|
Author
|
Topic: climb rate comparison
|
Nat Pilot
|
posted 01-27- 07:29 AM
Hey guys, ok, further to what I "thought" were my problems with the La5, I have the climb to 15,000ft now down to 5min 30, within 10 seconds of true, but because this bugged me so much as the climb speed seemed to high, I did exactly the same test with a 109k4 - runway start with full fuel load, AI takeoff and climb to 15,000ft. Time meassured fome AI hitting full throttle on takeoff.La5 - 5Min 30 secs, climbing at around 230mph (don't forget speed changes with altitude) 109k4 - 4 mins 3secs, climbing strangly at the lower speed of about 225mph. So, what seemd such a big problem because I thought that the climbspeed should be much lower, like around 200mph, is completely wrong when I compair to atleast 1 other top of the line aircraft, as such, my testing on climb and speed has now finished.... thank god for that lol ~Nat~ ------------------ 7./JG3 "Naturlich" "SDOE... What and where would you like to fly today?" http://members.nbci.com/naturlich/index.htm </B> IP: Logged |
Tailslide Pilot
|
posted 01-27- 06:08 PM
I don't use the AI to test planes.They don't necessarily attain the best climb rate. TS IP: Logged |
Tailslide Pilot
|
posted 01-27- 06:09 PM
Hmm.. or do they? How do you set your missions up for speed, waypoints, etc ? Maybe this would save me some time. TS
IP: Logged |
Nat Pilot
|
posted 01-27- 10:30 PM
mission start at Abbeville, waypoint1 right off the end of the runway, alt 15,000ft, spedd 450, waypoint 2 in a straight line not to far out to sea also set to 15,000ft speed 450. The AI takes off and starts climbing almost imediately and picks the highest sustainable climb and away it goes it actualy gains speed as it climbs, think thats due to the thinner air, but even so, it climbs to it's best ability and using the AI means each test is almost identical, just setup the mission with a full fuel load and let it run. To be honest Tail I thought this must be the way you guys test for climb rates, but yep, it's a lot better than trying to test it yourself cause unlike the AI we have trial and error, but the AI "knows" the best way to climb in any aircraft you put it in  ~Nat~ IP: Logged |
Tailslide Pilot
|
posted 01-29- 02:47 AM
After watching it try to land sometimes I get suspicious of its knowledge but I will give it a try see if it can do as good. Won't it just pick a climb angle that will get it at your set height in the set distance? TS
[This message has been edited by Tailslide (edited 01-29-2001).] IP: Logged |
Nat Pilot
|
posted 01-29- 03:27 AM
Hey Tail, after doing this lots of times with a number of different aircraft, I'm more convienced that this is right.What I maybe didn't mention above was to set the speeds higher than the aircraft can achieve, forcing it to climb as fast as it can. What the AI appears to be doing is basicaly this: It calculates from it's lift, power, drag ratios etc what the optimum "sustained" rate of climb will be, i.e, it's angle of attack to get to the altitude the fastest while maintaing a steady speed. For instance, I can get the La5 to climb very steeply and reach 15,000ft is a short distance, climbing at around 15/18 degrees, or, with a few little changes to the drag foil, it will climb at a higher speed, but not as quickly, at around 10 degrees, I've seen the same results while working on other FM projects. So it seems the AI picks the best angle to climb at to be able to sustain the climb and not lose speed, which from what I know is how a test for rate of climb should be carried out. I mean, OK, we could probably climb faster if we flew, we'd climb sharply then level out to regain speed, then climb again, the AI on th other hand looks to hold to best climb over the change in altitude, in this case from 500 ft to 15,000ft, and timing this gives you an accurate fate of climb, and climb time to a specific altitude. An aircraft with low lift will climb at a high speed, but a low rate. This all actualy makes FMs harder to do I think, atleast to get them accurate. Final test on the La5 with this gives me a climb time of 5mins 40 seconds, only 20 seconds to slow, at the same time, the climb angle is about 10/12 degrees at a rough overall speed of 235mph. This may sound fast, but compaird to the 109K4 which is greatly outclassed by the La5 it's not, it's hitting around 2,500ft/min, which is also about right. If you want I