Possible errors in Eagle Dynamics\DCS World OpenBeta\Scripts\Database\Weapons\warheads.lua (open beta) ?

[Mad_Shell]

I was looking a bit at some lua files in the game folders, and I perhaps found some questionable stuff in the Eagle Dynamics\DCS World OpenBeta\Scripts\Database\Weapons\warheads.lua
For most warheads, expl_mass is equal to the quantity of explosive in the warhead, and not to the total mass of the warhead:

 

warheads["C_5"] = -- S-5KO shaped-charge, fragmented
{
    mass            = 1.08,
    expl_mass        = 0.37, -- Warhead 1.08 kg, explosive 0.37 kg + fragments bonus
    other_factors    = { 1.0, 0.5, 0.5 },
    concrete_factors = { 1.0, 0.5, 0.1 },
    concrete_obj_factor = 0.0,
    obj_factors      = { 1.0, 1.0 },
    cumulative_factor= 3.0,
    cumulative_thickness = 0.1
};

warheads["C_8"] = -- S-8ÊÎÌ shaped-charge, fragmented
{
    mass            = 3.0,
    expl_mass        = 0.855, -- Warhead 3 kg, explosive 0.855 kg + fragments bonus
    other_factors    = { 0.5, 0.5, 0.5 },
    concrete_factors = { 0.5, 0.5, 0.1 },
    concrete_obj_factor = 0.0,
    obj_factors      = { 0.5, 1.0 },
    cumulative_factor= 5.0,
    cumulative_thickness = 0.3
};

warheads["C_8OFP2"] =  -- S-8OFP HE
{
     mass            = 9.2,
     expl_mass        = 2.7, -- Warhead 9,2 kg, explosive 2.7 kg + fragments bonus
     other_factors    = { 0.5, 1.0, 1.0 },
     concrete_factors = { 0.5, 1.0, 0.1 },
     concrete_obj_factor = 0.3,
     obj_factors      = { 0.5, 1.0 },
     cumulative_factor= 0.0,
     cumulative_thickness = 0.0
}; 

 

but for some other warheads, the expl_mass is equal to the total mass of the warhead
example:

 warheads["C_24"] = -- S-24 HE Warhead 123 kg, explosive 23.5 kg + fragments bonus
{
     mass            = 123.0,
     expl_mass        = 123.0, -- Warhead 123 kg, explosive 23.5 kg + fragments bonus
     other_factors    = { 1.0, 1.0, 1.0 },
     concrete_factors = { 1.0, 1.0, 0.1 },
     concrete_obj_factor = 1.0,
     obj_factors      = { 0.2, 1.0 },
     cumulative_factor= 0.0,
     cumulative_thickness = 0.0
}; 
 

Edited January 13, 2021 by Mad_Shell

[Northstar98]

I thought about this one too, unfortunately I have a sneaking suspicion that DCS doesn't actually model fragmentation (and definitely doesn't model the difference between different fragmentation warheads/patterns, such a continuous rod for instance) so to compensate for it, the explosive mass is inflated.

Edited January 18, 2021 by Northstar98

[Tippis]

I think a large part of it is explained by the section at the very end, where most of the stats are involved. They're not necessarily replicating real masses, but picking ones that achieve a specified desired effect when combined with different coefficients depending on what they hit. In some cases, this will mean fudging the numbers to have more or less “explosive mass” as a baseline because it yields the right combinations of effects on target without having to fiddle a whole lot with those coefficients.

 

The bit at the end, just quickly run through google translate, states:

- Description of the coefficients
- [[
expl_mass = 2.0,
The mass of the explosive in the warhead of the ammunition in kilograms

 other_factors = {HE1, HE2, HE3};
Coefficients of high-explosive action when hitting the ground:
HE1. - high-explosive striking effect (expl_mass * HE1)
HE2. - explosion effect size
HE3. - the size of the explosion funnel

 concrete_factors = {HE1, HE2, HE3},
Coefficients of high-explosive action when hitting concrete:
HE1 - high explosive striking effect (expl_mass * HE1)
HE2 - explosion effect size
HE3 - explosion funnel size

 concrete_obj_factor = CP,
Coefficients of concrete-piercing action when hitting concrete:
CP - concrete piercing effect for concrete ammunition (expl_mass * CP)

 obj_factors = {HE1, HE2},
High-explosive coefficient of action when hitting a ground object (equipment):
HE1 - high explosive striking effect (expl_mass * HE1)
HE2 - explosion effect size

 cumulative_factor = SC,
SC - cumulative effect for cumulative ammunition (expl_mass * SC)
 cumulative_thickness = TH
TH is the maximum thickness of the armor that the cumulative part of the demage penetrates (in meters).
     If the unit's armor is larger, then the cumulative damage is not applied.

