January 05, 2020

The Range of a Carrier Wing

In 2015, the Center for a New American Security published Retreat from Range, by Jerry Hendrix, a respected naval historian and former naval flight officer.1 In it, he argued that the range of US carrier air wings had fallen precipitously since the 1950s as it abandoned its deep-strike capabilities in favor of a focus on sortie generation and low costs. With the rise of weapons like the Chinese DF-21, this is unacceptable, and new aircraft need to be developed. This report has received extensive media attention, with several outlets taking up the cry for more range out of carrier wings.

Forrestal during her shakedown cruise in early 1956

Unfortunately, a large part of the report's thesis is wrong. Leaving aside that the threat of weapons like the DF-21 is grossly overrated, his analysis of the range of US carrier wings is dangerously flawed. Simply put, the numbers Hendrix gives for his 1956 carrier aircraft are completely wrong. They seem to have been generated by googling each of the aircraft in question, and then taking the values for "max weapons load" and "max range" and slapping them together. Anyone with any serious exposure to aircraft should immediately recognize that this is not the right way to do this. Payload affects range, particularly when it's being carried on the outside of the plane and displacing drop tanks. Not accounting for this speaks of at best extreme carelessness on the part of Hendrix and CNAS.

To prove this, I'm going to examine some of his claims in more detail. I'll be basing my work on the Standard Aircraft Characteristics Archive. The SACs are exactly what the name implies, half-dozen page documents that detail the aircraft's performance, including missions that have both bomb loads and ranges. This is level of detail even the best reference books often lack, and it lets us see exactly what the airplanes are capable of, and contrast that to the values he gives. His hypothetical air wing is the one that flew off the Forrestal class in the late 50s, and I was able to find SACs for all of the planes he mentions.

The AD Skyraider

“The AD-1 was both famous and infamous for the power of its radial piston engines. Capable of carrying 8,000 pounds of ordnance over 1,000 nm to a target,” (p.24)

AD-4 Skyraider

First, the AD-1 probably isn't the appropriate variant. Improved models entered service quickly, and Forrestal appears to have carried AD-6s on her shakedown cruise, so I doubt she ended up with a much earlier model later on. The SAC archive also doesn't have an AD-1, so I decided to go with the AD-4, the earliest available. This lists three missions, one of which involves a radar pod and can thus be ignored. Of the other two, one is a long-range mission with drop tanks, and carries 3,680 lb of ordnance with a combat radius of 540 nautical miles. The other is a short-range mission without drop tanks, and carries 5,680 lb of ordnance to 220 nm. The first is the sort of thing you'd expect for a deep-strike mission, and it doesn't come close to 1,000 nm or 8,000 lb of ordnance. Looking at the numbers, it seems like he might have confused radius and range, which is a perfectly understandable mistake for a random person on the internet, but much less so from a professional. It's also incredible that he assumed the max warload the plane was designed to carry was the relevant payload, as an AD loaded with that many bombs would probably have a combat radius about equal to the takeoff roll.

Error: Factor of 2 on payload and range.

The F2H Banshee

"The Banshee’s sturdy design and large fuel load gave it the ability to carry 3,000 pounds of ordnance to targets nearly 1,500 nm away." (p.25)

F2H-3 Banshee

Hendrix earlier identifies the model in question as the F2H3, which is clearly a misprint for F2H-3, so I pulled the appropriate SAC. There's only one ground attack mission listed, and it lists 1,580 lb of ordnance, 2,040 lb of external fuel, and a radius of 330 nm. The only range number on the sheet close to his 1,500 nm is the combat range (again not radius) for the fighter configuration (no air-to-ground ordnance) with two drop tanks, and this is obviously not a viable strike configuration.

Error: Factor of 2 on payload, factor of 5 on range.

The FJ-3 Fury

"The FJ3 represented an evolution in the aircraft’s design, shifting it from a pure fighter configuration, carrying only air-to-air missiles and four 20mm cannons, to a fighter/bomber role with the addition of two inboard pylons capable of carrying 1,000-pound bombs and two outboard pylons capable of carrying 500-pound weapons. These weight limitations meant that it was not nuclear delivery capable, as these bombs weighed more than 1,000 pounds at this point in their development. The Fury’s combat radius was only 650 nm, but with aerial refueling, its range could extend to nearly 1,250 nm. Aerial refueling, a relatively recent technological innovation, enabled the Fury to provide fighter escort services for the A3D Skywarrior..."

