Well hello! I had no idea I had so many fans. It's good to be loved.
Forgive me for responding to some of these points out of chronological order, but it made a little more sense to me to do it this way.
The examples in his writing was a loose screw on the Have Blue. If he were talking about smoothed out surface structures on the plane though he's really picking his cherries.
I can't find anywhere in my posting on SPF where I was talking about "smoothed out surfaces". I also did not mention air data instrumentation and the like.
The point is that the shape and size of these surface features, as well as its material composition and coating, are designed to limit radar return, and it's the overall effect of the airframe that counts and not its individual parts.
Actually, that is very much what I was illustrating.
Quellish's post argued that the J-20 had too many small protruding surface features that made it detectable with higher band radars, but it really doesn't have that many more surface features than the F-22 does, and like the F-22 those features are shaped.
The point of my post on SPF was to illustrate that the F-22 and J-20 are designed for very different threats, and very different missions. Comparing the two does not make sense. Please see below.
Ahh, Quellish's post on secretprojects. The only problem I have with his analysis is that he makes the claim that the J-20 has small details on its surface which hurt its stealth against higher band radars, but he doesn't back that claim with specifics, which poses a big problem in this case because all the pictures we've seen of the J-20 show that it has a very smooth flush surface.
There were some specifics in that post on SPF, and one of the points made was that the lower surfaces are very different from the upper surfaces (to quote myself: "Many of the surface disruptions seen on the upper half of the aircraft are not present on the underside."). This appears to be a result of optimizing the signature for specific aspects, which is not unusual. The F-22 is optimized for different aspects, because it has a different mission and faces different threats (that was the point of my post on SPF).
So if it has the "look", that means its shape has been designed to contain crucial elements that enhance its stealthiness.
Not until you measure and analyze it. The Testor's F-19 had "the look", but didn't actually test well.
Which makes His argument completely baseless.
Maybe I was not getting myself across clearly on SPF, but at the same time that post was written for a specific audience. Different readers may take away different things, I can't control that.
For reference, the post being discussed is here:
Can't post the link as I'm a new user, sorry.
To give readers some context lost here, in the thread on SPF there were comments asserting a belief that the J-20 stole or copied the canopy from the F-22. My contribution to that discussion pointed out that this would be futile, as the two aircraft are designed for very different threat environments. An F-22 is not designed for the same mission, and threats, as the J-20. An F-22 canopy would make no sense on the J-20, which I had thought pretty obvious. Nonetheless, I did provide information from an analysis of the J-20 which I had done as a project earlier in the year that illuminated the threats it was probably designed for.
The process of designing for reduced RCS is well defined. You can see a simple flowchart of that process here:
Can't post the link as I'm a new user, sorry.
Radar cross section on Google Books
("Radar Cross Section", 2nd Ed., Eugene F. Knott, John F. Shaeffer, Michael T. Tuley
Page 280, Fig. 7.1 "Shaping roadmap for the low observables design process")
To analyze a design, you simply reverse the flow of that diagram. Since shaping is the dominant contributor to RCS, by measuring and modeling the object you have nearly all of the information needed to determine the signature levels for different frequencies. From that you have a clear picture of the threats the object was designed for.
I collected a number of photos of the J-20 from public sources. The events up to the first flight and immediately after represented an interesting opportunity. The aircraft was at known points in space and time, as were the photographers. This made it relatively easy to use photogrammetry to analyze the photos of the aircraft from multiple viewpoints and construct a high fidelity model using OpenCV and other readily available libraries. The interesting part was dropping out improbably points from the model, which was done using k-means clustering after several other methods were tried.
The high fidelity model was then modified and used to perform an computational electromagnetics analysis. Methods of moments, finite element, and optical methods were used. Open source software such as Hedge can perform the required analysis with little modification, and are greatly accelerated by using GPU hardware. In my case I used Amazon EC2 instances with GPUs to perform multiple runs.
The end product is an RCS analysis from the source photographs, across different frequencies. Anyone with knowledge of the relevant physical laws and software processes can do the same.
The analysis showed that in some frequencies, the aircraft had a managed signature. In others it did not. Some features were clearly shaped, sized, and placed to reduce the signature for specific frequencies.
I had planned to go through the analysis process again, document it, and share some of it, but have not had the time. As it is now it may end up as paper or presentation at a conference later in the year. None of this is secret, or especially difficult to understand. "Radar Cross Section" is an excellent resource if you want to become knowledgable about radar cross section and the physical laws that determine it. There are many, many resources on the internet if photogrammetry interests you.
I do find it strange that no one has contacted me about this on SPF, either through a thread or private message. Really, if you have questions, I am more than happy to answer them. If I can't answer them, at least I can point you in the right direction. I won't be spending much time on SDF (just not something that interests me).
I hope that clears a up a few things. Have a great day!
