Conspicuous by Their Absence

Autopilot?

Hi Ivan,
I was trying to use an autopilot, but to no avail, and the CFS1 help files deny its existence. Manual test-flying for rate-of-climb is of course possible to a certain degree using the elevator trim, although there comes a point where it is insufficient. I managed to time the first 3281 ft though, and it´s bang on the dot as per specs: Exactly 10 minutes from sea-level to 3281 ft!
The other thing is that the plane climbs flying in 4-minute circles - the aileron and rudder trim settings only seem to be meant for FS2000. An Autopilot could correct this. Manual corrections interfere with the climb test.
Would there be a way to implement the Autopilot?
Cheers, and thanks in advance,
Aleatorylamp
 
Autopilot

Hello Aleatorylamp,

I did a little poking around on the autopilot settings for your Giant.
The autopilot DOES work, but the combination of settings is not optimal, so the result is unstable:

The Autopilot uses Trim to try to point the aircraft in the intended direction.
Sometimes, it cant adjust the trim fast enough to correct the direction before it has changed.

Example:
The autopilot is set for maintaining altitude.
The aeroplane is trimmed nose down and losing altitude
The autopilot adds nose-up trim and checks whether the nose is still pointed down.
If
The nose is level, the autopilot does nothing
The nose is pointed up, The autopilot tries to trim nose-down.
The nose is pointed DOWN, The autopilot tries to trim nose-up.

The problem is that on this aeroplane (and many others), the autopilot cannot change trim fast enough, so by the time the nose comes level, there is already a LOT of nose-up trim and the nose continues right on past level and goes nose-up.
As the nose goes UP, the autopilot cranks in nose-down trim but probably has it completely maxed DOWN before the nose comes back to level going down.

This situation repeats and the result is a divergent oscillation.

There could be a number of causes or a combination of these:
1. Your Longitudinal MOI is too high
2. Your Longitudinal Damping is too high (Less Likely)
3. The Trim Pitch Effect in Record 1101 is too low.

The Trim Pitch Effect field was -0.50
I adjusted it to -2.0 and removed the divergent oscillation.
The problem is that when the aircraft comes level, it cannot maintain altitude because the trim notches are too coarse and it either slowly gains altitude or loses it because it can't find a neutral trim setting to actually remain level..

There are a couple different solutions depending on what you are trying to accomplish.
My guess is that the Moments of Inertia are way too high. You might want to confirm that the numbers are correct.
Because of the engine arrangement, I would use a template for a twin engine aircraft.

Sometimes I purposefully set the trim this way to allow for very precise adjustment and just accept that autopilot (which these aeroplanes really didn't have) would only hold altitude when very close to level flight.

Hope this helps.
- Ivan.
 
No Autopilot!

Hi Ivan,
Thanks for your meticulous work, but that would be the next step. In my case, I am still one step behind. The Autopilot, even though activated in the .air file, is of no use. As soon as it is, I can try out the corrections you propose. These 4 records are present in the .air file: 317, 329, 1199, and 1203.

In your case, the Autopilot can be switched on, presumably reacts to the mouse, and can be ordered to do whatever the simmer wishes it to do.
In my case the instrument is a waste of space on the dashboard. It doesn´t work, can´t be switched on, does not react to the mouse, and has no little green lights. Not very decorative either...

Is it my version of CFS1? The help file clearly states that there is no Autopilot in CFS1.
Perhaps it is the 737-400 Autopilot Gauge that won´t work for CFS1 - only for FS98. The Lear45 Autopilot is even worse - it´s invisible except for its frozen numbers. There must be a special CFS1 add-on Autopilot gauge that I´m missing out on. I´ve been trying to find one without success. Any help in this direction will be appreciated!

Cheers,
Aletorylamp
 
Control Panel

My guess is that you have your panel Config set so that controls are not mo usable.
I sent you a copy of what I am using.
If it STILL doesn't work, let me know.

- Ivan.
 
Control Panel

Hi Ivan,
Thank you so much!
I don´t know what I would do without your help. The strange thing is, that the Test-panel stack works perfectly, and the throttle lever does react to the mouse.
OK, I´ll implement the test-system you sent. Thanks again!
Cheers,
Aleatorylamp
 
Control Panel and Autopilot working well!

