Wednesday, April 17, 2013

The F-35 and the Infamous “Sustained G” Spec Change

PART 1

Introduction

I can’t remember when I saw so many media outlets, bloggers, and just general ‘people’ with their panties in a wad over something they didn’t really understand. Honestly, who among those ‘critics’ bemoaning the Sustained G turn requirement changing from 5.3 to 4.6Gs even know everything they need to know as ‘inputs’ before they could even begin to formulate an informed opinion on the topic? There is STILL insufficient information in the public domain to come to any conclusions, but there’s a heck of a lot of presumption and assumption in spite of it.

I was neck deep in a home improvement project when Dave Majumdar at FlightGlobal’s DEWline blog put up his “What's the operational impact of reducing the F-35's performance specs?” piece. Majumdar gave a pretty big voice to an ‘anonymous’ but ‘highly experienced ‘ fighter pilot that had all sorts of negatives to pass on to the public. (I believe if you view the Majumdar/Flight Global posts and articles that followed on the topic you’ll conclude the voice in question almost certainly was coming out of a Boeing F-18 test pilot). This ‘anonymous’ fighter pilot’s views were counter-balanced at the end by someone who had actual knowledge of the F-35.

Note: I now notice that the entire post I commented on has been rewritten (down the memory hole, eh?) but the parts I lament in my comment seems to have been ‘reformulated’ elsewhere in an article here. (If I have a major complaint about FlightGlobal’s reporting, it is just this kind of ‘rewriting history’ stuff.)

I commented, in part, the following at that time:

The really disturbing aspect of this 'story' is how an anonymous "highly experienced" fighter pilot somehow is able to gin up some world-class doom and gloom from a few insufficient data points. The doom and gloom gets quoted here, and now this article is being passed around by the anti-JSF crowd-- zipping around the globe as some sort of authoritative source. However, all there is really are two insufficient sets of information that in all honesty should prevent someone from reaching any conclusion other than the program is being 'managed'.

First, let's deal with the changing of the maximum sustained g-force value for the different variant turn performance criteria.

All we know is that 1) The sustained turn g-force objective is a proxy for overall aircraft maneuverability (not even a KPP) and 2) For the particular speed, bank angle, weight, and altitude data points selected, the program office lowered the ‘g’ value after extensive testing and analysis.  



Aside from no real reason to assume the Program Office could or would do so without a good reason, or belief that the differences would not significantly impact operational performance, we should also consider the very real possibility that the F-35’s best sustained-g turn performance may have been found to lay outside the pre-selected test conditions. Perhaps by a little, perhaps by a lot. We don’t know. Not even “highly experienced” fighter drivers, unless they’re on the program, can divine the answer or what it means without more data.

On the one hand, I could have placed more emphasis on the ‘unknowns’: if only so a few of the others who were/are willfully ignoring them would have been perhaps a little less eager to jump on the ‘Lowering the Bar’ bandwagon. I COULD have typed “For the particular speed, bank angle, weight and altitude data points selected, [which are unknown to us] the program office lowered the ‘g’ value after extensive testing and analysis. On the other hand, I shed no tears for the self-identifying ignorami, and find the duping of a small number of media types merely... ‘unfortunate’.

Of course, the usual suspects took exception to my comments, and now the whole thread looks pretty silly as the article that I was critiquing, and others were defending (and OBTW also attacking me), is no longer even IN the article we all commented upon. But my prediction of the obvious, that the doom and gloom sound bites from a rather dubious source would spread like wildfire throughout the 'interwebs' while the factual counterarguments would remain alone and unloved proved all too true.

Fortunately, we have a few other data points that we can add to other information, including knowledge as to how programs and requirements ‘work’ and at least form one or more hypotheses as to what the relevance Sustained G spec change ‘means’-- based upon physics, publically released (vs. leaked) information (vs. unsubstantiated suspicions) about the F-35, and more importantly how everything relates to modern fighter design requirement priorities. We can make some assumptions (and identified as same) and use parametric modeling to more thoroughly understand the impacts of the changes, especially as they might either reflect or affect HOW the F-35 might ‘fight’ in the Within Visual Range (WVR) environment.

This post, as the title reflects, will concern itself with F-35 Sustained G-performance. I’ll get to the ‘Supersonic Dash’ spec change later.

