Why did the F-22 program not cater for a B or any subsequent upgrade version? I assume there could be minor upgrades to existing airframes, but this must surely be limited. Why was the program halted with such certainty?

The answer to all your questions is money. Building them and operating them was super expensive, and the F-35 was to be *much* cheaper.

T54A wrote:

Why did the F-22 program not cater for a B or any subsequent upgrade version? I assume there could be minor upgrades to existing airframes, but this must surely be limited. Why was the program halted with such certainty?

The stealth technology and upkeep for that era were expensive, and the threat the F-22 were meant to counter in Europe (the Russian Air Force) was no longer able to produce significant numbers of similar quality air superiority fighters. So the focus shifted to strike fighters like the F-35 and the Super Hornet.

The F-16 survived as an export model and because it had been adapted to the strike role. Same for the F-15, to a lesser extent. But the air superiority role of the F-22 had been met with the existing numbers.

Same would still be true today, if not for the rise of the Chinese capabilities. But the F-22 is not well suited for that conflict, in terms of range and forward deployment facilities. Neither is the F-35, really. So we really need the NGAD for that theater.

Big money without a true peer opponent. The F-22 (ATF) was initially planned to counter the soviets, the YF-22 and YF-23 flyoff concluded 4 months before the fall of the Soviet Union. As it became clear that Russia didnt have the economy to be a super power, there was little need for as many of those really expensive F-22s. 10 years later the JSF program had its flyoff and part of that program was to be cheaper to buy and operate than the F-22. Other than some more routine upgrades the F-22 hasnt changed much since introduction, and with the F-35 being newer and filling some of the gaps with the aging F-22, there is no reason to make a new version. Now 17 years in service, the F-22 is up for replacement, and seemingly the NGAD is closer to being a real plane than anyone though... leveraging F-35 materials and technology, the expensive to operate F-22 is likely looking at retirement in the not too distant future. There is very little that can challenge the F-35 right now (if a F-35 gets in a dog fight with something better, the F-35 pilot messed up), Russia and China's 5th gen are still in development so not a numerical threat... the USAF could still dominate without the F-22.

As everyone else said, $ was the biggie. As the USAF tends to make a B model a two-seater there wasn’t an appetite in the CAF for a two-seat replacement for a single-seat jet, especially for a jet so easy to fly and employ. Apart from all that, there’s the fuel problem: a second seat generally means less gas.

Oroka wrote:

Big money without a true peer opponent. The F-22 (ATF) was initially planned to counter the soviets, the YF-22 and YF-23 flyoff concluded 4 months before the fall of the Soviet Union. As it became clear that Russia didnt have the economy to be a super power, there was little need for as many of those really expensive F-22s. 10 years later the JSF program had its flyoff and part of that program was to be cheaper to buy and operate than the F-22. Other than some more routine upgrades the F-22 hasnt changed much since introduction, and with the F-35 being newer and filling some of the gaps with the aging F-22, there is no reason to make a new version. Now 17 years in service, the F-22 is up for replacement, and seemingly the NGAD is closer to being a real plane than anyone though... leveraging F-35 materials and technology, the expensive to operate F-22 is likely looking at retirement in the not too distant future. There is very little that can challenge the F-35 right now (if a F-35 gets in a dog fight with something better, the F-35 pilot messed up), Russia and China's 5th gen are still in development so not a numerical threat... the USAF could still dominate without the F-22.

The Chinese have apparently built 150 J-20's already.

Kiwirob wrote:

The Chinese have apparently built 150 J-20's already.

They may have 150 built (I cant imagine how they are keeping photos of them off the internet), but I really dont think they are fully baked yet. The F-35 has been physically complete for 15 years, but software has been a thorn to this day. Still, 840+ mature F-35s vs 150 J-20... I wouldnt be concerned right now about a gap between the F-22 and the NGAD, especially with reports of a prototype of the NGAD already flying.

(Rant, that is part of the reason why I dont consider the NGAD a '6th gen' fighter, more likely an optionally manned 5++ gen (F-22 5th gen, F-35 5+, NGAD 5++))

LyleLanley wrote:

As everyone else said, $ was the biggie. As the USAF tends to make a B model a two-seater there wasn’t an appetite in the CAF for a two-seat replacement for a single-seat jet, especially for a jet so easy to fly and employ. Apart from all that, there’s the fuel problem: a second seat generally means less gas.

Second seat needed to control the wingmen.

LTCM wrote:

Second seat needed to control the wingmen.

Now, sure. In the late 80s/early 90s that wasn’t the case.

Oroka wrote:

Kiwirob wrote:

The Chinese have apparently built 150 J-20's already.

They may have 150 built (I cant imagine how they are keeping photos of them off the internet), but I really dont think they are fully baked yet. The F-35 has been physically complete for 15 years, but software has been a thorn to this day. Still, 840+ mature F-35s vs 150 J-20... I wouldnt be concerned right now about a gap between the F-22 and the NGAD, especially with reports of a prototype of the NGAD already flying.

(Rant, that is part of the reason why I dont consider the NGAD a '6th gen' fighter, more likely an optionally manned 5++ gen (F-22 5th gen, F-35 5+, NGAD 5++))

Until they get them equipped with the WS-15, they are nothing but also-rans

T54A wrote:

Why did the F-22 program not cater for a B or any subsequent upgrade version? I assume there could be minor upgrades to existing airframes, but this must surely be limited. Why was the program halted with such certainty?

The problem was that the F-22 held onto legacy systems for way too long in development, so by the time the F-22 actually went into production, what was state of the art at the time of development was already obsolete and approaching end of life. By the time the F-22 actually went into production, the OEM's for some of those systems were telling the USAF, LM, and Boeing that they were going to discontinue production of those systems in a few years time. Hence the reason why production was capped, with no further production.

Now, add a programming language that has a very small user base, an avionics design that had a lot of technological dead ends caused by doing stuff the first time, small production numbers, and you have a recipe for major developmental issues down the line. Basically, any small changes to the hardware or software of the F-22 requires extensive development and recertification of the entire software suite and that costs ALOT of money and time to do, which you can't amortize across a large fleet.

