What you are asking require us to go back to the basics, which am certain is abhorrent to most people, here and on other forums elsewhere, because they usually come to these places with minds
ALREADY made up. So please bear with me...
The top line -- the smoothest one -- represent the sphere, or the diameter of a cylinder. Our Chinese members of this forum have taken this to mean that the sphere (shape) or the curvature (surface topography) are the worst in terms of RCS production.
NOTHING can be further from the truth.
What the sphere (or diameter on a cylinder) represent is
UNIFORMITY and
CONSISTENCY in terms of radiation. They are usually employed as standards for measurements and calibration of just about anything involving radar.
For the illustration above, we have one thing in common for those simple shapes: Surface wave inducers. And that the RCS graphs for each shape are non-rotational with the radar signal going 'left-right', if you will.
Tilt the plate enough and we will have an RCS that is far lower than the sphere when the plate will present only one edge diffraction signal: The edge facing the radar. But continuing in rotating the plate and eventually the plate will present the 'full Monty' to the radar, producing an RCS far far far greater than the sphere.
The same argument applies to all of the above shapes, not that all of them will produce an RCS greater than the sphere while under rotation, but the lesson remains: Uniformity and Consistency. In radar detection and data processing, variables and variations of any kind are natural attention attractants.
The ogive (oh-ghee-vee) may have a natural RCS lower than the sphere regardless of rotational aspect angles --
MAY because we have not touch size. But the ogive shape have three radiation modes: Surface, Specular and Edge, with four locations: two sides and two points.
Same thing with the plate except that when the plate is in 'full Monty' to the radar, there will be no surface wave behaviors whereas with the ogive, even if the ogive is completely perpendicular to the seeking radar, surface topography via curvature will induce surface wave behaviors and will deny the seeking radar some measure of detection.
For the double-rounded cone, we have three modes of radiation: Surface, Plate, and Edge. And these radiators: Two points, four plates, and two curvatures.
I will leave the other shapes as entertaining mental exercise for interested readers to figure out.
Keep in mind that the shape illustration is non-rotational. Now we will add in the '10-lambda' rule...
What the '10-lambda' (wavelength) says is that if the diameter (sphere or cylinder) is less than 10-wavelengths -- regardless of wavelengths -- then the 'creeping wave' behavior will occur. If the diameter is greater than 10-wavelengths, then the creeping wave behavior
WILL NOT occur. So for the diameter (sphere or cylinder) there will be a situation where the sphere will have only one radiation mode for the seeking radar: Specular. And this is regardless of rotational aspect angles.
Now apply the '10-lambda' rule to all of the above shapes while each is under rotation.
Now amplify
EVERYTHING above a million times because we are dealing with a complex body call an 'aircraft'.
An aircraft is a symmetrical but irregular body. People must understand this. Irregularity produces uncertainty which produces variations which will naturally attract attentions. This is inevitable for an aircraft whose surface topography contains rare instances of discrete of any of the above shapes but usually far worse: Combinations of those shapes.
The result is that
IF our goal is to control the behaviors of these radiation patterns, we must first understand the behaviors and we started with Ufimtsev. Since we cannot avoid the plate and its accompanying edges, aka 'wings' for example, we should try to contain their numbers and avoid placing them in clusters. This led us to the next rules: Containment of radiation modes. Avoidance of radiation clusters.
Now watch the bloodbaths between the radar and aerodynamic geeks.
Nowhere am I saying that the American 'stealth' aircrafts are the ones the world should go by. What I am saying is that if a foreign power want to enter the 'stealth' arena, it would behoove said 'stealth' fighter aspirant to return to the basics, study how we did it, self examine the technological capabilities, and give it a go. But do not think that the product is beyond critical examinations by observers especially when they have at least one generation of this technology as an unofficial standard to measure all 'stealth' aspirants.
These are burst data and while they can be intercepted, the bursts are so brief that at best they could be used as an warning, not as a locator.