I notice that several people want specific aircraft tested at various altitudes; I should mention that every test of three to four runs for a given aircraft can take 45 minutes for the actual test, another 45 minutes to get the raw data from the track and then another hour to reduce the information to a usable form.
If I weren't currently unemployed, I couldn't have accomplished what I have in the time I did it. In a sense, this sort of 'keeps my hand in' my professional calling; I spent several years writing technical test procedures aimed at proving whether an electronic or RF system or specific subunit met its required specs.
Anyone should be able to closely or nearly reproduce my results using the same methods with a little patience. Here's how I do it:
Acceleration testing is simple if you are consistent. Every test has to meet certain basic standards, and those standards have to be applied to every aircraft tested. 'Fairness' doesn't come into it; at altitude X, aircraft Y will fly in a straight and level course from starting speed Z until it can no longer accelerate (or the engine bursts into flames). Aircraft either reach a certain speed at a certain rate or they won't. Make sure that you record a track for every run, or series of runs--I have found that a maximum of four 'runs' will give you a consistent (and easily reproduced) result in a single track; after that, you have burned off a fair amount of fuel or exhausted the WEP additive's reservoir. The overheat light may also start coming on a bit sooner.
The Crimea map is assumed to be the ideal for testing conditions, and it has the added advantage of having a convenient large body of water for a consistent surface (none of those annoying 'bumps' as you cross rivers and lakes); by default, we set the time as noon, and in the QMB, you can select the nearest height to the desired test altitude, divert a bit south to get over the water, and then head due west, or 270°, obtain the desired test alt and speed. Trim for level flight at that speed and altitude; it can take time, but the idea is that every aircraft tested will start at the same speed and in the same flight condition. Make sure that your cooling flaps are in closed/auto position, your temps are in the normal range, and that your track is running. Then slam the throttle and prop pitch all the way forward and if available, add WEP.
Do your very best to maintain your altitude and course with stick and rudder until you can trim out and even then, some aircraft simply will NOT trim out in a dynamic (changing speeds) situation, so you must keep a close eye on your climb and turn & bank indicators and do your best to stay on the straight and level with the stick and rudder and try to keep as close as possible with your trim inputs. Keep going until the aircraft simply will not reach a greater indicated speed and/or the overheat gets critical (some aircraft are more sensitive to overheat than others, but the best indicator is the aircraft slowing down a bit instead of maintaining or gaining speed in level flight). End your run, reduce throttle and prop pitch (except LW fighters with their automatic controls), turn off WEP and open your radiator or cowl flaps all the way, and return to near your starting point at no more than 55% throttle; in all but the most extreme cases, this will cool your engine off to a reasonable level for the next run.
Rinse and repeat two or three more times. Make sure that you close your cowl or radiator flaps; these are very easy to forget, and they will reduce your acceleration by quite a bit in most cases.
Here comes the hard (and tedious) part: Reducing the Data.
You have flown your runs and recorded them, and you are satisfied that the 'runs' are as straight and level as you can get them (and I have found that this standard will evolve as you become more practiced in the test process). You have obtained the 'raw' data, which means that you've invested about 45 minutes to an hour and a half so far for each aircraft you test. Now you must extract the data by running the track and noting the times you reach each speed, along with relevant information.
Running the track immediately afterward you have recorded them should reproduce the results accurately, and you will be able to pause at every speed increase and note altitude and course as well as time: in many cases, seemingly slight variations of altitude or direction can raise or lower the time between 10kph intervals quite a bit. I also recommend watching the track in Wonder Woman view, but flying the tests in cockpit view; this will give you a clear indication of how accurate the instruments are--the needle and ball can be graded/compared against the vector ball, and the altitude and true speed are indicated in single digits, versus the basic 10 unit increments in the speedbar. This means that you must de-select Loop, select Manual Time Compression (this is a sanity preservation measure--you don't want to spend 10-15 minutes waiting for your aircraft to return to the start point after every run), Manual View Control and In-Flight Messages before starting the track.
