reported CBU 87/103 Damage issues.

[Terzi]

I have just watched Wags' video and I can say that for a 2 x 1000 pound area effect anti-armor munition, this damage is sooo little.

 

What you see in the video is just dust and explosion effects. What's the result? 2 to 5 BMP-2s killed? And the tents in the middle? Tents can be destroyed by a single bullet in DCS. Maybe if someone would throw 2000 lbs stone, it would be somewhat similar I was also reading comments on the video. Everybody is like 'wooow that's soooo good weapon, it destroyed everything from 70 miles." which I strongly disagree.

 

I guess it would be good to re-evaluate in DCS what BLU-97 based CBU-87, CBU-103 and the AGM-154A can do in reality. This is an anti-armor weapon (well not anti-heavy armor), and it should destroy light armor vehicles just fine. Maybe its time to check on that? And also implement the rotation speed model (rpm setting) for adjusting it spread vs lethality?

 

From globalsecurity.org:

"The BLU-97/B Combined Effects Bomb (CEB), effective against armor, personnel and material, contains a shaped charge, scored steel casing and zirconium ring for anti-armor, fragmentation and incendiary capability."

[Xavven]

Your sentiment is echoed on the reddit forums. As you pointed out, this is not really a JSOW problem as that's just the delivery vehicle, but really a BLU-97 issue, and maybe it's connected to the way area of effect damage is handled in DCS in general.

 

I also wonder if the tank treads of heavy armor would be vulnerable to a CEM in real life. I would love for DCS to implement maneuverability kill.

[Dragon1-1]

CEM wouldn't be much good against threads IRL, because they're on the bottom of the tank, and the bomblet comes in from the top.  Now, top armor, the AA MG, turret optics and so on could be damaged, but it would not be very effective against modern tanks. 

CEM in general seems to underperform in DCS.

[Blinky.ben]

I think this comes down to damage modelling more then weapon effect. Anti-armour weapons particularly bomblet types don’t have a big blast and fragment radius but a very narrow shaped charge and unless it strikes are area on the vehicle that is critical, then the weapon just leaves a small hole e.g if it hits the engine bay it may damage the engine under the armour but it doesn’t cause a nuclear catastrophic explosion that wipes out the whole vehicle. If it gets into the crew compartment it will most likely kill or injury most the crew but the engine and running gear will more then likely function just fine. As for DCS no ones happy until the vehicle is a complete smouldering mountain of flames which is not accurate at all but simply is should be a fairly intact vehicle that doesn’t function anymore or more likely some parts will no longer function while other parts do e.g it can’t move anymore more but the turret and weapons are still functional.

 

so I guess my long winded point is do we want a fairly accurate weapon that leaves little sign of damage meaning a realistic level of damage reporting will need to be conducted or a unrealistic weapon effect with a wall of flames and destruction making it easy to see that the vehicle is totally destroy being the unrealistic approach but a work around DCS shortages?

Edited July 11, 2021 by Blinky.ben

[Northstar98]

3 hours ago, Dragon1-1 said:

CEM wouldn't be much good against threads IRL, because they're on the bottom of the tank, and the bomblet comes in from the top.  Now, top armor, the AA MG, turret optics and so on could be damaged, but it would not be very effective against modern tanks. 

CEM in general seems to underperform in DCS.

 

AFAIK, the BLU-97/B has a RHAe penetration of ~120mm. Unless that tank has ERA on the roof or otherwise features additional protection, it's probably going through it. Most vehicles only have a few 10s of mm at most protecting their roofs, unless it's protected by ERA, or has an armour module directly underneath it.

 

Even if we assume the turret is out of bounds, you've still got engine decks, which are wholly unprotected.

 

But all of this is less relevant because the targets in the video were against BMP-2s - a lightly armoured IFV (and one of the more lightly armoured). It doesn't have more than a few tens of mm of RHAe protection at the very most and the roof has a whopping maximum of 12mm of RHAe over the hull and turret (which is mostly just 6mm). A shaped charge with over 100mm of RHAe penetration should have no trouble at all going through this.

 

As for the rest of the targets? They were tents...

