hornetfinn wrote:It's because the cockpit glass is such a shape that it redirects the reflections away from the radar.

... The reflecting area of the cockpit glass is extremely small compared to the geometrical area as only very small amount of the radar energy hitting it will go back towards the radar.

I think it's not quite true because we are not in the region where the laws of optic propagation act.

In general, the the waves of any length (not only optic band) undergo the Huygens–Fresnel principle

https://en.wikipedia.org/wiki/Huygens%E2%80%93Fresnel_principleAccording to it, every point which a luminous disturbance reaches becomes a source of a spherical wave; the sum of these secondary waves determines the form of the wave at any subsequent time.

This means that we may consider the canopy metal layer as a great antenna, which reradiates incident power.

It has its main lobe and side lobes just like any anttenna.

The mainlobe width is dependent on relatioship between antenna size (aperture) and wavelength.

The mirror in optic band, for example 1 meter long, will give the ratio near 2000000 times (optic waves are about 0.5 micrometer), and the beam is extremely narrow. And the sidelobes far from the main are extremely negligible (due to sin(x)/x function)

But what about X-band? For 1 m aperture and 0.03 m wave the ratio is only 33.

Then we get a picture like this:

And what is the exact value of reflection? Look at this picture.

For tilted flat plane here is the RCS formula:

It is taken from here:

http://books.google.by/books?id=XDs04HdQ4-gC&pg=PA183&lpg=PA183&dq=tilted+flat+plate+RCS&source=bl&ots=dol7wl54FC&sig=27W3hnnBtGpPkvwbuF5zzoA2oek&hl=ru&sa=X&ei=obZhVKDtG4HkOKL0geAC&ved=0CCcQ6AEwAg#v=onepage&q=tilted%20flat%20plate%20RCS&f=falseWhy flat? We just cut a plane from a frontal part of canopy - 1 m long and 0.1 m wide - such a strip I think can be well approximated by flat plane.

Then we may remove all the rest of the canopy and all the rest of the airframe (though they also have a non-zero reflections)

and put the values in the formula (it seems the incidence angle is near 60 deg. to normal, A = the square of the strip)

So, we get the value of 0.00058 which is 5.8 times greater than reported overall RCS.

I want to remind that we nullified all other reflections from the airframe (!) and took just a narrow strip from canopy.

Everybody can calculate the square of circle, and that was done in 2005.

But also everyone can put the values in formula given above.

What you think guys?

My guess is that we've been reported a true minimal RCS value

which is fair for very narrow angle range. And a true average value is a definitely classified information and it should be much greater evidently.