“Sustained Flight”



Let’s begin with defining Sustained Flight, Sustained Level Flight, and what makes a ‘Sustained G’ Turn... a ‘Sustained G’ turn. If you know basic aerodynamics you can skip this section, and I don’t want to hear about anything I ‘leave out’ or ‘over-simplify’ in the comments if you don’t skip it. I’m not trying to ‘dumb it down’, I’m trying to leave out stuff I don’t need to explain to get to the larger points I’m trying to make.

‘Sustained Flight’ is the condition where airspeed, altitude, and load factor ”n” are all held constant. To be ‘constant’ Thrust “T” and Drag “D” must be equal, and Lift ‘L’ has to equal Weight ‘W’ x n. Straight and level (coordinated) flight is ‘sustained’ when lifting surfaces are level, the aircraft is not climbing or descending, neither is it accelerating or decelerating, and the bank angle is ‘zero’.

When flying straight and level, the load factor equals 1 (n = 1) as in “1 gravity” or 1g. But it is not actually ‘gravity’. The load factor n is dimensionless: the ratio of Lift to Weight (L/W), and each component has the same unit of measurement and are thus cancelled out (pounds/pounds, etc). Load factor is referred to in terms of ‘g’ because it is "perceived" as some ‘multiple’ of the acceleration of gravity on board the aircraft.
 A ‘sustained turn’ is a turn where not only airspeed, altitude, and load factor ”n” are all held constant, but also the (non-zero) bank angle of the turn is held constant. To be ‘constant’, Thrust “T” and Drag “D” must still be equal, and Lift ‘L’ has to equal Weight ‘W’ x n, but ‘n’ is no longer equal to 1 because of the vector change of the aircraft as the turn is being executed.

For this exercise (and simplicity’s sake), we’ll assume the earth is ‘down’ and the aircraft is right side up (not inverted) and just say n is now greater than 1 (n > 1). For an aircraft with a typical wing-body-tail planform to enter and sustain the turn, the pilot/controller provides input to the control system that deflects the control surfaces to induce and then hold a bank angle, while increasing angle of attack needed to increase lift generated per unit of wing reference area to keep it equal to the load factor times the weight (n x W).

This means for any given set of airspeed, aircraft weight, and density altitude values, the load factor- accounted for in terms of ‘g’s- is a function of bank angle. Specifically, g= 1/cosØ. For example, a 60 degree bank angle (Ø=60), the ‘g’ value would be 1 divided by the cosine of 60, or 1/.5 = '2gs'.


60 Degrees Bank Angle and Resultant 2g Load Factor 

Increase the bank angle, and in a steady level turn the  g’s increase as well. As the use of a trigonometric function implies, and the chart below illustrates, the relationship is NOT linear.

'G's as a Function of Bank Angle
From the shape of the curve, we can easily observe that ‘g force’ begins increasing at a faster rate than the bank angle is increasing at around 3gs (if you remember your math, it is the point on the curve where the slope (m) of a line that is tangent to the curve = 1). By the time the typical ‘hot-fighter’ max g rating of 9g’s is reached, the bank angle has increased to about 82.6 degrees.

So what is ‘happening’ in the specific region of the curve where the F-35s ‘Sustained g’ spec change occurred?

Bank Angles: 4.6 vs. 5.3 Gs


The Only Conclusion: Bank Angles.

After that we need to start making 'assumptions'.

It doesn’t look like the airplane is doing anything too ‘different’ (minimal y axis delta) on the curve to get that ‘g’ difference, does it? That’s the first surprise waiting for people who haven’t thought much about ‘Gs’: the difference in bank angle between the two ‘levels of performance’ is about 1.6 degrees bank.

The difference looks like this:

Depending on airspeed, the bank angle could translate into a ‘small’ or a ‘large’ difference in turn rate and turn radius. Without knowing for certain what the weight, speed, and altitude is for the ‘performance standard’ at either 5.3g or 4.6g, the difference in bank angle between the two figures is all we can conclusively determine. Everything else depends on the missing data.

 

A Couple of ‘Likely Truths’

I’m pretty comfortable making some low-risk assumptions on top of the one conclusion. The first one is that F-35 max sustained G capability for the unknown flight conditions and configuration is actually somewhere in between the 5.3g and 4.6g values. I’m comfortable in doing so largely because the program has already demonstrated conservative programmatic behavior with the B model’s “take-off roll” spec change.