The F-22's avionics is a poster child for a case study for how NOT to design avionics for a combat aircraft. There was a lot of things they tried for the first time with the F-22 that were never repeated again for subsequent fighter designs because they turned out to be technological dead ends.

ThePointblank wrote:

T54A wrote:

Why did the F-22 program not cater for a B or any subsequent upgrade version? I assume there could be minor upgrades to existing airframes, but this must surely be limited. Why was the program halted with such certainty?

The problem was that the F-22 held onto legacy systems for way too long in development, so by the time the F-22 actually went into production, what was state of the art at the time of development was already obsolete and approaching end of life. By the time the F-22 actually went into production, the OEM's for some of those systems were telling the USAF, LM, and Boeing that they were going to discontinue production of those systems in a few years time. Hence the reason why production was capped, with no further production.

Now, add a programming language that has a very small user base, an avionics design that had a lot of technological dead ends caused by doing stuff the first time, small production numbers, and you have a recipe for major developmental issues down the line. Basically, any small changes to the hardware or software of the F-22 requires extensive development and recertification of the entire software suite and that costs ALOT of money and time to do, which you can't amortize across a large fleet.

The F-22's avionics is a poster child for a case study for how NOT to design avionics for a combat aircraft. There was a lot of things they tried for the first time with the F-22 that were never repeated again for subsequent fighter designs because they turned out to be technological dead ends.

The hardest thing with modern fighter development is keeping the electronics suite relevant and up-to-date.

It take YEARS to get the systems tested and validated. The lessons learned with F22 were put into play with the F35.

It effectively has a "electronics backbone" (for want of a better word) that is designed to allow upgraded systems to be swapped in as needed, as things change. However that is still a very difficult thing to with how fast things go obsolete in modern electronics and how long it take to propose, design, test then build and field them.

Add in cutting edge capabilities (such as the helmet integration in the F35) and the need to test and fix and redesign and you have a whole other level of pain to deal with.

Modern combat data and communication systems are a very difficult thing across the spectrum right now.

Tugg

42 of the planned 442 F-22 were intended to be two-seat F-22B.

Two were ordered in the pre-production batch of eleven airframes and were retained when that batch was reduced to nine in 1995. However in July 1996 the '22B was deferred for future production; it was never subsequently funded. Cutting it saved $150 million in development costs and $350 million in production costs.

In 2002 Lockheed-Martin proposed a stretched, delta-wing two-seat FB-22 Strike Raptor to replace the F-117 and F-15E, but there was no money in the coffers.

Can't say much about the F-22 in terms of performance.

Can provide insight on part of the reason why the F-22 became obsolete.

The F-22, F-117 and B-2 were part of the first generation of composite airframe. The composite design and manufacturing process were not as efficient as the next generation of composite design.

Things like composite wire brackets etc were expensive to produce. (Assumption here as I know the B-2 have some form of composite brackets, but not sure of the F-22). It takes something like the mass produced 787 and F-35 (and new composite manufacturing processes) to bring the cost of those components down.

As an asside, I had an opportunity to see one of the last F-22 wing panel skin go through non destructive inspection (NDI). Compare that to the 787 empenage structure box (which I also had the opportunity to see one going through NDI) is comparing order of magnitudes technology leap.

Titanium manufacturing also progress significantly because of and since the F-22.

bt

For those of us that prize looks over capability, it's important to note that the F-22 looks ten times cooler than the F-35.

wingman wrote:

For those of us that prize looks over capability, it's important to note that the F-22 looks ten times cooler than the F-35.

That’s a low bar! A London double-decker bus looks cooler than an F-35.

RJMAZ wrote:

Supercruise was always a marketing gimmick.

The F-22 combat radius is 590nm while subsonic. With 100nm of "supercruise" dash the combat radius reduces to 460nm.

That is a massive 28% reduction in combat radius for simply supercruising for 11% of the flight. 460nm combat radius is 920nm flight length. 100nm is 11% of 920nm.

The F-22 supercruise mission profile with a flight length of 920nm is then effectively 820nm subsonic and 100nm of supercruise.

The F-35 has a subsonic combat radius of an impressive 760nm in air to air config. The 669nm figure is with 2,000lb bombs which is far heavier than what the F-22 can carry internally.

The F-35 then has 1,520nm flight length at subsonic speed. To perform the F-22 supercruise mission profile it must fly 820nm subsonic and then fly an additional 100nm averaging Mach 1.5. The F-35 has 700nm of subsonic range remaining to put towards that 100nm supersonic part of the mission profile. With so much fuel remaining the F-35 can light the afterburner and comfortably match the speed of the F-22 supercruise mission profile.

https://en.m.wikipedia.org/wiki/Lockhee ... ghtning_II
https://en.m.wikipedia.org/wiki/Lockhee ... -22_Raptor

If only statistics could win conflicts.

Supercruise is a means to weapons delivery, not a measuring contest of which jet has the most comfortable fuel fraction to get there and stay awhile. The F-35 may be able to do what you say in relation to a Raptor, but when all the fuel weight and volume required to carry that weight is now being dumped into your burner (in relation to an F-22), that equals a sunk cost in aircraft weight in order to carry it all. With that in mind, the F-22's supercruise ability is more than "a gimmick". Especially since their altitude profiles are so very, very different.

As for the rest, It's almost like the development of one was guided by the mistakes of, and seeked to ameliorate the developmental weaknesses of, the other... Crazy.

On top of all that, their development cycles are separated by a good 20+ years and they're designed for completely different missions being accomplished in completely different ways. Still, the F-22 is nothing to scoff at, and potential adversaries certainly don't wipe their brow with a relieved "whew!" when they learn there's an F-22 nearby versus an F-35. Quite the contrary.

LyleLanley wrote:

RJMAZ wrote:

Supercruise was always a marketing gimmick.

The F-22 combat radius is 590nm while subsonic. With 100nm of "supercruise" dash the combat radius reduces to 460nm.

That is a massive 28% reduction in combat radius for simply supercruising for 11% of the flight. 460nm combat radius is 920nm flight length. 100nm is 11% of 920nm.

