Part 2: Top-Level Analysis of F-35 Transonic Performance
(Part 1 Here, Part 3 Here, Bonus Block 60 Comparison Here)Analytical Approach
To make the next part of our exploration as widely accessible and understood as possible, our goal will be to continue to minimize the amount of math and physics to the greatest practical extent: keeping things as simple as we can without doing violence to the phenomena. Think of it as a ‘studies’ approach.
Therefore, I will try to base analyses upon relative differences in ‘sum totals’ and relative percentages or fractions to some baseline. We can’t get away from a minimal discussion of the nature of acceleration versus drag, but we can illustrate the things about drag and acceleration that we need to grasp without crunching too many of the numbers. Which is a good thing, because without the need to know, we don’t have the actual F-35 numbers (ex: drag values) available to us in the public domain for plugging into any calculations anyway.
First we’ll summarize the relevant things we now know-- or think we know-- about the F-35’s Transonic Acceleration KPP and overall performance. We will focus on the F-35A model in examining the relevant physics of accelerating through the transonic region and how they affect all aircraft and in particular the F-35A. We’ll then follow with a discussion examining just one likely ‘workaround’ that the F-35A might operationally employ (if it is even needed), and discuss the ‘tactical’ impacts and whether or not what may obviate or mitigate whatever the ‘popularly-perceived’ operational challenges the F-35 variants endure passing through the transonic region.
What we know or think we know
Here is what we know or has been reported as stated by the F-35 Program Participants including the user communities that will help guide the discussion. Note: We’re numbering the references for easing later analysis and possible discussion.
1. The 2012 DOT&E report mentioned the following about the F-35 acceleration from .8M to 1.2M:
A.) A Model: Extended the time for acceleration from 0.8 Mach to 1.2 Mach by 8 seconds
B) B-Model: Extended the time for acceleration from 0.8 Mach to 1.2 Mach by 16 seconds
C) C Model: Extended the time for acceleration from 0.8 Mach to 1.2 Mach by “at least” 43 seconds
2. A Lockheed Martin representative was quoted in Air Force Magazine (cached) as stating the F-35 can maintain Mach 1.2 for a dash of 150 miles without using afterburners. Adding: "Mach 1.2 is a good speed for you, according to the pilots,".
(3, 4,and 5 are all from the same source)
3. “Based on the original spec, all three of the airplanes are challenged by that [acceleration] spec,” said Tom Burbage, Lockheed’s program manager for the F-35. “The cross-sectional area of the airplane with the internal weapons bays is quite a bit bigger than the airplanes we’re replacing.”…
4. …“We’re dealing with the laws of physics. You have an airplane that’s a certain size, you have a wing that’s a certain size, you have an engine that’s a certain size, and that basically determines your acceleration characteristics,” Burbage said. “I think the biggest question is: are the acceleration characteristics of the airplane operationally suitable?”…
5. …U.S. Air Force Lt. Col. Eric Smith, director of operations at the 58th Fighter Squadron at Eglin Air Force Base, Fla., and F-35 test pilot, said that flying the aircraft is a thrilling experience.
“I can’t even explain the adrenaline rush you get when you light the afterburner on that thing,” Smith said. “The acceleration is much better than an F-16.”
6. The F-35′s acceleration is “very comparable” to a Block 50/52 F-16. “Again, if you cleaned off an F-16 and wanted to turn and maintain Gs and [turn] rates, then I think a clean F-16 would certainly outperform a loaded F-35,”Kloos says. “But if you compared them at combat loadings, the F-35 I think would probably outperform it.” The F-16, Kloos says, is a very capable aircraft in a within visual range engagement–especially in the lightly loaded air-to-air configuration used during training sorties at home station. “It’s really good at performing in that kind of configuration,” Kloos says. “But that’s not a configuration that I’ve ever–I’ve been in a lot of different deployments–and those are the configurations I’ve never been in with weapons onboard.” –
7. The weight purge of a few years ago was so “intensive,” Crowley said, that “there’s not thousands of pounds” of weight left to be saved on the F-35. However, even with a three percent annual weight growth, the key performance parameters, or KPPs, won’t be affected.
