Hacker News new | past | comments | ask | show | jobs | submit login

The F-35 isn't that "bleeding edge" though. Most of the "innovation" comes from trying to accomplish a bunch of different roles with a single aircraft (shockingly, this has resulted in a plane that excels at none of them).



So VR helmets, EW antennas embedded in stealth skin, and crazy combat networking are all trivial features that we should expect a gaggle of government contractors to get right on the first pass? I'm not sure I buy that argument. The plane looks plenty fancy to me, at least in the context of its alternatives ("first lol" tech demonstrators don't count).


Both the B-2 and F-22 have embedded antennas. JshWright's point seems to make sense: the complexity comes from system's integration and software, compounded by the aircraft having to be too many things to too many people.

Basically, software, of which there's a ton to say regarding F-35 development. All the features required and component integration required them to adopt an agile methodology, moving away from the Ada-style development processes and, arguably hacking things together until they kinda worked. The predictable result is a long-tail of bugs that may never be fully eradicated.


What's with the "embedded antennas" references?

Microstrips have been used in planes for decades.


I think the point is "embedded in stealth skin".


> "VR helmets,"

AR, not VR, right? Pretty sure they were doing that in helicopters years ago. Doesn't the Apache have a 30mm autocannon slaved to the pilot's helmet display, such that it shoots where the pilot looks? I'm no helmet geek, but I think that stuff came out in the 80s.

I suppose the F-35's implementation is cooler, on account of being newer, but it doesn't seem earth shattering to me.


> AR, not VR, right?

Probably depends, but the way I see it, AR is when you pass the world through but draw stuff on top of it, whereas VR means replacing what you see wholesale. So for instance, a HUD is AR, and F-35 helmet is VR.


> Apache have a 30mm autocannon slaved to the pilot's helmet display, such that it shoots where the pilot looks

The same system also exists for the Eurocopter Tiger


Sure, they are both ar. But one is about as nice as a virtual boy, while the f35 is an occulous. Huge differince in tech.


You've lost me at 'crazy combat networking'.

But next time I talk to my CEO I will try this with 'we do some crazy networking'.


I mean having a look around at some of the technical innovation in this space seems to show that there are many firsts happening with the airframe.

https://www.wired.com/2016/06/course-f-35-comes-400000-augme...

They may not be firsts overall, but it seems like a pretty complicated environment to perform testing and development in.

I have to wonder what getting a software build out for a fighter jet must be like, and getting someone up in the air in different environments to test it must make the testing process daunting.


But would the alternative of designing and manufacturing many different aircraft be cheaper? And would it be as effective?

For it's faults, the F-35 is a weapon that can be configured and deployed at the drop of a hat for just about anything. That's a very powerful capability to have as a defensive and offensive weapon.


> But would the alternative of designing and manufacturing many different aircraft be cheaper? And would it be as effective?

Yes. The military has tried the multi service aircraft idea in the past, wasted a lot of money, and failed completely to develop one of the desired variants. That was the F-111, which was supposed to have a Navy and Air Force version.

The F-35 was supposed to feature 80% parts commonality between the Air Force, Navy, and Marine Corps versions. Instead, it ended up being 20%.

With that little parts commonality, the different services should've just developed their own planes. All three services would've ended up with a more capable aircraft for their requirements that wasn't so heavily compromised. And it in all likelihood would have been cheaper.

> For it's faults, the F-35 is a weapon that can be configured and deployed at the drop of a hat for just about anything. That's a very powerful capability to have as a defensive and offensive weapon

It can fly a multitude of roles but it is not as good at many of them as a more focused aircraft would be. Considering the massive costs it was a bad trade off.


The F-111 is the right comparison class. It was supposed to do everything; got relegated to the role every failed airframe does: attack aircraft. Basically a shitty expensive bomber. Just as the F35 is. Dipshit Generals who are bribed with Lockheed jobs afterwords will try to convince you that's really what they wanted; it ain't. It is an egregiously failed weapons system, though an excellent piece of modern political corruption. Yay we have a plane that costs $300 a troy ounce which can be used to blow up cavemen in Afghanistan.

Something a cloth winged Sopwith Camel biplane from WW-1 would probably actually be better at; certainly more cost effective.

Nerds who are touting this turkey because "woo woo cool half million dollar vidya game helmet" ... y'awl make me want to renounce my citizenship. This turd can be defeated by a Russian tractor mechanic detuning his radar slightly.