The F-22 supercruise mission profile with a flight length of 920nm is then effectively 820nm subsonic and 100nm of supercruise.

The F-35 has a subsonic combat radius of an impressive 760nm in air to air config. The 669nm figure is with 2,000lb bombs which is far heavier than what the F-22 can carry internally.

The F-35 then has 1,520nm flight length at subsonic speed. To perform the F-22 supercruise mission profile it must fly 820nm subsonic and then fly an additional 100nm averaging Mach 1.5. The F-35 has 700nm of subsonic range remaining to put towards that 100nm supersonic part of the mission profile. With so much fuel remaining the F-35 can light the afterburner and comfortably match the speed of the F-22 supercruise mission profile.

https://en.m.wikipedia.org/wiki/Lockhee ... ghtning_II
https://en.m.wikipedia.org/wiki/Lockhee ... -22_Raptor

If only statistics could win conflicts.

Supercruise is a means to weapons delivery, not a measuring contest of which jet has the most comfortable fuel fraction to get there and stay awhile. The F-35 may be able to do what you say in relation to a Raptor, but when all the fuel weight and volume required to carry that weight is now being dumped into your burner (in relation to an F-22), that equals a sunk cost in aircraft weight in order to carry it all. With that in mind, the F-22's supercruise ability is more than "a gimmick". Especially since their altitude profiles are so very, very different.

As for the rest, It's almost like the development of one was guided by the mistakes of, and seeked to ameliorate the developmental weaknesses of, the other... Crazy.

On top of all that, their development cycles are separated by a good 20+ years and they're designed for completely different missions being accomplished in completely different ways. Still, the F-22 is nothing to scoff at, and potential adversaries certainly don't wipe their brow with a relieved "whew!" when they learn there's an F-22 nearby versus an F-35. Quite the contrary.

Thank you.

In hindsight, would the F-23 have suffered from the same legacy issues we now see on F-22? Was it simply the best tech around at the time, or should the program have been managed better.

RJMAZ wrote:

T54A wrote:

In hindsight, would the F-23 have suffered from the same legacy issues we now see on F-22? Was it simply the best tech around at the time, or should the program have been managed better.

The YF-22 has an empty weight of 14,970 kg. The F-22 has an empty weight of 19,700kg. That is nearly a 30% weight increase and the F-22 didn't get any bigger in size. That would have played a big part in the reduced combat radius that effectively becomes the Achilles heel of the aircraft.

The F-23 would have suffered similar weight growth. The production F-23 also needed to be stretched to fit two short range missiles apparently. In hindsight longer range and lower cost was more important than super manoeuvrability. With no peer to peer conflict the ATF program should have become a strike fighter and the YF-23 might have been better in the long run.

If the production F-23 had a combat radius as good as the F-35 I think production would have lasted much longer. But this would have completely reduced the USAF demand for the JSF and we might not have the F-35 today.

With the F-22 the USAF soon realised that they had a short range, high maintenance air show machine and production ended suspiciously quick.

https://en.m.wikipedia.org/wiki/Lockheed_YF-22

No, production was curtained because a number of critical components in the F-22's avionics went out of production from various subcontractors, and there was no quick and cheap way of finding an alternative part or supplier to provide said components.

Basically, the F-22 has a very bad case of lock in caused by a number of design decisions early in the development of the F-22. This can be attributed to inexperience and trying to meet what turned out to be a flawed decision by the DoD.

In a nutshell, the F-22 avionics uses a Common Integrated Processor approach based on the Intel i960MX processor and is written in Ada. Later production F-22's have a hybrid PowerPC and i960MX architecture and some software code is ported over, which allowed the avionics to combine legacy software with newer software from the F-35 program, but that's fairly limited.

Specifically, the DoD had mandated in the 1980's that all future aircraft development use Ada as the primary programming language for the software. Good idea in theory; Ada is also very fault tolerant. Prior aircraft used a whole smorgasbord of different programming languages, some of which were proprietary which made development work difficult.

Problem is that Ada is a fairly challenging programming language to program for, with a very small programmer base with limited programming tools, and so is the number of platforms using Ada as the primary programming language. That's why today, Ada is really only used very sparingly in very specialized applications where failure is not acceptable.

Adding to the mess is that every other combat system onboard the F-22 would need to be redesigned or reprogrammed; the level of integration of the radar and ESM and it's capabilities and the onboard avionics is significantly deeper and higher than in previous aircraft. Touch one system, it affects other systems, and the entire software package for the avionics needs recertification and validation work.

Because the F-22's development took many years, and acquisition processes aren't known for being very fast and responsive, when the decision was made early on in the F-22's to use the i960MX chip, the chip was considered bleeding edge, state of the art. When the F-22 went into production, it was already obsolete and about to be discontinued because they held onto legacy platforms too long. And by the time the F-22 IOC'ed, you could literally find parts needed to keep the F-22's avionics and ground equipment going in a museum; there was no way, short of paying a subcontractor outrageous amounts of money to keep it in production for a small production run.

The i960 RISC was not a very common or popular platform to begin with (only the F-22 and the HAL Tejas, which is also a 1980s design that went IOC much later) which further complicated matters. Some of the newer F-22 CIP boards have more powerful PowerPC chips, along with FPGAs that have more compatibility with the F-35 and can share parts of the code base, but in order to replace the F-22's avionics, it will cost a whole lot of money even if they went off the shelf (say, took the avionics computers from the F-35), as they still need to redesign and reprogram everything else on the aircraft to be compatible.

