View Full Version : P38L climb performance via energy analysis

08-23-2005, 12:15 AM
Level Acceleration Performance Tests


In an effort to derive a climb schedule (the best speed and rate of climb to maximize climb for various altitudes) via techniques actually used for that purpose, I performed level acceleration tests to determine the specific excess power that may be used to determine best rate of climb. In the modern total energy perspective of flight theory it is known that power for acceleration or power for climb are essentially the two aspects of the same thing, a balance of kinetic and potential energy derived from lift and engine power and that of drag. Therefore one can merely fly level at a specific altitude and a specific power setting and determine the maximum specific excess power and the speed associated with it.

The total energy state of the aircraft is merely the sum of its potential and kinetic energy. Specific excess power is merely the derivative of the total energy state with respect to time. The power is excess in that it is power in excess of that needed for level flight or merely more than drag on the airplane. The conceptual formula is simple:

Ps = dh/dt + (Vt/g)(dVt/dt)

where Ps is specific excess power in ft/sec, h is altitude in feet, Vt is true airspeed in ft/sec, and g is the gravitational acceleration constant (32.3 ft/sec^2).

Any specific excess power determined via level acceleration tests is true only for the specific conditions it was tested under, i.e. altitude, temperature, weight, thrust, etc., however an adjustment can be made for other weight conditions that I will explain later.
The specific excess power is directly translatable into rate of climb and maximum Ps is the same as best ROC for the specific altitude and speed.

More detailed info on theory and procedures can be found here:




Missions were created in FMB for target test altitude. I used Crimea map, P38L, 75% fuel, 54€ of manifold pressure, auto radiator. Basically, the procedure is to establish level flight for target altitude at the lowest power setting needed and add throttle to desired level (here 54€ MP), and maintain level flight within 300ft of target altitude. Data collected with UDPGraph was IAS, TAS, MP, altitude, and time between readings. Poll time for UDPGraph was set at 500msec. Inputs (dh/dt and dVt) to the Ps equation were estimated via the difference between altitude and TAS between data collection intervals divided by the poll interval recorded by UDPGraph. I ran at least three trials for altitudes of 10,000, 15,000 and 25,000 ft. None for 35,000 ft as the P38L cannot maintain correct manifold pressure as listed in the pilot€s manual, and only 1 trial of 5,000 ft as climb performance is not terribly at issue at this altitude, and I got really bored with this process.

Mean results of the tests are as follows:

5,000 ft ROC = 3206.4 ft/min IAS = 183.13 mph

10,000 ft ROC = 3073.35 ft/min Weight Adjusted ROC (1.02 correction factor) = 3147.24 ft/min IAS = 175.36 mph

15,000 ft ROC = 2650.8 ft/min Weight Adjusted ROC (1.028 correction factor) = 2725.02 ft/min IAS = 173.8 mph

25,000 ft ROC = 2289.4 ft/min Weight Adjusted ROC (1.038 correction factor) = 2376.4 ft/min IAS = 153.41 mph

To explain the weight correction, Ps and thus the ROC were determined for a plane of 75% fuel weight, however the pilots€ manual shows fuel used to obtain the altitude, thus the plane is lighter at higher altitudes. I determined a weight for a P38L with 75% fuel with a base weight of 14,400 lbs + 6.5 pounds per gallon of fuel indicated by UDPGraph for 75% fuel, that weight being 16318.4 lbs. To determine the weight that the plane would be as per the pilots€ manual climb schedule I merely subtracted a weight equal to the weight of fuel used in the chart from the weight for the plane at 75% fuel. Weight Adjusted Ps was then determined by a simple proportional relation:

the calculated Ps/the 75% fuel plane weight = the weight adjusted Ps/the weight that the plane should be as per the climb schedule.


The values determined by these tests are fairly close to the pilots€ manual climb schedule for altitudes up to 15,000 ft, therefore I would suggest that the flight model may just be off at higher altitudes wherein TAGERT€s tests showed the P38L to be most lacking. Given the recent discussions about instrument accuracy in the game I believe use of 90% power for any climb test or level acceleration might be as valid as the 84% that seems to produce a MP of 54€. If the gauges are right, then the FM is off a bit at higher altitude where most people do not fly and, I feel, is relatively inconsequential. Most fights are 5000m and below on most servers I have been on. If gauges are off, there is no way to be sure of anything.

I can supply tracks and data if anyone wants it, but I am pretty much done with testing and this issue in general.

The FM is mostly off at higher altitudes. Don€t fly there until it is fixed or don€t expect real life performance. No one else seems to be up there, so it hardly matters.

08-23-2005, 01:20 AM
Energy Analysis of TAGERT's climb check for P38L 54" Hg manifold pressure, 75% fuel


Some time ago TAGERT challenged the idea that the P38L Late was basically not an overcharged uber plane and made a good demonstration of the lack in climb performance for the P-38L and the P-38L Late.

His thread can be found here:


In that thread TAGERT provides a reference for real life P-38L performance as evidenced in the actual pilots' manual for the plane. He also provides several test conditions for the L and the L Late as well as tracks to demonstrate his performance of climb checks to test if the plane can climb as it should.