[History Sidebar: For the takeoff roll spec change, the JSF Program Office didn’t just ease the requirement to make it so the B model would ‘pass’. They changed the requirement such that it both met military need AND would allow for further ‘bad’ surprises without having to revisit the issue. Because of this past program decision, and just using common sense, I suspect the actual performance difference between the original spec and current performance is even less than all the complainers realize.]

The second assumption I’m willing to make--with only slightly less confidence--concerns which ‘limit’ was hit going for 5.3 sustained ‘g’s under those mystery (can’t repeat it enough) conditions. Max Sustained Turn limits are either “lift limited” or “thrust limited”. I believe the ‘thrust limit’ was hit versus lift limit, for a couple of reasons, and it is important to note now that ‘thrust limited’ can be viewed as either insufficient thrust for the drag experienced at the specified conditions (weight, speed, and altitude) or higher than expected drag at the specified conditions.

I would almost bet, but have no information to confirm, that the drag rise was higher than expected for the selected set of 'spec conditions'. I remind readers again, we have no direct information as to what those flight conditions were.

Exercise: Exploring Comparative Sustained Turn Rate Performance


If we are to do ANY comparison of the JSF performance with any other aircraft benchmarks we are going to have to make what some (not me) might call a small leap of faith and presume the weight, armament, and fuel loads as well as the altitude and airspeed are the same as what is commonly referred to at F-16.net as the ‘Bowman' Paper or Brief.

[Personal Note: If there is any Cosmic Justice in this world I predict it will befall CDR Bowman for ‘phoning it in’ on his Air Command and Golf paper and the superficial analyses and ‘pronouncements’ within. If he is still active duty when the time comes, I look for him to standup one of the early F-35 squadrons, if Navy assignment desks are half as evil as Air Force assignment desks.]

For the purposes of our exercise therefore, let us ASSUME (and we all know how that word parses) that the flight conditions and the aircraft configuration for the Sustained G spec is as follows:
60% Internal Fuel  
2 AMRAAMS  
15000 Ft Altitude
Mach .8 Airspeed

I will proceed only in evaluating the ‘G-Spec’ change for the F-35A model and let others make their own analyses on other variants and other comparisons than the ones I will make. I also caution against assuming that the results for those conditions and aircraft configuration translate linearly to any other set of conditions (they don’t) and against assuming that the actual performance at that set of conditions/configuration was the ‘best’ possible at any one of the given conditions. Just one of the weight, speed or altitude parameters could vary slightly and sustained turn performance could go up or down in a manner out of proportion to the change. I’ve also noticed that while the specs usually look at a .8M sustained turn, from at least the ‘F-15 forward’ the best sustained G for US fighters at 10,000-20,000 feet altitude seems to reside somewhere between .85M and .9M.

 Keep these curves and data in mind when we move the discussion forward in the next post:

F-15 Turn Rates


These diagrams come from around the web and a personal reference I picked up at a used book store near Carswell JRB/Plant 4. The web sources are of uncertain provenance, but I found a good ref for the F-15 at a little different weight at 10K ft that correlates well to the F-15 data above. The F-18 data smells like public relations and is more nebulous. I can't tell you how many empty weight values I found for the F-18s, and early in the program the Navy was absolutely anal about couching internal fuel weights as fuel fraction percentages instead of just how much fuel would be carried internally.

The F-16 Blk15 is a good reference-probably the best available-as it has all that vaunted maneuverability the 'reformers' bemoan as ruined with later, heavier versions. The F-4 is a good data point because we have a distinct configuration attached to the performance numbers, and some disparage the F-35 as 'F-4 like'. The Mirage was interesting to me because it is a contemporary of the F-16 and represented the pinnacle of the delta-winged dogfighter (I remember reports of its debut at the Paris Air Show quoting a USAF General as saying "The French have finally perfected the F-106") until the Euro-canards started rolling off the line. In a perfect world, Sukhoi and Eurofighter will e-mail me their E-M diagrams before I go to press on the next post.

 I expect the next post to cause a furball all of its own.  

