Now, on the topic.. There is no proof.. for and against whether the Chinese tested the Stealth Characteristics of the aircraft on a pole model. I certainly doubt the Chinese found a way to avoid the whole pole test step in the design process so it is likely that they did go through that procedure. However, to say that the US is the ONLY country in the world that can do these pole tests for RCS would be inaccurate. The British have the facility, the Russians have it... and most likely the Chinese do too.
The main reason for isolated anechoic chamber testing is to have a baseline RCS measurement of the body where that baseline is as free of contaminant radiation as possible, such as those from manmade sources like TV and natural like cosmic background radiation (CBR). The drawback is limited frequency and power range. So then we developed the outdoor testing regime where we can use a greater frequency and power range. Some contaminant radiation are inevitable and that is why these facilities tends to be in physically isolated areas, such as away from large population centers.
There was a mocking post about US still using full scale models instead of computer simulation. I do hope Chinese scientists and engineers takes after that individual and not conduct full scale outdoor testing of the J-20.
One must not also forget that the F-117 was the result of computer simulations based on a program to find that hopeless diamond configuration. A simulation that did not involve pole tests or large infrastructure as such. So the technology to get estimates on what the RCS would be of a particular shape before even giving it a physical form and verifying any ideas on stealth has existed since the late 70s. Would that program not be available for study by now or even replicated??
After all, it is at the end a combination of mathematical equations which someone(or a team) with a good understanding of EM and Wave theory should be able to replicate.
The F-117 did undergo full scale model outdoor RCS testing. There were some computer simulations but because of the available technology at that time, planar angled faceting was the only technique for the F-117, hence its famous appearance today. The majority of the engineers on the program used the classical sliderule. Nevertheless, even in math one can be sloppy, especially when dealing with unpredictable variables, such as search freq from the enemy radar and/or maneuvering rate of the aircraft that will present different aspect angles to the seeking radar that will produce high/low frequency response (reflections) from the aircraft. If the responsible engineer or team of engineers fail to consider these variables down to the decimal points, the full scale model may not be as 'stealthy' as originally hoped for.
Recently, (I believe it was either History channel or National Geographic) there was a program on the Horten 229 and its purported stealth features. The documentary was undertaken in cooperation with Northrop Grumman and it was their team that built a pole model for RCS testing. However, during the building of the model the engineers also ran a simulation Estimating the RCS before the pole test.
The estimate was lower than the actual test as there were places that created a spike which the simulation program did not take into account. Yet, there was a way to measure RCS of a design before any pole test had occurred.
@
gambit
The H-229's low RCS was coincidental, not deliberate.
The flying wing design was known to all major aviation powers since the 1920s, from Europe to Russia and to America. Jack Northrop knew about it. But radar detection, as in the exploitation of reflections which was known since the 19th century, was not fielded until WW II. The flying wing design was known for its long range capability, not of its natural lower RCS when compared to other flying designs. Reimar and Walter Horten knew about radar and they adopted radar spoiling measures from submarine snorkels into the H-229, but they never knew how effective those measures could be because the war ended before they could test the -229 specifically against radars.
Anyway...If the logical (simulated) measurement of a body produced an estimate that is lower than the physical measurement, then the original math failed and always failed because the engineers either missed a contributor somewhere or they did not think a contributor would have any significance. Even with today's computer technology, data out is only as good as data in, but bombarding the body with real signals will always be the final arbiter.