The F-35 took many of the lessons learned from the F-22's development to avoid many of the pitfalls:
1. A decision was made early on to use C++ instead of Ada as the programming language to broaden the programmer base and to use standard development tools that are much more readily available;
2. A decision was also made to use the much more commercially popular PowerPC platform, which has multiple vendors making chips for, and thus a larger supplier base;
3. The F-35's avionics approach is to use “virtual machines” (VMs) extensively to avoid having core specific functions;
4. The F-35's avionics software uses a middleware so that a change in hardware does not require a complete rewrite and recertification of the avionics suite. Likewise, a change to the one part of the software does not require a complete rewrite and recertification either;
5. All major avionics components and weapon stores are connected via fiber-optic interconnects which allows for future upgrades without rewiring the jet;
6. The F-35 isolates, in discreet VMs, the individual processing of avionics data then fuses the results. This isolated each VM from any bugs or problems another VM faces. The F-22 ran multiple functions in a single computer instance so if there was a problem, it affected multiple functions;
7. The F-35 was planned for all IOC functions in its current hardware state and only every other Block requires significant hardware changes, which involve replacing, or adding new processor cards

ThePointblank wrote:

RJMAZ wrote:

T54A wrote:

In hindsight, would the F-23 have suffered from the same legacy issues we now see on F-22? Was it simply the best tech around at the time, or should the program have been managed better.

The YF-22 has an empty weight of 14,970 kg. The F-22 has an empty weight of 19,700kg. That is nearly a 30% weight increase and the F-22 didn't get any bigger in size. That would have played a big part in the reduced combat radius that effectively becomes the Achilles heel of the aircraft.

The F-23 would have suffered similar weight growth. The production F-23 also needed to be stretched to fit two short range missiles apparently. In hindsight longer range and lower cost was more important than super manoeuvrability. With no peer to peer conflict the ATF program should have become a strike fighter and the YF-23 might have been better in the long run.

If the production F-23 had a combat radius as good as the F-35 I think production would have lasted much longer. But this would have completely reduced the USAF demand for the JSF and we might not have the F-35 today.

With the F-22 the USAF soon realised that they had a short range, high maintenance air show machine and production ended suspiciously quick.

https://en.m.wikipedia.org/wiki/Lockheed_YF-22

No, production was curtained because a number of critical components in the F-22's avionics went out of production from various subcontractors, and there was no quick and cheap way of finding an alternative part or supplier to provide said components.

Basically, the F-22 has a very bad case of lock in caused by a number of design decisions early in the development of the F-22. This can be attributed to inexperience and trying to meet what turned out to be a flawed decision by the DoD.

In a nutshell, the F-22 avionics uses a Common Integrated Processor approach based on the Intel i960MX processor and is written in Ada. Later production F-22's have a hybrid PowerPC and i960MX architecture and some software code is ported over, which allowed the avionics to combine legacy software with newer software from the F-35 program, but that's fairly limited.

Specifically, the DoD had mandated in the 1980's that all future aircraft development use Ada as the primary programming language for the software. Good idea in theory; Ada is also very fault tolerant. Prior aircraft used a whole smorgasbord of different programming languages, some of which were proprietary which made development work difficult.

Problem is that Ada is a fairly challenging programming language to program for, with a very small programmer base with limited programming tools, and so is the number of platforms using Ada as the primary programming language. That's why today, Ada is really only used very sparingly in very specialized applications where failure is not acceptable.

Adding to the mess is that every other combat system onboard the F-22 would need to be redesigned or reprogrammed; the level of integration of the radar and ESM and it's capabilities and the onboard avionics is significantly deeper and higher than in previous aircraft. Touch one system, it affects other systems, and the entire software package for the avionics needs recertification and validation work.

Because the F-22's development took many years, and acquisition processes aren't known for being very fast and responsive, when the decision was made early on in the F-22's to use the i960MX chip, the chip was considered bleeding edge, state of the art. When the F-22 went into production, it was already obsolete and about to be discontinued because they held onto legacy platforms too long. And by the time the F-22 IOC'ed, you could literally find parts needed to keep the F-22's avionics and ground equipment going in a museum; there was no way, short of paying a subcontractor outrageous amounts of money to keep it in production for a small production run.

The i960 RISC was not a very common or popular platform to begin with (only the F-22 and the HAL Tejas, which is also a 1980s design that went IOC much later) which further complicated matters. Some of the newer F-22 CIP boards have more powerful PowerPC chips, along with FPGAs that have more compatibility with the F-35 and can share parts of the code base, but in order to replace the F-22's avionics, it will cost a whole lot of money even if they went off the shelf (say, took the avionics computers from the F-35), as they still need to redesign and reprogram everything else on the aircraft to be compatible.

The F-35 took many of the lessons learned from the F-22's development to avoid many of the pitfalls:
1. A decision was made early on to use C++ instead of Ada as the programming language to broaden the programmer base and to use standard development tools that are much more readily available;
2. A decision was also made to use the much more commercially popular PowerPC platform, which has multiple vendors making chips for, and thus a larger supplier base;
3. The F-35's avionics approach is to use “virtual machines” (VMs) extensively to avoid having core specific functions;
4. The F-35's avionics software uses a middleware so that a change in hardware does not require a complete rewrite and recertification of the avionics suite. Likewise, a change to the one part of the software does not require a complete rewrite and recertification either;
5. All major avionics components and weapon stores are connected via fiber-optic interconnects which allows for future upgrades without rewiring the jet;
6. The F-35 isolates, in discreet VMs, the individual processing of avionics data then fuses the results. This isolated each VM from any bugs or problems another VM faces. The F-22 ran multiple functions in a single computer instance so if there was a problem, it affected multiple functions;
7. The F-35 was planned for all IOC functions in its current hardware state and only every other Block requires significant hardware changes, which involve replacing, or adding new processor cards

Well, that’ll about cover the fly-bys…

ThePointblank wrote:

No, production was curtained because a number of critical components in the F-22's avionics went out of production from various subcontractors, and there was no quick and cheap way of finding an alternative part or supplier to provide said components.

Basically, the F-22 has a very bad case of lock in caused by a number of design decisions early in the development of the F-22. This can be attributed to inexperience and trying to meet what turned out to be a flawed decision by the DoD.

In a nutshell, the F-22 avionics uses a Common Integrated Processor approach based on the Intel i960MX processor and is written in Ada. Later production F-22's have a hybrid PowerPC and i960MX architecture and some software code is ported over, which allowed the avionics to combine legacy software with newer software from the F-35 program, but that's fairly limited.

Specifically, the DoD had mandated in the 1980's that all future aircraft development use Ada as the primary programming language for the software. Good idea in theory; Ada is also very fault tolerant. Prior aircraft used a whole smorgasbord of different programming languages, some of which were proprietary which made development work difficult.

