Future damage modeling maybe?

[Tank50us]

So, an idea I have been tossing around is to do with the actual damage models of units in the game, namely ships and ground vehicles.

 

Now, as much as we all love DCS, and its accuracy, we all know that tanks and ships do not have health bars, and as a result, it can be a bit jarring to see them (yes I know we can turn them off, but this thread isn't about that). What I'm proposing is a possible fix to the DM that might work, and to at least discuss it in a civil manor.

 

Tanks and AFVs

The way I see it, vehicles should be broken into several parts to have a proper module, this will apply to ships as well, but I'll get to that when I get down there. For tanks, the breakup should be: Hull, Suspension, turret (if applicable), weapons, engine, and crew.

 

Hull: the most basic part of a tank or AFV. Everything 'attaches' to it, and thus forms the core part of the new damage model. If it's destroyed, the vehicle is destroyed.

 

Suspension: The tracks or wheels of a vehicle. Each vehicle has a minimum of two 'suspension' hit boxes, and when 'destroyed', the vehicle is temporarily immobilized. The amount of time it takes to fix is likely fairly quick, since changing a tire or re-connecting a track pin is something crews can be trained to do with just the tools they have on hand, but until it's ready to move, that unit isn't going anywhere.

 

Turret: The big rotaty thing that brings the shooty things onto the target. If 'destroyed', the gears are out, and the turret is stuck in whatever direction is was facing, or, in some cases, will start moving much slower than normal (as the gunner now has to hand-crank the thing)

 

Weapons: Pretty simple, if knocked out, the vehicle loses that weapon.

 

Engine: Although in the hull, hits to the engine will cripple the vehicle, and prevent it from moving at all. This is effectively a mobility kill, as engines are not the sorta thing to be easily fixed in the field. Although it's unlikely that an M1 will knock out a T-80s engine from the front, and A-10 strafing run can get a few shots through the engine deck, and knock the engine out.

 

Finally, the crew: If the tank is destroyed, there's a good chance the crew may bail out (exact odds depend on the tank btw), But if the crew is ever killed or incapacitated, the vehicle will not move or fight any longer.

 

Ships:

 

Like tanks, ships would be broken into several sections, including: Hull sections (at least six), the main deck, super-structure, weapons, and propulsion.

 

Hull section: Each ships hull is broken into at least 6 sections. As each section takes damage, the ship will start to suffer more and more, and if a section is completely destroyed, the ship will begin to sink towards that section. The sections are: Bow Starboard, Bow Port, Midships Starboard, Midships Port, Stern Starboard, Stern Port. To reiterate, if any of these sections are destroyed, the ship will begin to sink. The logic is that if a solid chunk of the middle of the ship no longer exists, the captain will likely order the abandonment of the ship.

 

Main deck: While not important on most ships, damage to the main deck on carriers will result in being unable to perform flight operations.

 

Superstructure: The upper works of a ship, home to the bridge, radar masts, etc. If damaged, the ship may be left uncontrollable, or unable to fight at all.

 

Finally propulsion: If a ship has a steam or diesel electric propulsion system, any damage to the engines, or funnels will render the ship unable to move at full speed, or at all.

 

 

So what do you guys think?

[sirrah]

Sure, I don't think anyone will oppose to your idea, but... I expect this will cause a big hit on performance.

 

I'll probably not make myself popular saying this, but in times where an up to date GPU costs more that a kidney: "Stop adding new goodies and eye candy. Shift focus completely on optimization, bug fixing and completing existing modules"

 

 

So, yeah, great idea, but imho not something ED should spend resources in now.

Edited May 10, 2021 by sirrah

[Northstar98]

I love the idea and several of us have been going on about it for some time now.

 

As for the performance concerns, if it's to the level of the WW2 damage model for aircraft that would be perfect. The main concern for me is penetration and fragmentation/spalling for tanks.

 

I would love to see some proper ship physics, which includes sinking (things like Cold Waters handles this perfectly IMO, as well as having ships that interact with the water properly).

 

I've already talked about vehicles in another thread but with regards to ships:

• Definitely divide the hull into compartments - if we can get the actual compartments that would be ideal, but if we can't we can approximate them.
• All sensors and weapons.
• Engines + Propulsion systems.
• Ammunition.
• Fuel.
• Steering system.
• Bridge.
• Command Centre/CIC/Ops
• Other miscellaneous components such as elevators, hangar doors, communications, NAVAIDs etc.

[shagrat]

Actually there is already some "more" detailed modeling going on than "health bars"  under the hood. Still this could be improved.