Hi Ivan,
OK, I used your Test Panel, and of course the Autopilot worked perfectly, but with my Panel.cfg it didn´t!
Comparing the entries, I found the culprit! It was the line "render_3d_window=1".
I use this line to prevent the black windows my graphics card causes, and this apparently blocks the mouse clicks to push the Autopilot buttons, although strangely enough, the mouse does work on other gauges!
Anyway, eliminating the "render_3d_window=1", the Autopilot works now.
Thank you very much indeed, I would have never found the problem otherwise!
Cheers,
Aleatorylamp.
 
Render 3D

I was pretty sure it was something in the config file that disabled mouse input but didn't remember what setting controlled that.

Glad it worked for you.
- Ivan.
 
Of performances and altitudes

Hi Ivan,

Our use of the turbocharger parameter in the .air file to mimic early German high-compression, altitude-rated, throttle gated engines with their low rpm, gives striking results:

Performance increases noticeably as the aircraft reaches rated altitude, and then decreases again after that. There is a definite peak at critical altitude. Also, the increase in power during the first climbing leg, 0 to 3281 ft, is fine. Speeds and times are bang on! Level flight speeds for the two altitudes are of course also correct.

Higher up, climb rate is only about one-third too high, and I´m still trying small changes to the engine torque graph and the propeller efficiency/thrust tables in the higher range RPM/speed areas, but they always seems to affect level-flight speed - to be expected, as you said!

Also, different power/load combinations for testing also yield very similar results, so I think that given the circumstances of the very narrow margin for adjustment allowed by the low RPM and relatively low altitude levels, the approximation that we are getting is very much acceptable! Would you agree? It´s like the doctors who often ask for a second opinion!

Cheers,
Aleatorylamp
 
Hello Smilo,

Can you sticky the Engine Performance Tuning Thread listed below. I find that I keep pointing people at it and it isn't all that easy to find unless you know what you are looking for.

Thanks.

http://www.sim-outhouse.com/sohforums/showthread.php?77148-Engine-Performance-Tuning-Tutorial


Hello Aleatorylamp,

I personally like performance to be a touch higher than advertised, but you may have your own preferences.
If you are getting the proper performance at and below critical altitude but something higher up is not behaving right, you do have the option of adjusting both the Engine Torque Curve AND the Friction Loss Curves.

As a quick summary of the Engine Tuning Tutorial listed above:
The result is that if the Torque is raised and the Friction is also raised, the engine output at and below critical altitude will be unchanged BUT.....
Above critical altitude, the Torque drops while Friction remains the same, thus the net output of the engine is reduced.
Thus high altitude performance is reduced.

Conversely, you can do the opposite to increase high altitude performance.

I thought this was a cool thing when I was experimenting with it a while back.
I am not sure all the numbers match up well with reality, but they DO let us get closer to the performance we want.

- Ivan.
 
Higher altitude tuning

Hi Ivan,
I did read the FW190 engine-tuning thread, but I´m afraid I found it a tad over my head... due to which I am immensely grateful for your useful summary of the inter-relationship between the friction and torque curves and their results!
A few minutes ago I just got a slight improvement by adjusting the Torque Curve a bit more. The testing that I´m doing is with 80% fuel and 80% standard load, so this 30% excess power above critical altitude would be worse with lighter loads. I´d be quite happy with about 10% excess, so
I´ll follow your suggestion and try doing both curves together now.
Thanks a lot!
Actually, I really don´t know what we all would do without your help!!
Cheers,
Aleatorylamp
 
Last edited:
More friction, more torque

Hi Ivan,
It works! (You will think "Of course it does!").
For the 267 Hp engines, looking at the 2 graphs and their scale, I decided to start increasing torque by 0.01 and friction by 5, and it´s almost incredible how performance stays put below critical altitude and gets lower higher up! Now at +0.2 and +10, I´m almost there. Now for another timed climb! Let´s see what happens.
It is as you say, really cool!
Thanks again,
Aleatorylamp
 
Thanks Aleatorylamp,

Actually I think you all would get along just fine without any help from me.

I am just saving you a bit of time because with the exception of the propeller tuning, I have pretty much done everything that you are trying with the Giant. I can tell you though that all the factors I have commented about in engine and propeller and general AIR file tuning took me years to figure out and I am still not certain about the meaning of a LOT of the parameters.

I have also never tried to do everything you are doing in a single AIR file. (Which should tell me that there are a few older projects I should probably update to more current standards.)