Problem is that Ada is a fairly challenging programming language to program for, with a very small programmer base with limited programming tools, and so is the number of platforms using Ada as the primary programming language. That's why today, Ada is really only used very sparingly in very specialized applications where failure is not acceptable.

Adding to the mess is that every other combat system onboard the F-22 would need to be redesigned or reprogrammed; the level of integration of the radar and ESM and it's capabilities and the onboard avionics is significantly deeper and higher than in previous aircraft. Touch one system, it affects other systems, and the entire software package for the avionics needs recertification and validation work.

Because the F-22's development took many years, and acquisition processes aren't known for being very fast and responsive, when the decision was made early on in the F-22's to use the i960MX chip, the chip was considered bleeding edge, state of the art. When the F-22 went into production, it was already obsolete and about to be discontinued because they held onto legacy platforms too long. And by the time the F-22 IOC'ed, you could literally find parts needed to keep the F-22's avionics and ground equipment going in a museum; there was no way, short of paying a subcontractor outrageous amounts of money to keep it in production for a small production run.

The i960 RISC was not a very common or popular platform to begin with (only the F-22 and the HAL Tejas, which is also a 1980s design that went IOC much later) which further complicated matters. Some of the newer F-22 CIP boards have more powerful PowerPC chips, along with FPGAs that have more compatibility with the F-35 and can share parts of the code base, but in order to replace the F-22's avionics, it will cost a whole lot of money even if they went off the shelf (say, took the avionics computers from the F-35), as they still need to redesign and reprogram everything else on the aircraft to be compatible.

The F-35 took many of the lessons learned from the F-22's development to avoid many of the pitfalls:
1. A decision was made early on to use C++ instead of Ada as the programming language to broaden the programmer base and to use standard development tools that are much more readily available;
2. A decision was also made to use the much more commercially popular PowerPC platform, which has multiple vendors making chips for, and thus a larger supplier base;
3. The F-35's avionics approach is to use “virtual machines” (VMs) extensively to avoid having core specific functions;
4. The F-35's avionics software uses a middleware so that a change in hardware does not require a complete rewrite and recertification of the avionics suite. Likewise, a change to the one part of the software does not require a complete rewrite and recertification either;
5. All major avionics components and weapon stores are connected via fiber-optic interconnects which allows for future upgrades without rewiring the jet;
6. The F-35 isolates, in discreet VMs, the individual processing of avionics data then fuses the results. This isolated each VM from any bugs or problems another VM faces. The F-22 ran multiple functions in a single computer instance so if there was a problem, it affected multiple functions;
7. The F-35 was planned for all IOC functions in its current hardware state and only every other Block requires significant hardware changes, which involve replacing, or adding new processor cards

Thank you for this enormously informative post, I learned a great deal, and the USAF F-22 decision is much more understandable in these terms. As well as many aspects of the F-35 program. Nice job!

scbriml wrote:

wingman wrote:

For those of us that prize looks over capability, it's important to note that the F-22 looks ten times cooler than the F-35.

That’s a low bar! A London double-decker bus looks cooler than an F-35.

Just be glad that the Boeing contender, the X-32 wasn't chosen.

ThePointblank wrote:

No, production was curtained because a number of critical components in the F-22's avionics went out of production from various subcontractors, and there was no quick and cheap way of finding an alternative part or supplier to provide said components.

Basically, the F-22 has a very bad case of lock in caused by a number of design decisions early in the development of the F-22. This can be attributed to inexperience and trying to meet what turned out to be a flawed decision by the DoD.

In a nutshell, the F-22 avionics uses a Common Integrated Processor approach based on the Intel i960MX processor and is written in Ada.

I don't believe this.

If the air force wanted to continue to build more F-22s, it would have been easy to buy a few thousand i960s and other ICs as the production was winding down. In fact, I rather hope they did because the existing fleet need to be supported. Seriously, if Intel calls up and says "we're shutting down the i960 line, what's your last order" grabbing enough to finish aircraft production is not hard.

I assume there are more discontinued chips in that plane than the i960, but the same technique works.

These chips are a small part of the total cost of the aircraft.

They discontinued the F-22 because (1) it cost too much and (2) it cost to much to operate and (3) there was no near peer for it to fight and (5) early visions of the F-35 were dancing in their heads.

kitplane01 wrote:

ThePointblank wrote:

No, production was curtained because a number of critical components in the F-22's avionics went out of production from various subcontractors, and there was no quick and cheap way of finding an alternative part or supplier to provide said components.

Basically, the F-22 has a very bad case of lock in caused by a number of design decisions early in the development of the F-22. This can be attributed to inexperience and trying to meet what turned out to be a flawed decision by the DoD.

In a nutshell, the F-22 avionics uses a Common Integrated Processor approach based on the Intel i960MX processor and is written in Ada.

I don't believe this.

If the air force wanted to continue to build more F-22s, it would have been easy to buy a few thousand i960s and other ICs as the production was winding down. In fact, I rather hope they did because the existing fleet need to be supported. Seriously, if Intel calls up and says "we're shutting down the i960 line, what's your last order" grabbing enough to finish aircraft production is not hard.

I assume there are more discontinued chips in that plane than the i960, but the same technique works.

These chips are a small part of the total cost of the aircraft.

They discontinued the F-22 because (1) it cost too much and (2) it cost to much to operate and (3) there was no near peer for it to fight and (5) early visions of the F-35 were dancing in their heads.

You actually have no idea what you are talking about. No you can't "just buy" a few thousand extra parts of whatever commercial part you need. Well... you can, but than you need to warehouse them, ensure they are usable for whatever time frame you need them (do you have the next congressional budget allotment approved?). Then when you pull them you have to test them and you needs spares for future maintenance needs (which is why most DMS is purchased and warehoused). And when you do actually buy them, you need to negotiate with the vendor because you actually don't have an unlimited budget because you already have hundreds or DMS items to purchase, warehouse and use. Each of which is costing the program hundreds and thousands of dollars extra every year because the warehouse that keeps them is a secured, temperature controlled facility that cost a lot because it is already holding thousands... millions of other DS items.