Let us assume that the warhead of a conventional bomb has an explosive mass of 10 kg.

The bomb has the following coefficients:

obj_factors = {0.5, 1},
concrete_factors = {0.8, 1, 1},
other_factors = {0.9, 1, 1},
cumulative_factor = 5,
concrete_obj_factor = 3
cumulative_thickness = 0.05

then,

1.if a bomb falls into the ground, then the high-explosive effect will be 10 * 0.9 = 9
2.If a bomb falls on a concrete object, then the high-explosive effect will be 10 * 0.8, plus an additional 3 * 10 = 30 concrete-piercing damaging effect is transferred to this object
3.If a bomb hits a car, then the high-explosive effect will be 10 * 0.5,
    plus an additional 5 * 10 = 50 cumulative lethal action is transferred to this object if the vehicle's armor is less than 5 cm.

In order not to write large tables for each warhead, there is a simple_warhead function that takes one parameter - a mass of explosives. The output is a conventional warhead with a high-explosive fragmentation effect.
-]]

 

Run through a few scenarios with the different warheads and see if the numbers make more sense with those parameters in mind.

[Cmptohocah]

21 hours ago, Northstar98 said:

I thought about this one too, unfortunately I have a sneaking suspicion that DCS doesn't actually model fragmentation (and definitely doesn't model the difference between different fragmentation warheads/patterns, such a continuous rod for instance) so to compensate for it, the explosive mass is inflated.

 

There is no fragmentation in DCS. It's collision based: the missile needs to hit the target.
Something like this:

Edited January 19, 2021 by Cmptohocah
Added an example.

[Northstar98]

On 1/19/2021 at 3:12 PM, Cmptohocah said:

There is no fragmentation in DCS. It's collision based: the missile needs to hit the target.
Something like this:

 

 

That's not what I'm talking about, you can still have fragmentation effects from impact only detonations - but fragmentation effects aren't modelled, there's only 1 'explosive effect' that tries to achieve a more expected effect - like what Tippis said.

 

But to give you an answer, there are definitely proximity fuses in DCS, I've seen them many times for missiles that have them. In MP desync complicates things, which I imagine is what's going on in that tacview screenshot.

 

What I'm talking about here are actual warhead effects, not fusing. And so far DCS only models blast/high-explosive effects, not fragmentation. What it does do however, is inflate the explosive mass of warheads to approximate fragmentation effects (as much as 5x for missiles with long fuse ranges).

Edited January 22, 2021 by Northstar98

[Shadow KT]

On 1/19/2021 at 5:12 PM, Cmptohocah said:

There is no fragmentation in DCS. It's collision based: the missile needs to hit the target.
Something like this:

 

There is no fragmentation, just blast radius. You don't need a missile to hit something directly to kill it. Missiles can proximity fuse and still cause damage. How exactly the proximity fuze works.... I don't know as it is not consistent, neither in SP, nor MP.

Fragmentation -> NO
Blast Radius -> YES

[Northstar98]

1 minute ago, Shadow KT said:

There is no fragmentation, just blast radius. You don't need a missile to hit something directly to kill it. Missiles can proximity fuse and still cause damage. How exactly the proximity fuze works.... I don't know as it is not consistent, neither in SP, nor MP.

Fragmentation -> NO
Blast Radius -> YES

 

I think he was referring to the fact that the R-27 in the screenshot should've probably fused and then took this to suggest that proximity fuses don't exist in DCS, when they do (just with their own issues).

 

But apart from that fragmentation effects aren't modelled in DCS, in the warheads.lua there's just explosive mass, and this is inflated to compensate for the lack of fragmentation modelling.