Exactly, every ground attack air plane from WW2 with added flares would work fine.

Some in the Airforce recognized this and it lead to the light-attack aircraft competition.

But currently it's stopped, I guess there was not enough money in it and was killed by a competing faction in the Pentagon.


See also the disastrous development process of the Bradley Fighting Vehicle.

https://en.wikipedia.org/wiki/Bradley_Fighting_Vehicle#Histo...


I'm baffled that the largest airforce in the world is so desperate for a jack-of-all-trades aircraft. It makes a lot of sense for a small airforce. The F-16 was incredibly popular with smaller airforces because it could do everything. Never the best at anything, but good enough for a small airforce.

But the US can afford specialised aircraft. Betting the house on a jack-of-all-trades means you end up with a master of none. The US should be master of everything, and that means more specialised fighters, bombers, CAS, etc.

If the A-10 isn't good enough anymore, by all means design a replacement. But don't expect it to also take of vertically and perform air superiority. Everything comes at a trade-off. Having an airframe designed for VTOL and then use it for planes that don't VTOL is just wasteful.


> I'm baffled that the largest airforce in the world is so desperate for a jack-of-all-trades aircraft.

A big motivation was that developing combat aircraft is stupidly expensive and extremely prone to cost overruns, which then make programs vulnerable to cancellation. So the defense industry and DoD got together to make this project very difficult to cancel. They sold the virtues of cost savings by sharing components, then they staked the future of US and allied air power on this one platform. On top of that they spread the production of the aircraft across FORTY-FIVE states. It is not only the most expensive weapons program in history, it is the most politically engineered.

> But the US can afford specialised aircraft. Betting the house on a jack-of-all-trades means you end up with a master of none. The US should be master of everything, and that means more specialised fighters, bombers, CAS, etc.

Yes, but one of the critical weaknesses of the US military's philosophy is that if it's worth doing, it's worth overdoing. The DoD frequently capitulates to the desire to gold plate weapons systems and change requirements to the point that projects get cancelled. So, the JSF is in a way, a response to that tendency, just the totally wrong one.

> If the A-10 isn't good enough anymore, by all means design a replacement. But don't expect it to also take of vertically and perform air superiority. Everything comes at a trade-off. Having an airframe designed for VTOL and then use it for planes that don't VTOL is just wasteful.

I think the STOVL requirement is key factor that really effed up the JSF program. And that entire requirement is born out of an irrational institutional fear the Marine Corps has about being left without air support that they harbor from WWII.


I can totally understand the Marines wanting a VTOL/STOVL aircraft. Just don't make it the same aircraft that the airforce and navy use for non-VTOL missions. That's just putting totally irrelevant restrictions on fighters that have nothing to do with VTOL.


The F111 was designed as an Air Force aircraft. period.

They later built two prototypes to test for the Navy, decided a common airframe wasn't going to work, and built the F14 using the lessons learned from that development effort instead.

The three variant JSFs are essentially three separate aircraft run by one program office.

The USAF and USN used a common aircraft in Vietnam to great effect -- the F4.

The JSF per aircraft cost is going to be less than the F18 by the time the program is finished.


> The three variant JSFs are essentially three separate aircraft run by one program office.

They're actually shutting down the joint program office because it is now in fact, three separate aircraft. Which again, puts lie to the whole rationale of the program in the first place.

> The USAF and USN used a common aircraft in Vietnam to great effect -- the F4.

The F4 did decently, and if we use it as a guide, the Air Force and Navy variants should have been their own program, and the USMC should've had a separate airframe to accommodate their STOVL requirement. But that's not what we got, and that STOVL requirement is a large reason the platform has some big fundamental compromises.

> The JSF per aircraft cost is going to be less than the F18 by the time the program is finished.

People keep promising that, but the DoD and Lockheed keep laying games with how they quote costs in order to make it seem like cost parity is closer than it really is. Flyaway costs for the F-35 keep getting quoted against procurement costs for 4.5 gen aircraft to make the F-35 look better than it is on cost.


The cost actually is. It's not a thought about something in the future, it's a fact.

The USMC used the F4 alongside the USN as they always have. USMC aviators are trained as Naval aviators, deploy with USN ships, and support USN missions as part of the CVW. The ARG concept is different, and USMC STOVL fixed wing wasn't a thing in Vietnam. It's a more modern concept. You're projecting things that didn't exist into the past.