Since we’re working with ‘pictures’ and NOT real data, I’ve had to do some translating which may have brought associated minor errors with it. I don’t see anything remarkably out of place at the moment:


Aircraft Performance and Configuration Data Translated From Curves 





















If anyone has problems with the table I’ve assembled below using the charts above, and can come up with either better authoritative released data or good reasons why I shouldn’t use this data, I’ll leave the door open to changes for a couple of days while I finish doing the 'turn rate' and other math.

I predict that analysis of the data, combined with certain ‘truths’ about what level of significance should be attached to differences in sustained turn rates, and other things we know about the F-35 and have already covered concerning Energy-Maneuverability in an earlier post will clearly indicate we should be thinking of the F-35 as probably being a ‘competent’ if not ‘solid’ “kinetic” dogfighter and definitely NOT ‘a Dog’ (as people who have 'agendas' or little understanding of 21st Century air combat and aircraft design would lead us to believe).

Next: Part 2  

Tuesday, April 09, 2013

Charter Cable: Media Malpractice

Charter News, 'Isn't'

(Still working on a lengthy 'aircraft/F-35 maneuverability' post, but this HAS to go up tonight.)

Charter Cable is my cable provider. NO complaints about the internet speed or connectivity, not even though I suspect their move to 'all digital' last week wreaked havoc with signals (off and on) as thousands of users finally added even more thousands of cable boxes and cards to the network in just a few days. It now seems to have stabilized, so 'no problem'.

But Charter Cable's 'homepage' has a section with rotating 'news' headline pictures and captions. All too often the caption and photo make it appear that some tragedy has happened in the US or even just the 'Modern' world, and you click on the link talking about a school being bombed with what appears to be a typical American elementary school (they've done school buses too if I recall correctly) and the story is about a school in some war zone in a 'turd world' country. The Chief and I just chalked it up to lazy web content developers and editors.

Today, they went beyond 'lazy' and deep into 'media malpractice' . I got home and booted up the laptop to check the web and this is what greeted me (left headline):


Charter Home Page 9 April 2013 ~1920 Hrs CST
 


WTFO? "Veteran Kills 13"?


I clicked on the link, and this is what popped up:

Charter 'Article' 9 April 2013 ~1920 Hrs CST

Oh. A Serbian 'vet' in Serbia loses it and goes on a rampage. Tragic in it's own right. Why the 'trick' headline?

You would have to be either incompetent or agenda-driven to put this one up.  Either way it doesn't 'inform' but misleads and distorts several issues in one nice swoop.

Besmirching veteran's mental health? Check!

'Tragedy' as background for upcoming 2nd Amendment legislation? Check!

The Chief likes to try and calm me down when some unthinking slug nearly kills us because they're doing something clueless in traffic. She says something like "I'm sure they didn't see us". She forgets what makes me the MOST angry is the fact that they probably were clueless as to what was going on around them. If I assume Charter was just being 'brain dead' in this, it just p*sses me off more. tell me again: What business are they in? Do they have any standards?
 
Either way, Charter's website is Media Malpractice writ large in a Low Information Consumer world.

Are There ANY Adults At Charter Cable?

Monday, April 01, 2013

A Backgrounder on Energy-Maneuverability

Because if you learned all you know about E-M from Boyd, you’re only about a half-century behind the learning curve

Housekeeping

The first two things to remember in this kind of discussion are:
1. Performance metrics are used as PROXIES for what is important in a Weapon System under acquisition or already fielded. The metrics are NOT important in and of themselves. They are only as important as the degree to which they inform developers and operators on the system’s true versus desired capabilities.

and...

2. Weapon system specifications are initially established based upon what is believed to be needed and what is believed to be feasible within the projected budget and schedule before the development is given the ‘go-ahead’. There are only varying degrees of assumed confidence in the ability to achieve what is seen as feasible, and this depends much on perceived technology maturity. Only after the project is underway will the need and/or feasibility, given the actual maturity and budget, begin to be revealed, and it may be only truly knowable towards the end of development. Adjusting the specifications as new information is acquired, while still meeting Warfighter needs is sound engineering and management. It is NOT (as someone in the POGO crowd or ‘low information media’ might claim) “cheating” or a sign of “failure”.