It is not an easy or simple thing. And when combined with an analysis that shows the war-fighting equipment they are destined for is has other issued which have diminished its relevance then it becomes a moot action.

DMS sucks and is a constant battle. Even the F-35, which was designed to be able to manage the high-speed obsolescence of modern electronic technology, has trouble as it operates on a three to five year procurement to the production floor time frame.

Tugg

kitplane01 wrote:

ThePointblank wrote:

No, production was curtained because a number of critical components in the F-22's avionics went out of production from various subcontractors, and there was no quick and cheap way of finding an alternative part or supplier to provide said components.

Basically, the F-22 has a very bad case of lock in caused by a number of design decisions early in the development of the F-22. This can be attributed to inexperience and trying to meet what turned out to be a flawed decision by the DoD.

In a nutshell, the F-22 avionics uses a Common Integrated Processor approach based on the Intel i960MX processor and is written in Ada.

I don't believe this.

If the air force wanted to continue to build more F-22s, it would have been easy to buy a few thousand i960s and other ICs as the production was winding down. In fact, I rather hope they did because the existing fleet need to be supported. Seriously, if Intel calls up and says "we're shutting down the i960 line, what's your last order" grabbing enough to finish aircraft production is not hard.

I assume there are more discontinued chips in that plane than the i960, but the same technique works.

These chips are a small part of the total cost of the aircraft.

They discontinued the F-22 because (1) it cost too much and (2) it cost to much to operate and (3) there was no near peer for it to fight and (5) early visions of the F-35 were dancing in their heads.

It's a matter of DoD and Capitol Hill budgeting processes. Not very good at getting long term lead items purchased or acting quickly, and there were constant attempts at developing a new avionics core for the F-22 throughout the program which all in the end lead no where. So there was a constant thought that the avionics system on the F-22 would be replaced by something new down the line, which never happened due to budgetary issues and technical difficulties.

It even got to a point where in 2008, the USAF was said to have been looking for a contractor who could install a second "brain" into the avionics system. Apparently the USAF was so disgusted with how follow-on development was proceeding that it wanted to basically keep the current architecture running core functions, and then graft a second one over top to provide other functions like sensor fusion, weapons control, ect. That went nowhere as well.

Just for comparison, the early 1970's Mig-25M (2 prototypes) which had R15BF2-300 engines which fixed the overspeed issues, gave legitimate mach 3.3 top speed, and had a lot more power (32k lbs thrust).
Demonstrated 1200 nmi combat radius at mach 2.35 cruise with 4x1000 lb AAM's, 80,000 ft combat ceiling, and mach 3.3 top speed
(This is the aircraft that set the zoom climb record of 123,500 ft, and beat the F-15's time to climb records)

44,000 lb empty weight and 32,000 lbs internal fuel

Sure, no stealth but beats everything else in the interceptor category

Tugger wrote:

kitplane01 wrote:

ThePointblank wrote:

No, production was curtained because a number of critical components in the F-22's avionics went out of production from various subcontractors, and there was no quick and cheap way of finding an alternative part or supplier to provide said components.

Basically, the F-22 has a very bad case of lock in caused by a number of design decisions early in the development of the F-22. This can be attributed to inexperience and trying to meet what turned out to be a flawed decision by the DoD.

In a nutshell, the F-22 avionics uses a Common Integrated Processor approach based on the Intel i960MX processor and is written in Ada.

I don't believe this.

If the air force wanted to continue to build more F-22s, it would have been easy to buy a few thousand i960s and other ICs as the production was winding down. In fact, I rather hope they did because the existing fleet need to be supported. Seriously, if Intel calls up and says "we're shutting down the i960 line, what's your last order" grabbing enough to finish aircraft production is not hard.

I assume there are more discontinued chips in that plane than the i960, but the same technique works.

These chips are a small part of the total cost of the aircraft.

They discontinued the F-22 because (1) it cost too much and (2) it cost to much to operate and (3) there was no near peer for it to fight and (5) early visions of the F-35 were dancing in their heads.

You actually have no idea what you are talking about. No you can't "just buy" a few thousand extra parts of whatever commercial part you need. Well... you can, but than you need to warehouse them, ensure they are usable for whatever time frame you need them (do you have the next congressional budget allotment approved?). Then when you pull them you have to test them and you needs spares for future maintenance needs (which is why most DMS is purchased and warehoused). And when you do actually buy them, you need to negotiate with the vendor because you actually don't have an unlimited budget because you already have hundreds or DMS items to purchase, warehouse and use. Each of which is costing the program hundreds and thousands of dollars extra every year because the warehouse that keeps them is a secured, temperature controlled facility that cost a lot because it is already holding thousands... millions of other DS items.

It is not an easy or simple thing. And when combined with an analysis that shows the war-fighting equipment they are destined for is has other issued which have diminished its relevance then it becomes a moot action.

DMS sucks and is a constant battle. Even the F-35, which was designed to be able to manage the high-speed obsolescence of modern electronic technology, has trouble as it operates on a three to five year procurement to the production floor time frame.

Tugg

Question: How are F-22s supported? Where do they get the now-not-produced chips to support them? And if the government had wanted to continue F-22 production, could they not have used that mechanism? Or just funded long-lead-items?

It seems like you're talking politics, and there were plenty of politicians who's districts had F-22 building employees. And I believe LM has lobbyists.

And couldn't LM have pre-spent $millions to keep a $billlions project making profits?

I really don't see a practical problem here, and warehousing and testing need to be done anyway for maintenance and support.

744SPX wrote:

Just for comparison, the early 1970's Mig-25M (2 prototypes) which had R15BF2-300 engines which fixed the overspeed issues, gave legitimate mach 3.3 top speed, and had a lot more power (32k lbs thrust).
Demonstrated 1200 nmi combat radius at mach 2.35 cruise with 4x1000 lb AAM's, 80,000 ft combat ceiling, and mach 3.3 top speed
(This is the aircraft that set the zoom climb record of 123,500 ft, and beat the F-15's time to climb records)

44,000 lb empty weight and 32,000 lbs internal fuel

Sure, no stealth but beats everything else in the interceptor category

It gets to the place where it is shot down faster?

