Redesigning the turret is a good idea, but wether the results are good or bad depends on the final design.
The T-90 does not have an isolated ammunition compartment and also does not have blow-out panels; although it is theoretical possible to use different autoloaders like that of the "Black Eagle" or the Ukranian autoloader used on the T-72-120 and T-84 Oplot, which are mounted at the turret rear and feature blow-out panels.
But the T-90 is not that bad in terms of ammunition storage, because all ammunition is stored in the hull floor. A German tanker once caluclated that with mid-1980s ammunition (in terms of accuarcy, firing from a standing tank at a standing tank) the probability to hit the ammunition directly is less than 20% from 1,500 - 2,000 m distance.
The performance of the T-72s during the Gulf War with exactly the same ammunition storage was very poor, but there might be other reasons for this (like exaggerated numbers of cases where the ammunition cooked off, lack of spall-liner, etc.).
I don't think that a tank below 50 tonnes is a real option. In the end the requirement "max. 50 tonnes" would likely lead to another T-90-like tank. For reducing the weight you need to reduce the volume, which makes ammunition compartimensation impossible and also will lead to the adoption of an autoloader. If you want to have thicker side armour than the T-90, the only possibility would be to reduce the front armour.
If you ask me, the tank should weigh above 50 tonnes, but less than 65. However my view of the "perfect" modern MBT would require adopting advanced technology which often hasn't been fielded until know. One nice example are unmanned and low-profile turrets, which have been projected since the 1950s. Take a Arjun hull, put a Falcon II low profile turret on it and the resulting tank will weigh less than 50 tonnes. Further weight reduction can be achieved by using the EuroPowerpack, which reduced hull length by 1 to 1.5 meters - the resulting tank would not be perfect, but it would weigh as much as a T-90, while offering better performance than the Arjun (depending on turret front armour).
No, that's not really possible. I don't know how effective the Kanchan armour is, but with Kontakt-5 or Relikt at the turret front an Arjun should have similar frontal protection. But the Arjun weighs too much.
The "top notch" Western tanks have increased side-protection (more than 2 tonnes weight), roof armour (more than 1 tonne weight), mine protection (~2 tonnes weight) compared to the basic Leopard 2A5/M1A2 SEP/Challenger 2 - these weigh already more than 60 tonnes.
A 40 - 50 tonnes tank like the T-90 can have similar frontal armour, but has less mine protection, less side armour and no ammunition compartimensation. Add ERA to the sides (depending on coverage and type of ERA 2 to 4 tonnes), then the side armour should be similar against shaped charges (HEAT, RPGs, ATGMs). But other problems cannot be solved on the T-90 with the given weight limit.
The question is what they want to deliever. Do they really want thick frontal armour resistant to 120 - 125 mm APFSDS? Do they really want protection against large tandem charge ATGMs?
One nice example should be the Puma IFV - it weighs 42 tonnes with full armour (against RPGs and small ATGMs), so putting a tank gun in overhead mount or in an unmanned turret on it without exceeding 50 tonnes weight shouldn't be a problem. But the frontal sector is less armoured than a tank. Alternatively they could want a tank with less frontal armour than the Arjun/T-90/Leopard 2 - the Lince (a scaled down Leopard 2 tank made for Spain) has a combat weight of around 50 tonnes - but it had less armour than a modern Leopard 2 and far less side armour than the M1A2 TUSK, Challenger 1/2 TES and Leopard 2A7/PSO. The Western requirement leading to 65 - 75 tonnes tank is (and was) "frontal armour against modern ammunition, mine protection and side armour against RPGs/ATGMs" in some cases also increased roof armour. If the Indian army has different requirements then they could make a 50 tonnes tank - but this would probably look like a T-90 with Indian armour and Indian gun.
One other way how the Indian tank designers could achieve the goal is to gamble, i.e. saying "this is the current Pakistani ammunition, we only focus on this" - as the current ammunition used by Pakistan is on a performance level of ~1985, it could be possible to create a tank weighing somewhere around 50 tonnes protected against this ammunition. When Pakistan would upgrade their ammo or India would fight against China (which has pretty good ammo), then the new 50 tonnes tank would be obsolete.
I don't think that it is a realistic goal. All tank design programmes I know ended up with a higher weight than originally expected, like the Leopard 1, Leopard 2, MBT-70, M1, Leclerc etc.
The Isreali suggestion however is also not optimal in my opinion. I am not a big fan of the current Arjun tanks and personally would prefer the T-90 over the Arjun Mk.1 at least, because the Arjun has quite a few design problems (too wide turret, too large gun mantlet, weakened zone in the side protection, not ideal hull armour layout, gunner's sight is misplaced etc.). Improving the Arjun is not a bad idea, but not like the Arjun Mk.2 - current photos of the prototypes didn't show any fixed flaws. They should improve the design per se, and not use the old design with new ERA and sights. I hope you are right about the turret redesign.
Different concepts? You know that for sure? We believe that Burlington ("Chobham") is also a type of reactive/semi-reactive armour, like the Russian armour used on T-72B/T-90. Yes, there are design differences, but the concept and the working mechanism should be similar. I didn't argue with thickness, but with volume.
When it gives "more or less same protection" while being more or less as thick as the other armour, I don't see any reason for your point 1. I also did not mention that the T-90S uses the same armour as the Arjun.
I was talking about the currently fielded Arjun tanks. What future brings you might be different from what currently is being tests. In Austria there was a report that new upgrades would be developed for the Leopard 2, which was the reason for some discussions in the internet. The result was that one tank was armoured gun shields for the turret MGs and one tank equipped with a anti-riot fence... noone expected this in the internet, but some deeper improvments. One other nice example is the Swiss upgrade for their Panzer 87, which was one of the most advanced at it's time - but it costed too much and most parts (like the improved turret front and side armour, the RWS and the anti-bomblet protection) weren't fielded. The main differences between normal Pz 87 and upgraded one are the commander's sight and the exhaust smoke/thermal signature reduction kit.
I was writing about the current situation. Saying "the Arjun is better because it has advanced rounds", while the current ammunition used is HESH and a "not-so-good"-APFSDS, is simply wrong. I didn't mention any advanced rounds per se, because there are currently none in the Indian arsenal.
I do think that I know quite a view things about metallurgy. But what I wrote is that there seems to be a misunderstanging in the semantic of the word. "Metallurgy" means "composition of an alloy and treatment during manufacture", this has nothing to do directly with armour protection and weight. Secondarily the steel alloy used will influence armour protection, but the major protection is a result of what armour design is used.