FJ-3 Fury

First, we have another butchering of the designation system, as the correct form is FJ-3.2 This one avoids explicitly linking payload and range, and even suggests that he's looking at range in a fighter configuration instead of an attack configuration. And the SAC gives numbers a lot closer to his values in that case, although it's still not quite right. The fighter (400 gallons external, no other ordnance) configuration gives a radius of 560 nm unrefueled and 1075 nm with an inflight refueling. So he's only off by 10-15% (although still rather optimistic), but this is also the only case where there is no ordnance, the real source of problems in his numbers so far. For the record, the SAC does have one ground-support configuration, with 1,147 lb of ordnance, 400 gallons external, and a radius of 245 nm.

The A3D Skywarrior

"The Skywarrior was the culmination of the ADR-42 heavy bomber program. It was capable of carrying 12,800 pounds of ordnance, nuclear or conventional, 1,826 nm."

A3D-2 landing aboard Forrestal

This is almost certainly the A3D-2/A-3B Skywarrior, as the earlier A3D-1 was only used for development. The SAC gives a maximum takeoff weight of 73,000 lb from a carrier's catapult,3 and a radius at this weight with a pair of 2,050 lb stores of only 1,200 nm.4 Technically speaking, it's possible that Hendrix was speaking in this case of range instead of radius, in which case he would have actually underestimated the A3D's range, but this is unlikely, as he holds out the A3D, quite rightly, as the longest-ranged member of the carrier wing. As such, I can only treat his number as a radius claim, and note that even an unarmed A3D fully loaded with fuel couldn't make it to a target 1,820 nm away and back, nor could a carrier-launched A3D with an inflight refueling.

Error: Factor of 3 on payload, factor of 1.5 on range.

The A4D Skyhawk

"The Skyhawk’s max speed was 550 knots. Its top altitude was just over 40,000 feet and combat range was 550 nm unrefueled (it did not have external fuel tanks) while carrying up to 9,000 pounds of ordnance of all types, including nuclear weapons."

Even while being hoisted aboard, the A4D-1 Skyhawk retains its signature drop tanks.

The first thing that springs out at me from this is the claim that the Skyhawk does not have external fuel tanks. This is hilariously wrong, as operational Skyhawks almost always carried a pair of large external tanks. The lead-in to the parenthetical seems to suggest that he meant to say the A4D had no in-flight refueling capability and was merely sloppy. This looks to have been true for the A4D-1, which entered service in October 1956 and will thus form the basis of this analysis, but incorrect for later models. The SAC gives several different missions, two of which have no external tanks. If we humor him and use these, then the radius is approximately 175 nm for about 2,000 lb of ordnance. If we instead base it around the Skyhawk as actually used, we have a choice between 575 nm and 1,050 lb of ordnance and 385 nm and 3,500 lb of ordnance. The SAC also provides data on some other attack altitudes. 1,050 lb of ordnance gives a radius of 625 nm if delivered from 15,000 ft,5 but only 410 nm with a 50 nm run-in at sea level. But I'm going to score this based on the text sample, and ignore the external tanks.

Error: Factor of 4.5 on payload, factor of 3 on range.

The A-5 Vigilante

"the Vigilante could fly nearly 1,300 nm unrefueled, delivering a 1,700-pound nuclear weapon while traveling twice the speed of sound."

An A-5A Vigilante in flight

Hendrix does acknowledge that the Vigilante wasn't part of his 1956 airwing, but although it didn't enter service until 1961, we'll take a look at it anyway, based on the SAC for the A3J-1/A-5A. It gives two supersonic high-altitude attacks with Mk 28 store (1,885 lb, and I'm not sure why he underestimates this), which is exactly the sort of strike Hendrix seems most interested in. The one with a pair of drop tanks has a radius of 945 nm, and the highest speed I can find in the SAC is around Mach 1.7, not Mach 2. If we assume that 1,300 nm is the range instead of the radius, then the Vigilante actually exceeds his performance specifications, but as with the A3D, this seems like an overly generous reading, and I will instead treat the claimed value as a radius.

The tail of a Vigilante, showing the back end of the linear bomb bay, here plugged by the temporary tail cone of an RA-5C reconnaissance model.

