The Weapon System Concept...

[pathology_doc]

...as explained in Buttler's book on US Secret Bomber Projects, is a wonderful thing. There's no doubt that it particularly applied to the aviation industry in the late 1940s and onwards, at least in terms of integrating structure, propulsion, avionics and armament, as various things (engines, the enormous amounts of fuel jets required, bulky avionics, internally-stored missiles where relevant, etc) began to compete for the previously cavernous internal spaces of new aircraft.

I've just finished reading Norman Friedman's book on British cruisers in and after the world wars, and have long been a fan of D.K. Brown, and it struck me that really, the whole "weapon system" concept was well understood by the battleship builders, except that of course they were also required to integrate armour protection into the design. On the other hand, there was compensation inasmuch as the sensor/fire control aspect didn't really come into it until the end of the pre-dreadnought era. Friedman mentions the frantic rejigging of designs in order to get the equipment fit deemed necessary into cruisers that had to obey treaty (and also affordability) limits, and the inevitable overloads (and X-turret sacrifices) that were required as more and more war equipment had to be loaded on.

This and the ever-fluid design process surrounding the British interceptors that died with the Sandys review drove me to hypothesise that the weapon system concept is a really marvellous thing when your technology (and the doctrine that surrounds it) is mature. It is, however, potentially disastrous to the development of a project (especially when said project is on shaky ground anyway) when one or more of the doctrine, weapons, sensor/fire control systems and propulsion wind up in what tvtropes.org calls "Development Hell" or drags others into it, or is evolving too quickly for a design to be frozen for integration. Get your doctrine wrong, or find yourself caught short by a newly evolved requirement, and suddenly the integrated approach comes crashing down like a house of cards.

It's all very well to design a neatly wrapped package; but unless you remember to include stretch ("space and weight reserved" matters as much for post-war aircraft as for ships), what you end up with might be no better than the lash-up you could have had years earlier (e.g. by shoving AI23 into a slightly redesigned Fairey FD2 nose and pinning a pair of Firestreaks to the wingtips), or - if the government axe swings - it might be nothing at all. To say nothing of what happens when the best (fire-and-forget) becomes the enemy of good enough: e.g. a call for SARH missiles in the intercept role, obliging radar manufacturers to integrate illuminators into fighter radars from the start of the design phase, or the delays and cost overruns inherent in the Avro-Canada Arrow's fire-control system and proposed missile armament, possibly providing the straw that broke the camel's back, when what they could have had from the start (a FCS using Falcon) was more or less what they eventually got when they bought the CF-101.

Sometimes it really does make sense to build a nicely shaped empty shell with the biggest engines, and then fit the best systems you have at the time inside of or onto it. At least now, whatever weapons we might develop, we can only expect the electronic systems we depend on to drive them to get more and more compact.

 

[Nik]

"...we can only expect the electronic systems we depend on to drive them to get more and more compact."

And much more expensive, with a lead-time longer than the air-frame ??

 

[sferrin]

"...we can only expect the electronic systems we depend on to drive them to get more and more compact."

And much more expensive, with a lead-time longer than the air-frame ??

Not if they were smart about it and standardized. Look how freqently PCs get upgraded, and how much they've increased in power over the years.

 

[royabulgaf]

I know, Sferrin. Why can't they have some sort of standardized rack setup? Each new big project, including civilian projects (how many airlines are flying the 787? Has the A380 backlog been cleaned up yet? ) seems to have a problem. When you have a new engine, new airframe, new electronics, new materials, new fabrication techniques, at least one is guaranteed to be a non-starter.

 

[Grey Havoc]

Not if they were smart about it and standardized. Look how freqently PCs get upgraded, and how much they've increased in power over the years.

Problem is, that systems using COTS components (both hardware and software) make major tradeoffs in terms of things such as microprocessor reliability (a far greater problem than you might think) in order to be so cheap and to meet (often crude) performance requirements. It one of the reasons that PC (and COTS in general) based systems are not recommended for mission critical applications, i.e. if the system dies, people tend to die or get injured.

Just because you have a COTS spec system that has seemingly more processor cycles available than a MILSPEC system, does not not make it a better system than an older, but better designed, more reliable and often, more capable, MILSPEC system.

 

[pathology_doc]

On that note, I vaguely remember reading something (I think it was late 80s or early 90s) about a plan to stretch out the life/extend the capabilities of late-war/early post-war American DDs and DEs sold to developing nations, by equipping them with a data processing net built around off-the-shelf PCs?

I think at least Bill Gunston and possibly some others have made the point that (to paraphrase) yes, your average MILSPEC processor is no faster than an IBM 8086; on the other hand, how many high speed PC motherboards will keep on working when nuclear explosions with the attendant EMP bursts are going off all around them?

 

[GeorgeA]

Also don't forget the often overlooked parameter of software reliability. The most advanced silicon possible is not as useful as software that has been tested and has mission critical levels of reliability. That's where the labor and the schedule slips really come from these days. A hardware upgrade that requires recompiling and retesting of code is a huge money sink and schedule killer.

 

[pathology_doc]

Fascinating to come back to this topic after exactly a year (!!), and yet again a Friedman book ("Modern Warships") was the cause. Contrary to what Friedman posits, I think the weapon system concept was well understood and successfully applied by the designers of battleships, armoured and protected cruisers; except of course back then the system was armament plus protection plus propulsion rather than armament plus sensors plus propulsion as it is today. I can't think of a more integrated approach than the protected cruiser concept, which in essence used the fuel tanks (and the fuel itself) as part of the armour system.

The rise of the sensor element came about with centralised fire control, of course, and the first battleships to be designed with e.g. Dreyer fire control tables inbuilt rather than fitted could be argued to be the first true weapon systems - a vehicle designed to meet a specific threat with weapons, sensors, structure, protection, and propulsion incorporated in a design package before the first plate is laid or rivet is fired. Perhaps this honour would best have gone to the G3 battlecruisers, whose design was IIRC already beginning to incorporate rapid-fire AA armament (and would have had space for more). Everything after that was compromised by the need to meet treaty restrictions; and by the time capital ship design was freed from those, the battleship age was already at an end.

Back then, of course, the enemy's weapons were impossible to jam, spoof or destroy without destroying the platform anyway, and you had to count on being hit and being able to survive those hits, a state of affairs which arguably persisted until the end of the Second World War (if you lost your radar you could fight still; just nowhere near as effectively). All sensors were passive sensors, whereas nowadays we agonise about EMCON. The fact that the most potent weapons a surface ship possesses (excluding air groups) can be shot down, diverted or confused was always going to change the focus of the systems approach, especially since one hit can (depending on the missile and the target) be the end of you.