Then again, there are certainly a lot of folks who build pretty decent AIR files without all this detail we are getting into. I am just the kind that likes to understand the reasons for what I am doing.

Keep us updated.
- Ivan.
 
Narrow margins to work within

Hi Ivan,
I realize I am getting the benefit of your experience - whatever tip or suggestion comes my way, so fresh and useful, is thanks to the effort you have gone through in the past.

The several different aspects we have been working on with this engine, have the added difficulty of the narrow RPM and altitude margins within which it operates, added to the abstract aspect of the numbers and graphs in the .air file. The better one understands how they work, the more successful one will be.

In my last post I mentioned some significant results I was getting with small increases in Torque and Friction. Actually, the significance was rather more numerical than practical, as a difference of 25 FPM looked like a lot, but translated into climbing time differences only meant very little.

From the initial 35.0 friction entry (approx 12% Horsepower), in order to get any real benefit to slow down climbing time higher up, I had to more than double Friction to 80.0, compensating Torque, of course. My initial estimation of 5 friction units for 0.01 torque was more or less OK, but had to be increased slightly more for higher friction values. Perhaps 0.15 is better.

Now instead of being 1/3 too fast, it has improved to about 1/4, which is quite good given the circumstances.
Another matter is trying to imagine what´s going on inside this engine with twice the friction and augmented torque! I wonder if it is suffering or bitterly complaining!! Friction at 80.0 ... Oh!, what fun!

Of course... we are trying to make the simulator, which is excellent at modern, fast, high-flying supercharged piston engines, do something it was not exactly designed for: To make a 100-year-old slow, rumbling engine loaf around within lower altitude and speed margins - and one of the ways to make this work is to increase friction and torque really "unreally", to get a more real result.

I think I won´t raise friction and torque any further... in case something starts smoking and groaning!
Thanks again for your support!
Cheers,
Aleatorylamp
 
Climb Rate versus Time to Altitude

Hello Aleatorylamp,

Just as a note, I reduced the Friction by quite a lot when I first started on the Giant's Engines.
Don't worry about mechanical concerns. There really isn't anything mechanical or practical about what we are doing here.
The numbers are going to come out a little weird because of the way we are using engine performance limitations to reproduce a procedural limitation.

If I were you at this point, I would work on Climb RATES instead of TIME to Altitude. It will end up being a lot more consistent even if the times are too short, because climb rate and acceleration are quite related.
You can use the Autopilot to ease the work load.

- Ivan.
 
Climbs

Hi Ivan,
OK, so a value of 80 is within limits. The way I do a flight test is set the wing-leveller of the autopilot, set elevator-trim to a little under max. and later regulate it a bit to stabilize at a reasonable rate of climb (295-310 fpm) at a reasonable speed (46-48 kt) for this plane, set the red printout at the top of the screen to give me KTAS, and give full power. Here the fun starts...
I´ve printed out a form-table where I first write down the parameters I have altered, then lift-off time, the times and KTAS, FPM, RPM and HP for different altitudes: 500, 1000, 2000, 3281, 4300, 5000, 6562, 8000 and 9843 ft. Then I do the level flight tests for 500 and 4300 ft.
This gives me a comprehensive table which I can look at and see how things vary. Upto now I´ve done about 18 tests, 6 of them with parallel increases on friction and torque settings. I can also note down the fuel consumption to set the fuel for a given altitude if I only want to do a partial test.
I don´t use the autopilot for other than the wing-leveller (I wish I could eliminate the leftwards yearning) - I don´t really need it, and the MOI´s won´t allow it anyway, as reducing them starts a lot of vibration and things, so I thought I´d give it a miss. Also, according to several sources, the slow response to joystick movements is realistic for this plane.
I´ll keep you posted!
Cheers,
Aleatorylamp
 
Performance tests

[h=2]Performance tests[/h]
Hi Ivan,
I thought I´d dare and set up some numbers, for the climbing tests, at the cost of being perhaps boring... but they do illustrate how this is working in spite of our narrow manoeuvering margin.
For example, a summary of the last test with normal (12% Horsepower) friction at 35.0 - Times are in minutes and seconds:
-------------------------------------------
Ft Alt:...500.....3281....4300.....6562.....9843
KTAS......46.......48........49.......50.........52
FPM......287......312......322......268.......191
RPM.....1264....1307....1333.....1352.....1344
HP.........204.....223......227......208.......182
Times:...1:45...10:30...14:00...21:15....35:30
aim:................11:00.............27:00....55:00
time diff:.........- 0:30...........- 5:45.....-19:30