Basically a vehicle already has different "zones" and from testing and discussion, a "damage" beyond certain values reduce mobility (half max speed). There is definitely room for improvement and a more detailed damage representation (opposed to modeling) would improve this tremendously. One thing for example is that there is no need for a real-time representation of the damage effects. Let me explain: a hit on one of the existing hitboxes (front, rear, L/R side, top, turret...) may penetrate the armor and cause internal damage. Depending on hitzone different damage options from non-vital to critical system/effect can be calculated/randomized. This does not need to happen in real time and can be delegated into a separate thread as the "effect" of the damage could actually happen even 2-5 seconds after the hit. A damaged hydraulic line, or engine, scraped mechanical part may take a moment to actually "break". So we can easily offload these computations to a separate thread/CPU core provide a table detailing the behavioral effects for the AI to be looked up by the simulation when convenient. Even a delay of a couple seconds should be acceptable.

This "effect code" will determine things like reducing speed/immobilization (set AI hold), crew is incapacitated (AI off), weapon system disabled (ROE hold fire) and so on. Instead of always exploding the unit...

Of course in a perfect world, we will get a similar detailed damage model as in the warbirds for vehicles and ships, but in the meantime and until the CPU performance has evolved to support it on that level, this would be a good compromise.

 

Btw.  damage model for infantry is currently "strange". I mean, if you are hit by anything akin to a bullet or bomb fragment, grenade parts, etc. you may not be dead, but still out of the actual fight and need evacuation instead of gunning at low flying planes and helicopters.

Edited May 10, 2021 by shagrat

[Northstar98]

35 minutes ago, shagrat said:

Actually there is already some "more" detailed modeling going on than "health bars"  under the hood. Still this could be improved.

Basically a vehicle already has different "zones" and from testing and discussion, a "damage" beyond certain values reduce mobility (half max speed).

 

Last time I looked at the .lua the damage model zones on vehicles (which are on the minority of vehicles) only serve as a damage modifier - and it was based around the RHAe protection for KE projectiles.

 

Things like mobility were still tied to thresholds on a health bar - same for ships.

 

Personally if we get something equivalent to the WW2 damage model, that would be fantastic. The issue we run into here though is we need to facilitate round types better. I think the main issue is going to be penetration mechanics and fragmention/spalling/sharpnel.

 

It's more of a dream, but what GHPC or WT has would be ideal.

[shagrat]

12 minutes ago, Northstar98 said:

 

Last time I looked at the .lua the damage model zones on vehicles (which are on the minority of vehicles) only serve as a damage modifier - and it was based around the RHAe protection for KE projectiles.

 

Things like mobility were still tied to thresholds on a health bar - same for ships.

 

Personally if we get something equivalent to the WW2 damage model, that would be fantastic. The issue we run into here though is we need to facilitate round types better. I think the main issue is going to be penetration mechanics and fragmention/spalling/sharpnel.

 

It's more of a dream, but what GHPC or WT has would be ideal.

The effects of reduced mobility were introduced long time ago. The zones are already in the game (for armored stuff, as unarmored basically takes penetration damage from any angle) and thus ED could simply add the "damage effects" on top of the current model.

The threading of that effects would offload the calculation for even large clusters of vehicles/units and minimize the performance impact.

If we wait for the ray tracing based modelling of shrapnel and fragmentation with internal system modeling and the CPU performance on par with that, I may be dead and buried before that happens... So for me a "simpler" approach where the effects of damage are modeled in n more detailed manner, instead of a true to life bullets/fragments passing through a dozen vehicles and systems, would be appreciated.

Sometimes less is more...

 

If ED finds a way to balance a raytracing damage model for larger amounts of vehicles with performance, I am all for it, but I personally can easily enjoy an easier solution, if just the effects we see on the vehicles are modeled, based on a hitzone.

[Tank50us]

1 hour ago, shagrat said:

If ED finds a way to balance a raytracing damage model for larger amounts of vehicles with performance, I am all for it, but I personally can easily enjoy an easier solution, if just the effects we see on the vehicles are modeled, based on a hitzone.

 

And ships would be a good way to test it. After all, they're significantly larger than a tank, and as a result, easier to test against. Although for a ship, I do still think that if a hull section is completely destroyed, that should be enough to sink the thing since irl, having a solid chunk of your ship blown away usually results in rapid and uncontrolled flooding which sinks you eventually. Of course, we can certainly accelerate it by continuing to hit the ship until it rolls over (alaBismark), but I do think it would be a good place to start, if for no other reason than the sheer size of some of these ships in DCS.