Beyond Boyd: A Quick Survey of the Evolution AFTER the E-M Revolution

I’ve picked two AIAA papers on the subject to illustrate the post-Boyd reality: one from the 1980’s and one from the 1990’s. There are a fairly large number of scholarly papers concerned with the inadequacy of the traditional Boyd ‘E-M diagram’ that lead right on up to the present day, but as T. Takahashi observed just last year in a paper proposing yet another ‘better way’ to visualize E-M:
For aircraft whose missions demanded combat maneuverability, the United States Air Force required production of Energy-Maneuverability (E-M) plots. The design of the F-14, F-15, F-16 and F-18 was tailored by designers working to better the E-M capability of Soviet aircraft. In this context, a government oversight driven tradition developed where industry was required to plot many aerodynamic and aircraft performance parameters as functions of speed and altitude.
To begin our review, I think one of the better overviews of how E-M factors evolved post-Boyd is encapsulated in a paper by Skow, Hamilton and Taylor (1985). They observed:
“In the late 1940's and early 1950’s, with the advent of jet propulsion, radical new wing designs and greatly expanded flight envelopes, a corresponding need for more definitive measures of merit for aircraft performance comparisons was generated. When the "century series" fighters were developed and rear-aspect IR missiles became the principal air-to-air combat weapon, point performance comparisons were found to be inconclusive and insufficient to predict superiority. Out of this need, energy-maneuverability (E-M) concepts were formulated and developed. In the 1960's. E-M came into widespread use by aircraft designers and fighter pilots. E-M provided an analog picture of a fighter's performance capabilities over a range of velocities and altitudes. It gave quantifiable credit to measures of merit which allowed the advantages of speed (energy) and turning (maneuverability) to be balanced. These measures of merit were shown to be dominant in determining the outcome of an air battle at that time.  
 ...
But, as they say, time marches on, and in the past 10-12 years, several significant advancements have been made in the capabilities of fighter aircraft and air-to-air weapons. Three of these advancements; the all-aspect IR missile, greatly improved weapons delivery systems, and high thrust-to-weight engines have dramatically altered the character of the air battle, especially the close-in fight. The modern air battle is characterized by (1) time compression – shorter duration maneuvering required and (2) harder maneuvering - nose position at the expense of energy vs. nose position with energy conservation…
Air combat trends have expanded to ever increasing altitudes and speeds for beyond visual range (BVR) combat and conversely have tended to a lower and sometimes slower arena for close-in, within visual range (WVR) combat. … …This changing complexion of air combat, primarily due to the all-aspect IR missile, has altered the relative significance of the various performance characteristics with which we judge relative merit. Table 3 depicts the more common agility characteristics with some relative rankings.”  

Table 3 Reconstruction from “Advanced Fighter Agility Metrics
In 1992, Cox and Downing featured Befecadu (BF) Tamrat’s 1988 metric of Combat Cycle Time in the proposed system of “functional agility metrics” they evaluated for use in measuring aircraft maneuverability: Combat Cycle Time, Dynamic Speed Turn Plots, and Aircraft Energy State. The impetus for searching for these new metrics?
During the Korean War and the Vietnam conflict, jet fighter aircraft emerged with greatly expanded altitude and Mach ranges. This era also saw the advent of the short range heat seeking missile. These missiles required maneuvering to achieve a rear-aspect firing position. The measures of merit, altered to match the increase in aircraft capability, were advanced point performances (thrust-to-weight ratio, maximum Mach number, sustained turn rate, etc.) and the energy-maneuverability performance method. In recent years, the level of fighter aircraft and weapon system technological sophistication has reached new heights; the most critical advance being the development of the all-aspect infrared missile. This missile negates the requirement of maneuvering to achieve a rear-aspect firing position and concurrently has caused the traditional point performance measures of merit to become deficient for determining the combat effectiveness of a fighter aircraft. To remedy this deficiency, new measures of merit are being investigated which examine aircraft maneuver and control capabilities not previously quantified.
We could dive deep into discussion on each metric proposed in this paper, but I would like to focus on Tamrat’s Combat Cycle Time, as it is what differs most from the pioneering but simpler Boyd-Era POV of E-M, and the authors ably describe how Tamrat builds on same:
Traditionally, the need to achieve a rear-aspect position for a gun or missile firing opportunity led fighters to engage in battles lasting several minutes. These engagements were characterized by sustained maneuvering. This type of combat made the turn rate verses Mach number diagram (doghouse plot) useful for determining one fighter's advantage over another. When maneuvering for a rear-aspect firing solution, the interior of the doghouse plot is important. This interior region is represented by the sustained turn rate line and is typically dictated by available thrust and the lift to drag ratio of the aircraft. The compressed time scales of today's air combat arena have made sustained maneuvering less critical. The desire for a first shot opportunity leads to fights dominated by transient load factor/lift limited maneuvers. The emphasis then is shifting toward more dynamic maneuvering or the exterior of the doghouse plot. Figure 1 shows these critical regions on a conceptual doghouse plot.