“All of our predictions for performance are based on an end-of-life, worst-case” scenario relative to the F135 engine’s power capacity, “so the true performance of the jet, throughout its life, will be much better.”
8. PARLIAMENTARY JOINT COMMITTEE ON FOREIGN AFFAIRS, DEFENCE AND TRADE 20 MARCH 2012 (Tom Burbage) Transcript:
For any performance-related requirements, we artificially penalise [sic] the engine by five per cent fuel flow and two per cent thrust. Those margins are given back as we mature the design and get more and more solid on exactly what it is going to do. They are there for conservative estimation up front. We have not taken back any of those margins yet so, when those margins are taken back, the airplane will continue to be well in excess of its basic requirement.
Information from ‘other’ sources.
9. The baseline transonic acceleration specifications for the three variants prior to the spec change are believed to be:
A) A Model: The original “threshold” Key Performance Parameter (KPP) specification time for transonic acceleration (.8 to 1.2 Mach) was ≤ 55 seconds at 30Kft Altitude.
B) B Model: The original “threshold” Key Performance Parameter (KPP) specification time for transonic acceleration (.8 to 1.2 Mach) was ≤ 65 seconds at 30Kft Altitude.
C) C Model: The original “threshold” Key Performance Parameter (KPP) specification time for transonic acceleration (.8 to 1.2 Mach) was ≤ 65 seconds at 30Kft Altitude.
These times were asserted in an oft-referenced Air Command and Staff College paper authored in 2008 by a then-Lt Commander Geoffrey P. Bowman (USN). These numbers have been repeated in the popular aviation press (example) and Australian Air Force Air Vice Marshall Osley came close to ‘anecdotally’ confirming the F-35A KPP as asserted in the ‘Bowman Paper’ in testimony before an Australian Parliamentary Committee in 2013. The aircraft gross weights and fuel loads for the KPPs have not been revealed, though the Bowman Paper observes they are for internal-only weapons load configurations.
***********
APC Quotes added 20 Apr 15 (because the Australian Parliamentary Committee link seems to be a 'moveable' feast):Dr JENSEN: Air Vice Marshal Osley, in a previous hearing you responded to APA's criticism of the F35's aerodynamic performance and you said that it is inconsistent with years of detailed analysis undertaken by Defence, the JSF program office, Lockheed Martin and eight other partner nations. Given that the Director of Operational Test and Evaluation has indicated that the JSF program office, the JPO, has asked JROC to reduce the sustained turn and the acceleration performance essentially to exactly the numbers that APA was predicting years ago, what does that say about the detailed analysis by Defence, the JSF program office, Lockheed Martin and the eight partner nations?
Air Vice Marshal Osley: The points that the Director of Operational Test and Evaluation made there about the manoeuvrability, as you point out it was the sustained turn and the transonic acceleration. He pointed out that the targets that have been set for those parameters were not going to be met by the F35. The figure of I think it was 55 seconds for transonic acceleration, the F35 was going to take 63.9 seconds to do that. That is obviously at a certain altitude, I think it was 30,000 feet, and a range of mach 0.8 up to mach 1.2.
The point to make about those is that that acceleration by the F35 is in a combat configuration. If you look at the legacy aircraft and we talk about comparable performance, a legacy aeroplane would require weapons and, obviously, external fuel tanks to be in combat configuration.
**********
10. …This next reference is included because I believe the article that it is from was at the root of the comment that provided me with the impetus for these posts. It also hints at a way forward within
An interesting factoid, one of the USMC test pilots mentioned this little tidbit—they have to use a modified Rutowski profile in order to get the F-35B and C up to Mach 1.6. Basically, you do one push over, unload the jet and accelerate, get up to 1.2, turn and repeat until you hit 1.4 Mach, turn and repeat till you hit Mach 1.6. It just barely gets there and barely has any gas left over afterwards. The kinematics are basically F/A-18C-like, though that was apparently exactly what was expected….
