drag works, they are defined by transforms that are attached to the models and which you point in an appropriate direction. If so the thrust could be offset to align with the centre of mass, I'm pretty sure that since they will have to account for that in the real Skylon, it can be achieved in KSP. If nothing else, thrust vectoring might be able to compensate. I can look into that if you like, and if so, then you can make an accurate model. But if there is anyone out there who is more familiar with Unity who can explain this, and how to do it, I'd love to know for sure. They probably haven't figured out the RCS yet, the images and videos published are promotional materials, Reaction Engines is, as the name implies, interested in building a hybrid engine. I would assume that the Skylon and LAPCAT images are just concepts to show what it could be used for. Where the thrusters will go is going to have to be based on where the Centre of Mass ends up. Some non-intrusive RCS units could be another thing to add to the list of models you'll need to build Skylon. Like the engines themselves, I'd love to have those - it offends my sense of aesthetics to make a sleek, well proportioned space plane and then have four objects that look like four-way lawn sprinklers sticking out where they would screw up the airflow. Take a look at this image - http://ca.photos.com/royalty-free-images/space-shuttle-discovery-ov-103-nose-detail-port-left-side/150495422 As you can see, on the space shuttle the thrusters are streamlined, just holes in the fuselage, not "lawn sprinklers". I'm pretty sure that's the way they'll be going for the Skylon. If you look at some design images for the shuttle it should help you to figure out where to place them. Hope this helps.

As mentioned B9 does do the Sabre engine which will hopefully one day power the Skylon, although it is a big pack to download if you only want the Sabre (70MB, 176 parts, a handful of .dlls). You can read my rather lengthy thoughts about that on a Sabre thread here - http://forum.kerbalspaceprogram.com/showthread.php/23728-REQUEST-Skylon-Saber-Engine-Reaction-Engines-Ltd?p=647091#post647091. But B9 has a lot of nice stuff so by all means get it if you have a use for all the other parts. There is also a rather nice Skylon model already produced by a guy named Lando, available on the Space port here - http://kerbalspaceprogram.com/skylon-pack/ although you should note the developer's comments about the engine still needing some work. The engine's also a bit too effective as a rocket engine, it has the same Isp as the LV-N at 1/3 the weight and with 3 times the thrust, and that Isp is almost double what the real Sabre (and any other chemical rocket) can produce. But if that doesn't bother you, his model looks pretty nice.

I'm not a big fan of doing all sorts of math to figure out the launch angles but I do use two tools. First is a phase angle chart that is out there. The second is the Alarm clock addon. It can be programmed to tell you when a transfer window is coming; I like to leave myself a day's warning. When that pops up I launch my ship and while I wait for the window, I use the maneouvre node thingy to figure out the trajectory. Like going to the Mun you want to leave orbit when Duna is coming over the horizon to gain the advantage or Kerbin's orbit and rotation, so place the node at that point in your orbit. Then pull the prograde arm until your planned trajectory reaches Duna orbit. You won't likely get an encounter right away but you will probably be close enough that you will get the closest approach markers. At that point you can start adjusting the prograde/retrograde arms and the two arms that point towards and out from your orbital plane and see what effect that has on the closest approach markers. Keep fiddling until you get an encounter. Any encounter will do, you can fine tune doing course correction when you are out of Kerbin's SOI. And don't forget you can also grab the maneouver node and drag it prograde and retrograde in your Kerbin orbit to adjust your trajectory. Once you have it set, you can do your ejection burn using the navball to guide you. Or if you use Mechjeb, just hit "execute next node" and away you go.

I usually have the opposite. When I get into the SoI of the planet I often see a dog leg trajectory that bends well inside the planet surface. I have to burn to get my trajectory up to 50 or 100 km.

It's hard to tell from the picture but there seem to be a lot of struts for something that is essentially stationary and shouldn't undergo much stress. I believe using the docking port sr (the big one) would reduce wobbliness allowing you to remove many struts. Removing the RCS tanks and thrusters is also a good suggestion I count 9 sets so that alone is 45 parts gone. Anything that is duplicated or in higher multiples is a good candidate for cutting back.