[Mad_Shell]

Bump! As Tippis wrote, there are various coefficients applied, but I still don't see why for some HE fragmentation warheads expl_mass is the total mass of the warhead, while for others it's just the mass of explosives in the warhead. Would be great if an ED dev could confirm it's working (or not) as intended.

[Chizh]

For some warheads, for which the destructive effect is mainly bring out by fragments, we have increased the mass of the explosive to compensate of the fragments lack.
Our plan is to develop a system of warhead fragmentation, when we implement this, all the masses of explosives will be brought to real value.

[Mad_Shell]

5 minutes ago, Chizh said:

For some warheads, for which the destructive effect is mainly bring out by fragments, we have increased the mass of the explosive to compensate for the lack of fragments.
Our plan is to develop a system of warhead fragmentation, when we implement this, all the masses of explosives will be brought to real value.

 

Thx for the answer Chizh. Is the plan to bring this new fragmentation system when the new damage model is applied to modern units, or that will be for later?

Edited April 6, 2021 by Mad_Shell

[Chizh]

1 hour ago, Mad_Shell said:

 

Thx for the answer Chizh. Is the plan to bring this new fragmentation system when the new damage model is applied to modern units, or that will be for later?

 

I don't want to promise precise dates, but yes, we want to tie the development of the fragments system to the further development of the damage model.

[Avimimus]

6 hours ago, Chizh said:

I don't want to promise precise dates, but yes, we want to tie the development of the fragments system to the further development of the damage model.

 

I suppose this will only be limited to air-to-air missiles?

 

I always thought that a fragmentation model could be added for rockets - although large salvos might require automatically turning the system off. It'd be great to have a better rocket fragmentation model for aircraft like the Gazelle - even if it might be a bit too much for large Su-25 salvos.

 

[GGTharos]

It isn't necessary to turn the system off for any amount of salvos.  Fragmentation can be relatively simply implemented as a function of probability for the number and size of shrapnel received with both of those, as well as fragment velocity/damage controlled by the range from the explosion.

Multiple weapons means running this calculation multiple times, but it wouldn't need to simulate something like an actual explosion with say 400 fragments per warhead (or more for larger warheads).

 

This could even apply in a stop-gap way to the current damage models by randomly choosing the impact location on a target and applying the damage as if from a gun round (but instead of a 'bullet' you use the shrapnel's damage characteristics).

 

There are some challenges, like ensuring that masked areas are not hit so, after computing how many pieces of things should hit in and in which parts of the target:

 

If you imagine an aircraft target, the shrapnel will only hit within a certain volume - for some missiles this is spherical but for most it's a ... conical donut or some such shape.  So, if a missile's fuze covers the engines (say, coming from the rear) you don't want to apply damage to the wings or cockpit for example.  The current DM does take this into account.

 

Similarly if you consider a tank, or soldier, you want to run some rapid raytracing after deciding the damage and hits per area, randomly assign where the target is hit, then raytrace the LOS and apply the hits where they belong accordingly to any cover or the target itself, if there's some cover between them.

 

There are some challenges here, like at certain ranges there are so many fragments that it's not worth calculating:  Simply figure what's covered and what's not, then figure out penetration/bounces as a probability function and apply damage.

Edited April 8, 2021 by GGTharos

[Fri13]

12 minutes ago, GGTharos said:

It isn't necessary to turn the system off for any amount of salvos.  Fragmentation can be relatively simply implemented as a function of probability for the number and size of shrapnel received with both of those, as well as fragment velocity/damage controlled by the range from the explosion.

Multiple weapons means running this calculation multiple times, but it wouldn't need to simulate something like an actual explosion with say 400 fragments per warhead (or more for larger warheads).

 

This could even apply in a stop-gap way to the current damage models by randomly choosing the impact location on a target and applying the damage as if from a gun round (but instead of a 'bullet' you use the shrapnel's damage characteristics).

 

Just cheat. Check the line of sight to unit to get the sector that unit model occupies. And then take the fragmentation pattern (varies by warhead, angeles, fuses etc) and divide the matching sector from it. Then run a ray tracing (on CPU, not on GPU if not supported) for the remaining fragments on the target. This can be done early before impact as when flight path is known like unguided rockets or free falling bomb. 

 

Guided munitions are more challenging but does not require so much either in the timely presented processing. As after the explosion there is extra time to calculate hit points and damage before smoke and fire disappears.