From "Evaluation Of Functional Agility Metrics For Fighter Class Aircraft"
The first region is the dynamic pressure limit line (A) along which pitch rate is used to load the aircraft to enter a turning situation. Next, flight path and nose pointing maneuvers occur along the limit load factor and lift limit boundaries (B). The time to unload the aircraft is shown as the region in between the lift limit line and the 1-g line (C). The last critical region shown is the 1-g acceleration (D) in which an aircraft regains lost energy to leave the engagement or enter a turn to pursue another target. The Combat Cycle Time metric, developed by B.F. Tamrat combines these four critical regions into a single parameter. The time for each maneuver segment is calculated and then summed to obtain a time for the entire maneuver cycle. The critical regions, segment times, and corresponding maneuvers are:  
A t, pitch up to load factor limit  
B t, turn along load factor and lift limit 
C t3, unload from elevated AOA/load factor  
D t4, regain original energy level  
The Combat Cycle Time metrics offers an advantage over traditional measure of merit for turning maneuvers. The doghouse plot shows the sustained and instantaneous turn rates for only one flight condition (i.e. specified power setting, mach, altitude, altitude, weight, etc.). The Combat Cycle Time, however, is characterized by a continually changing flight condition constrained within structural, lift, and power limits.

What it All Means

If there are recurring themes over the years in all the proposed metrics that move beyond Boyd/ Christie’s E-M diagram paradigm, it is 1) All-aspect missile attack changed the game and 2) the proposed metrics try to account for the missing values of the speed of ‘transition’: the time needed to point the aircraft in a desired direction from another, the ability to decelerate as well as accelerate in vertical and the horizontal, to change pitch, roll and yaw.

 The lingering deference to standard E-M charting in the operational world springs more from their relative ease of understanding by non-engineer aircrew and government bureaucrats than from their utility. The E-M diagrams benefit from the fact that what really counts, is harder to explain. So we lumber along with the deficient E-M charting and rely on the ability of the untrained/uninterested to recognize the important nuances.

One of the biggest impacts on aircraft E-M performance that isn’t in the vehicle’s dynamics vis a vis the external environment (but sometimes dealt with in the OODA loop) is the man-machine interface: how fast the pilot perceives a change/threat and reacts to or preempts the effect of the change/threat by changing the vehicle dynamics. How well can the crew execute the mission in light of all the factors that could diminish the pilot’s ability to manage change and threats? Factors such as: How much attention does he have to pay to not bending or breaking the airplane or keep from departing controlled flight while engaged in the fight?





Limitations of Boyd-Era E-M Proxies and the F-35

A good example of the limitations of the traditional metrics is found in the caterwauling over the recent change in the F-35’s sustained-g turn standard from 5.3 to 4.6gs at some pre-selected altitude and speed. So much hang-wringing and finger-pointing—coming from those who not only do not know how important or unimportant the change was to needed/desired overall aircraft performance, but for the most part also have no idea what the change meant in and of itself, in ‘performance’ terms.

That, friend Reader, is a good lead-in for my next post (or two). Examine the reconstructed ‘Table 3’ above and ponder how the F-35 (especially the F-35A) performance stands up in the post-Boyd air combat world.

We'll soon engage in some speculative analysis on the F-35, but the speculation will be based upon logic, facts, physics, and experience.


References:

AIRCRAFT CONCEPT DESIGN PERFORMANCE VISUALIZATION USING AN ENERGY-MANEUVERABILITY PRESENTATION; Timothy T. Takahashi; AIAA 2012-5704

ADVANCED FIGHTER AGILITY METRICS; Andrew M. Skow, William L. Hamilton, John H. Taylor; AIAA-A85-47027

EVALUATION OF FUNCTIONAL AGILITY METRICS FOR FIGHTER CLASS AIRCRAFT; Brian W. Cox, Dr. David R. Downing; AIAA-92-4487-CP