 

Overall, the main issue as I see it are as follows:

• Super simplistic damage model for vehicles (the most advanced are broken up into different armour values, but this just acts as a multiplier for damage done, so long as it's past a threshold, and even so - those armour values don't seem to differentiate between KE and CE equivalent protection). There's no subsystem damage modelling of external or internal components whatsoever, let alone things like volumetric armour or even simulation of different armour types (especially ERA).
• Low fidelity of weapon warhead effects (and not talking about graphical effects). I'm not sure what difference there is, if any between how HEAT, HE, HESH and HEF ammunition is modelled in DCS, apart from maybe a penetration factor for HEAT. There's certainly no fragmentation modelling at all, and the steel casing of the BLU-97/B is supposedly enough to create fragments able to damage soft and lightly armoured targets (for instance, like the BMP-2) out to around 100 feet, and to at least injure personnel out to 500 feet.

Edited July 11, 2021 by Northstar98

[Tippis]

…and that's before we get into the effect on infantry, which in this day and age are more threatened by sneezes than having a half-ton, area-effect, anti-soft-armour glide bomb dropped on them.  

[Frederf]

CEMs are pretty weak sauce, takes about 7 direct hits to kill a T-72. I think 1-3 would be a lot better. It would also make a big difference between a 500' HOF and 1800' HOF on hard targets.

[Dragon1-1]

5 hours ago, Northstar98 said:

AFAIK, the BLU-97/B has a RHAe penetration of ~120mm. Unless that tank has ERA on the roof or otherwise features additional protection, it's probably going through it. Most vehicles only have a few 10s of mm at most protecting their roofs, unless it's protected by ERA, or has an armour module directly underneath it.

That's why I said modern tanks, which had generally wizened up to that sort of tricks, and do have rooftop ERA and the like. Even the older Russian ones are sometimes retrofitted to be plastered with tiles all over. Older models (and BMP2s) would definitely get a hole punched in them, hence my comment about the -97s underperforming. I suspect this also has to do with how the Roceye isn't nearly as useful as it should be.

Edited July 11, 2021 by Dragon1-1

[Northstar98]

2 hours ago, Dragon1-1 said:

That's why I said modern tanks, which had generally wizened up to that sort of tricks, and do have rooftop ERA and the like.

 

It's only really Russian tanks that do that though, the only other tank that up-armours it's roof is the Leopard 2A5/2A6/2A7 where there is an additional composite roof protection package - though we don't have that tank.

 

Though there are still areas it doesn't have protection over.

 

Of course there's also hard-kill APS (not sure how many protect against things travelling vertically downwards though) but we don't have any of those modelled at all, nor do we have tanks with them fitted.

 

And even so, it still leaves engine decks - which are completely vulnerable, and you take the engine out, you've got yourself a mobility-kill and you've also caused that tank to loose power.

Edited July 11, 2021 by Northstar98

[Terzi]

8 hours ago, Northstar98 said:

 

Overall, the main issue as I see it are as follows:

• Super simplistic damage model for vehicles (the most advanced are broken up into different armour values, but this just acts as a multiplier for damage done, so long as it's past a threshold, and even so - those armour values don't seem to differentiate between KE and CE equivalent protection). There's no subsystem damage modelling of external or internal components whatsoever, let alone things like volumetric armour or even simulation of different armour types (especially ERA).
• Low fidelity of weapon warhead effects (and not talking about graphical effects). I'm not sure what difference there is, if any between how HEAT, HE, HESH and HEF ammunition is modelled in DCS, apart from maybe a penetration factor for HEAT. There's certainly no fragmentation modelling at all, and the steel casing of the BLU-97/B is supposedly enough to create fragments able to damage soft and lightly armoured targets (for instance, like the BMP-2) out to around 100 feet, and to at least injure personnel out to 500 feet.

 

 

 

9 hours ago, Blinky.ben said:

so I guess my long winded point is do we want a fairly accurate weapon that leaves little sign of damage meaning a realistic level of damage reporting will need to be conducted or a unrealistic weapon effect with a wall of flames and destruction making it easy to see that the vehicle is totally destroy being the unrealistic approach but a work around DCS shortages?

 

Totally agree with those two statements. Even if the real-like damage modeling is currently not the top priority thing for ED, there should be a mid-state between destroyed and alive, which may be put as "out of service" or "out of use" or "critically damaged" or "light damage" so what this means is that, the owning faction will not be able to use this vehicle unless it goes through repair.

 

Maybe one of those states will be "it can move but cannot shoot" or "it can shoot but cannot move" or "it can only shoot weapon 1 and 3 and can move". This "repair" what I am talking about would somehow be connected to maintenance crew, possibly to be included in the planned DCS: Dynamic Campaign.

 

As for the damage assessment, this isn't yes or no. Your car's battery might have been run out. So your car isn't destroyed but still cannot move unless you do something.

 

So there should be some probability based damage effects. Maybe an APC is 50m from the explosion center but takes no critical damage, while another at 200m takes a critical shot to its fuel tank or ammunition so it gets burned.

[Tippis]

8 minutes ago, Terzi said:

Totally agree with those two statements. Even if the real-like damage modeling is currently not the top priority thing for ED, there should be a mid-state between destroyed and alive, which may be put as "out of service" or "out of use" or "critically damaged" or "light damage" so what this means is that, the owning faction will not be able to use this vehicle unless it goes through repair.

 

Maybe one of those states will be "it can move but cannot shoot" or "it can shoot but cannot move" or "it can only shoot weapon 1 and 3 and can move". This "repair" what I am talking about would somehow be connected to maintenance crew, possibly to be included in the planned DCS: Dynamic Campaign.

 

As for the damage assessment, this isn't yes or no. Your car's battery might have been run out. So your car isn't destroyed but still cannot move unless you do something.

 

So there should be some probability based damage effects. Maybe an APC is 50m from the explosion center but takes no critical damage, while another at 200m takes a critical shot to its fuel tank or ammunition so it gets burned.

The thing is, those damage states already exist. The problem is that they're very badly spaced out in terms of when they trigger, and the effect they have on the unit behaviour is next to invisible unless you're constantly monitoring it through the external view.

 

 

I'll just use my regular copypasta

 

Fundamentally, the problem is that the damage application is completely backwards.

 

Even a simple hit point system can be made to work while they chip away at more intricate systems modelling in all vehicles, but only if that hitpoint pile is treated properly. Right now, it isn't.

 

At the moment, ground vehicle damage application basically consists of three different components:

• A hitpoint pile — the bigger the vehicle, the more hitpoints it has, and the tougher it is.

• A damage mitigation stat — an abstraction of armour to simply deflect some smaller amounts of damage application, including an aspect calculation whereby, depending on the vehicle and the angle of attack, the damage mitigation is scaled up or down.

• A four-(and-a-half)-tiered damage state: fine(ish), system-crippled, movement-crippled, (burning, soon to be) dead.

 

It's that last one that is set up horribly. In particular, the thresholds are nonsensical in relation to the full hitpoint pile, although the order is also questionable. Essentially, it's a case of, at 50% HP, the unit stops working; at 25% (or thereabouts), it starts moving slowly; at 10% it starts burning and will slowly lose its remaining hitpoints; and at 0% it dies and explodes. Not a single one of those are where they should be.

 

By all means, units should probably explode at 0% HP, but they should start burning a lot sooner (and and stay burning a long time after), and in particular they should be dead long before that. The reason this matters is that the only event you can reliably automate without scripting up every single unit in a mission (say goodbye to your CPU) is death. It's what scores point in the kill screen; it's what most mass triggers (“group dead”, “group alive” and the “…less than” versions of the same triggers) use to do their thing. To make that happen, and to make the attack actually count from a game-mechanical perspective, you end up having to hit individual trucks with 500lbs bombs, where a 0.5lbs bomblet should really be able do the same job: in this case, to reduce the hitpoint pile to 0 to trigger the “death” state.

 

Similarly, somewhat depending on exactly what kind of unit we're talking about, movement should probably be lost long before the system as a whole is gone, unless we're talking about something flimsy (eg. radar antennas and the like on anti-air), in which case the systems should be gone the moment something sneezes in their general direction.