FYI and FWIW, here’s one of the more decent explanations of a “Rutowski maneuver” that I’ve found online:
If the final speed is near the aircraft’s maximum speed, the large speed increase necessary renders the conventional method of using the peaks of the Ps curves useless. However, the energy method works well. Note in this example the optimum climb path includes an acceleration in a dive. This optimum energy climb path is also known as the Rutowski climb path, after its developer. The path (Figure 7.14) [at the link] consists of four segments to reach energy state E in minimum time. Segment AB represents a constant altitude acceleration from V = 0 to climb speed at state B. The subsonic climb segment follows a path similar to the one illustrated to the tropopause at state C. This subsonic climb is usually a nearly constant Mach number schedule. An ideal pushover or dive is carried out at constant Eh from C to D. The acceleration in the dive is actually part of the optimum climb path.
(It's better at the source with the graphic)
11. There’s a Block 50/52 (F110-GE-129 or F100-PW-229 powered) Dash-1 supplement (Hellenic Air Force) out on the web that provides significant insight as to the F-16 performance baseline used to set the original F-35 transonic KPP.
Deduced Transonic Acceleration KPP Times Analysis of the Variants
The most important performance ‘data’ in our hands are the believed-to-be-actual acceleration times
From References #1 and #9 above, we find evidence that the F-35 variants are currently expected to perform the transonic dash from .8 to 1.2 Mach in the following times:
Model Original KPP Time (secs) Exceeded by (secs) New KPP Time (secs)
F-35A 55 8 63
F-35B 65 16 81
F-35C 65 43 (minimum) 118
Why the differences, and why by ‘that’ much?
IMHO these differences would make an excellent case study for an introductory course on aircraft design if only to just illustrate how much even small differences in design can affect aspects of performance in large ways. I can’t think of any other case where there are three ‘operational’ combat designs that are so close to each other but still different enough to tease out likely drivers of performance differences. (I am excluding aircraft that have had all sorts of various bulges and blisters scabbed on post-manufacture). The first question that springs from observing these differences is how much of the variation is due to variation in aerodynamic shape, propulsion efficiency, and weight?
To help us investigate, we now need to take a look at the possible relevant differences. Those would be the differences between aerodynamic shapes, the engine installations, and relative weights: elements affecting the drag equation (Figure 3 from Part 1) which we now modify to remind us that the drag coefficient is but one element in the drag equation:
 |
| Figure 8 (Numbering Continued From Part 1 |
As the range of airspeed from .8 Mach to Mach 1.1 is the same, and the same air density (same altitude) are common to all variants, this reduces the number of variables on the drag side of the thrust-drag equation to two: the Drag Coefficient and the Cross-Sectional area. On the thrust-side of the equation, we will have less to work with but not so much less as to prevent us from making several observations and ‘educated’ estimations. Since this little exercise is about ‘understanding’ and not ‘quantifying’, I am comfortable working within these limitations.
Relative Weights
We’ll model some likely weights (based upon the ‘Bowman Paper’ and acknowledged fuel and weapons carriage plans). Why ‘weight’? Because it is an excellent proxy for examining relative lift differences between the variants which “weight”, in level flight, must equal “lift”—the major constituent driver of ‘drag’.
If we take the empty weights, add half-internal fuel weight and the basic ‘day-one’ combat weapons load weight (with a wedge ‘rounding-up’ to account for the weight of weapon interfaces of each of the three variants), we arrive at a ‘likely’ set of aircraft weights for the Transonic Acceleration KPP:
Empty Wt 50% Fuel Wt Weapons Load (2 JDAMS, 2 AIM-120s) Total
F-35A 29300 9240 5000 43540
F-35B 32300 6750 3000 42050
F-35C 34800 9875 5000 49675
These configurations, whether exactly those used for the Transonic Acceleration KPP or not, are at least representative of the aircraft weights before release of air-to-ground stores and at the mid-mission point. They are therefore ‘realistic’ for use as a baseline for our explorations. They also cannot be too far off relative to each other to make too much of a difference in our search for understanding, as the only variable is fuel weight and the lower F-35B model JDAM carriage capability.