There are no compromises between the platforms for STOVL -- the JSF-B is a separate aircraft as it has been since the beginning and as you say in the sentence above. JSF-A and JSF-C are different airframes with different requirements and different capabilities.


The airframe designs of the F-35A and F-35C were heavily-compromised by the F-35B's wide lift fan. That single requirement makes ALL F-35's wider, heavier, and with worse drag than they would be otherwise.


They're separate airframes. There are literally only a very small number of components (mostly the engine) that are the same across all three variants, and all three variants have different dimensions and flight characteristics.


The fuselage across all three variants is the same, as the other poster stated, the result has been that the F-35 has worse aerodynamic performance, is heavier, and has a wider radar cross section than it would have had if it didn't have to accommodate the STOVL requirement.

So while the different variants have different wing surfaces there is no denying that the Air Force and Navy versions would've been designed differently if the airframe didn't have to fit a lift fan in it.

So then you're back to the fundamental question of why even bother. Parts commonality is 1/4th of what it was supposed to be. Costs are nowhere near what they were originally billed as, even with the most optimistic estimates.

Really the only virtue of the JSF approach was that it made the program too big to fail.


The fuselage between the three is not the same. The A and C model fuselage is not the same as the B.

The size of the aircraft is driven by the need to carry large amounts of internal fuel and all internal stores, not the need for a lift fan for the B model. This allows all three models to operate with a reduced cross section for an expanded set of missions, which is the purpose of the aircraft.

Parts commonality is essentially non existent except for the engine and parts of the avionics -- that does make a difference and is a cost savings, but it's not the issue you make it out to be.

Costs are better than they were sold as, and their availability rates are better than they were sold as, and their accident rate is so low as to be non existent. It's a phenomenal aircraft and a capability that's going to completely change the balance of power in some regions of the world.


But if single purpose aircraft are more capable at that purpose, the F-35 still loses when it goes into battle. The fact that it doesn't suck at anything in particular only counts on a spreadsheet. No F-35 is going to be used for different purposes by the same pilot or military service. The best tool for the job — that's something the F-35 is not now and never can be.

When on a bombing run radar stealth and bomb load matters, but acrobatic agility and VTOL / STOL doesn't. On a carrier, durability matters every time you slap the deck and VTOL /STOL might be very handy, but stealth or high altitude capability, not as much.

Designing a fighter/bomber to please every service is like training every soldier, sailor, airman, and marine to have the same fighting skills. It's going to get them all killed.


It might not be cheaper, though I suspect it would be just because white elephants will white elephant. And besides, what's the likelihood that three separate projects will fail? But the likelihood that a jack of all trades will be a master of none, and thus perhaps ineffective, is pretty high.

> at the drop of a hat

Ha! There are still multiple variants, and you cannot reconfigure instances of one variant to be instances of another, and certainly not at the drop of a hat.


>But would the alternative of designing and manufacturing many different aircraft be cheaper?

Ultimately yes.

You can be much more agile when you aren't trying to satisfy a large collection of armed services, nations, and mission roles many of whom are mostly interested in putting their mark on something instead of making it better.

>And would it be as effective?

Not if your definition of effective is ability of one aircraft to do everything. If your definition of effective is forces ability to achieve goals, then yes a set of aircraft specialists can be better than one generalist.


Just basic post-flight download of data takes more than the flight. The reconfiguration happens only in two or three places in the world and can take weeks if not months and then you need to have high-speed Internet access to get it.

Quick turnaround is something yiu get from Soviets/Russians or Swedes, where the doctrines involved make it priority feature.


I suspect most of the innovation is actually in the acquisitions side of the house. They could burn money for decades on this project and still come out ahead compared to running half a dozen acquisitions projects for multiple planes.


Could you explain this argument in a bit more detail please. Why and how is this way better, what's the comparison, what are the main cost drivers/differences?


Sure, take this chart[0] and multiply it times the number of air frames one would anticipate needing to accomplish various missions. Not only are the bureaucratic meetings amplified, but the O&M long tail starts to kill you with the logistics of moving and storing parts, training specialists on each aircraft, administering the additional specialists... it keeps going and going.

As an aside, I didn't say it was "way better", I just said they would come out ahead even burning money for decades.

0: https://www.dau.mil/tools/t/Department-of-Defense-Acquisitio...




Registration is open for Startup School 2019. Classes start July 22nd.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact

Search: