Subsystem damage ideas

[klauss]

Ehm... I guess I'm showing English isn't my mother language.
I meant: "Ferrius should know what I don't, and should be able to clarify".
I guess what I wrote reads as: "Ferrius should have known better".



BTW: "Dumbfires are armed upon launch" - interesting... I'd like to model that ingame. Would they explode if hit?

PS: I may have left too much implied. A hornet on grass would be an out-of-order hornet. I don't think those landing gears are able to handle uneven terrain, so landing/taking off/rolling on grass would wreck them.

[Privateer Ferrius]

Yes. Any missile can be destroyed by fire when they are in flight. This is the purpose of the Phalanx cannon on Canadian naval ships

[klauss]

I meant while on the hardpoint.
I'm thinking of something along the lines of "Internal ammo explosion" on BattleTech.

[Privateer Ferrius]

This could not be done, no. Only when it is in flight could a missile be destroyed.

(I never understood that in BattleTech, but CASE II made it pretty impossible once that came around anyway )

[klauss]

Ok... pretty much what I thought.

Now, imagine you fire your missile, and it gets stuck (doesn't leave the hardpoint). It would be armed, and ready to blow... right?

Would WC benefit of such modelling? Or would hardcore WC fans kill me?

[Privateer Ferrius]

I think the question is more: how much would you gain for the volume of effort you'd put into doing that?

[klauss]

Much, I think.
I've been thinking about adding mechanisms for complex subsystem failures in almost every place.
The problem is that damage should be noticeable... not just a color in the armor indicator. When you get hit, you get damaged. That means part of your ship will not work or become unreliable.

The missile thing would be a particular case of the former.

Hm... split?

[Privateer Ferrius]

Hrm. Splitting is a good idea.

Also, the chances of a missile getting stuck are infinitesmily small. Pretty much nonexistant.

[klauss]

Not even under heavy damage situations?
I find that hard to believe. So many things could go wrong.

[Privateer Ferrius]

In the real world, there have been 15 missile hardpoint malfunctions of that matter. In the history of military aviation. If you've taken enough damage to cause that kind of problem, you're probably dead anyways.

So yes, it is pretty improbable. The chance pretty much goes into statistical insignificance.

[klauss]

I see...
...so... what kind of malfunctions are probable enough, then?

[Privateer Ferrius]

Targeting malfunctions are much more probable. There are tales of heat-seeking missiles locking on to the heat signatures of friendly craft, or even the pilot's own.

Also, radar doesnt stop working unless its broken - sometimes damage causes it to report incorrect data as opposed to not reporting data.

As well, a common malfunction in high damage situation is the FOFI (friend or foe identification) system reporting craft incorrectly as friend or foe.

[klauss]

As well, a common malfunction in high damage situation is the FOFI (friend or foe identification) system reporting craft incorrectly as friend or foe.

Interesting... I would imagine that one would have been designed to avoid that. You know... it could be disastruous. And being a cryptography dude, I know it can be done.

[Privateer Ferrius]

All you have to do on a basic levels is have voltage high instead of low, and bad things happen The software can only interpret those 1s and 0s the electronics give, after all.

[klauss]

IFF is done on software - it checks whether a message exchange between the two systems decrypts correctly, if what I know has any accuracy.

What could go wrong is the connection to the IFF light - but since on modern craft that goes to the PPI radar screen, then such malfunction would be really odd - what kind of hardware malfunction turns an icon on a CRT into another?

It would be rather straightforward to design the system to interpret a malfunction so that it shows all craft as neutral, rather than friend or foe - at least, that way the pilot knows the system is malfunctioning.

[Spaceman Spiff]

Don't know if that might help you. Here is a list of possibel subsys fallouts i could imagine:

Schields:
-complete fallout
-fallout in one sector

Radar/Auototracking:
-complete fallout
-fallout friend/foe function (all vessels might turn up friendly)
-fallout targeting fungtion
-complete fallout of the autotracking funktion
-partial fallout of the autotracking funktion (might always aim too much on the right, for example (this would make it necessary to switch the AT offline... *3)

Engines:
-lower energie levels
-lower highspeeds
-afterburner can't be switched off
-damage might effect flying attitute (ship could aways roll slightly left)

Weapons:
-damaged would have a complete faltout
-damaged guns would need a huge amount of energie (would make it necessary to be able to switch singel guns offline *1)
-missle hardpoint fallout which leads to heavy damage or destruction of the own ship (this would make some sort of testprogramm necesarry, which the pilot can run if the ship is heavy damaged and will tell him if the hard points are still operational. *2)
-fallout friend/foe missles; image missles (might be radar missfunction)

Armor:
-damaged pieces of armor might limit cockpit view or lock other systems (like missles or turrets).

If we introduce such damage effects,I'd consider it important to also introduce imergency actions which will LIMIT the effect of the damage. take *1-3 as examples...

EDIT: Somebody ought to check the fallouts in Privateer...

[Privateer Ferrius]

IFF is done on software - it checks whether a message exchange between the two systems decrypts correctly, if what I know has any accuracy.

What could go wrong is the connection to the IFF light - but since on modern craft that goes to the PPI radar screen, then such malfunction would be really odd - what kind of hardware malfunction turns an icon on a CRT into another?

It would be rather straightforward to design the system to interpret a malfunction so that it shows all craft as neutral, rather than friend or foe - at least, that way the pilot knows the system is malfunctioning.

I don't think you understand the underlying electroincs. The 1's and 0's correspond to a high or low electronic pulse in the underlying circuitry. All you have to do is have a malfunctioning capacitor somewhere and the brown stuff hits the fan. Software has checks and balances against one or two capacitors acting up, but when they all are, this data checking causes bad data, because the software is pretty much guessing at that point what data is good.

[chuck_starchaser]

Sure, Ferrius; like you have a perfectly deep understanding of the underlying electronics... LOL! Capacitors...
What Klauss was saying, FYI, is that the friend/foe signal would NOT be communicated between one subsystem and another as a single bit, 1 or 0, but as part of a larger packet of info that would include error detection and possibly error correction codes.
Capacitors...
The types of capacitors that would in any likelihood fail would be the bulky and crappy electrolytics, used for the power supply. If the power supply fails, everything fails; you wouldn't get friend and foe confused; you'd get a blank screen.

[Privateer Ferrius]

Sure, Ferrius; like you have a perfectly deep understanding of the underlying electronics... LOL! Capacitors...
What Klauss was saying, FYI, is that the friend/foe singnal would NOT be communicated between one subsystem and another as a single bit, 1 or 0, but as part of a larger packet of info that would include error detection and possibly error correction codes.
Capacitors...

You do know what it does, right?

Error correction codes aren't what would be sent. More likely an MD5 checksum (or whatever is the WC equivalent) The problem lies in when the error-checking generation works properly but the circuitry or software generating the actual information is malfunctioning.

Regardless of the mechanincs of the matter, it HAS happened before IRL. And it's happened in no damage situations too - Patriot missiles, anyone?

[chuck_starchaser]

I was talking "between subsystems". But same thing applies. If you have a ship with a friendly transponder that sends a CRC or MD5, if the signal is not received correctly, there's a chance in billions the signal would be wrong, yet pass the checksum checking. Now, if the signal doesn't pass the check, the heuristics wouldn't assume that it is therefore a foe; it just says "bad signal".
And before you say "what if the circuit that checks the checksum fails?", well, that's not a "circuit", but the same processor that does a lot of other computations for that subsystem. And processors don't get partial damage very easily; either they work or they don't. Now, I won't say it's "impossible" to get ff confusion, but it IS highly unlikely. One possible scenario is, if your transponder has stopped working altogether, and you're in the middle of hostilities, and your ship's computer is using a purely heuristic, backup identification of friend or foe, that's a lot more likely to cause confusion.

That Patriot, btw, what I think I read was that the British plane hadn't been given the right codes, and the Patriot was programmed to launch on any unidentified. We're talking about damage; not about low IQ -related problems

[Privateer Ferrius]

I was proving an exxagerated example. Things like that happen all the time when systems are damaged.

By the way, do remember that if you're not broadcasting FOFI that you're painting a big red bullseye on you for FOFI-based missiles.

[chuck_starchaser]

Yep, missiles are too expensive to waste on uncertainties. It's true. Even the lowliest of them run into the tens of G's.

[Privateer Ferrius]

Another idea: ammo-feed failures. For the mass drivers, that is.

[klauss]

Also... chuck... attacking my own argumnet, military systems get produced by lowest-bid contractors - they're not necessarily top-notch designs

For instance, I think the military is still using LFSRs on their cryptographic devices. Say what you want about them, that you could design secure LFSR-based cryptosystems, but still there are too many ways of attacking them. Yet, they use them because "you can implement them with ASICs and have them be totally secret" - ya... like secrecy is the best way to achieve cryptographic security...

[Privateer Ferrius]

Another idea: ammo-feed failures. For the mass drivers, that is.

Quoted for emphasis