Using the F-35A Model weight as the basis, we find that over a range of mid-mission fuel loads, the F-35C weight, and therefore lift needed to be generated in level flight to be around 13-14% higher than the F-35A.
Using the F-35A Model weight as the basis, we find that over a range of mid-mission fuel loads, the F-35C weight, and therefore lift needed to be generated in level flight to be around 13-14% higher than the F-35A.
 |
| Figure 9. Relative f-35 Variant Weights |
We also see that largely due to a smaller total fuel load and lower weapons payload, the F-35B needs to generate approximately 3-4.5% less lift to maintain level flight. (This also again illustrates a point made in the Sustained G discussion about the lighter an aircraft, the more sensitive its performance is to weight changes.)
We can therefore conclude that the wing wave drag coefficient percentage contribution to the total drag coefficient that is due to lift for the F-35B is less than the F-35A (under the same flight conditions in the speed region we are looking at of course), and that the F-35C’s wing wave drag coefficient contribution due to lift is significantly higher than for the F-35A. Further, we can conclude there is little, if any, difference between the wing wave drag contribution due to volume for the F-35A and B because the F-35A and F-35B wings are identical in area: they have the same fixed length (chord), and span. We also know the area of the A and B wings are identical and their cross-sectional volumes very nearly so (more on this in the next section).
We can therefore conclude that the wing wave drag coefficient percentage contribution to the total drag coefficient that is due to lift for the F-35B is less than the F-35A (under the same flight conditions in the speed region we are looking at of course), and that the F-35C’s wing wave drag coefficient contribution due to lift is significantly higher than for the F-35A. Further, we can conclude there is little, if any, difference between the wing wave drag contribution due to volume for the F-35A and B because the F-35A and F-35B wings are identical in area: they have the same fixed length (chord), and span. We also know the area of the A and B wings are identical and their cross-sectional volumes very nearly so (more on this in the next section).
We can also conclude that there is a large difference between the F-35A and F-35C in wave drag contribution as a percentage of the total drag due to both lift and volume. This is because the F-35C has a much bigger wing and control surfaces (more on this too in the next section). The overall coefficient of drag for the F-35C might theoretically be smaller than the A or B at some points along the curve, but that would have to come from reduction in the total drag via other means. For all we know, the total volume distribution of the F-35C comes closer to an ideal Sears-Haack distribution than the other two variants. We cannot definitively rule this out without more data, but given the differences between the F-35A and C in transonic acceleration, I believe we can assume with some relative confidence the F-35C has an overall drag coefficient much, much, higher than the A model, and even if the F-35C drag non-wing coefficient was lower than the A model, it certainly does not overcome the F-35C’s larger cross-sectional area. If the differences in drag coefficient due to wing lift were the only effective differences between F-35 models, we would expect the F-35B to have better transonic acceleration more on a par with the F-35A due to it weighing less when loaded, and we would expect the F-35C to be much slower in accelerating in level flight than either the A or B model. Since the A model has the best transonic acceleration, we should assume there is some other factor (or factors) affecting the F-35B performance beyond wing wave drag. Also, while it appears the C is the worst performer in transonic acceleration, it is almost certainly NOT due solely to wing wave drag contribution to total drag coefficient. This leads us to the next discussion, Aerodynamic shaping.
Aerodynamic Shaping
Even this kind of comparison without hard data would be impossible if we were comparing completely different aircraft, or if we were seeking to quantify the differences instead of just understanding them. The good news is our limitations will make this section mercifully short.
Cross-sectional area comparison
First, we find that the cross-sectional differences between the F-35A and C are different in nature and scale than the differences between the F-35A and F-35B.
 |
| Figure 10. Cross-sectional Are Differences |
The differences between the F-35A and B are more in the 'non-direct' lift regions of the cross-sectional area and are obviously due to the B model lift fan installation (#7). The wing profile is slightly increased at the lower wing root (#6) due at least in part to the STOVL roll-control system installation.
The shape differences between the F-35A and F-35C are far greater. First, wing (#3) and horizontal stabilizer (#2) extensions as well as the taller vertical stabilizers (#1) add significant cross sectional area to the C model. Further, it appears the C model wing cross-section is indeed thicker (#5) than the F-35A’s to support the greater wingspan, but this could be an illusion from a longer curved under-surface and/or wing twist (either way, this presents an increased cross-section perpendicular to the airstream versus the A or B model). The F-35C wings also present an effectively increased cross-section to the airstream at the wing-folds (#4). The F-35A and C do have a slightly increased cross-section versus the B model in the area of the tail hook enclosure (#8) but unlike the other cross-sectional areas, it is unclear how much of this is area that is presented perpendicular to the airstream in level flight. Not known is if there are unseen relative differences in inlet design that affect not only installed thrust but also effective cross-sectional area, bur I believe we've captured the dominant difference drivers.
The Impact of Lift Surface Area on Wave Drag
As already noted, Wave Drag due to ‘volume’ as a contributor is characterized in terms of a cross-sectional area for some fixed length. The F-35C is overall slightly longer than the F-35A and B, but its lift surfaces also have a significantly longer “fixed length” than the near identical F-35A and B surfaces. The F-35C’s wing and tail surface wave drag contribution due to ‘volume’ is therefore significantly greater than that for the F-35A and B. As a result we would expect the total drag of the F-35C in level flight to be much greater than the F-35A or F-35B due to having both a higher wing wave drag coefficient and greater cross-sectional area.
 |
| Figure 11. Extended Vertical surfaces on C model not shown |
Engine Installations
Whereas the first two topics of weight and aerodynamic shape dealt with the drag side of the acceleration equation, the engine installation aspect is relevant more on the thrust side of the equation. This is actually a pretty simple thing to compare on the F-35, as all three engines are essentially the same, with the major differences (performance wise) being between the F-35A/C engines and installations being highly common and the F-35B engine installation being very different from the other two. The physical differences of the F-35B installation-- the integrated lift-fan at the front of the system, the articulated exhaust duct with a different afterburner and nozzle installation at the other end, as well as a roll control nozzle system tapping flow off of the core engine in the middle cannot ‘help’ the conventional-mode of operations. Any significant performance differences between the F-35A and C engines would have to be due to any installation differences that are not apparent/acknowledged.
For all practical purposes, we should then expect the F-35A and F-35C engines to have “the same” installed thrust and efficiencies. On the other hand, with the F-35B we would expect installed thrust penalties due to the added mass of the lift fan drive shaft and open roll-control ducting, the shaft passing through the bifurcated air inlets where they meet, as well as the different exhaust nozzle. We don’t know for certain, but it makes sense to believe the F-35B engine installation may be somewhat less efficient than the A or the C model in the conventional operation mode.
A Summary of Thrust vs. Drag
To summarize the relative differences between the variants that would affect transonic acceleration in a table:
 |
| Figure 12. Factor Differences between variants, F-35A Baseline |
Transonic Acceleration: What is it good for?
The first thing to understand about the transonic acceleration KPPs is that they are proxies for the reasons WHY a fighter aircraft would want to accelerate quickly through the transonic region in the first place. The (obvious) answer is to gain a position advantage against an enemy in the air or on the ground by reaching that position to achieve an advantage before an enemy can position themselves to counter it. In the Air-to-Ground scenario this might mean simply getting to a high speed sooner to reach a time-critical target before it disappears or for some similar reason. We will ignore the air to ground aspect in thinking about the F-35s transonic acceleration KPPs and assume that because they are set at a 0K ft altitude, there wasn't a lot of Air-to-Mud consideration behind them. In the Air-to-Air combat it is about either closing some distance to gain a superior position on an opponent or opening some distance to gain position or keep the opponent from gaining a position on you. So the key thing to remember, and I hope the reader finds this an irritatingly obvious point, is that an acceleration KPP is about the ability to cover some distance over a set time when starting at a relatively lower speed. A transonic acceleration KPP is about being able to cover some distance as quickly as possible while moving ever faster through the particularly challenging transonic speed region. This seems to beg the question: how significant it is really that the F-35A is projected to exceed its KPP by 8 seconds? (Note: We’ll get to the B and C KPPs in passing as we proceed.)