So shouldn't the score be something like "Factor of 1.5 on range, correct on payload"? Well, no. The problem is that the Vigilante wasn't really capable of delivering bombs. It used a unique linear bomb bay, storing the Mk 28, attached to a pair of drop tanks, in a tunnel between the engines. This was intended to allow the Vigilante to drop bombs while supersonic, a problematic task for a conventional bomb bay, but in practice, it didn't work. The ejection mechanism was unreliable and occasionally a catapult shot would leave the so-called "stores train" sitting on the ship's deck. But the worst failing had to do with aerodynamics. Stores separation tests discovered that the Vigilante had an uncomfortable tendency to "tow" the stores train behind it after separation, ruining accuracy, as well as (probably) the aircrew's pants. On the whole, the Vigilante was a failure, and its career as a bomber lasted only a few years before the bulk were converted to the RA-5C reconnaissance configuration.6

Error: Factor of 1.5 on range, couldn't drop bombs.

I've picked on the 50s air wing because it was the easiest to evaluate, but his abuse of numbers continues throughout the paper. Checks of the F-4 and A-7 SACs reveal similar exaggeration to the rest of the numbers given, with bombloads that are twice what the aircraft actually carried and range numbers that require at very least a substantially lighter bombload. Unfortunately, the SACs for modern aircraft are not available to the public, so I can't settle the interminable debate over how far a Hornet or Super Hornet can actually go.7

An F-35C landing aboard USS George Washington

Nor are the numerical problems limited to aircraft performance. On page 57, Hendrix claims that the F/A-18E/F cost $85 million less per aircraft than the F-35C, a value that doesn't even match the numbers given elsewhere in the paper of $85 million for the Super Hornet and $130 million for the F-35C, to say nothing of the actual values. The current values are something like $66 million for the F/A-18E/F, and $110 million for the F-35C, although figuring out values for this kind of stuff is always complicated. Moreover, the F-35C is not particularly mature, and the cost is likely to fall further as production ramps up, with the latest contracted batch running around $94 million each. It's also worth pointing out that the Navy is rushing the MQ-25 Stingray into production to plug the tanker gap that Hendrix discusses, although the program was still in the early stages in 2015, when the paper was written.

Overall, Hendrix's claims on the future of the American carrier force fail to stand up to scrutiny. Retreat From Range is riddled with math errors of the sort that should not be made by anyone who has a basic acquaintance with military aviation, and the story of falling range is much less dramatic when accurate numbers are used. Hendrix and CNAS as a whole should have done better.

1 This means that he was someone who operated aircraft systems instead of flying planes, specifically on P-3 Orions.

2 The Fury is unusual in that the FJ-1 was a straight-wing predecessor to the famous F-86 Saber, while the FJ-2, FJ-3 and FJ-4 were navalized derivatives of the F-86.

3 Two important footnotes to this number. First, Hendrix claims that the max takeoff weight of the A3D is 83,000 lb, a number completely unsupported by the SAC. Second, the weight for a field (non-catapult) takeoff is indeed 5,000 lb higher, and the SAC lists a second set of range figures for this configuration. But since we care about the carrier's strike range, those are irrelevant.

4 The SAC provides some interesting details on this high-altitude number. If a sea-level approach is made at cruise throttle, radius is reduced 60 nm for every 100 nm flown at SL. If the approach is made at military (max sustained) thrust, it drops 120 nm for every 100 nm of approach.

5 The baseline involves diving to sea level for deliver from cruise.

6 See here for more information on this fascinating airplane.

7 I may investigate this later, using CMANO/CMO, although that is obviously not as good as using official figures. I suspect that the range changes are heavily confounded by changes in how aircraft operate. In particular, low-level strike takes a lot more fuel than high-level strike, and modern aircraft are optimized more for that regime at a cost in efficiency in high-level cruise.


  1. January 05, 2020Mike Kozlowski said...

    ...Thoughtful and well done critique - bravo!

    One thought about the weight discrepancy on the A-5C's Mk28 - is it possible that the empty fuel tanks were included in the initial weight, and he simply counted the weight of the device itself?


  2. January 05, 2020bean said...

    Possible. Swords of Armageddon gives the 1,885 lb weight for the A5J bomb assembly including fuel tanks, but I can't find a reference to any other B28 variant in SoA that is lighter than that value. The wiki article lists a minimum weight of 1,700 lb, but doesn't give a source. It's not an implausible minimum weight, although 185 lb sounds pretty light for a pair of 400-gallon tanks. Based on a quick google, that seems right for a single tank of that size, although that's an external drop tank, not an internal one. I wouldn't be hugely surprised if 1,885 lb is the actual weight of the weapon and Hansen misread the NATOPS manual he cites.

  3. January 05, 2020Eric Rall said...

    The Fury’s combat radius was only 650 nm, but with aerial refueling, its range could extend to nearly 1,250 nm.

    What is the limiting factor of range with in-air refueling? The only things I can think of are how far out the refueling craft can go, crew endurance, or the plane only being able to refuel a small number of times without maintainence.

  4. January 05, 2020Lambert said...

    Trick is to refuel the refuelling craft.
    (Leading to tyranny-of-the-rocket-equation-esque exponential blowups.)

    Over half of the refuellings needed for Black Buck I were victor-victor.

    Range of a Hornet: Time to break out the Reynolds numbers and get crunching?

  5. January 05, 2020Philistine said...

    @Eric Rall The limit is the lesser of crew endurance and tanker availability. If you have lots of tanker support, you can fly 12+ hour missions in F-16s ("Can." Not, as a rule, "should."); but most non-US air forces would probably run out of tankers well before their fighter pilots reached the point of absolute exhaustion.

  6. January 05, 2020bean said...


    In principle, there's no limit on the range of an aircraft using aerial refueling. (See Black Buck). In practice, you have several issues. Besides crews and tankers, there are some bits that simply can't be filled in flight. Engine oil is the most likely one. In this case, the SAC has a specific profile with a single refueling, which is what I used and what I assume Hendrix meant in the quoted paragraph.

  7. January 05, 2020cassander said...

    As someone who literally sells numbers about aircraft for a living, I can assure you that most numbers relating to aircraft are, if not outright lies, as least potentially misleading. Aircraft are complicated things, and the correct answer to almost any question, even simple things like wingspan or the number in service will start with "Well, it depends, are you counting..."

    Any single number for an aircraft range is a straight up lie. How much fuel it carries, how quickly it burns the fuel, what sort of routes the aircraft has flown in the past, those might be real, but the correct answer for range is always, always starts with "Well, it depends..."

  8. January 05, 2020cassander said...

    Also Bean, just FYI, flyaway cost is USUALLY a better figure to use when you're quoting from DoD budgets.

    Of course, as I say that, I'm also deeply skeptical of the current figure of 66 million per for super hornets. That's basically basically on the level with what F-16s are costing these days, but the navy has been reporting something like that figure for a few years now, so what are you going to do?

  9. January 06, 2020bean said...

    In retrospect, I'm kicking myself for sorting through the FY20 budget instead of just emailing you. You actually know what you're looking for, while I saw a confusing mass of numbers, and decided to grab the highest one, just to be safe. If I'd seen a clear label of "flyaway", I definitely would have picked it, but I didn't, and so decided to grab the highest numbers I saw just to be safe. If they're buying Super Bugs for 66 million, what are F-35Cs going for?

  10. January 06, 2020VQG said...

    What faction of the military do you think this was written to support? Team heavy bomber, or just the anti F-35 wing of the navy?

  11. January 06, 2020Alexander said...

    @VQG I don't know if it was the goal of the writer, but it sounds pro MQ-25, and perhaps pro integration of standoff weapons, or conformal fuel tanks for the Super Hornet.

  12. January 06, 2020bean said...

    I think Hendrix is generally on the anti-carrier wing, and he's advocated for heavy bombers in the past.

  13. January 06, 2020cassander said...


    The aircraft they're taking delivery of this year will cost about 110m, but they've signed contracts for the next couple years that see the price dropping down to 94m in lot 14. And it's anyone's guess what the Block 3 Super Hornets will cost so really, it depends on what you count. : )

  14. January 06, 2020bean said...

    I've edited in those values, as well as the $66 million for the Super Hornets. Thanks.

  15. January 06, 2020cassander said...

    I thought I added in an addendum, but it seems to have been swallowed by the aether. In 2015, when the article was written, the cost for an F-18 was ~60m, and the projected average for the whole run of F-35C was about 145m. I'm not sure why it was so high, it wasn't unexpected or unplanned that the costs would fall, but it does mean that the 85 mil difference in 2015 was at least grounded in the official figures. Of course, there's still no excuse for inconsistently quoting the unit price elsewhere.

Comments from SlateStarCodex:

Leave a comment

All comments are reviewed before being displayed.

Name (required):

E-mail (required, will not be published):


You can use Markdown in comments!

Enter value: Captcha