 

Ideally, the whole thing would be set up something like:

 

• The hitpoint pile is still there because it's too much effort to get rid of it.

• The damage states are set by unit type, and all happen a lot sooner. Eg. for a tank, it's mobility loss at 80%, system loss at 70%, death at 50%; for a mobile SAM, it might instead be system loss at 95%, mobility loss at 80%, death at 50%. The only unit where death should happen at 0% HP is infantry, and they should still lose their ability to fight long before that.

• For added bonus funtime: have system loss also affect mobility so that units that lose their offensive capabilities run away really fast, until mobility damage sets in and they instead have to run away really slow… (or just have two stages of reduced mobility if you're boring).

• Tie triggers into not just the revised death limit, but also to the “non-operational” and “immobile” thresholds so those can be used as mass triggers to score points and achieve objectives with ease.

 

ED have already indicated that they're working on a ground vehicle damage model update Later™, so this kind of stopgap isn't likely to happen and thus not worth a full wishlist thread, but at least that last point will still need to happen.

[Shadow KT]

Waiting on the video.

[Xavven]

5 hours ago, Shadow KT said:

Waiting on the video.

 

Looks like it's up

 

It does seem to be a fairly direct response to threads like this one, and a particular one on reddit.

 

[Tom Kazansky]

It's hard not to be a ED fanboy.

[PanzerPunisher]

'SA-10s can engage them...but in this engagement I've turned that off in the Mission Editor'

Really?? Might as well have used static objects.

[Tom Kazansky]

5 minutes ago, PanzerPunisher said:

'SA-10s can engage them...but in this engagement I've turned that off in the Mission Editor'

Really?? Might as well have used static objects.

The video is about the damage of the 154A. How many of the JSOWs reach the site is up to us to examine.

Edited July 11, 2021 by Tom Kazansky

[Silver_Dragon]

The Sa-10 appears has impacted by four JSOWs, and a detail, the Right "big bird" adquisition radar (at right of HQ trailer) appear as redone, as a new 3D model and the damage model.

Edited July 11, 2021 by Silver_Dragon

[Frederf]

CEMs are effective anti-armor weapons which can kill main battle tanks. A video of a salvo of cluster weapons with moderate submunition density destroying soft targets arranged in a compatible array is no surprise. Show a test where a dev tool detonates a single BLU-97 at  0, 1, 2, 3, 4, 5, etc. meter away from target types X Y Z A B C and post the damage values. Then analysis in earnest can begin.

[nighthawk2174]

There 

1 hour ago, Frederf said:

CEMs are effective anti-armor weapons which can kill main battle tanks. A video of a salvo of cluster weapons with moderate submunition density destroying soft targets arranged in a compatible array is no surprise. Show a test where a dev tool detonates a single BLU-97 at  0, 1, 2, 3, 4, 5, etc. meter away from target types X Y Z A B C and post the damage values. Then analysis in earnest can begin.

Take note of the truck at 1:25, if there were people in that truck they would be dead.  Same thing should apply to the missiles at the SA2 site and SA10 site.  And anyone in the command vehicles in the site.

[Frederf]

I started to do an analysis but quit because it's a lot of work. My assumptions were 300 frag lethal for 20m originating at the BLU-97 in elevation -0 to +60, 360 in azimuth. One could figure out with enough effort what density of BLUs are required to have a 20-50-80% chance of a 1m cube being intersected by one of those trajectories statistically. I've seen quotes of 6-12mm armor penetration.

[nighthawk2174]

Just now, Frederf said:

I started to do an analysis but quit because it's a lot of work. My assumptions were 300 frag lethal for 20m originating at the BLU-97 in elevation -0 to +60, 360 in azimuth. One could figure out with enough effort what density of BLUs are required to have a 20-50-80% chance of a 1m cube being intersected by one of those trajectories statistically. I've seen quotes of 6-12mm armor penetration.

That amount of pen for the frag within that distance seems fair as most lightly armored targets have that amount of armor or less.