Lev.flight 4300 ft: 70kt, 257 hp, 1436 RPM.
Aim:.................. 70kt, 267 hp, 1450 RPM.
500 ft:............... 65kt, 217 hp, 1320 RPM.
--------------------------------------------------
Now, the best so far with Friction increased to 80.0, Torque correctly compensated:
--------------------------------------------------
FT Alt:......500.....3281.....4300.....6562.......9843
KTAS........46.........48.........49........50...........53
FPM........292........308.......308......246..........164
RPM.......1280......1333......1348....1340........1322
HP...........215........223.......223.....200..........172
Times:.....1:50.......10:50....14:00...22:10......38:20
aim:.....................11:00...............27:00.....55:00
time diff:.............- 0:10...............- 4:50...-16:20

Lev.flight 4300 ft: 70kt, 257 hp, 1448 RPM.
Aim:.................. 70kt, 267 hp, 1450 RPM.
500 ft:............... 64kt, 216 hp, 1319 RPM.
----------------------------------------------------
Results are very similar below rated altitude, but readings above that, are better by 1 and 3 minutes. A slight torque reduction gave even better climb results there, by another 1 and 2 minutes, but at the cost of level flight, which was 3 Kt too low because of insufficient RPM for the turbo to set in.

Well, now I´ll see what happens with Friction at 100!!
Cheers,
Aleatorylamp​
 
Last edited:
Choosing What to Measure

Hello Aleatorylamp,

Interesting numbers you have there. I see what you are aiming for but can't quite figure out how the climb rate matches up with the time to altitude.

Without knowing that, any advice I can give will be general at best.

I still suggest you go for climb rates rather than time to altitude. It is a much faster measurement and you also know if your rates are very non-linear and how they are so.
I generally mess with fighters and am more concerned with the available surplus power (climb rate) at an altitude more so than how long it takes to reach an altitude at the 30 minute rating. Are you SURE that on the real aeroplane your time to altitude was measured with absolute full power? That is an odd situation.

I believe that you should be figuring out your power requirements at
Sea Level
Critical Altitude
and
Service Ceiling

This will save you time in the long run and give a better feel for the aeroplane than a an absolutely correct time to height. It is the handling and maneuverability people will notice rather than a climbing time that is 20 seconds too fast.

Once you know what your target HP is at each altitude, it only takes seconds to check it and tune again if needed.

BTW, if your climb rate is too high and everything else is nearly in line, perhaps changing your Induced Drag (Wing Efficiency) might be worth a try again.

- Ivan.
 
Climb rates

Hi Ivan,
Thanks for the post! It´s difficult to arrange the columns and lines for "tables" in the post format, as spaces seem to appear and disappear ad lib.
I´m afraid I don´t really understand your comment, for the climb rates are mentioned in the table, and is what I go by. The info I have of the 11, 16, and 27 minutes it takes to get to 3281, 6562 and 9843 ft, give me average FPM climb rates, and are the performance parameters I go by. The critical altitude is 4300 ft.

In the table, where I had to put in dots instead of spaces, the top line is the altitude, the next one, the speed TAS Knots, then comes FPM which is the rate of climb I get at each altitudes. Then, the times are how long it takes to get there from lift-off, the aim is what I´m looking for and the difference is how far off I am. Below that are the level flight parameters and what I´m aiming at.

Meanwhile, I´ve done the 100 friction test, and this time I took it up to the ceiling.
On one hand, the borderline between the turbo being able to set in or not, is much narrower because of the RPM needed, which is "braked" by the friction, and on the other, slight over-compensation here causes Hp numbers to wildly overshoot and fluctuate. Obviously, some parameter combination limit exists here.

Anyway, I get two options here:
1) A "performance" tuning, where there is an interesting effect: The FPM and HP peak at rated altitude is more pronounced and really noticeable as you climb, and the beginning and end results of the test are lower. Timing results are practically identical to the 80 friction test.
2) A more conservative tuning, which is less exciting, where I can squeeze out another minute slower getting to 6562 ft and another 2 minutes slower getting to 9843 ft.
Personally, I prefer the first!
Overall, instead of taking the specified 150 minutes to the 12467 ft ceiling, it took 58, and the climb rate drops to 93 fpm.

