An Example SSTO Design Process
This document is an attempt at a walk-through for a work-flow when building craft with FAR. The author is a novice and this may require some editing.
Your basic craft should pay attention to the basic needs of static stability described in various forum threads.
That is: you should place your center of lift a short distance behind your center of mass. You should also place your undercarriage in such a way that the back wheels can pivot near the center of mass. Be mindful of the impact of fuel consumption. If possible, set up your fuel tanks so that the center of mass will tend to move forward as tanks empty. Using tweakable fuel tanks, you can temporarily empty tanks and check how the center of mass moves in the SPH. Finally, set up your control surfaces, but you may want to reduce the amount of control so you don't make wild movements that will result in your craft disassembling from aerodynamic stresses.
Here is an image of the ship I created with this basic work-flow.
Once the basic set of components is placed on the ship, it's time to start tweaking parts with guidance from the FAR stability windows.
The first things to check are whether the craft will be stable over a range of sub-sonic speeds. If you can't get stability here, things are unlikely to be easier at supersonic speeds. You are basically looking for the yellow stability curve to always be negative at low (M=0.2) and high subsonic (M=0.8) speeds. You can also click the fuel empty button to make sure stability improves or stays the same as you empty the tanks.
There is some additional information in the dynamic stability window that is very useful early on: these are the angle-of-attack needed for level flight at different speeds. If you don't want to be trying to keep the plane straight at 180 m/s on the runway, it's a good thing to check these values. When you click "calculate stability", the velocity corresponding to your entered Mach number will be displayed (e.g. M=0.2 -> v=68m/s) and the AoA needed for level flight is displayed. At low speed, this could be quite high and you'll see ">25" listed. As you increase the velocity, you should see that AoA decrease. Now is a good time to try the "flap" buttons to see what happens when you actuate your craft flaps. Compare as well to full and empty tanks. If you can get your angle of attack down to 5 degrees and your velocity not much higher than 100m/s, you should have no problem taking off.
If you're having trouble with this, then you should consider adding more lifting surfaces to your craft.
Once you know you have something that works for low speed, you want to look at what happens over a range of Mach numbers. Switch up the numbers to shift from lower 0 to upper 5 and enter something like 3-5 in the Mach/AoA area then click "Sweep Mach". Watch out for the region around Mach 1 as this is often where things get hairy. Check different angles of attack (e.g. 10-20) to get a sense of what angles your craft could be happy at. You don't want an excessively small stability window or else you'll never be able to keep the craft going. Don't forget to check the empty values!
The dynamic stability window is kind of mysterious. Green is good; red is bad. Ideally, you will see nothing but green including when you run your flaps if you have them. If you get some red, then you have to determine if it is bad or tolerable. This is where the "simulation" window comes in.
When you see some red values, you have to find out if the craft can recover from the perturbations or if tends toward flat spin. Click on "simulation" and you'll see either the longitudinal or lateral stability values that were previously calculated. These are the values from your last stability calculation. Now, find the first red value and look at the second letter in the designation (e.g. if Mw is unstable, then you want to look at perturbations of w). In the entry windows, find your desired perturbation and put in a value of 5 or so. Then run the simulation and look at the results. If your curves are wildly diverging then the perturbation is not tolerable and you'll need to fix the instability. If the perturbation dies down after some oscillations, then you can probably live with it. If you aren't sure what happens after the initial perturbation because things are still happening, you can increase the time for the simulation.
If you do have to fix instabilities, there's a lot of tweaking required to find the right things to change. Often the pitching moments can be fixed by subtly changing the relative distance from the CoM and CoL. This can sometimes be identified quickly by clicking the full/empty fuel button if your CoM moves to a new spot when empty. Yaw instabilities can sometimes be addressed by a larger tail, more vertical tail, or moving the tail rearward. If you're just starting, look at the number that is red before you make the change and then after. If you see it change toward the correct value, you'll start building intuition as to what works and what doesn't.
One aspect of this is to change the density of the air for the simulation. As you increase altitude of the plane, there is less air for the various surfaces to press against and this can cause instabilities to arise. You will also want altitude to reduce the total drag on the vessel and achieve greater velocities before losing air-breathing engines. The density of the air can be adjusted in the dynamic stability window. This process will start to tell you how you will want to fly the craft as well. For example, you may want to avoid supersonic flight below 10km because stability works better with thinner atmosphere.
Fixing dynamic stability is where a lot of time can be spent. Instabilities at supersonic velocities can be tough to diagnose and adjust sometimes. There's a lot of trial and error in this process. You should go through your entire, expected Mach and altitude range to make sure you avoid any unwanted surprises on the way to orbit.
Flaps can be added to the control surfaces via their right-click tweakables menu. An action group will have to be set to actuate the flaps with one button increasing deflection and one button decreasing. The same goes for spoilers.
Flaps are generally added to increase lift. The dynamic stability window will show you the needed AoA for level flight and if you select various flap settings, you should see this number go down. Flaps increase lift at the expense of increased drag so on the static stability window, you could see the L/D ratio stay constant or even decrease even though the absolute amount of lift is increasing.
If you are unhappy with forward pitching when actuating flaps, you can set a canard to also work as a flap and use the tweakable flap strength to try and balance the main control surfaces. When you look at the CoM and CoL indicators, if you click a new flap setting and the CoL arrow rotates forward, then you will pitch downward. If, instead, it stays mostly positive, you are balancing the flaps correctly.
Spoilers decrease lift and increase drag. Their main use seems to be as part of a braking system after landing. You can set control surfaces to the middle of a wing to only act as spoilers by disabling the pitch/roll/yaw control for the surface and enabling spoiler only.
Be sure to check dynamic stability again with flaps at sea-level densities and landing speeds both full and empty. You don't want to flip over right after leaving the surface of the runway or on approach!
Here are some things to check out when you think you've got a viable vessel:
If you are picking up speed but the craft pulls or wobbles: your craft is not strong and needs stiffness. Struts can help provide this. Alternatively, you'll have to check how the parts are connecting to each other to see if there's a "weak" point in the chain that's ruining things. Be sure to fix wobbly gear otherwise you might be able to get off the ground, fly around and have a grand old time only to eat dirt when you try to land.
If you are turning too fast and your craft is breaking up, then you should get back in the SPH and use the tweakable control systems to decrease the amount of control. Note: it's easier to break up at low altitude as the density is greater and there's more air to bite into. It seems that if you can fly in such a way as to avoid 5-6g of acceleration, then you should be ok. Much more than that will likely result in aerodynamic failure.
If you feel bad about Kerbal test pilots getting used up consider installing Vanguard ejection and parachute system!
Be patient and have fun!