DO NOT START WITH A BLANK SHEET OF PAPER. I recommend using a notebook of lined paper, with your start speed and target speeds listed from top to bottom in descending order on the left, with room for at least six columns: alt/course and time for each run, with space at the bottom for rows noting the overheat and end run times. Record the 'raw' time as each speed is reached, along with the alt/course; the best way to do this is to Pause the track immediately. Some aircraft transition across 10kph points in 2 seconds or less, so most of us will be unable to keep up with the data without pausing. It is a good idea to note the aircraft type on the top of the page, just in case you can't get around to reducing the raw data until later.
By 'raw' time, I mean the track time noted in the lower right hand corner; you can convert the start time to 0:00 and measure from that point for each run later.
I have been recording data from 270/280 to 350 kph, 350 to 370, and then in 10kph increments thereafter. I use Indicated Air Speed, because in-cockpit that is all you will have, even with the Speedbar. You can get the True Air Speed from Wonder Woman view if you like, but the conversion at a given altitude is the same for all aircraft. The starting speed is more or less arbitrary, the fact is that some of the faster and heavier American types are hard to get stable at that speed; you might prefer to start at the aircraft's economy cruise speed or at the speed that you believe that WWII squadrons flying that aircraft would get to before entering the combat area (pilots flying slower accelerating types like the P-40 often made sure to be already 'fast', or over 350kph indicated, as they entered the area of likely enemy contact--units that did this usually suffered fewer casualties as a result).
Simply put, if you aren't quick, you had better already be fast when the feces starts.
So far, I have limited testing to 10,000 ft/3050m, but anyone can choose to test at any altitude as long as he compares all of the aircraft he tests at the same altitude--no apples to oranges comparisons allowed.
Okay, we have the raw data for three or four runs written down--now we reduce it to basic units.
I prefer to simply subtract the start time from the next measuring point, which gives us the time from 'zero' to the first marker. If your run's start time is 1:17 and you reach 350 kph at 1:33, the interval time would be 0:16, or 16 seconds. For the sake of simplicity, I measure the times between markers, so the next interval is from 350 to 370. If you reach 370 at a raw time of 1:39, you record a time of 0:06, or six seconds. Continue in this manner for each of your runs. Each interval is recorded in terms of the time between designated speeds, and your times between those speeds should be fairly similar for each run. If they are very different, we will want to refer back to the alt/course columns for that interval; if you climbed 40-60 meters during a given interval on one run, that time will likely be a bit slower than the others, and if you were in a steady shallow climb or shallow diving condition the effect can be cumulative. If run #1 and run #3 are fairly straight & level and have a near identical time for a given interval while run #2 is in a bit of a climb and is clearly slower, or in a bit of a dive and a lot shorter, you may choose not to use run #2 for that interval.
What I try to do is get a reasonable average time for each interval, and then add them up for an 'ideal' average for the whole range of start to finish speeds. You can then add the times to get a clear idea of how long it takes most pilots to reach a given speed from the starting speed in that aircraft at that altitude. This allows you to see that by the time an FW 190A-5(1.65 ata) reaches 400kph at 3050m (45 seconds from start at 270kph), a Spitfire Mk Vb (CW) is between 450 and 460 kph and has been doing over 400kph for almost 20 seconds, which would confer an almost insurmountable lead and the initiative to the pilot of the Spitfire at that altitude.
You may notice that the accuracy with which you can fly straight and level is at least partly dependent on the cockpit instruments' clarity and honesty (for lack of a better word); this is very much a factor in how much you can squeeze out of a given aircraft. A steady slight climb or dive can result in the addition or loss of two or three hundred meters in a matter of seconds at high speeds and a change of vital seconds in an interval time. If it is a factor, make sure to note it; that is partly why I suggest switching back and forth from Wonder Woman view while viewing the track.
I have created a series of Word documents with tables listing the results and notes, but they are too big right now to attach; I'll see if I can just copy the data from the tables to get under the 19.5 kb limit. I also plan to create a chart in Excel in a week or so (again, because it helps keep my hand in).
In any case, I believe that using my methods, anyone can create a fairly accurate picture of what plane A can do versus plane B in terms of level acceleration. I welcome others to try it and report their results.
cheers
horseback
|