(Seriously, I know speed and altitude reduces and AA missiles engagement window, but the USAF seems to think stealth reduces this window even more.)

kitplane01 wrote:

Tugger wrote:

You actually have no idea what you are talking about. No you can't "just buy" a few thousand extra parts of whatever commercial part you need. Well... you can, but than you need to warehouse them, ensure they are usable for whatever time frame you need them (do you have the next congressional budget allotment approved?). Then when you pull them you have to test them and you needs spares for future maintenance needs (which is why most DMS is purchased and warehoused). And when you do actually buy them, you need to negotiate with the vendor because you actually don't have an unlimited budget because you already have hundreds or DMS items to purchase, warehouse and use. Each of which is costing the program hundreds and thousands of dollars extra every year because the warehouse that keeps them is a secured, temperature controlled facility that cost a lot because it is already holding thousands... millions of other DS items.

It is not an easy or simple thing. And when combined with an analysis that shows the war-fighting equipment they are destined for is has other issued which have diminished its relevance then it becomes a moot action.

DMS sucks and is a constant battle. Even the F-35, which was designed to be able to manage the high-speed obsolescence of modern electronic technology, has trouble as it operates on a three to five year procurement to the production floor time frame.

Tugg

Question: How are F-22s supported? Where do they get the now-not-produced chips to support them? And if the government had wanted to continue F-22 production, could they not have used that mechanism? Or just funded long-lead-items?

It seems like you're talking politics, and there were plenty of politicians who's districts had F-22 building employees. And I believe LM has lobbyists.

And couldn't LM have pre-spent $millions to keep a $billlions project making profits?

I really don't see a practical problem here, and warehousing and testing need to be done anyway for maintenance and support.

The DLA (Defense Logistics Agency) handles support for all programs and has warehouses across the country. For each production block, spares are included and the DLA then manages them. The various defense contractors and agencies then request needed items to be released for repairs and maintenance operations.

As to LM or any defense contractor spending their own money? (Sorry, don't take this as anything against you...)

Sadly, as recently seen in space related activities, the defense contractors are all trained to only spend the governments money and spend it as long as they can. If the government wants something, it has to pay for it.

Tugg

kitplane01 wrote:

Question: How are F-22s supported?

An example from another defense program.

They would gave bought extra sets of LRUs as spares. How many they buy would depend on the mean time between failure prediction for each component. They will use up the spare until they have not enough to support deployment, then they would pony up the money to design new LRU with new, available components.

This is done on a as need basis as each time they do this, it costs lots of money.

Again this is coming from an active production aircraft where you have fresh money coming in for future buy. For out of production aircraft, the calculation is more difficult.

bt

It is not F-22 but as an example here is the DLA's 2019 announcement of their support for F-35 spares:
https://www.dla.mil/About-DLA/News/News ... are-parts/

Tugg

RJMAZ wrote:

kitplane01 wrote:

I don't believe this.

If the air force wanted to continue to build more F-22s, it would have been easy to buy a few thousand i960s and other ICs as the production was winding down. In fact, I rather hope they did because the existing fleet need to be supported. Seriously, if Intel calls up and says "we're shutting down the i960 line, what's your last order" grabbing enough to finish aircraft production is not hard.

I agree. The chips played no part in production ending. The chip manufacturers would have continued production of any chips required. The i960 chip for example cost around $100 when new. Even if the price per chip went up to $10,000 due to a low production run it would barely change the cost of aircraft.
[...]

I just have to comment on this ignorant comment... NO the chip manufacturers would have continued production of any of the chips needed. Do the math. At $10,000 per chip,even needing 10 per plane that is 2000, so $20 million for the manufacturer.... and that is not profit.... they need billions in revenue,so perhaps a few hundred million in profits to meet shareholder expectations.

And no the military will not pay $100,000.00 per chip.

Tugg

Tugger wrote:

RJMAZ wrote:

kitplane01 wrote:

I don't believe this.

If the air force wanted to continue to build more F-22s, it would have been easy to buy a few thousand i960s and other ICs as the production was winding down. In fact, I rather hope they did because the existing fleet need to be supported. Seriously, if Intel calls up and says "we're shutting down the i960 line, what's your last order" grabbing enough to finish aircraft production is not hard.

I agree. The chips played no part in production ending. The chip manufacturers would have continued production of any chips required. The i960 chip for example cost around $100 when new. Even if the price per chip went up to $10,000 due to a low production run it would barely change the cost of aircraft.
[...]

I just have to comment on this ignorant comment... NO the chip manufacturers would have continued production of any of the chips needed. Do the math. At $10,000 per chip,even needing 10 per plane that is 2000, so $20 million for the manufacturer.... and that is not profit.... they need billions in revenue,so perhaps a few hundred million in profits to meet shareholder expectations.

And no the military will not pay $100,000.00 per chip.

Tugg

However, LM has lots of warehouses, if you want to buy a few for future needs.

I still want to know how the current F-22s are being supported.

kitplane01 wrote:

However, LM has lots of warehouses, if you want to buy a few for future needs.

I still want to know how the current F-22s are being supported.

I told you, the DLA. They hold all legacy components for program sustainment. and LM uses that material per awarded contracts to sustain the aircraft.

https://www.lockheedmartin.com/en-us/wh ... nment.html
https://www.airuniversity.af.edu/News/D ... rce-goals/
https://www.rand.org/content/dam/rand/p ... _TR763.pdf
https://ndiastorage.blob.core.usgovclou ... Agenda.pdf

Lockheed Martin is the leading industry partner for the Defense Logistics Agency across a broad spectrum of supply chains. We
are the DLA’s largest commercial customer and have enjoyed superior agency support for our Kelly Aviation operation. We pro-
actively partner with the DLA on leading sustainment programs such as the F-22,

Tugg

Really hard to pick sides, here.

On one side is the cogent, well-reasoned argument made by well informed people citing logical hardware, coding, and supportability flaws in the program which prevented “future-proofing” and prevented future growth. Flaws which are so well known they are industry poster children of how not to run a program.

