Cfs1 furball

[Ivan]

Hello Rince33,

The 1% Flight Model is probably better than the stock flight model though I can't say I have flown it.
The Hurricane wasn't really the most agile aircraft around. If you are interested, there are several flight evaluations available online of modern restored Hurricanes though I don't believe any of them are Mk.I types. There are a few surprising gotcha;s with the Hurricane.

As for the 1% concept, in my opinion, the idea wasn't too bad though it had its faults in implementation, but some of the data that it was based on was not particularly accurate. I spent a LOT of time trying to address some of the issues of faulty data especially with armament. Their engine tables are an interesting read, but are not particularly accurate when compared to the aircraft manuals for the aircraft for which I have done a comparison. The spreadsheets also do not give nearly enough information to actually build an engine with reasonable power curves.

As for an example of exactly why I do not have faith:
CFS1 only seems to handle Single Speed Superchargers.
For a Single Speed Supercharger, there is the power at Sea Level at maximum Manifold Pressure and power at Critical Altitude above which the Supercharger cannot maintain Sea Level Manifold Pressure. At that critical altitude, the engine output will be pretty comparable to its sea level output.
At Intermediate altitudes, its power may be slightly higher than at Sea Level.

With a Two Speed Supercharger, things are a bit different.
I will use the Shvetsov M-82FN engine from the La-5FN as an example.
At Sea Level, Maximum non WEP manifold pressure is 1000 mm Hg giving 1650 HP.
Critical Altitude for Low Blower is 1650 Meters. Above this, Low Speed cannot maintain Sea Level boost pressure.
Switching to High Blower doesn't work because that would supply more boost than the engine can tolerate.
IIRC, the Blower shift altitude is around 4000 Meters.
The engine power still increases slightly after the shift point and reaches its maximum at 4650 Meters (about 15,200 feet)
Above that, the Supercharger cannot maintain Sea Level Boost.
The problem though is that High Blower draws more power from the engine than Low Blower did and instead of 1650 HP with 1000 mm Hg of boost, it now only gives 1450 HP with the same boost.

From 1650 Meters to 4000 Meters, the real M-82FN engine will provide something BELOW 1650 HP.
It will exceed 1650 HP but only for a few hundred meters above Sea Level and a couple hundred meters above 4000 meters.

With our single speed CFS1 supercharger, There is a HUGE altitude range in which the engine power is higher than it should be: Probably from about 2500 feet to 15000 feet if we get the boost pressure correct.
If we try to get engine output correct, the excessive power probably happens from around 2500 feet to 12500 feet.

If we match speeds at Sea Level and at Critical Altitude, then we will exceed maximum speed at intermediate altitudes and probably by quite a lot.
My belief is that I would much rather have speed at Sea Level and Critical Altitude a bit lower than they should be and not have speeds at intermediate altitudes be quite as excessive.

If someone knows of a better method, please let me know.

- Ivan.

[rince33]

Hi Ivan.
Thanks for the detailed and illuminating reply. You've prompted me to read up a bit on hurricane engine performance.

I'll probably stick with the 1% model for a while and see how it plays out in the campaign.

How accurately does Cfs1 portray real world physics? If it's not that accurate then whatever flight data is used by the model, authentic or not, will be inacurate. Just wondering about this.


Cfs1 flight model does seem to fly differently to most older sims, there seems to be a bit more lateral movement in turns when trying to get on the tail of enemy ai which makes the process more difficult and more authentic I think. This is more pronounced in the 1% version where the rudder effects seem weaker.

[Ivan]

Hello Rince33,

How accurately does Cfs1 portray real world physics? If it's not that accurate then whatever flight data is used by the model, authentic or not, will be inacurate. Just wondering about this.

From what I have heard and from my own limited experience, the simulator in CFS1 which isn't greatly different from later sims does a pretty fair job as long as the attitude of the aircraft is "Normal": More or less with the nose in the direction of travel. As soon as the aircraft starts tumbling, going sideways, backwards and presenting some other undefined profile to the "Airflow", the calculation of movement becomes questionable.
ALL desktop class flight simulators have this problem.
If you want to do better than this, you need to get the software that engineers and the military uses and that stuff from what I have heard is generally about the cost of a new car.

Within the capabilities of the desktop simulators, the accuracy of the flight model has a pretty great effect on how "correctly" the virtual aeroplane will mimic the handling and performance of the real aircraft. Often the data that is needed isn't available so we as designers need to take a best guess. There are also many ways to get the same effect. This is why I commented that Aleatorylamp and I didn't agree on Superchargers or Propeller Tables. I believe I am correct and he probably believes he is correct.

As for "Data" on handling, the P-39 Airacobra that we each designed is a pretty good example. We both have heard that the aircraft had a CoG that was too far aft when unloaded and that it was unstable. We each took a different approach.
I went through a lot of reports and believe my flight model mimics the behaviour pretty well with characteristics I expect to find.
He probably did something similar. The big question here is what does "unstable" mean to each of us? None of us has actually flown an Airacobra.
The differences in interpretations means the results we each got were quite different.

Cfs1 flight model does seem to fly differently to most older sims, there seems to be a bit more lateral movement in turns when trying to get on the tail of enemy ai which makes the process more difficult and more authentic I think. This is more pronounced in the 1% version where the rudder effects seem weaker.

The AIR file in CFS1 has a lot more ability to control stability and moments in various directions than prior flight models such as FS98. Whether it is done correctly for a particular aircraft is another story entirely. I have heard that various stock aircraft were intended to be flown with different simulator settings. The "Easy" setting makes aircraft much more stable and I believe the Hurricane was actually intended for that. When stability is generally reduced in the "Hard" simulator setting, perhaps the flight model (AIR File) isn't a good match with what the simulator is doing.

Just about all the issues you are describing can be pretty seriously altered by editing the AIR file in the correct places.
As for the 1% business, all their system really addresses is straight line performance. They do not make any claims to addressing handling and I believe that handling affects the perception of agility and maneuverability much more so than straight line performance. (How many players complain that the Me 109G is 100 MPH too fast???)
I can give you two aircraft that have the same maximum speeds, climb rates, stall speeds, ceiling, etc. with one that you can aim instinctively, maneuver in tight places, and is a killer in a close in dogfight and another that can't do any of those just by tuning the handling a bit.
This is actually what I spent a couple hours doing for the La-5FN two days ago. The original version was much too responsive and agile and didn't have the gotchas that the real one did. I am still not done and even when I am, the result will be subject to debate: When the report says the ailerons get stiff at about 370 MPH, HOW much roll rate is lost???? I don't really know so I have to guess a bit.

- Ivan.