The chart above gives the expected climb performance for the P-38L. We are concerned with the figures relevant to the 17,400 weight configuration.


I submitted TAGERT's P-38L, 54" Hg, 75% fuel climb check track to an energy analysis and these are my results.

In the leg of the climb to 5,000 ft, TAGERT's average ROC and IAS were 2977.65 ft/min and 189.27 mph (well above the required 180 mph) respectively. These averages are based on the period of time when his climb actually met the 180 mph requirement for the climb schedule. In fact, once 180 mph was reached in the climb TAGERT continued to accelerate in his climb which violates the provisions of the test.

A specific excess power analysis showed that over that same climb interval, the plane had an average Ps of 54.14 ft/sec or 3248.36 ft/min. This is the climb potential that could have been used to meet the test criteria but was not, thus adversely affecting climb performance and time to climb figures. Essentially, there was power that could have been used to climb but was not.

In the 5,000 - 10,000 ft leg, average ROC and IAS were 2800.31 ft/min and 185.22 mph, respectively. The target speed for this altitude should be near 178 mph. It is not. The energy state of the plane at this point averaged Ps = 48.28 ft/sec or 2896.89 ft/min. In this leg, there appears still to be sufficent energy to approach but not meet the pilots' manual figure of 3100 ft/min. Had velocity been converted to climb here, the Ps value (energy potential for climb and therefore rate of climb) would have been higher as the equation is more sensitive to changes in rate of climb than rate of acceleration.

In the 10,000 - 15,000 ft leg, average ROC and IAS were 2548.26 ft/min and 176.84 mph. Average Ps = 42.83 ft/sec or 2570.00 ft/min. Here, climb performance is well below 2900 ft/min while speed is still above but within acceptable limits for the test.

As I stated in my previous post, climb performance above 15,000 ft does not appear to be correct so further analysis of the energy in TAGERT's climb check is not really necessary as there is no expectation on my part that it will be as stated in the pilots' manual.


The specific excess energy (or climb potential) of this climb check seems to decrease at a rate greater than that expected from the climb schedule. The plane's ROC in the climb check fails to meet milestones for ROC and IAS for a good deal of a test. In fact, it fails to meet the requirements of the test, which is supposed to be evaluating the plane's performance against the real life climb schedule, at a crucial time in the climb check. Yet, can it be a fair test if the test requirements are not met when they are easiest to meet, at the beginning of the climb check?

In the leg to 5000 ft, the plane's thrust is used for speed instead of climb. A consequence of this is that more energy is lost to drag, because as speed increases drag increases, and drag is highest at lower altitudes. The idea of the climb schedule is that it is the best way to fly to conserve energy for climbing. That was not done in this climb check when it mattered most. Energy is lost to drag in the climb and climb performance suffered due to it.

Note that in my own level acceleration tests in the post above, I was able to derive a climb performance that mirrored the real life schedule for altitudes of 5,000 and 10,000 ft and pretty close for 15,000 ft. This supports the idea that the in-game P-38 L has the potential to climb close to real life performance at altitudes below 15,000 ft. The energy analysis of TAGERT's climb check leads to a similar conclusion. Had the requirements of the climb schedule been met, TAGERT's test might have been closer to real life values in the pilots' manual for altitudes of below 15,000 ft. Above 15,000 ft, all bets are off and the in-game P-38L does not seem to perform as it should under the conditions that it was tested.

08-23-2005, 03:13 AM
nice analysis, would be interesting to see somthing like this for the 109G6/G6AS

after all the latest discussions I gave the 109G6/AS a try at 10.000 Meters and was surpriced by the bad handling compared to P38, P47 and P51B they fought up there RL.


08-23-2005, 02:12 PM
One thread is enough dont you think? Why not keep it in tagerts thread?

08-23-2005, 08:35 PM
I thought since I had done so much work that it needed its own thread as opposed to being buried in page 24 of TAGERT's thread amongst many other posts that involve sniping among parties.

If anyone new was going to read what I had done, I thought it best to put it into a new thread. That's about it.

08-26-2005, 07:39 AM
So your solution to this problem is not to fly over 15,000' if you expect accuracy? Well, maybe you fly around in the weeds, but many of us fly at, and much higher than 15K and would like all planes to perform to their specs up there.

08-26-2005, 08:26 AM
Q.E.D. eh Tachyon http://forums.ubi.com/groupee_common/emoticons/icon_smile.gif
Great analysis and I agree with your logic in creating a new thread.

Next, of course, is to actually jump through the hoops and meet the climb sched.

09-05-2005, 04:24 PM
I agree that the climb schedule should be flown, but meeting it or not is not the real issue. Whether a plane can meet a predicted climb schedule has more to do with reality than performance in a sim in my opinion.

I chose to derive a climb schedule through energy analysis as it was impossible to meet the pilots' manual climb schedule in many instances and it is certainly easier to fly level than to climb at an exact ROC and speed.

In fact the energy analysis yields more information than merely trying to meet a climb schedule. It tells you a bit more where things are breaking down in the sim and is not dependent on the pilot's performance and flight skills.