On the other side is the “just throw them in a box above the fridge” and “just buy a warehouse or something” crowd using cold, hard fiscal analysis like “add a mil to the price” while vaguely alluding to design flaws known only to him and brave future authors.

This is a tough choice.

RJMAZ wrote:

The chip making machines can make many different chips on the one machine.

Chip machines of today yes, chip machines at the time whether to refresh F-22, not so much. We are talking about 25 years ago.

RJMAZ wrote:

More lines of code than any other fighter in the year 2005

Yes but I don't think the different systems have the common architecture they have today. Refreshing one LRU is not an issue, you are looking at complete system refresh as well.

Note that every new chip require new certification testing at the component level and maybe up to the LRU level, some cert by commonality is posible, but vibration and functional testing will have to be re-done.

As noted. There are many factors related to why the F-22 is expensive. Cost of the chip may be debatable. Cost of the chip certification as well as any new components is not.

bt

LyleLanley wrote:

Really hard to pick sides, here.

On one side is the cogent, well-reasoned argument made by well informed people citing logical hardware, coding, and supportability flaws in the program which prevented “future-proofing” and prevented future growth. Flaws which are so well known they are industry poster children of how not to run a program.

On the other side is the “just throw them in a box above the fridge” and “just buy a warehouse or something” crowd using cold, hard fiscal analysis like “add a mil to the price” while vaguely alluding to design flaws known only to him and brave future authors.

This is a tough choice.

It is hard to tell!

One side cancels multi-billion $$$ programs because they don't know how to store a few thousand chips that can live in room tempature and won't decay with age, and the other side reminds them that these problems are super solvable. And "politics stopped me from buying ahead" isn't really a good answer.

Of course, "your fighter costs too much to build and operate" is a harder problem to solve.

Tugger wrote:

kitplane01 wrote:

However, LM has lots of warehouses, if you want to buy a few for future needs.

I still want to know how the current F-22s are being supported.

I told you, the DLA. They hold all legacy components for program sustainment. and LM uses that material per awarded contracts to sustain the aircraft.

https://www.lockheedmartin.com/en-us/wh ... nment.html
https://www.airuniversity.af.edu/News/D ... rce-goals/
https://www.rand.org/content/dam/rand/p ... _TR763.pdf
https://ndiastorage.blob.core.usgovclou ... Agenda.pdf

Lockheed Martin is the leading industry partner for the Defense Logistics Agency across a broad spectrum of supply chains. We
are the DLA’s largest commercial customer and have enjoyed superior agency support for our Kelly Aviation operation. We pro-
actively partner with the DLA on leading sustainment programs such as the F-22,

Tugg

I honestly don't understand this. I feel like we're writing past each other. I'd realy like to understand what you're saying.

I *think* you're saying the DLA buys chips ahead for sustainment purposes, but cannot buy chips ahead for production purposes. Why could they not also buy chips ahead for production? Is the answer "it's not in their mission statement".

And why cannot LM buy chips ahead? Even with their own $$$. Because they should find it profitable to spend $millions to save $billions contract.

Last question (sorry to ask so many): Is your claim that it's not reasonably physically possible to buy and store chips ahead of need, or are you saying the organizations involved are not organzied to do so?

kitplane01 wrote:

… "politics stopped me from buying ahead" isn't really a good answer.

Unfortunately, the government is full of “not good” answers. It’s just how the system is set up.

No one is saying what you’re saying isn’t physically doable - it is. But it’s absolutely not economically doable. The system is not conducive to updates without spending an exorbitant of money the DoD didn’t (and still doesn’t) have. The F-22 is the aviation equivalent of buying an iMac that is physically unable to have its RAM, hard drive, and CPU updated and then coming back years later to update it. Doable if you spend big $$$ but otherwise cheaper to buy a new computer.

Rewind to 2009 for a moment: it’s the middle of the worst economic crisis since the Great Depression; you’re fighting two (!!!) counter insurgencies at the same time - neither of which had any use whatsoever for a stealthy, super maneuverable air-to-air only fighter; you have massive investment bills coming due for an extremely expensive tri-service fighter that is overweight, behind schedule, and over budget. Finally, you have an Air Force seemingly spending more time fighting to buy 375 F-22s for the next big war rather than supporting the other branches fighting current wars (slow rolling the 50 CAP ISR mandate). This is the same branch that also caused repeated international embarrassment by shipping nuclear weapon parts to Taiwan by mistake and flying 6 nuclear warheads halfway across the country without anyone noticing for two days.

Now with all that, the service says they need many billions of dollars more to modernize the F-22 - just to enable it to be truly modernized.

Kind of a hard sell.

kitplane01 wrote:

And why cannot LM buy chips ahead?

It's not necessarily the chip but the whole circuit card and components that goes in to the LRU. LM can buy the chip, but they do not own the circuit card or the LRU design.

I've actually ran in to this situation recently. We had a notice from one supplier that they will stop production of a push to test button that we use as part of of our smoke detection system. We had the opportunity to do one last buy. The part probably costs no more than $100. It took more than a month to get the power that be to approve an order to cover the rest of our projected production run plus lots of spares.

If we had to redesign the assembly to accept a new/similar to push to test button, it would have cost us 100 X the cost of the lot buy to redesign/installation paperwork.

This is just for a simple push to test button!

bt

bikerthai wrote:

kitplane01 wrote:

And why cannot LM buy chips ahead?

It's not necessarily the chip but the whole circuit card and components that goes in to the LRU. LM can buy the chip, but they do not own the circuit card or the LRU design.

I've actually ran in to this situation recently. We had a notice from one supplier that they will stop production of a push to test button that we use as part of of our smoke detection system. We had the opportunity to do one last buy. The part probably costs no more than $100. It took more than a month to get the power that be to approve an order to cover the rest of our projected production run plus lots of spares.

If we had to redesign the assembly to accept a new/similar to push to test button, it would have cost us 100 X the cost of the lot buy to redesign/installation paperwork.

This is just for a simple push to test button!

bt

If it's not sensitive info, I'd be really interested in understanding *why* this supplier is ceasing production of the part?