[shagrat]

On 5/10/2021 at 2:45 PM, Tank50us said:

 

And ships would be a good way to test it. After all, they're significantly larger than a tank, and as a result, easier to test against. Although for a ship, I do still think that if a hull section is completely destroyed, that should be enough to sink the thing since irl, having a solid chunk of your ship blown away usually results in rapid and uncontrolled flooding which sinks you eventually. Of course, we can certainly accelerate it by continuing to hit the ship until it rolls over (alaBismark), but I do think it would be a good place to start, if for no other reason than the sheer size of some of these ships in DCS.

I guess one thing about modern warships is the resilience they provide. I am no expert on naval warfare, but from my limited knowledge typically you don't sink warships any more, but incapacitate/immobilize/take them out of the fight by destroying sensors, weapons, propulsion or causing damage that needs imminent attention... Especially with our typical Harpoon, Maverick or even Exocet missiles you need a lot of hits to actually sink a ship. Torpedoes below the waterline may work better, but from a flight simulator perspective the focus should be on sensors, weapons, propulsion and how damage to these affects combat readiness.

Anyway, a majority of DCS is focused on aircraft attacking ground troops, specifically providing CAS, since DCS: Black Sharks and DCS: A-10C Warthog. For me this is still the core of DCS as most other aspects (CAP, SEAD, Escort, Intercept, etc.) focus on preparing the attrition of enemy ground forces, striking enemy infrastructure or support of friendly troops trying to do the same...

Thus it would improve the core aspect of DCS if the damage modeling and effects on the ground troops get enhanced and AI reaction to these effects and threats improved.

 

[Dragon1-1]

Actually, a cruise missile hit at the waterline, or even a Maverick F in ship mode (which also causes it to go for the waterline) definitely can sink a ship. HMS Sheffield learned this the hard way after being hit by an Exocet in the Falklands. The only reason it took so long to sink was that the warhead didn't detonate. It may take a while, and sometimes it'll be possible to save the ship, but even a single missile can kill it.

[Northstar98]

9 minutes ago, Dragon1-1 said:

Actually, a cruise missile hit at the waterline, or even a Maverick F in ship mode (which also causes it to go for the waterline) definitely can sink a ship. HMS Sheffield learned this the hard way after being hit by an Exocet in the Falklands. The only reason it took so long to sink was that the warhead didn't detonate. It may take a while, and sometimes it'll be possible to save the ship, but even a single missile can kill it.

 

With the Sheffield, I'm pretty sure the warhead did actually detonate - at least that was the findings of "The Loss of HMS Sheffield - A Technical Re-assessment". Though the damage the ship suffered was mostly by fire - which gutted the ship, so it would've been a more or less complete mission kill/near catastrophic kill and only sinking while under tow, with water ingress  through the entry hole, in rough seas.

 

But even above the waterline - you can rupture water mains, as well as buckle and crack bulkheads.  

 

The problem with going on SINKEX exercises is that they have ships in their most survivable configurations.

[Tank50us]

4 hours ago, Northstar98 said:

 

With the Sheffield, I'm pretty sure the warhead did actually detonate - at least that was the findings of "The Loss of HMS Sheffield - A Technical Re-assessment". Though the damage the ship suffered was mostly by fire - which gutted the ship, so it would've been a more or less complete mission kill/near catastrophic kill and only sinking while under tow, with water ingress  through the entry hole, in rough seas.

 

But even above the waterline - you can rupture water mains, as well as buckle and crack bulkheads.  

 

The problem with going on SINKEX exercises is that they have ships in their most survivable configurations.

 

And if memory serves, Sheffield suffered from a case of overconfidence as well. If I understand correctly, the hit and damage by that hit was something that she could have, and honestly, should have been able to survive, but from what I understand some of her water/fire doors were left open when she went to battle stations, and that contributed to the spread of the fire that would ultimately claim the ship. Of course, other issues were present that lead to it, like the lack of a CAP to intercept the incoming aircraft before they were in Sea Dart range, as well as some failures in the Sea Dart system that exposed why an all-missile defense isn't as good an idea as some people thought (this is why CIWS is a thing). There was also a lack of ECM, and even a lack of training was cited as an issue.

 

For proper DM on a ship, these would have to be things that are modeled in as well. For example when a ship goes to action stations (from green state to red state), it will take time for the crews to get where they need to be, and that all depend on the size of the ship. Obviously tanks wouldn't have to worry about this, as their crew is usually in the seat already, so irl, switching from a calm state to a 'significant emotional event' state would take about as long as the TC reaching behind his gunner and slapping him with a canteen or throwing that annoying pebble from his boot at the driver.