can give you the mission I use for this test, but I've no doubt you can create it yourself. But, the key word in all this is "sustained" we can climb faster, but can't sustain that climb whereas the AI does. ~Nat~ Oh, forgot to answer your point on the AI just climbing to the waypoint. Thats not what it does, see I set waypoint 1 very close to the runway to force the AI to climb straight away, but I set it at 15,000ft, waypoint 2 can be any distance away then aslong as it's not to close, or waypoint 3 is also set to 15,000ft and in a straight line with the others, the AI is then straight from take off trying to get to the altitude and ignores the waypoints other than for direction changes, so it might get to 15k midway between 1 and 2, or closer to 1 or 2 etc or even beyond 2 It's just trying hard to get up there. [This message has been edited by Nat (edited 01-29-2001).] IP: Logged |
Nat Pilot
|
posted 01-29- 03:39 AM
I think this fits:High speed, low AoA = low climb rate low speed, high AoA = high climb rate Problem is this is all affected by engMaxhp, propRadius, propPitch, propElements, wing Airfoils, drag foils and so on, you could make such a mess of all this that it climbs fast at a high AoA with a high rate of climb, but then all your speeds at altitudes would be a mess aswell. Right now the only part I can't get right is low level speed, I'm down by about 25mph, but I now honestly think that this is all down to air thickness not affecting the MaxHP of an engine, if it did I'm sure it would come in spot on since changes I make to increas the low level top speed, also badly affect high level, I think it can only be down to the engine at high alt if pumping out to much power (500hp to much) IP: Logged |
wakeup tailgunner Pilot
|
posted 01-29- 09:58 AM
A little while back, I was involved in a thread which oputlinedsome possible 'standard' test missions to put planes through to assess plane-pack readiness, and to supply information on relative frmae-rates. Nat has hit on another! If Nat's 'climb to altitude' mission, plua a few others were packed together, us FM 'no-nothings' would have some baseline figures to work towards. If the AI flies the missions, then all the numbers wil be relative. So...test the AI take-off test the fr's in controlled conditions (alone / squadron / dogfight )state PC spec too! test the planes top speed under AI control test the climb to altitude under AI control Some testing will have to be 'seat of the pants' stuff, but a set of standard missions, under control of the AI, will give us relative data of how planes perform. Then we could publish a set of relative data in the read-me.
IP: Logged |
Nat Pilot
|
posted 01-29- 10:56 AM
thats a good idea TG, right now I have the following test missions:Climb to 15,000ft from take off (timed from when AI hits full throttle on the runway {full fuel load ) 20,000, 15,000, 5,000 and 1,000ft max speed tests (full fuel) Air Combat 4v4 (use a comparative enemy and watch general AI handling) Strike mission (max loadout, full fuel) Takeoff with hill at end of runway (thought this needed to be tested lol) Each mission is pure AI from the start, on the climb and strike mission you should at the same time pay attention to how the AI's ground handling is, and it's takeoff ability. On the 15,000ft climb you should let the mission run and check the landing ability, but only if you're happy with the climb time ofcourse. There's probably other test missions you could run, and some of these depend on the data you have for the aircraft. IP: Logged |
Tailslide Pilot
|
posted 01-29- 12:25 PM
I usually also test takeoff on that german base on the dover map with a cliff at the end the AI often has problems with that one. TS
IP: Logged |
Nat Pilot
|
posted 01-29- 06:38 PM
Aberville field? the one right to the south on the coast? Thats where my missions start other than the hillside one.. although with that field aircraft have trouble landing, they tend not to lock the brakes after landing and roll over the cliff LOLIP: Logged |
Whirlwind Pilot
|
posted 01-31- 11:41 AM
Finding top speed at a given altitude is easy - just set up a 3 way point mission - 1 way point to line the AI up, the second to start the run, and the third to end it. Way point 1 should set the altitude you want to test at, way point 2 should have speed set a 1000 and give the AI lots of time to fly between wp 2 and 3.IP: Logged | |