I think it is satisfactory and acceptable, and the flying feeling is fine!
Well, I´ll try the induced drag setting now! It may help to slow down the plane higher up then.
Thanks a lot once again!
Cheers,
Aleatorylamp
 
Last edited:
Hi Ivan,
In the documentation I have, the fact being that these airplanes had no fighter-plane characteristics, on their normally long bomb runs, throttles were mostly left alone after passing rated altitude.
A cruise-speed also existed, but was only a bit lower than max. speed, to allow flight in formation. This was often done with the Gotha G-planes, but not with the Giants, who usually flew individual missions, and if 2 or 3 were involved, there was no formation flying.
Not being fighters, there was no need for WEP. Neither was there a 5-minute max. take-off power because throttles were gated on the ground anyway. The pattern seemed to have been: Take off at 75% power, circle to cruising altitude, gather there if flying in formation, and then give full power to head for their destination.
They were all heavily armed, had protective plating for all crew positions, could take a lot of punishment due to their sheer size, and had no escorts. If attacked, pilots did some slow evasive manouevers but not much more, and gunners gave effective and precise defensive machine-gun fire in all possible directions.
Thus, maximum power was equal to maximum continuous power, and all had auto-mixture. There was no need for WEP or 5 minute-full power. Also, to safeguard the engines that broke if pushed beyond these limits, this was not a criteria.
This seems to apply to both Mercedes and Maybach D.IVa engines, which despite their greater power, were both not used on fighters because of their weight and slower RPM. Lighter fighter aircraft engines of the DIII class also had an "over-compressed" DIIIa high-altitude engine type, (Mercedes and Maybach) with automatic carburettors to limit low altitude power. Here there definitely was a maximum time limit for full power during combat, a maximum continuous power to head for a destination, and a cruising speed for returning home or loitering.
Incidentally, the Maybach engines were more powerful at higher altitude than the Mercedes ones, as they were more "over-compressed", in spite of being slower lower down.

So, now we have some more numbers:
Here is the 100-friction test, torque compensated for performance.
Timing Results are almost identical to the previous 80-friction test, but the balance is much more delicate;
Slight under-compensation causes turbo failure, slight over-compensation causes Hp surge and fluctuation.
Peak Hp, RPM, and RoC(FPM) at 4300 ft critical altitude is more apparent, and very noticeable during flight!Also, level-flight Hp is more exact. I like it!
--------------------------------------------------------------
FT Alt:......500....3281...4300.....6562....9843....1 1300.....12467

KTAS..........46....48......49........50.......53. ...........55.........56
RoC(FPM)..294....308.....357......244......156.... ....113.........93
RPM........1276...1338....1388....1342....1320.... ..1307.....1300
HP............210...225......242.....202......171. .......158.......148

Min:sec..1:45.....10:40...13:40...21:43...38:25... ..50:50....58:00 >time to altitude
aim:.................11:00.............16:00....55 :00..............150:00 >available spec. times
time diff:...........-0:20............-4:57....-16:35..............-92:00 >how many sec. too fast

Lev.flight 4300 ft: 70kt 266 hp 1436 RPM -Correct!
Aim: 70kt 267 hp 1450 RPM -goal performance
500 ft: 64kt 213 hp 1308 RPM -definitely acceptable
----------------------------------------------------------------------------
A slight reduction in engine-torque will prolong time to 6562 ft by another 1 minute, and to 9843 ft another 2minutes, but at 2 prices:
1) Loss of power all round, including the prominence of the peak at critical altitude.
The difficulty here is that the affected RPM range covers climb speeds both above and below critical altitude, and also level flight below it.
2) There´s lower level flight performance at critical altitude. Although this can be corrected by adjusting Zero lift Drag, it messes up lots of stuff elsewhere.

Now I´ll try out the induced drag adjustment you suggest, but I think that this is actually already quite OK now.
Cheers,
Aleatorylamp
 
Last edited:
induced drag

Hi Ivan,

Just a short note - I had already lowered induced drag when you mentioned it as a possibility to help to reduce performance higher up. There´s no improvement with further reductions (simultaneous balancing-out increase in Zero Lift Drag included), so methinks that perhaps the point of "as good as it can get" may have been reached.

What would you think? I could now upload the plane with the present .air file, which flies very nicely (after all our efforts it should, too!!), and further tweaks, if any, could be included in a possible FD update if need be.

Cheers,
Aleatorylamp
 
You must log in or register to reply here.
Back
Top