F-35A Transonic Acceleration Performance
It has been determined by the F-35A program’s analyses that the original Transonic Acceleration KPP for the modeled F-35A will be exceeded by 8 seconds, taking 63 instead of 55 seconds. But considering we have multiple sources (#5, #6 above) citing excellent acceleration characteristics compared to an F-16 that imply superior initial acceleration, and the analyses are based upon degraded engine performance (#7, #8 above) how likely is it that the 8 second difference is even operationally relevant? Take a look at a second hand on your watch or clock. Watch it for 8 seconds. If you were going M1.2 at the end of eight seconds, how fast were you going at the start of the eight seconds? M1.19? M1.18?
Just from the drag equation and the drag coefficient charts we know that acceleration will be much greater at the start of the acceleration run than at the end. Is it possible the F-35A exceeds the initial predicted acceleration but just barely misses the full expectation? Answers to these questions could have bearing as to whether or not extending the KPP by 8 seconds was even operationally relevant. As it stands, it should still be viewed as superior to a Block 50/52 F-16. As we will see, Kloos (Ref #6) understates the F-35A’s abilities compared to the Block 50/52, at least at the 30K ft KPP altitude (Ref #10).
How the F-35A new KPP standard stacks up against the F-16 Block 50/2
If we examine the F-16’s ‘Combat Max AB’ transonic acceleration data (Ref #11, Table A8-12), and compare it to the F-35A’s newest transonic KPP time (63 seconds) we find the F-35A loaded with two AMRAAMs and two 2KLb JDAMs has better acceleration than the F-16C/D in 20 of the 30 possible weight/drag index combinations shown in the tables (weights from 20K lbs to 41K lbs, and Drag Indexes from 0 to 250). |
| Figure 13. F-35A KPP Superior to most F-16C Weight/Drag Combinations |
 |
| Figure 14: External Stores start adding drag quickly |
This is about as much as a Block 50/52 F-16C can weigh/carry without increasing the drag count:
 |
| Figure 15. Max 'Clean' Block 50/52 weight.( Source Ref 11) |
 |
| Figure 16: Empty Fuel Tanks and Store Stations Still Add Drag |
On the other hand, for 6 of the 20 load-outs where the F-35 KPP performance meets or beats the F-16C, the F-16 can’t even GET to Mach 1.2 (for three combinations the F-16 performance is ONLY subsonic).
In short, even if the F-16 is running on fuel fumes carrying wingtip AMRAAMs and
LANTIRN pods, it can only carry just a little more internal fuel before the F-35A's latest transonic acceleration KPP standard can be said to be ‘better’ than an F-16C Blk50/52 in transonic acceleration.
In short, even if the F-16 is running on fuel fumes carrying wingtip AMRAAMs and
LANTIRN pods, it can only carry just a little more internal fuel before the F-35A's latest transonic acceleration KPP standard can be said to be ‘better’ than an F-16C Blk50/52 in transonic acceleration.
 |
| Figure 17: F-35A KPP performance vs. F-16 w/ DI=100 |
 |
| Figure 18: Only 2 AIM-9s greater load than F-35 Internal Load Out. It'll need those AIM-9s long before te F-35 will. |
Even the F-35B and F-35C transonic KPP times meet or beat a similarly-loaded but much lighter weight Block 50/52 F-16s:
 |
| Figure 19: With comparable war loads: F-35A and F-35 B beats all & F-35C beats most F-16 configurations. |
The ‘meaning’ we can derive from the revised F-35A transonic dash KPP is this: It still represents the stated ‘F-16 Like’ performance goals and overall, it exceeds them. Other than that, everything else is guesswork.
BUT! The ubiquitous 'some' might ask....
‘What If’ Operational needs require the F-35A/B/C to get rid of all or part of those extra darned seconds enroute to Mach 1.2?
THAT topic we will address in closing the series in Part 3.
