There could be other designs of flying wings, contrary to most people's belief B-2/B-21's shape is not the most optimal for stealth as it's lambda wing structure from the back forms converging trailing edges which under certain condition could form corner reflectors from the rear, this also is why B-21 only has a single large V shaped trailing edge instead of B-2's many V shaped trailing edge.
View attachment 154568
This planform is the most optimal design for radar stealth, of course the intakes and exhaust could be further refined like the B-21 which has very low profile intakes effectively hidden when viewed from the front:
View attachment 154570
Exhaust could take the form similar to this ADVENT proposal which has a S-duct not just in the intake engine but also in the exhaust to fully hide the engine. Along with a very low-profile exhaust that could blend into the trailing edge for maximum radar and IR stealth.
View attachment 154571
All of these are new technologies for China and probably the rest of the world as well, hence the long development time H-20 is taking along with the fact that it has to be large enough to have sufficient range to strike CONUS, enough space to have atleast a crew of 4 and their amenities for the duration of the mission, carry enough cruise missiles to cause significant damage with a single sortie, probably also some technology onboard for it to fit into the next generation combat ecosystem.
Supersonic H-20 right now doesn't make sense; this isn't the 50s where flying at supersonic speed and high altitude could save you from SAMs and interceptors. Not to mention the range of such an aircraft is questionable, XB-70 was the closest thing we've got to sustained supersonic bomber and that thing had terrible range despite its massive size and the fact that it has 0 stealth what-so-ever. Even if you could make such an aircraft stealthy, you'll never reach the same level as even the B-21, you also now have to deal with massive IR signature from sustained supersonic cruise. Before someone says why not just have it cruise subsonically before doing a supersonic dash, back in the 50s when the XB-70 was designed that was exactly what they thought but they later found that optimising the aircraft for supersonic cruise was in fact more efficient than trying to do both. Optimising for extreme subsonic efficiency(required if you want intercontinental range) while still being able to go supersonic is extremely difficult and is what led to Tu-160 using heavy/non-stealthy variable wings and even today this is still the best if not the only solution.
Personally, I think H-20 will neither be a supersonic aircraft nor a B-2/B-21 style flying wing but something more unique in form and even more stealthy than the B-21 radar wise while also having a lower IR signature to evade modern long range EO sensors and future satellite based early warning systems.
The B-2 was also meant to have a V-shaped trailing edge. The sawtoothing was a late addition / scope creep to give it low altitude bombing capability.
Technology is moving towards credible space-based ISR, there are advances in anti-stealth radar, advances in IR targeting, advances in drone technology to create kill webs with multiple attritable LO UAVs, etc. and etc.
So speed starts becoming a factor again, now that technology is catching up to stealth (not enough and in all areas to make stealth obsolete, but enough that you can’t wholly rely on it against a peer or more advanced foe).
To be clear, I’m not talking about any mach 3 nonsense, or even mach 2. A supercruise speed of mach 1.2 to 1.6 would suffice, with maybe mach 1.8 max speed on reheats. The only possible VLO planform options would be different flavours of tailless cranked kite
(well, if it were hypersonic speed then other options are available, VLO design would be mostly pointless).
Also related to advances in technology, LO standoff munitions may not survive against the dense IADs of a peer / more advanced foe. Hence the move to hypersonic missiles (even though some nations aren’t quite ready to publicly admit this yet). So the kind of munitions you’d need to strike CONUS would become long, air-launched HGVs and the like (also provides additional benefit of increased standoff range). The only possible VLO planform options would be different flavours of tailless cranked kite.
Honestly, if I was at CAC and had just dreamed up the J-36 after a hotpot induced hallucination — I’d immediately pick up the phone to warn XAC like “hey, so about that big teapot of yours…”
You don’t even need variable geometry to achieve a supersonic-capable bomber, again, you can just use a cranked-kite. Just look at the SR-71 and XB-70’s wings (including LEXs), they’d actually work for a VLO supersonic aircraft (after some refinements - and for the avoidance of doubt, obviously not the fuselages, engine nacelles etc.).
You can just essentially build a bigger (perhaps more elongated) J-36. I’d imagine PLAAF looked at initial H-20 and realised it might still work on 2035 US (though relying on your adversary’s failures, or decline in this case, is foolhardy). They then realised it might not be all that successful against 2030 China, let alone 2035. Or they could’ve just asked themselves “what if the US had J-36s”
(yes, and of course all the other elements of a robust system of systems and kill webs)… Either the PLA is at the complete mercy of the B-21 and have done nothing to counter such aircraft, or it’s the opposite and through their work, they’ve discovered increasingly numerous and effective approaches to countering subsonic VLO flying wings.
If you really think about it, provided we don’t go endoatmospheric manned hypersonic combat aircraft, and prior to aerospace combat aircraft — everything will start to look like a J-36… the airborne cruiser needs an airborne battleship, and maybe even airborne aircraft carriers one day.