[Floydii]

Is it just me, or are people just assuming that because the CBU87 and JSOW-A have lots of bomblets that they are going to fall on all targets everytime? What is the evidence of real-world effectiveness against multiple armoured targets? 

 

The blast/frag effect is only effective against unarmoured targets (trucks, radars, tents, stores, aircraft, people etc.) Anything with actual protection needs a bomblet to land directly on it to have any effect beyond flat tires or busted optics. This is a probability game that has a pretty wide range of possible results.

 

The meta of DCS has AFVs as the primary focus of most ground attacks (likely as a result of everyone simming tank busting in the A10C and KA50 for 10+ years). In reality there are 10 unarmoured targets for every AFV out there, and 100 targets for every MBT. CBUs become alot more attractive (due to increased target effect) when your targets aren't always tanks. 

[nighthawk2174]

3 minutes ago, Floydii said:

Is it just me, or are people just assuming that because the CBU87 and JSOW-A have lots of bomblets that they are going to fall on all targets everytime? What is the evidence of real-world effectiveness against multiple armoured targets? 

 

The blast/frag effect is only effective against unarmoured targets (trucks, radars, tents, stores, aircraft, people etc.) Anything with actual protection needs a bomblet to land directly on it to have any effect beyond flat tires or busted optics. This is a probability game that has a pretty wide range of possible results.

 

The meta of DCS has AFVs as the primary focus of most ground attacks (likely as a result of everyone simming tank busting in the A10C and KA50 for 10+ years). In reality there are 10 unarmoured targets for every AFV out there, and 100 targets for every MBT. CBUs become alot more attractive (due to increased target effect) when your targets aren't always tanks. 

Sure but even in wag's video the frag from the bomblets probably would have obliterated the SA2 site and SA10 site launchers potentially causing fires/explosion's.  And on the convoy the BTR's probably would have been shredded and the BMP's the same as while their front would be resistant the lower sides and rear are not and probably would have killed the crew.  Their weakest armored points would probably resist the frag if they were outside the main pattern area but inside I have doubts.  Now ofc a direct hit would probably severally injury or kill crew.  At best it only kills the engine which eats the shell and the crew bails out.  Bailing out is something that should happen more often as is if the tank is penetrated the crew is probably going to bail out.  There are even cases in Syria of poorly trained crews bailing out after a hit even though they did not got penetrated.

[Tippis]

That's basically the problem in a nutshell: only direct hits kill, but not all the things that direct hits should kill, and indirect hits don't kill at all, not even the things that it should kill.

 

All of that is because what qualifies as a “kill” in DCS is to reduce the hit points to zero, but almost by analogous-abstraction-and-simplification-definition that is not what cluster submunitions do. Unless we're talking about meat targets, where a single frag could be equivalently lethal, bomblets are not intended to make anything go up in a big ball of fire — they're meant to punch holes in things that don't work if they have holes in them.

 

Even besides the absolutely lacklustre representation of frag and blast damage, cluster bombs as they are represented in DCS are inherently incompatible with the ground vehicle damage model. They incapacitate and “kill” targets in a way that the damage model is simply incapable of representing unless the submunition damage is boosted to tacnuke levels.

[NineLine]

Hey guys, I appreciate a good bug thread, but I am not sure where we are going with this one, what are we expecting? HE submunition kills of tanks? 90%-100% kill ratio on all vehicles no matter their armor? The damage model in its current state to do more than it was ever designed to do with out upgrading it?

 

Please include tracks and examples back with documentation, also keep in mind the damage model is planned to be improved, as well even then, the cluster weapon isn't a kill all weapon.

 

Keep in mind, things DCS needs improvements on that we have noted:

• More Robust DM for ground vehicles
• Better reactions to enemy fire (suppression, improved dispersion, etc)

  43 minutes ago, Tippis said:

  only direct hits kill

Things like this are too vague and have no sense, a tank needs a direct hit with the proper shell to score a kill. I need specifics, I need good reports, or this might as well be in Chit Chat. Please don't make flashy statements without backing tracks and such, otherwise its just chest beating and wont help exact any sort of change if needed.